The Cambridge Economic History of China: Ebin A

the cambridge economic history of
CHINA
China’s rise as the world’s second-largest economy surely is the

most dramatic development in the global economy since the year
2000. Volume II, which spans China’s two turbulent centuries
from 1800, charts this wrenching process of an ancient empire
being transformed to re-emerge as a major world power. This
volume for the first time brings together the fruits of pioneering
international scholarship in all dimensions of economic history to
provide an authoritative and comprehensive overview of this
tumultuous and dramatic transformation. In many cases, it
offers a fundamental reinterpretation of major themes in
Chinese economic history, such as the role of ideology, the rise
of new institutions, human capital and public infrastructure, the
impact of Western and Japanese imperialism, the role of external
trade and investment, and the evolution of living standards in
both the pre-Communist and Communist eras. The volume
includes seven important chapters on the Mao and reform eras
and provides a critical historical perspective linking the past with
the present and future.
D E B I N M A is Professor of Economics at Hitotsubashi University,

Tokyo, Japan.
R I C H A R D V O N G L A H N is Distinguished Professor of History at
the University of California, Los Angeles.
Published online by Cambridge University Press

t h e ca m b r i d g e e c o n o m i c h i s t o r y of
CHINA
Building on a wide array of recent scholarship, the two volumes
of The Cambridge Economic History of China bring together the
fruits of pioneering international studies in all dimensions of
economic history, past and present. Exploring themes including
political economy, agriculture, industry and trade, technology,
ecological change, demography, law, urban development,
standards of living, consumption, financial institutions, and
national income, the two volumes together provide broad
temporal coverage across all of Chinese history, including
recent developments in contemporary China.
VOLUME I
To 1800
edited by debin ma and richard von glahn
VOLUME II
1800 to the Present

THE CAMBRIDGE
ECONOMIC HISTORY OF
CHINA
*
VOLUME II
1800 to the Present
*
Edited by
DEBIN MA
Hitotsubashi University, Tokyo
and
RICHARD VON GLAHN
University of California, Los Angeles

University Printing House, Cambridge C B 2 8B S, United Kingdom
One Liberty Plaza, 20th Floor, New York, N Y 10006, USA
477 Williamstown Road, Port Melbourne, V I C 3207, Australia
314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre,
New Delhi – 110025, India
103 Penang Road, #05–06/07, Visioncrest Commercial, Singapore 238467
Cambridge University Press is part of the University of Cambridge.

It furthers the University’s mission by disseminating knowledge in the pursuit of
education, learning, and research at the highest international levels of excellence.
www.cambridge.org
Information on this title: www.cambridge.org/9781108425537
D O I: 10.1017/9781108348485
© Cambridge University Press 2022

This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First published 2022
Printed in the United Kingdom by TJ Books Limited, Padstow Cornwall
A catalogue record for this publication is available from the British Library.
Two-Volume Set I S B N 978-1-107-14606-8 Hardback
Volume I I S B N 978-1-108-42557-5 Hardback
Volume II I S B N 978-1-108-42553-7 Hardback
Cambridge University Press has no responsibility for the persistence or accuracy of
URLs for external or third-party internet websites referred to in this publication
and does not guarantee that any content on such websites is, or will remain,
accurate or appropriate.

CHINA


Tokyo, Japan.

CHINA
VOLUME I
To 1800
VOLUME II
1800 to the Present

THE CAMBRIDGE
ECONOMIC HISTORY OF
CHINA
*
VOLUME II
1800 to the Present
*
Edited by
DEBIN MA
and
RICHARD VON GLAHN


www.cambridge.org
D O I: 10.1017/9781108348485


CHINA


Tokyo, Japan.

CHINA
VOLUME I
To 1800
VOLUME II
1800 to the Present

THE CAMBRIDGE
ECONOMIC HISTORY OF
CHINA
*
VOLUME II
1800 to the Present
*
Edited by
DEBIN MA
and
RICHARD VON GLAHN


www.cambridge.org
D O I: 10.1017/9781108348485


CHINA


Tokyo, Japan.

CHINA
VOLUME I
To 1800
VOLUME II
1800 to the Present

THE CAMBRIDGE
ECONOMIC HISTORY OF
CHINA
*
VOLUME II
1800 to the Present
*
Edited by
DEBIN MA
and
RICHARD VON GLAHN


www.cambridge.org
D O I: 10.1017/9781108348485


Contents
List of Figures page viii

List of Maps xi
List of Tables xii
List of Contributors to Volume II xv
Acknowledgments xvii
Note on Citations xix
Introduction to Volume II 1
debin ma and richard von glahn
part I
1800–1950
1 . Ideology and the Contours of Economic Change 15

debin ma
2 . Economic Transition in the Nineteenth Century 48

william t. rowe
3 . Agriculture 87
debin ma and kaixiang peng
4 . Handicraft and Modern Industries 124

linda grove and tōru kubo
5 . The State and Enterprises in Late Qing China 167

chi-kong lai
6 . State Enterprises during the First Half of the Twentieth Century 184
morris l. bian

Contents
7 . Money and the Macro-economy 208

dan li and hongzhong yan
8 . Public Finance 244

elisabeth kaske and may-li lin
9 . Financial Institutions and Financial Markets 280

brett sheehan and yingui zhu
10 . Chinese Business Organization 324

madeleine zelin
11 . The Economic Impact of the West: A Reappraisal 354

james kai-sing kung
12 . Foreign Trade and Investment 414

wolfgang keller and carol h. shiue
13 . Transport and Communication Infrastructure 457

elisabeth köll
14 . Education and Human Capital 496

pei gao, bas van leeuwen, and meimei wang
part II
1950 TO THE PRESENT
15 . The Origin of China’s Communist Institutions 531

chenggang xu
16 . China’s Struggle with the Soviet Growth Model, 1949–1978 565

dwight h. perkins
17 . Living Standards in Maoist China 606

chris bramall
18 . The Political Economy of China’s Great Leap Famine 642

james kai-sing kung
19 . China’s External Economic Relations during the Mao Era 685

amy king
vi

Contents
20 . The Chinese Economy in the Reform Era 722

barry naughton
21 . China’s Great Boom as a Historical Process 775

loren brandt and thomas g. rawski
Index 829
vii

Figures
1.1 Real index of Chinese and Japanese per capita GDP (1850 = 100) page 19
1.2 Real value of Chinese imports and exports at 1933 constant prices
(thousand yuan) 21
1.3 Western impact and Chinese response: number of newly established
modern firms and banks 22
1.4 Real production and GDP indices (1912 = 100) 23
1.5 Metal currencies (silver and copper) and money supply in China,
1911–1936 (million yuan) 25
3.1 China’s historical crop yield 91
3.2 Per capita output of raw grain, 1500–1950 92
3.3 Distribution of Qing crop yields based on local gazettes 94
3.4 Wage index relative to grain price in Beijing and Hebei province 96
3.5 Major disasters with a death toll above 1,000 (1840–1949) 99
3.6 Malthusian and Boserupian models 101
3.7 Percentage of idle time (left axis) and wage rates in units of copper
cash (right axis) 105
3.8 Agricultural sideline production 106
3.9 Trends of agricultural commercialization 110
3.10 Diffusion of scientifically produced silkworm varieties in Japan and the lower
Yangzi 116
4.1 Industrial production in mainland China, 1933 126
4.2 Imports of machinery, iron, and steel, 1886–1915 129
4.3 Silk reeling filatures 135
4.4 Silk exports by production source 139
4.5 Industrial production index, 1938–1945 154
4.6 Capital of newly built mills, classified by industry 155
4.7 Industrial production index, 1912–1949 161
7.1 The capital power of China’s banking industry in 1894, 1925, and 1936 217
7.2 Banknote issuance in China from 1890 to 1936 221
7.3 The modern-oriented gross domestic fixed-capital formation
in China from 1903 to 1936 236
7.4 The interbank annual interest rate in Shanghai from 1872 to 1936 237
8.1 Lijin revenue, 1858–1908 251
8.2 Foreign and Native Customs 253
11.1 Expansion of the treaty port system 361
viii

List of Figures
11.2 Growth in the number of treaty ports and self-initiated ports 362
11.3 Geographical distribution of treaty ports and self-initiated ports 365
11.4 Foreign direct investment and debt 367
11.5 Foreign investment and foreign debt by country 368
11.6 Geographic distribution of foreign private firms 369
11.7 Foreign firms, by country of origin 370
11.8 Sectoral distribution of foreign firms 371
11.9 Domestic trade network 372
11.10 Numerical growth of foreign firms and banks 374
11.11 Geographical distribution of foreign firms and banks (1920s) 375
11.12 Rise of domestic modern firms and banks in China (total trend) 377
11.13 Geographical distribution of domestic modern firms and banks (1920s) 378
11.14 Development trend of domestic modern firms and banks 380
11.15 Development of domestic modern firms in non-port regions 381
11.16 Spatial distribution of treaty ports opened before 1865 and munition factories 382
11.17 Growth of electrical power plants in China, 1903–1931 385
11.18 Electricity generation in China, 1929 386
11.19 Adoption of the steam engine and machinery in modern Chinese firms 387
11.20 Growth of missionary activities in China 391
11.21 Missionary activities and the primary schools they founded (1922) 392
11.22 Growth of universities in China 395
11.23 Growth of universities in China, by type/funding source 395
11.24 Distribution of universities in 1937 396
11.25 Distribution of engineers in 1937 400
11.26 Number of Chinese students studying abroad and returning, a partial sample 402
11.27 Geographic distribution of constitutionalists and revolutionists (1900–1911) 407
12.1 China’s foreign trade, 1865 to 1940 434
12.2 Predicted versus actual bilateral trade volume for Shanghai 438
12.3 Foreign firms in China: the case of Shanghai, 1872–1921 441
12.4 Foreign firms in China by country of origin 442
12.5 Export and import flows to and from Shanghai 443
12.6 The size of bilateral trade between regions 445
12.7 The impact of foreign influence in China: geographic effects 449
12.8 New and disappearing goods: China’s exports 454
14.1 The school system before 1905 499
14.2 The civil service examination system and its associated degree titles 501
14.3 The education system after 1905 510
15.1 The institutional genes of the Chinese empire: an institutional trinity 535
15.2 The institutional genes of the imperial junxian system 537
15.3 The institutional trinity of the RDT/RDA system 539
15.4 Stylized governance structure of China’s RDT/RDA central–local regime 540
16.1 Urban population and employment growth 1952–1990 590
17.1 Gini coefficients for rural per capita income, 1934–1978 628
17.2 Per capita rural consumption of meat and grain in Guizhou, 1938–1978 632
17.3 The urban–rural gap 637
18.1 Average GDP per capita (in constant US$), 1950–1960 655
ix

List of Figures
19.1 Chinese total trade, 1950–2006 687

19.2 Chinese exports to and imports from the Soviet Union, 1950–1978 695
19.3 China’s trade with Communist and non-Communist countries, 1950–1965 700
19.4 China’s bilateral trade with leading trade partners, 1950–1978 706
19.5 China’s imports of rolled steel and chemical fertilizer, 1950–1960 714
19.6 China’s trade with Canada and Australia, 1955–1965 717
20.1 Consumer inflation (1979–2019) 756
20.2 Budgetary revenues and expenditures (share of GDP) 760
20.3 Industrial workers (total and state) 768

Maps
3.1 Agricultural areas of China page 89

7.1 Cities with a direct money-transferring link with the major
financial centers of Shanghai, Hankou, and Tianjin in 1924 218
9.1 Locations of the branches of six large remittance houses, 1870s–1880s 288
9.2 Bank offices (foreign and Chinese), 1915 289
9.3 Banks in 1947 290
9.4 Cash shops in 1947 291
9.5 Insurance company and co-operative treasury offices, 1947 293
12.1 China Maritime Customs stations and treaty ports 431
12.2 The impact of foreign influence on local capital markets 450
13.1 China’s railroads, c. 1900 468
13.2 China’s railroad network, c. 1935 482
18.1 Famine mortality by province, prefecture, and county, 1959–1961 651
xi

Tables
2.1 Net flow of silver to China, 1818–1850 page 74

2.2 Silver revenue collected at Board of Revenue treasury, 1802–1850 80
3.1 Mean and distribution of agricultural yields (shi per mu) in Qing China 93
3.2 Farm size and productivity 103
3.3 Tenancy rates in the 1930s 118
3.4 Explaining interest rates across regions 121
4.1 Market share of domestic products, 1920–1936 141
4.2 Production and trade of machine-made cotton yarn and cotton pieces 143
7.1 M2/GDP in China, the UK, and the US in the late nineteenth and
early twentieth centuries 239
8.1 Budget for the 26th Year of Guangxu (1900) in kuping taels 256
8.2 National revenue for 1911 (Ministry of Finance) 257
8.3 Revenue structure of the central government, 1913–1945 262
8.4 Income and expenditure of the Nationalist government, 1928–1937 269
8.5 Income and expenditure of the Nationalist government, 1937–1945 276
9.1 Basic functions of financial institutions in the late Qing period 282
9.2 Forms and ownership governance of financial institutions in the late Qing
period 283
9.3 Estimated numbers of financial institutions in China 285
9.4 Establishment of county/city co-operative treasuries, 1937–1944 301
9.5 Growth of the capitalization of co-operatives, 1937–1945 302
9.6 Survey of rural finance in 1943 303
9.7 Survey of rural finance across China (cash lending) in 1941 305
9.8 Uses and percentage distribution for co-operative loans in
Guangxi province, 1938–1941 306
9.9 The China Merchants Steam Navigation Company assets and liabilities prior to
1880 308
9.10 Loans from Shanghai’s Fukang (福康) cash shop to industrial
enterprises, 1899–1907 310
9.11 Index of domestic bond transactions and volume of transactions, 1926–1937 314
9.12 Stock index and volume of stock and corporate bond transactions, 1931–
August 1937 315
9.13 Native provinces of bankers in 1936 318
11.1 Locations of self-initiated ports 366
11.2 Impact of foreign firms and banks on domestic modern firms and banks 379
xii

List of Tables
11.3 Location of munition factories 383

11.4 Adoption rate of the steam engine and machinery 388
11.5 Long-term effects of treaty ports and self-initiated ports 389
11.6 Comparison of curricula in the 1900s 393
11.7 Determinants of university locations and engineers 397
11.8 Determinants of overseas studies 403
11.9 Effect of overseas studies on political participation 408
12.1 Average trade shares, 1865–1900 435
12.2 Major sources of Chinese imports, 1900–1946 436
12.3 World merchandise trade by country 436
12.4 Business investments in China by country 439
12.5 Geographical distribution of the direct business investments
of four countries, 1931 440
12.6 Lower geographic barriers and welfare 446
12.7 Measuring the appearance and disappearance of goods 451
13.1 Freight turnover by modern means of transportation (in million ton-miles) 491
14.1 Estimation of the enrollment ratio of sishu 506
14.2 Number of post-elementary schools in China proper 508
14.3 Regulatory curriculum models for primary schools (percentage) 511
14.4 Enrollment rates per 1,000 school-age population, 1900–1950 516
14.5 Composition of persons by education level for selected counties, 1880–1920
(percentage) 517
14.6 Composition of secondary schools by type (percentage) 518
14.7 Number of engineers by birth cohort 520
14.8 Gender composition in secondary schools (1912–1946) 523
16.1 Annual plan completion record 571
16.2 Impact of the Great Leap Forward on the economy 585
16.3 Wages and the urban cost of living (1957–1964) 591
16.4 Price distortions in GDP growth estimates 597
16.5 Sources of growth (1952–1990) 598
16.6 Per capita rural economic performance 600
16.7 Urban employment and consumption estimates 601
16.8 Urban–rural per capita consumption ratio 604
17.1 Per capita net peasant income, 1954–1978 611
17.2 Alternative estimates of national per capita peasant income 612
17.3 Selected per capita urban incomes, 1943–1980 614
17.4 Trends in food consumption during the Maoist era 617
17.5 Contrasting estimates of food consumption, 1954–1978 621
17.6 Rainfall deviations in 1976–1978 and 1982–1984 from the 1951–2000 norm 623
17.7 Life expectancy and infant mortality in China, 1952–1981 625
17.8 Coefficients of variation (CVs) for provincial per capita net
peasant income, 1954 and 1978 630
17.9 Per capita grain output by agricultural region, 1977 634
18.1 Weather shocks and grain output, 1953–1966 663
18.2 The effects of grain output, political rank, and CC membership on grain
procurement, resale, and excess procurement, 1957–1965 672
xiii

List of Tables
19.1 Chinese export levels, 1913–2003 687

19.2 China’s five largest trading partners, 1950–1952 699
20.1 Development indicators, 1978 724
20.2 Twelve key provisions of the 1993 Resolution on Creating a Market System 764
xiv

Contributors to Volume II
M O R R I S L. B I A N is Professor in the Department of History at Auburn

University.
C H R I S B R A M A L L is Professor in the Department of Economics at SOAS
University of London.
L O R E N B R A N D T is Noranda Chair Professor in the Department of
Economics at the University of Toronto.
P E I G A O is Assistant Professor of Economics at Yale–NUS College,
Singapore.
L I N D A G R O V E is Professor Emerita, Sophia University.
E L I S A B E T H K A S K E is Professor of Modern Chinese Society and Culture at
Leipzig University.
W O L F G A N G K E L L E R is Professor of Economics at the University of
Colorado.
A M Y K I N G is Associate Professor at the Coral Bell School of Asia Pacific
Affairs, Australian National University.
E L I S A B E T H K O¨ L L is Professor and William Payden Collegiate Chair in the
Department of History at the University of Notre Dame.
T Ō R U K U B O is Professor Emeritus at Shinshu University.
J A M E S K A I - S I N G K U N G is Sein and Isaac Souede Professor in Economic
History at the Faculty of Business and Economics, Hong Kong University.
C H I - K O N G L A I is Reader in Modern Chinese History at the University of
Queensland.
D A N L I is Professor of Economics at Fudan University, China.
M A Y - L I L I N is an Associate Research Fellow at the Institute of Modern
History, Academia Sinica.
D E B I N M A is Professor of Economics at Hitotsubashi University, Japan.
B A R R Y N A U G H T O N is So Kwanlok Chair of Chinese International Affairs in
the School of Global Policy and Strategy at the University of California at
San Diego.
xv

List of Contributors to Volume II
K A I X I A N G P E N G is Professor of Economics in the School of Economics and

Management, Wuhan University.
D W I G H T H . P E R K I N S is Harold Hitchings Burbank Research Professor of
Political Economy (Emeritus), Department of Economics, Harvard
University.
T H O M A S G. R A W S K I is Professor of Economics and History (Emeritus) at
the University of Pittsburgh.
W I L L I A M T .R O W E is John and Diane Cooke Professor of Chinese History at
Johns Hopkins University.
B R E T T S H E E H A N is Professor of History and East Asian Languages and
Cultures at the University of Southern California.
C A R O L H. S H I U E is Professor of Economics at the University of Colorado.
B A S V A N L E E U W E N is Senior Researcher at the International Institute of
Social History, the Netherlands.
M E I M E I W A N G is Assistant Researcher at the Institute of Economic
Research, Chinese Academy of Social Science.
C H E N G G A N G X U is Honorary Professor at the Asia Global Institute,
University of Hong Kong.
H O N G Z H O N G Y A N is Professor of Economics at Shanghai University of
Finance and Economics.
M A D E L E I N E Z E L I N is Dean Lung Professor of Chinese Studies and Professor
of History at Columbia University.
Y I N G U I Z H U is Professor of History (Emeritus) at Fudan University.
xvi

Acknowledgments
We are very grateful in the first place to our nearly fifty contributors, whose
collective efforts and dedication made this project possible. The timely
progress of this project was greatly facilitated by our ability to hold
conferences to assemble the contributors to each volume, exchange ideas,
and settle the division of labor among the forty chapters that ultimately were
included in the published work. We are deeply indebted to the benefactors
who provided financial support for the conferences. We want to thank Bas
van Leeuwen for hosting the conference for the contributors to Volume II at
the University of Utrecht on December 15–16, 2017, with generous financial
support from the European Research Council under the European Union’s
Horizon 2020 Programme/ERC-StG 637695 – HinDI (as part of the Historical
Dynamics of Industrialization in Northwestern Europe and China ca. 1800–
2010: A Regional Interpretation project). We are thankful for organizational
assistance from Robin Philips and Zipeng Zhang. The volume benefited from
the participation and external discussion of Maarten Prak, Keetie Sluyterman,
Jan Luiten van Zanden, and Peer Vries. Debin Ma also acknowledges some
editorial-assistant support from the School of Economics of Fudan University
in China.
The conference for contributors to Volume I was held on August 27–8,
2018, at the University of California, Los Angeles. We are grateful to the
Chiang Ching-Kuo Foundation for International Scholarly Exchange, which
provided the main financial support to underwrite the conference expenses.
Additional financial support was provided by the Center for Chinese Studies
and the Division of Social Sciences at UCLA, and we especially want to
acknowledge the UCLA Center for Chinese Studies, its director Yunxiang
Yan, and its assistant director Esther Jou for their logistical support. We also
want to thank Sunkyu Lee and Kayoko Fujita for their invaluable assistance in
facilitating the UCLA conference.
Lucy Rhymer at Cambridge University Press has instigated, encouraged,
and guided this project at every step, and we surely would not have been able
xvii

Acknowledgments
to bring it to completion without her unflagging support. We are also

grateful to the sharp-eyed Emily Sharp at Cambridge University Press for
all of her help in the final production of these volumes.
xviii

Note on Citations
The two volumes of the Cambridge Economic History of China differ in their
citation of Chinese and Japanese works in the footnotes and the “Further
Reading” bibliographies, reflecting differences in conventions between schol-
arship on premodern and modern periods.
In Volume I, book and article titles for secondary scholarship in Chinese
and Japanese are given in the original language and romanization (Pinyin
romanization in the case of Chinese). Names of authors are given in full
(including first names).
In Volume II, Chinese and Japanese book and article titles are given in the
original language and English translation without romanization. First names
of authors are abbreviated.
In Volume I, footnote references to primary sources in Chinese are
abbreviated as acronyms. A comprehensive list of Chinese primary sources
with full bibliographic information for primary sources appears at the end of
the volume.
In Volume II, full bibliographic information for references to primary
sources is given in the footnote in which they are cited.
xix

Introduction to Volume II
China’s rise as the world’s second-largest economy surely is the most dra-
matic development in the global economy since the year 2000. But China’s
prominence in the global economy is hardly new. Since 500 B C E,
a burgeoning market economy and the establishment of an enduring imperial
state has fostered precocious economic growth. Moreover, contrary to the
view that China’s economy withered under the dual constraints of Western
colonialism and Chinese tradition after 1800, recent scholarship has identified
the onset of modern economic growth in response to new incentive struc-
tures, investment opportunities, ideas, and technology, laying the foundation
for the post-1978 economic miracle. China’s combination of market-led
growth under the firm hand of the state has produced a model of economic
development that challenges conventional theories of capitalism and eco-
nomic growth. The spectacular growth of the contemporary Chinese econ-
omy has also spurred deeper investigation into the Chinese economy – long
a neglected field of study, at least in the Western academy. Scholarship on
Chinese economic history has now developed to the stage where
a Cambridge History devoted to the subject is appropriate and feasible.
These volumes, a collaborative effort by nearly fifty scholars, bring together
the fruits of pioneering Western, Japanese, and Chinese scholarship in all
dimensions of economic history, past and present.
Early studies of the Chinese economy focused on China’s distinctive
philosophical and political traditions.1 In his published 1911 Columbia Ph.D.
thesis, The Economic Principles of Confucius and His School), Ch’en Huan-chang
sought to introduce the basic concepts and practices of Chinese political
1
For historiographic surveys of twentieth-century scholarship on Chinese economic and
social history, see T. Brook, “Capitalism and the Writing of Modern History in China,”
in T. Brook and G. Blue (eds.), China and Historical Capitalism: Genealogies of Sinological
Knowledge (Cambridge, Cambridge University Press, 1999), pp. 110–57; R. von Glahn,
“Imagining Pre-modern China,” in P.J. Smith and R. von Glahn (eds.), The Song–Yuan–
Ming Transition in Chinese History (Cambridge, MA, Harvard University Council on East
Asian Studies, 2003), pp. 35–70.
https://doi.org/10.1017/9781108348485.001 Published online by Cambridge University Press

economy, highlighting the pre-eminence of the Confucian tradition.2 Of

course, Ch’en was writing at the moment of the dissolution of the Chinese
empire and the apogee of the modern world-system defined by industrial
capitalism and Western political hegemony. Already in Ch’en’s book – and
for long afterwards – the study of Chinese economic history was fixated on
the question of the apparent lack of economic progress throughout two
millennia of imperial history. Many of the answers proposed by Ch’en –
Confucian disdain for moneymaking, a rigid and inert social structure,
overpopulation, and isolation from the outer world – recurred in subsequent
works.
The fall of the Qing empire in 1911 – and with it, the end of China’s imperial
past – inspired hope for China’s rapid transformation into a modern, progres-
sive nation. But the failure of Republican institutions to thrive in the wake of
the empire’s demise gave rise to doubts about China’s ability to “learn from
the West.” Some intellectuals advocated wholesale repudiation of Chinese
cultural traditions and embrace of Western culture as defined by cosmopol-
itanism, Enlightenment values, Republican government, scientific reasoning,
and capitalist economic institutions. Others sought to reinvigorate China’s
“national essence” by reviving authentic Chinese values that had been
attenuated by Western spiritual pollution and Manchu overlordship. The
Japanese scholar Naitō Konan, writing in 1914, envisioned a future in which
Japan supplanted China as the ascendant center of a reinvigorated “oriental
culture” that would eclipse the spiritually vacuous materialism of the West.
Naitō advanced the novel thesis that East Asia’s “modern age” had actually
begun centuries earlier, in the transition from the Tang (618–907) to Song
(960–1279) dynasties. This Tang–Song transition had witnessed the triumph
of autocratic monarchy over aristocratic rule and engendered a vibrant
commoner culture liberated from feudal domination. But China’s incipient
modernity proved premature; after the Song dynasty, China’s “modern age”
degenerated into senility, and in contemporary times the dynamic center of
oriental culture had shifted to Japan.3
The emergence of history as an academic profession in China and Japan
during the 1920s was accompanied by skepticism toward received historical
2
H.-C. Ch’en, The Economic Principles of Confucius and His School (New York, Columbia
University Press, 1911).
3
Naitō Konan 内藤湖南, 支那論 (On China) (1914), rpt. in 内藤湖南全集 (The
Complete Volumes of Naitō Konan) (Tokyo, Chikuma shobō, 1972), vol. 5, pp.
291–482; see also H. Miyakawa, “The Naitō Hypothesis and Its Effects on Japanese
Studies of China,” Far Eastern Quarterly 14.3 (1955), 533–52; S. Tanaka, Japan’s Orient:
Rendering Pasts into History (Berkeley, University of California Press, 1993).

traditions and more critical empiricism in historical methodology. But by the

late 1920s historical scholarship in China became enveloped by disputes over
the trajectory of Chinese history from antiquity to the present, most notably
the so-called Social History Debate, which dwelled primarily on questions of
feudalism and capitalism in China. Marxism and the Russian Revolution
loomed large over these controversies. Left-wing scholars and writers were
keen to demonstrate that China’s historical experience conformed to the
universal categories of historical development as defined by Western philo-
sophers. In his Study of China’s Ancient Society (1930), Guo Moruo was the first
to apply to Chinese history the five-stage theory of human history (from
primitive communism to socialism) formulated by Soviet Marxist scholars.4
Guo argued that the onset of the Iron Age in China in the first millennium B C E
inaugurated feudal relations of production that endured even under the
centralized bureaucratic empires.
The anomalous character and persistence of Chinese “feudalism” – which
bore little resemblance to the fragmented political order of medieval
Europe – posed a vexing problem. Some historians in China sought to resolve
this incongruity by espousing an economic definition of feudalism based on
the antagonism between the landowning class and the peasantry, and thus
postulated a feudal epoch that stretched from the ancient Zhou era through-
out imperial history and continued even during the post-1911 Republican era.
In Europe, Max Weber had traced the apparent stagnation of Chinese
historical development to the special character of the imperial bureaucratic
state, which Marxist scholars reformulated as a distinctive species of “bur-
eaucratic feudalism.” Drawing on the theories of both Marx and Weber, Karl
Wittfogel, a leading figure in the Frankfurt school of Marxism, proposed that
the Chinese imperial state was founded on an “Asiatic mode of production”
which hindered the dynamic forces of class struggle that motivated historical
change.5 Although Soviet Marxists firmly repudiated the idea of an Asiatic
mode of production, it gained considerable currency among some Chinese
and many Japanese scholars in the 1930s. Characterizing China as the arche-
type of an “oriental society” trapped in the Asiatic mode of production
provided a succinct explanation for the immobility of Chinese history, and
4
Guo Moruo 郭末若, 中國古代社會研究 (A Study of Ancient Chinese Society)
(Shanghai, Lianhe shudian, 1930).
5
K.A. Wittfogel, “The Foundations and Stages of Chinese Economic History,” Zeitschrift
für Sozialforschung 4.1 (1935), 25–58; Wittfogel, “Die Theorie der orientalischen
Gesellschaft,” Zeitschrift für Sozialforschung 7.1–2 (1938), 90–122. In the postwar period
Wittfogel developed a more elaborate version of this thesis. See his Oriental Despotism:
A Comparative Study of Total Power (New Haven, Yale University Press, 1957).

for Japanese historians reinforced Japan’s singular role in both “escaping from
Asia” and leading it toward modernity.
To be sure, other scholars distanced themselves from these theoretical
debates and focused instead on empirical study and honing analytical
methods of economic history. In China, this trend coalesced around the
journal Food and Money Semi-monthly (Shihuo banyuekan 食貨半月刊),
inaugurated in 1934 by Tao Xisheng, which featured contributions by
Tao, Ju Qingyuan, He Ziquan, Quan Hansheng, and others. The “Food
and Money” group (the name derived from the title given to chapters
on fiscal administration in traditional dynastic histories) sought to bal-
ance economic theories (both Marxist and non-Marxist) with empirical
evidence and rigorous methodology.6 Much of their scholarship centered
on the early imperial era, especially the Tang dynasty.7 The 1930s also
witnessed a profusion of pathbreaking Japanese scholarship on Chinese
economic history, likewise centered on the Tang–Song dynasties, led by
Katō Shigeshi,8 Hino Kaisaburō,9 Sogabe Shizuo,10 and Miyazaki
Ichisada.11 The inspiration for these studies can be traced back to
Naitō’s hypothesis of the Tang–Song transition as the beginning of
“East Asia’s modern age,” as Miyazaki – Naitō’s successor as professor
of Chinese history at Kyoto University – entitled his influential
6
For a brief introduction to the Food and Money group and its methodological
approaches to economic history, see Su Yongming 苏永明, “‘食货派’的经济史研究
方法探讨” (An Exploration of Economic History Methodology in “Shihuo pai”), 史学
史研究 (Historical Research) 2007.3, 77–83.
7
Most notably, Ju Qingyuan 鞠清遠, 唐宋官私工業 (Government and Private Industries
in Tang and Song) (Shanghai, Xin shengming shuju, 1934); Quan Hansheng 全漢昇, 中國
行會制度史 (History of the Chinese Guild System) (Shanghai, Xin shengming shuju, 1935);
Tao Xisheng 陶希聖 and Ju Qingyuan 鞠清遠, 唐代經濟史 (Tang Economic History)
(Shanghai, Shangwu yinshuguan, 1936); Ju Qingyuan 鞠清遠, 唐代財政史 (Tang Fiscal
History) (Shanghai, Shangwu yinshuguan, 1940).
8
Katō’s seminal essays on Tang–Song (and some Qing) economic history were pub-
lished in a two-volume posthumous work: Katō Shigeshi 加藤繁, 支那経済史考証
(Research on Chinese Economic History) (Tokyo, Tōyō bunko, 1952).
9
Hino’s prolific research on Tang–Song history has been reprinted in his twenty-volume
collected works: Hino Kaisaburō 日野開三郎, 日野開三郎東洋史学論集 (Collected
Works of Hino Kaisaburō on East Asian History) (Tokyo, San’ichi shobō, 1980–1988).
10
Sogabe’s research primarily focused on Song fiscal institutions and monetary history;
see Sogabe Shizuo 曽我部静雄, 宋代財政史 (Song Fiscal History) (Tokyo,
Seikatsusha, 1941); Sogabe, 中国社会経済史研究 (Research on Chinese Social and
Economic History) (Tokyo, Yoshikawa kōbunkan, 1976).
11
Miyazaki’s first monograph studied the monetary history of the tenth century:
Miyazaki Ichisada 宮崎市定, 五代宋初の通貨問題 (Monetary Issues in the Five
Dynasties and Early Song) (Kyoto, Hoshino shoten, 1943). His body of work, which
extended over the entire breadth of Chinese history, has been reprinted in his collected
works, which run to twenty-four volumes: 宮崎市定全集 (Complete Works of
Miyazaki Ichisada) (Tokyo, Iwanami shoten, 1991).

synthesis.12 The priority of the Tang also can be seen in the first European-
language monograph explicitly devoted to Chinese economic history, the
Hungarian-born scholar Étienne Balazs’s 1932 doctoral thesis for Berlin
University on the economic history of the Tang dynasty.13 Although Balazs
championed empirical research over both theoretical formulations and nar-
rowly construed philological study, his scholarship was explicitly couched in
the Weberian project of comparative study within universal categories of
historical development and contributed to the conception of China as – in his
words – a “permanently bureaucratic state” that obstructed the emergence of
an independent merchant class and, by extension, capitalism.14
The prewar generation also pioneered the use of quantitative data for the
study of Chinese economic history. Particularly noteworthy in this regard were
the contributions of Quan Hansheng, the forerunner in the study of price history,
international trade, and national revenue (initially focused on the Song, and later
extended to the Ming–Qing periods as well), and Liang Fangzhong, who
published a pathbreaking essay on Ming population, land, and taxation statistics
in 1935 and a landmark study of the sixteenth-century Single-Whip tax reform in
1936.15 In addition, scholars such as Chen Hansheng and Fei Xiaotong, trained in
economics and anthropology in the US and the UK respectively, published
monographs on the contemporary rural economy based on extensive field
research that became foundational studies.16 During the Japanese occupation
12
Miyazaki Ichisada 宮崎市定, 東洋的近世 (Early Modern East Asia) (Osaka, Kyōiku
taimusu sha, 1950).
13
Published as É. Balazs, “Beiträge zur Wirtschaftsgeschichte der T’ang-Zeit (618–906),”
Mitteilungen des Seminars für Orientalische Sprachen 34 (1931), 1–92; 35 (1932), 93–165; 36
(1933), 1–62.
14
See his 1957 essay “China as a Permanently Bureaucratic Society,” in É. Balazs, Chinese
Civilization and Bureaucracy (New Haven, Yale University Press, 1964), pp. 13–27. For
Balazs’s impact on Chinese studies in France, where he spent most of his career, see H.
T. Zurndorfer, “Not Bound to China: Étienne Balazs, Fernand Braudel and the Politics
of the Study of Chinese History in Post-war France,” Past and Present 185 (2004), 189–221.
15
Quan’s studies were published in a two-volume collection, Quan Hansheng 全漢昇,
中國經濟史論叢 (Research on Chinese Economic History) (Hong Kong, Xinya
shuyuan Xinya yanjiusuo, 1972), and a subsequent three-volume collection: 中國經
濟史研究 (Research on Chinese Economic History) (Hong Kong, Xinya shuyuan
Xinya yanjiusuo, 1976). A collection of Liang Fangzhong’s writings on economic
history was published as Liang Fangzhong 梁方仲, 梁方仲文集 (Collected Works
of Liang Fangzhong) (Guangzhou, Zhongshan daxue chubanshe, 2004). Liang also
supervised the compilation of what remains the most authoritative collection of
Chinese historical statistics: Liang Fangzhong (ed.), 中国历代户口田地田赋统计
(Statistics on Chinese Population, Land, and Land Taxes) (Shanghai, Shanghai renmin
chubanshe, 1980).
16
Chen H.-S., Landlord and Peasant in China: A Study of the Agrarian Crisis in South China
(New York, International Publishers, 1936); Chen, Industrial Capital and Chinese Peasants
(Shanghai, Kelly and Walsh, 1939); H.T. Fei, Peasant Life in China: A Field Study of Country

of China in the 1930s–1940s a legion of Japanese scholars, operating under the

auspices of the Investigation Department of the South Manchuria Railway
(Mantetsu Chōsabu, founded in 1907), compiled hundreds of reports on the
Chinese economy that have been extensively mined by later scholars.17 The
pronounced influence of Keynesian economics among young Chinese econo-
mists also produced the first efforts to compile national income data and measure
GDP in the late 1940s, efforts that would be jettisoned along with non-Marxist
economics after the founding of the People’s Republic of China in 1949.18
In the postwar era, Japanese scholarship on Chinese economic history was
especially robust. The ascendancy of Marxist analysis in the Japanese acad-
emy – an explicit repudiation of the prewar imperialist project – established
a new paradigm for interpreting Chinese economic history. Rejecting both
the thesis of “oriental stagnation” and Naitō Konan’s ideas about China’s
precocious modernity, economic and legal historians such as Sudō Yoshiyuki
and Niida Noboru portrayed the Tang–Song transformation as the formative
phase of a feudal society based on the subordination of serfs to a landlord
class.19 These servile relations were reproduced by the patriarchal social
institutions of family, lineage, and guild that inhibited the emergence of the
“rational” legal and economic institutions which the great German sociolo-
gists denoted by the term Gesellschaft (impersonal social relations). The
density of patriarchal communal social relations (Gemeinschaft, rendered in
Japanese as kyōdōtai 共同体) in Chinese society precluded the formation of
an independent bourgeoisie and the transition to capitalism. More import-
antly, these studies – and a plethora of non-Marxist scholarship as well –
generated a wealth of new empirical research on Chinese economic history,
Life in the Yangtze Valley (London, Kegan Paul, 1939), Fei, Earthbound China: A Study of
Rural Economy in Yunnan (Chicago, The University of Chicago Press, 1945). Foreign
researchers also contributed to the compilation of social and economic statistics, most
notably J.L. Buck, Chinese Farm Economy (Chicago, The University of Chicago Press,
1930); Buck, Land Utilization in China (Shanghai, Commercial Press, 1937); S.D. Gamble,
Peking: A Social Survey (New York, George H. Doran Co., 1921).
17
For overviews of this scholarship, see J. Young, The Research Activities of the South Manchuria
Railway Company, 1907–1945: A History and Bibliography (New York, Columbia University
Press, 1966); Matsumura Takao 松村高夫, Yanagisawa Yū 柳沢遊, and Eda Kenji 江田憲
治, 満鉄の調査と研究―その「神話」と実像 (Investigation and Research on the
South Manchuria Railway Company: Its “Myth” and Reality) (Tokyo, Aoki shoten, 2008).
18
Wu Baosan 巫寶三 (ed.), 中國國民所得 (Chinese National Income) (Shanghai,
Zhonghua shuju, 1947); T.-C. Liu, China’s National Income, 1931–36: An Exploratory
Study (Washington, DC, Brookings Institution, 1946).
19
For a synopsis of the voluminous studies by Sudō and Niida (and the prodigious
postwar Japanese scholarship on Tang–Song economic history generally), see the state-
of-the-field essay by P. Golas:“Rural China in the Song,” Journal of Asian Studies 39.2
(1980), 291–325.

including a growing body of work centered on the Ming–Qing period and the
twentieth century.
In China, of course, orthodox Marxist historiography prevailed after 1949.
However, numerous studies attesting to the vitality of the market economy
in the late Ming dynasty gave rise to the idea that the “sprouts of capitalism”
(zibenzhuyi mengya 資本主義萌芽) had begun to emerge by the late six-
teenth century, if not earlier.20 Shang Yue became the most prominent
exponent of the controversial thesis that an incipient bourgeoisie had formed
in the late Ming period and China thus was already beginning the transition
to capitalism before it was derailed by the Manchu conquest of the Ming in
1644.21 Japanese historians quickly joined this debate. Scholars who focused
on the urban economy and merchants, such as Fu Yiling and Tanaka
Masatoshi, tended to underscore the potential for indigenous capitalist
transformation.22 Historians who concentrated on relations of production
in the agrarian economy and handicraft industries, such as Nishijima Sadao,
were far more skeptical.23 However, by 1960 Shang Yue’s thesis on the
“sprouts of capitalism” was deemed heretical within the Chinese academic
establishment.
Although Shang Yue’s contention that China was on the verge of
a breakthrough to capitalism before the Opium War was repudiated, most
scholars assented to the proposition that rising commodity production in the
Ming–Qing era attested to an “advanced” form of feudalism in China,
challenging the idea that the Western European historical experience exclu-
sively defined the archetype of the feudal economy. According to this line of
thought, the intrusion of foreign imperialism in the nineteenth century and
China’s subjugation as a “semi-colony” warped the development of Chinese
capitalism, precluding the formation of an autonomous national bourgeoisie
20
For a brief and regrettably tendentious introduction, see A. Feuerwerker, “From
‘Feudalism’ to ‘Capitalism’ in Recent Historical Writing from Mainland China,”
Journal of Asian Studies 18.1 (1958), 107–16.
21
Shang Yue 尚鉞, 中國資本主义关系發生及演变的初步研究 (Preliminary Research
on the Origin and Evolution of Chinese Capitalist Relations) (Beijing, Sanlian shudian,
1956).
22
Fu Yiling 傅衣凌, 明代江南市民經济試探 (An Exploration of Jiangnan Citizens)
(Shanghai, Shanghai renmin chubanshe, 1957); Tanaka Masatoshi 田中正俊, “中国歴
史界における「資本主義の萌芽」研究” (Research into the “Sprouts of Capitalism”
in Chinese History Studies), in Suzuki Jun 鈴木俊 and Nishijima Sadao 西島定雄 (eds.),
中国史の時代区分 (The Periodization of Chinese History) (Tokyo, Tōkyō daigaku
shuppansha, 1957), pp. 219–52.
23
Nishijima Sadao 西島定雄, “中国古代社会の構造的特質に関する問題点” (Issues
Related to the Special Features of Ancient Chinese Social Structure), in Suzuki and
Nishijima, 中国史の時代区分, pp. 175–208.

and instead fostering state-led “bureaucratic capitalism.” Thus China

remained a “semi-feudal, semi-colonial” society until the Communist Party
initiated a proletarian revolution beginning in the 1930s. Notwithstanding the
ideological bent of PRC scholarship in this era, a vast array of source
materials, statistical series, and valuable monographic studies were published
that stimulated research both within China and abroad.24
By contrast, in the immediate postwar era economic history was a virtually
untouched subject in Western scholarship on China, which was deeply
adverse to the Marxist cast of the Chinese and Japanese studies mentioned
above. In an essay on the state of the field at the dawn of the 1960s, American
scholar Albert Feuerwerker dourly observed that
(1) Monographic studies in Chinese rarely come up to the standards expected
of European economic historians. (2) Few detailed investigations have
appeared in European languages; and their quality is very uneven. (3)
Perhaps both in quantity and quality the most significant body of mono-
graphic work has been done in Japan, though here too there are serious
limitations growing out of the strong hold of Marxist ideology in Japanese
academic circles. (4) There is, to my knowledge, no satisfactory synthetic
treatment of Chinese economic history in any language to which a non-
specialist might go for a substantive introduction to this subject.25
Feuerwerker’s citations belie these grim conclusions to some extent; for

example, he mentions Peng Xinwei’s Chinese Monetary History, a magisterial
survey that remains unsurpassed to this day.26 Mention also should be made
of the numerous publications during these years by Lien-sheng Yang, perhaps
the first scholar to write a Ph.D. thesis at a US university (Harvard, 1946) on
Chinese economic history.27 Regrettably, the ideological struggles that con-
vulsed China during the Cultural Revolution of the 1960s–1970s shuttered the
universities, inflicted enormous personal hardship on many scholars, and
effectively suspended serious scholarship.
24
A. Feuerwerker, “China’s Modern Economic History in Communist Chinese
Historiography,” China Quarterly 22 (1965), 31–61.
25
A. Feuerwerker, “Materials for the Study of the Economic History of Modern China,”
Journal of Economic History 21.1 (1961), 42.
26
Peng Xinwei 彭信威, 中國貨幣史 (Chinese Monetary History) (Shanghai, Shanghai
renmin chubanshe, 1958).
27
Yang’s work includes Money and Credit in China: A Short History (Cambridge, MA,
Harvard University Press, 1952) and Les aspects économiques des travaux publics dans la
Chine impériale: Quatre conférences (Paris, Collège de France, 1964); and studies gathered
in his essay collections Studies in Chinese Institutional History (Cambridge, MA, Harvard
University Press, 1961) and Excursions in Sinology (Cambridge, MA, Harvard University
Press, 1969).

In any event, the decade of the 1960s would prove to be a watershed in

Western scholarship on Chinese economic history. The rise of social science
research and its application to historical study provided the catalyst for
a series of landmark studies that would shape Western scholarship on
Chinese economic history for a generation: Ping-ti Ho’s incisive dissection
of historical population statistics;28 Albert Feuerwerker’s study of state-led
industrialization efforts in the late nineteenth century;29 the meticulous
reconstruction of Tang fiscal administration by Denis Twitchett, which
amply demonstrated the value of Japanese scholarship;30 Robert Hartwell’s
provocative findings on the precocious development of the coal and iron
industries in Song China;31 the insights of economic geography applied by
G. William Skinner to generate a new paradigm of market structure and
marketing systems in China;32 the application of quantitative analysis to
agricultural production pioneered by Dwight Perkins;33 Ramon Myers’s
revisionist analysis, drawing on quantitative data from the Mantetsu surveys,
of economic performance in rural China in the Republican period;34 Shiba
Yoshinobu’s empirically rich and analytically sophisticated tour de force on
commerce and merchant enterprise in the Song, which became accessible to
a wider audience through Mark Elvin’s abbreviated translation;35 crowned by
Elvin’s own theoretically innovative paradigm of the course of economic
development in China across the imperial era, another work deeply informed
by Japanese scholarship.36
28
P.-T. Ho, Studies on the Population of China, 1368–1953 (Cambridge, MA, Harvard
29
A. Feuerwerker, China’s Early Industrialization: Sheng Hsuan-huai (1844–1916) and
Mandarin Enterprise (Cambridge, MA, Harvard University Press, 1958).
30
D. Twitchett, Financial Administration under the T’ang Dynasty (Cambridge, Cambridge
University Press, 1963)
31
R.M. Hartwell, “A Revolution in the Chinese Iron and Coal Industries during the
Northern Sung, 960–1126 A D,” Journal of Asian Studies 21.1 (1962), 153–62; Hartwell,
“Markets, Technology, and the Structure of Enterprise in the Development of the
Eleventh-Century Chinese Iron and Steel Industry,” Journal of Economic History 26.1
(1966), 29–58; Hartwell, “A Cycle of Economic Change in Imperial China: Coal and Iron
in Northeast China, 750–1350,” Journal of the Economic and Social History of the Orient 10
(1967), 102–59.
32
G.W. Skinner, “Marketing and Social Structure in Rural China,” Journal of Asian Studies
24.1 (1964), 3–43; 24.2 (1965), 195–228; 24.3 (1965), 363–99.
33
D.H. Perkins, Agricultural Development in China, 1368–1968 (Chicago, Aldine, 1969).
34
R.H. Myers, The Chinese Peasant Economy (Cambridge, MA, Harvard University Press,
1970).
35
Shiba Yoshinobu 斯波義信, 宋代商業史研究 (Tokyo, Kazama shobō, 1968), trans-
lated as Commerce and Society in Sung China (Ann Arbor, University of Michigan Center
for Chinese Studies, 1970).
36
M. Elvin, The Pattern of the Chinese Past (Stanford, Stanford University Press, 1973).

By the 1970s, then, Western scholarship on Chinese economic history had

achieved a new level of maturity. The wealth of new empirical studies since
then has fostered vigorous, indeed contentious, debate on issues such as the
impact of Western and Japanese imperialism, the nature of the peasant
economy, regional systems and market networks, and what has come to be
called the “Great Divergence” debate. In particular, the past three decades
have seen significant revisionist scholarship on Chinese economic perform-
ance in the tumultuous nineteenth and twentieth centuries following the
intrusion of Western imperialism in the post-Opium War era. Contrary to
the once dominant pessimistic interpretation of a Chinese economy wither-
ing under the dual constraints of Western colonialism and Chinese tradition,
the new scholarship has identified the onset of modern economic growth in
this era – at least in some crucial regions and sectors – as a powerful response
to new incentive structures and investment opportunities, as well as the
inflow of new ideas and technology, laying the foundation for China’s
economic takeoff in the post-1978 reform era. Within China, perhaps the
most significant development is the publication of the three edited volumes
of the History of Chinese Capitalist Development in 1983.37 Under the leadership
of one of the coeditors, Wu Chengming of the Chinese Academy of Social
Science, these three volumes, although framed within a Marxist framework,
brought together a generation of devoted senior and junior scholars to
provide a comprehensive economic history of modern China from the
early modern era to 1950. Wu, himself an economist who received
a master’s degree from Columbia University in the 1940s but was banished
during China’s Cultural Revolution era, emerged as an intellectual leader in
economic history within China throughout the 1980s and 1990s.
Although much of the new scholarship on Chinese economic history has
focused on the post-1800 period, for which quantitative data are much more
abundant, there have been significant advances in the study of China’s
premodern economy, particularly for the ancient period. Important new
data generated from archaeological research in China, ranging from textual
and artifactual evidence to urban morphology and settlement studies, have
yielded fresh insights into social and economic livelihood in ancient China
and enable us to trace the course of economic change with much greater
temporal and geographic precision. Although Western scholarship on
37
Xu Dixin 許滌新 and Wu Chengming 吳承明 (eds.), 中國資本主義發展史
(Developmental History of Chinese Capitalism), 3 vols. (Beijing, Renmin chubanshe,
1985). Volume 1 was translated into English as Xu Dixin and Wu Chengming (eds.),
Chinese Capitalism, 1522–1840 (New York, St. Martin’s Press, 2000).
10

China’s premodern economic history remains modest, quantitatively speak-

ing, compared to the prodigious output of Chinese and Japanese scholars,
Western historians have done pioneering work in many aspects of the
premodern economy, including environmental history; demography; legal
institutions and economic organization; kinship, gender, and the household
economy; political economy; and economic sociology.
These scholarly developments have coincided with the unfettering of Chinese
scholarship from shopworn Marxist–Leninist ideological blinders since the early
1980s and an enormous surge in new scholarship on Chinese economic history
within China. With the opening of new archives and the improvement of the
academic infrastructure in China, new generations of Chinese scholars have
begun to make important methodological and theoretical contributions to the
study of Chinese economic history ranging from agriculture and demography to
finance and law. The rapidly growing presence of Chinese scholars within the
global economic history community (as seen, for example, at meetings such as
the triennial World Economic History Congress) also testifies to the rising
impact of Chinese economic history within the profession internationally.
A significant shift in the scholarly landscape has been the gradual rebalancing
of the community working on Chinese economic history. Twenty years ago, the
few scholars working on economic history were largely based outside China.
The last decade has seen the steady growth of a young generation of researchers
who returned to China having gained training in quantitative and economic
approaches from North American and European Ph.D. programs. In conjunc-
tion with the rising stock of foreign-trained Ph.D.s in economic history, univer-
sities in China and Hong Kong now produce a steady stream of Ph.D.s trained in
quantitative economic history, many of whom are making careers in China. The
sheer quantity of the new scholarship on Chinese economic history since the
1970s defies adequate summary in this brief essay, but it will be cited copiously
throughout these volumes. In addition, synthetic surveys of Chinese economic
history have now begun to appear.38 Along with these surveys are two commis-
sioned special journal issues devoted entirely to Chinese economic history.39
38
L. Brandt, D. Ma, and T.G. Rawski, “From Divergence to Convergence: Reevaluating
the History behind China’s Economic Boom,” Journal of Economic Literature 52.1 (2014),
45–123; R. von Glahn, The Economic History of China from Antiquity to the Nineteenth
Century (Cambridge, Cambridge University Press, 2016); Okamoto Takashi 岡本隆司
(ed.), 中国経済史 (Economic History of China) (Nagoya, Nagoya daigaku shuppankai,
2013).
39
D. Ma (ed.), Money, Finance and Commerce in Chinese History, special issue of Frontier of
Economics in China 13.3 (2018); K.J. Mitchener and D. Ma (eds.), A New Economic History of
China, special issue of Explorations in Economic History 63 (2017).
11

The Cambridge Economic History of China is divided chronologically into two

volumes, with the first volume devoted to the period before 1800 and
the second volume to the period from 1800 to the present. Each volume is
further subdivided into two broad chronological sections, but within these
divisions the chapters are organized topically rather than chronologically.
Part I of Volume I (in six chapters) covers the period from 1000 B C E to 1000
C E, with Part II (in twelve chapters) devoted to the period from 1000 to 1800.
The unequal portions assigned to the pre-1000 and 1000–1800 periods reflect
differences in the depth and breadth of the scholarship at this point in time.
The usual periodization of Chinese history posits a sharp break between the
middle imperial period of 750–1500 (often subsumed under the Tang–Song
transition rubric discussed above) and the 1500–1800 era (whether this period
should be defined as China’s “early modern” era remains controversial). The
scholarship usually reflects this divide as well. However, given the topical
structure of the volume, we believe that the 1000–1800 period should be
treated as an integral whole.
Volume II similarly is divided into two broad chronological parts that are
subdivided into thematic chapters. Part I covers the period from 1800 to 1950,
encompassing the last century of the Qing dynasty and the Republican era
(1911–1949), in fourteen chapters. Part II, in seven chapters, examines the
dramatic transformations of the Chinese economy since the founding of the
People’s Republic of China in 1949.
As the first attempt in the Cambridge History series to focus on Chinese
economic history, our two volumes will remedy a large lacuna in the
discipline of economic history and respond to the increasing demand from
both specialists and the general public for a comprehensive introduction to
the subject. These volumes will provide an authoritative survey incorporat-
ing up-to-date research at the frontiers of knowledge, including quantitative
data that are accessible to a general economic history audience, as well as
addressing some of the most important current debates in Chinese and global
economic history. We also hope that these volumes will serve both as
a standard reference and as a resource for teaching.
12

part I
1800–1950

1
Ideology and the Contours of Economic
Change
debin ma
When visited by the British trade mission led by Lord George Macartney, who
aimed to show off the best of Western trade and technology, the Qianlong
Emperor of Qing China was known to have famously replied in 1792, “Our
Celestial Empire possesses all things in prolific abundance and lacks no product
within its borders. There is therefore no need to import the manufactures of
outside barbarians in exchange for our own produce.”1 Qianlong’s statement
came at the height of Qing’s glory, overseeing a remarkable tripling of popula-
tion and a doubling of territory between the fifteenth and eighteenth centuries.
No single political entity at the time achieved such size in both geography and
population with such stability and cohesion.
The Qing’s self-conceit was shattered only five decades after Qianlong’s
statement by China’s humiliating military defeat to the much smaller polity of
Britain in the infamous 1842 Opium War, which in turn reduced the mighty Qing
to semicolonial status through the treaty port system. The mid-nineteenth-
century confrontation symbolizes the clash between a highly centralized singular
Chinese empire and a fragmented and competitive European state system
empowered by centuries of waves of commercial, financial, political, techno-
logical, and scientific revolutions. More importantly, during the tumultous
century after China’s forced opening in the mid-nineteenth century, the mixed
Chinese record, especially in relation to neighboring Japan, raises the more
pertinent question of why it took so long for China to industrialize after the
Industrial Revolution had been well underway elsewhere.
With China’s unprecendented economic growth during the last four
decades and her emergence as the world’s second-largest economy, the tide
I want to thank Joel Mokyr, Kaixiang Peng, Jared Rubin, and Rui Wang for comments and
encouragement. But my special gratitutde goes to Thomas Rawski, who provided meticu-
lous comments and feedback that greatly improved the chapter. I remain responsible for
all errors.
1
Quote from https://en.wikipedia.org/wiki/Macartney_Embassy, accessed August 14, 2016.
15

debin ma
has turned with the myth of the once mighty Qing making a curious
comeback in the increasingly popular claim that current Chinese economic
success marked only a triumphant return to China’s eighteenth-century
glory. The claim is as misleading as a misconstrued conflation of total GDP
with per capita GDP. In per capita terms, eighteenth-century China probably
remained a poor agrarian economy, which could still be the largest economy
in the world thanks to her sheer population size and because per capita
income differences in the preindustrial world were relatively small before the
onset and spread of the Industrial Revolution. It might have been the largest
but not the leading economy, with limited impact in external trade and
investment globally.2
Writing before the onset of China’s current boom, earlier generations
of scholars had to explain China’s historical economic stagnation, often
due to the prevalence of nepotism, corruption, and other elements of
Chinese social structure and behavior that prevented a dynamic response
of the sort attained during Japan’s Meiji era (1868–1912). But China’s recent
reversal of fortune seems to have generated a tide of scholarship that now
advances the opposite view, attributing recent Asian prosperity to the
same “cultural values” formerly thought to have obstructed economic
dynamism, and arguing that there were no fundamental shortfalls in
China’s political, legal, or other institutional traditions. Pomeranz’s influ-
ential book on The Great Divergence (2000) further argues that Britain’s
head start in industrialization arose from its domestic supplies of cheap
coal and because its colonies provided superior access to land-intensive
goods rather than from any advantage linked to political, legal, or other
institutional factors.3
For the 1850–1949 period under study, this chapter emphasizes the fact that,
ultimately, industrial revolution in modern East Asia started with borrowed
institutions and ideology. This takes us back to the much older conceptual
framework of modernization known as “Western impact and Chinese
response” in the Chinese or East Asian context. In particular, this chapter
reasserts the peculiar and unusual importance of the external or Western
2
It is true that there was substantial trade in the export of Chinese silks and tea and an
inflow of Latin American silver ingots and coin as well as New World crops such as
maize and potatoes.
3
L. Brandt, D. Ma, and T. Rawski, “From Divergence to Convergence: Re-evaluating the
History behind China’s Economic Boom,” Journal of Economic Literature, 52.1 (2014),
45–123; K. Pomeranz, The Great Divergence: China, Europe, and the Making of the Modern
World Economy (Princeton, NJ, Princeton University Press, 2000).
16

Ideology and the Contours of Economic Change
influence on a large country like China. Developing a new conceptual

framework, I argue that given the dual monopoly of ideology and power
under the Qing, political and economic changes often had to be initiated from
outside the empire. Openness and external influence could act as a constraint
in a polity lacking internal checks and balances. The external represents a new
source of alternative power that breaches the monopoly of power and
ideology. However, how much the external could exert an impact depends
on how much it elicits formal institutional and ideological change in the
domestic context.
In this regard, I show that rather than resource endowments such as coal,
or even the discovery of New World resources, what impeded China’s
progress in the globalized world is what some historians of China have called
the scarcity of “intellectual resources” 思想资源, or what sometimes is
referred to as ideology.4 However, the scarcity of intellectual resources in
mid-nineteenth-century Qing China is of a particular sort, that of confronting
the rapidly advancing West transformed by the Industrial Revolution. This
Chinese scarcity could have been overcome through massive borrowing,
learning, and importing of Western ideology and institutions, which is partly
what Meiji Japan succeeded in doing. These intellectual resources allowed
the late industrializer to construct an entirely different set of political govern-
ance, economic systems, and social organization.5
I argue that both the capacity and the willingness to borrow and learn
themselves are endogenous to China’s pre-existing political institutions and
geopolitical position in East Asia. In this regard, China’s highly centralized
and absolutist political regimes and traditional dominance in a China-
centered world order have led not only to a closure of mind to new
4
These ideological or intellectual resources represent our ways of understanding, inter-
preting, and theorizing about the world we live in and the ways we interpret the past,
construct the present, and imagine the future. They both support and constrain our
institutions, policies, and day-to-day decisions. See Wang Fansen 王汎森, “戊戌前后思
想资源的变化：以日本因素为例” (Changes in Intellectual Resources before and
after the Hundred Days Reform), 二十一世纪 (Twenty-First Century) 45 (February
1998), 47–54.
5
See Wang Fansen 王汎森, 权力的毛细血管作用，清代的思想、学术与心态 (The
Penetrating Role of Power, Ideas, Academics, and Moods in Qing) (Beijing,
Beijing daxue chubanshe, 2015). Jin Guantao 金观涛 and Liu Qingfeng 刘青峰, 兴盛
与危机，论中国社会的超稳定结构 (The Cycle of Growth and Decline: On the
Ultrastable Structure of Chinese Society) (Beijing, Falu chubanshe, 2011). Jin Guantao
金观涛 and Liu Qingfeng 刘青峰, 开放中的变迁，再论中国社会超稳定结构 (The
Transformation of Chinese Society (1840–1956): The Fate of Its Ultrastable Structure in
Modern Times) (Beijing, Falu chubanshe, 2011). Ge Zhaoguang 葛兆光, “十八世纪中
国的盛世危机” (Crisis in China’s Glorious Eighteenth Century), February 19, 2019,
www.ftchinese.com/author/%E8%91%9B%E5%85%86%E5%85%89.
17

debin ma
intellectual resources, but also to failure to recognize or perceive impending

crisis and threats. So the process of change is not automatic, requiring
complicated and sometimes risky feedback processes between social and
political experiments, institutions, and ideology. This chapter links this
complicated process of Western impact and Chinese response by connecting
the changing incentives to economic agents on the ground with the shifting
ideologies of elites at the top. It matches the contours of economic change
with the specific timing of intellectual and ideological transformations during
this period and embed our narrative in two specific cases of commercial and
financial development. The first section lays out the quantitative profile of
Chinese economic change from 1850 to 1950. The second section turns to
historiography and builds a new analytical framework linking ideology with
economic changes. The third section examines the three phases of economic
change based on the new analytical framework.
The Chinese Economy, 1842–1949: Stagnation

or Takeoff?
Stagnation or Takeoff?
Our analysis of the Chinese economic record starts with the aggregate
economic indicator of GDP. However, despite a new wave of historical
GDP research, estimates before the 1930s still need to be treated with caution
and used with hesitation given the controversies on long-term statistics as
basic as population and agricultural acreage.6 Nonetheless, few would dis-
agree that China of the nineteenth and twentieth centuries is a traditional
agrarian economy with a 60 to 70 percent share of GDP in the agriculture
sector.
6
See S. Broadberry, H. Guan, and D.D. Li, “China, Europe, and the Great Divergence:
A Study in Historical National Accounting, 980–1850,” Journal of Economic History 78
(2018), 955–1000. Y. Ma and H. de Jong, “Unfolding the Turbulent Century:
A Reconstruction of China’s Historical National Accounts, 1840–1912,” Review of
Income and Wealth 65.1 (2019), 75–9. X. Yi, Z. Shi, B. van Leeuwen, Y. Ni, Z. Zhang,
and Y. Ma, “Chinese National Income, ca. 1661–1933,” Australian Economic History Review
57.3 (November 2017), 368–93. A Maddison, Chinese Economic Performance in the Long Run,
2nd ed, rev. and updated (Paris, Development Centre of the Organization for Economic
Co-operation and Development, 2007). For problems with data, see discussion in D. Ma,
“Economic Growth in the Lower Yangzi Region of China in 1911–1937: A Quantitative
and Historical Perspective,” Journal of Economic History 68.2 (2008), 385–92. T.G. Rawski,
Economic Growth in Prewar China (Berkeley, University of California Press, 1989).
18

Figure 1.1 plots China’s real per capita GDP index against that of Japan. The
figure reveals the sluggish performance of Qing and Republican China relative
to Japan’s transformation. Although the indexing of the level of the per capita
GDP of the two economies in 1850 both at 100 seems arbitrary at first sight, it
actually turns out to be consistent with a comprehensive study based on the
reconstruction of the 1930s benchmark purchasing-power parity estimates
which places the Japanese per capita GDP level at about three times that of
China during the 1930s, a ratio that broadly matches that in Figure 1.1.7
Projecting this 1930s three-to-one ratio back to 1850 with the trend of growth
rates of these two countries leads to a comparable level of per capita GDP for
the two countries around the mid-nineteenth century in Figure 1.1. Hence our
implicit estimate of a common starting point for these two countries in 1850
presumes that the divergence between Japan and China only started only after
their being forced open by Western imperialist forces rather than before. This
contradicts the recent claim of the so-called “Little Divergence” within Asia
before the mid-nineteenth century.8
400
350
China Japan
300
250
200
150
100
1850
1872
1875
1878
1881
1884
1887
1890
1893
1896
1899
1902
1905
1908
1911
1914
1917
1920
1923
1926
1929
1932
1935
1938
Figure 1.1 Real index of Chinese and Japanese per capita GDP (1850 = 100)
Sources: Maddison Project database, at www.rug.nl/ggdc/historicaldevelopment/maddison/
releases/maddison-project-database-2018?lang=en, but database reindexed at 100 in 1850
7
K. Fukao, D. Ma, and T. Yuan, “Real GDP in Pre-war East Asia: A 1934–36 Benchmark
Purchasing Power Parity Comparison with the U.S.,” Review of Income and Wealth 53.3
(2007), 503–37.
8
According to a recent estimate by J.-P. Bassino, S. Broadberry, K. Fukao, B. Gupta, and
M. Takashima, “Japan and the Great Divergence, 730–1874,” Explorations in Economic
History 72 (April 2019), 1–22, Japanese per capita GDP in 1990 dollars stood at 904 against
China’s 600 in 1850. This led to the claim of the “Little Divergence”“ within East Asia
before it was opened to Western colonialism. This would be inconsistent with the
estimated series in Figure 1.1 which used growth rates data back-projected from the
1930s.
19

debin ma
But was the Chinese economy as stagnant as revealed in Figure 1.1, and
specifically, how well did aggregate GDP statistics capture the overall pattern
of economic change during this era? Thomas Rawski noted the jarring
mismatch between the rapid pace of growth in modern sectors such as
trade, modern industry, banking, and monetary aggregates and the relatively
stagnant GDP profile in Figure 1.1. The reason behind this mismatch is that
the share of modern sectors is a tiny part of GDP dwarfed by the traditional
agriculture and handicrafts sectors. For example, modern factory produc-
tion – which recorded 8 percent annual growth between 1912 and 1936 – was
only 3.4 percent of the total GDP even in the 1930s.9 The attempts by Rawski
to raise the agricultural growth rate in order to make a case for overall
economic growth between 1912 and 1936 were – as he himself admitted –
based on strong assumptions. In another attempt at reconciliation, Ma
derived regional GDP estimates and argued that increased per capita output
and structural changes of the sort associated with Simon Kuznets’s concept of
modern economic growth only occurred in two major regions of the lower
Yangzi and the Northeast (Manchuria), where foreign investment and insti-
tutions were quite important.10 In this chapter, rather than attempt to
reconcile the aggregate with the sectoral statistics, I intend to reconstruct
a new narrative as revealed by the statistics to capture the timing and multi-
faceted nature of economic change in modern China.
External Trade, Industry, Infrastructure, and Capital

Accumulation
We first start out with one of the most reliable and continuous economics
statistics for the entire 1865–1949 period – the import and export data as recorded
by China Maritime Customs.11 In contrast to aggregate GDP statistics, Figure 1.2
reveals a remarkable expansion of China’s international trade, with real Chinese
imports and exports increasing eight- to tenfold between 1867 and 1932.
However, external trade did not take off right away after China’s forced opening
in the mid-nineteenth century. While the increase in external trade was con-
tinuous, there was clearly an acceleration from the 1880s and 1890s onward,
raising the Chinese share in world trade to a peak of more than 2 percent of
global trade flows in the late 1920s, a ratio that was not regained until the 1990s.
So the dream of China as a huge market for British manufacture – “every
9
Ma, “Economic Growth,” 364. 10 Ma, “Economic Growth.”
11
See below in this chapter and the chapters by Shiue and Keller, Kaske and Lin in this
volume on the organization and data of the Maritime Customs system.

200
180
160
Imports Exports
140
120
100
x 10000
80
60
40
20
0
1867
1870
1873
1876
1879
1882
1885
1888
1891
1894
1897
1900
1903
1906
1909
1912
1915
1918
1921
1924
1927
1930
Figure 1.2 Real value of Chinese imports and exports at 1933 constant prices (thousand yuan)
Source: Minami Ryoshin 南亮進 and Makino Fumio 牧野文夫 (eds.), アジア長期経済統
計 3 中国 (Long-Term Asian Statistics, vol. 3, China) (Tokyo, Tōyō Keizai Shinpōsha,
2014), Table 6.1.9
Chinaman wears a cotton cloth of the Lancashire factories” – barely came to

fruition until three decades after Britain’s forced imposition of free trade.
It turns out that the 1880–1890 turning point in Chinese foreign trade was
not entirely coincidental. In the absence of comprehensive economic statis-
tics for the entire period, we can gauge economic activities from the meticu-
lous firm and banks data compiled by a generation of scholars, which
culminated in the works of Du Xuncheng.12 Figure 1.3 plots the number of
newly established Chinese and Western firms and banking institutions from
the 1840s. Du classfied his firms as the so-called nationalist capitalist enter-
prises, which are basically Chinese modern firms connected with the use of
Western technology and production methods. Not surprisingly, following
the signing of the Nanjing Treaty in 1842, Western firms and banks took the
lead in establishing trade and financial enterprises in the designated treaty
port areas, with a steady pace of growth marked by a small uptick from the
1890s for Western firms. While Western firms and banks were leading the
way, Chinese follow-up or catch-up in terms of the number of firms remained
modest until the 1890s, when an outburst of growth completely overtook
Western establishments. The peak of the establishment of modern Chinese
12
I want to thank James Kung for alerting me to the use of this set of data for this purpose.
21

debin ma
300
250 Foreign banks Chinese banks
200
Foreign firms Chinese firms
150
100
50
1901
1865
1897
1913
1873
1869
1917
1921
1889
1925
1841
1845
1909
1857
1861
1877
1885
1905
1893
1849
1853
1881
Figure 1.3 Western impact and Chinese response: number of newly established modern
firms and banks
Sources: Du Xuncheng 杜询诚, 民族資本主义与旧中国政府, 1840–1937 (Chinese
Capitalism and the Old Chinese Government, 1840–1937) (Shanghai, Shanghai shehui
kexueyuan chubanshe, 1991). Huang Guangyu 黃光域, 外国在华工商企业辞典
(Dictionary for Foreign Industrial and Business Enterprises in China) (Sichuan, Sichuan
renmin chubanshe, 1995)
banks occurred during the 1910s and 1920s as modern banks were only
permitted around the turn of the century following Qing legislation.
Although a very problematic measure of actual economic activities, the
number of firms and banks turned out to capture quite accurately both
the timing and the turning point of Chinese industrialization. Calculations
from the data compiled by Du Xuncheng showed that the nominal annual
industrial investment by Chinese nationals in the 1914–1925 period was eleven
times that of the 1840–1911 period.13 Similarly, Rawski’s figure for “modern-
oriented” fixed investment (calculated from consumption of cement, steel, and
machinery) grew at an average annual rate of 8.1 percent between 1903 and 1936,
outpacing Japanese gross domestic fixed capital formation in mining, manufac-
turing, construction, and facilitating industries, which advanced at an annual
rate of 5.0 percent. Likewise, between 1902 and 1931, inflows of foreign direct
investment also grew at annual rates of 8.3 percent, 5 percent, and 4.3 percent for
13
Du used 1911 as the cutoff period; the contrast of industrial expansion versus stagnation
would be even sharper if the cutoff period was the mid-1890s. Throughout this period,
the Chinese monetary standard was silver-based with moderate inflation. See Ma,
“Economic Growth.”
22

900
Cotton yarn
800
Industrial production
700
Fixed capital formation
600
Linear (real GDP)
500
400
300
200
100
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
Figure 1.4 Real production and GDP indices (1912 = 100)
Sources: Minami and Makino, アジア長期経済統計, Tables 4.D.1 for cotton yarn and
4.2.1 for industrial production. T.G. Rawski, Economic Growth in Prewar China (Berkeley,
University of California Press, 1989), Table 5.2, p. 245, for fixed capital formation
Shanghai, Manchuria, and the rest of China respectively.14 Transport develop-

ment measured by China’s railway track length increased sharply from a mere
364 kilometers in 1894 to over 21,000 by 1937.15
Figure 1.4 plots the real production indices of machine-produced cotton yarn,
modern industrial production, and capital stock. They grew by between six- and
eightfold between 1912 and 1936 against a largely stagnant real GDP for the same
period. These statistics were consistent with the historical narrative. Beginning at
the very end of the nineteenth century, activity in mining and manufacturing
accelerated sharply from its small initial base. Factory production, initially
focused on textiles, food processing, and other consumer products, concentrated
in two regions: the lower Yangzi area, where both foreign and Chinese entre-
preneurs pursued factory expansion in and around Shanghai, and China’s
northeast or Manchurian region, where Japanese initiatives predominated. By
14
See L. Brandt, D. Ma, and T.G. Rawski, “From Divergence to Convergence: Re-
evaluating the History behind China’s Economic Boom,” Journal of Economic
Literature 52.1 (2014), 45–123.
15
See Elisabeth Köll’s chapter in this volume for railways and telegraph lines.
23

debin ma
1935, Chinese factories, including some owned by British or Japanese firms,

produced 8 percent of the world’s cotton yarn (more than Germany, France,
or Italy) and 2.8 percent of global cotton piece goods production. Despite the
importance of foreign investment in Shanghai and especially in Manchuria,
Chinese-owned companies produced 73 percent of China’s 1933 factory output.
Growing production of light consumer and industrial goods, combined with the
accumulation of experience in operating and repairing modern machinery,
generated backward linkages that spurred new private initiatives in machinery,
chemicals, cement, mining, electricity, and metallurgy.16
A Financial Revolution
The spectacular growth of Chinese banks from the 1900s, as revealed in Figure 1.3,
was the outcome of a remarkable triumph of a free-banking version of the silver
standard championed by largely privately held Chinese banks, foreign financial
institutions, and traditional money shops. Figure 1.5 shows that while total specie,
as measured by silver bullion, dollars, and copper cash, barely registered any
increase during 1911–1936, total money supply (M1), which includes specie,
increased at an annual rate of 5 percent between 1911–1916 and 1931–1936. This is
only possible because the bank notes and deposit components of M1 surged at
a remarkable annual rate of 9.5 percent during the same period. As a result, the
estimated share of notes and deposits in M1 money supply rose from between 22.3
and 34.6 percent in 1910 to a minimum of 40.4 percent in 1925 and 83.2 percent in
1936, with the turning point marked by an uptick in M1 around 1917–1918.17
Overall, Chinese banks – mostly modern but also including native banks –
accounted for the lion’s share of the growth in this period. The ratio of deposits
held by Chinese modern banks relative to foreign banks increased from about
two in the 1910s to about four in 1930–1935. The same ratio for banknotes held by
Chinese over foreign banks increased from 1.5 to about three for the same period.
Remarkably, price levels remained stable, whereas total money supply nearly
tripled between the 1910s and 1930s, while annual GDP growth registered no
more than 2 percent during the same period, indicating a heightened degree of
monetization and financial deepening. These growth rates translate into what Ma
observed as nothing short of a financial revolution, as demonstrated in
Figure 1.5.18
16
See Rawski, Economic Growth in Prewar China, Chapter 2.
17
Rawski, Economic Growth in Prewar China, p. 157.
18
See D. Ma, “The Rise of a Financial Revolution in Republican China in 1900–1937:
A Survey and New Interpretation,” Australian Economic History Review 59.3 (November
24

10000
9000
8000 Specie (silver and copper)
7000
6000 Total money supply (M1)
5000
4000
3000
2000
1000
0
1911
1913
1915
1917
1919
1921
1923
1925
1927
1929
1931
1933
1935
Figure 1.5 Metal currencies (silver and copper) and money supply in China, 1911–1936
(million yuan)
Source: D. Ma, “The Rise of a Financial Revolution in Republican China in 1900–1937:
A Survey and New Interpretation,” Australian Economic History Review 59.3 (2019), 242–62
Nonetheless, the fact that most of the transformations we have described so

far were connected with China’s modern and foreign sectors has led to the rise
of what many called a dualistic economy characterized by a sharp divide
between the small pockets of modern cities or treaty ports and the vast rural
hinterland, a phenomenon famously described by R.H. Tawney as “small
islands of privilege at the seaports and on the great rivers . . . a modern
fringe . . . stitched along the hem of the ancient garment.”19 Rhodes
Murphey went further, to describe the impact of treaty ports on China as
being like “a fly (who) could ultimately irritate its host enough to provoke
a violent counterreaction, but not to change the elephant’s basic nature.”20
However, as I will argue below, because of political-economy and institu-
tional spillover, the impact of the external and modern sector cannot be
captured by quantitative effects alone. To understand this peculiar pattern of
economic change, namely an economic takeoff around the end of the
nineteenth century after a hiatus of four decades after being opened to
2019), 242–62. See the chapter by Li and Yan in this volume for a longer period of money
supply.
19
R.H. Tawney, Land and Labour in China (London, Allen and Unwin, 1932), p. 13.
20
Cited in E. Motono, Conflict and Cooperation in Sino-British Business, 1860–1911: The Impact of
the Pro-British Commercial Network in Shanghai (New York, St. Martin’s Press, 2000), p. 166.
25

debin ma
Western imperialism, I turn to a review of historiography and build a new

conceptual framework to link intellectual evolution with economic cycles.
Paradigms and Frameworks

Western Impact and Chinese Response: A Re-assessment of
Historiography
Among the most important modernization paradigms that explains the
modern Chinese development is the so-called Western-impact-and-Chinese-
response framework championed by the sinologist John Fairbank. Writing in
1954, Teng and Fairbank remarked,
Since China is the largest unitary mass of humanity, with the oldest continu-
ous history, its overrunning by the West in the past century was bound to
create a continuing and violent intellectual revolution . . . Throughout this
century of the “unequal treaties,” the ancient society of China was brought
into closer and closer contact with the then dominant and expanding society
of Western Europe and America. This Western contact, lent impetus by the
industrial revolution, had the most disastrous effect upon the old Chinese
society. In every sphere of social activity the old order was challenged,
attacked, undermined, or overwhelmed by a complex series of processes –
political, economic, social, ideological, cultural – which were set in motion
within China as a result of this penetration of an alien and more powerful
society. The massive structure of traditional China was torn apart . . . The
old order was changed within the space of three generations.21
To some degree, for China, as for many other non-Western countries,

modernization is nearly equivalent to westernization, and in this regard the
Chinese record is a partial failure in comparison with Japan, the only non-
Western country to succeed at that time. However, this Fairbankian impact-
and-response framework itself was challenged by Fairbank’s own student,
Paul Cohen. In his widely acclaimed book Discovering History in China, Cohen
argues that the impact-and-response framework may inadvertently lead to an
amplification or simplification of the Western influence in China, neglecting
the role of Chinese agency and China’s internal dynamics. Indeed, as Cohen
argued, much of the so-called Chinese response to or embrace of the West
may well be superficial, and reflect more the domestic dynamics. There were
also regional and temporal variations of the Western influence, which could
21
Written by Ssu-yu Teng and John K. Fairbank, cited in P. Cohen, Discovering History in
China: American Historical Writing on the Recent Chinese Past (New York, Columbia
University Press, 1984), pp. 9–10.
26

be highly visible or even predominant among certain social groups and the
treaty port zones, or more specifically in what Cohen referred as the
Hong Kong–Shanghai corridor, but largely absent from the vast hinterland.
In short, Cohen calls for the discipline of sinology to return to China and to
viewing China on her own terms.22
Cohen’s approach of “discovering history in China” offers a powerful
corrective to what some would see as the Eurocentric tinge in the Fairbank
framework and a possible inspiration to the recent California school’s
emphasis on China’s own superior indigenous initial conditions in living
standards, commercial and contractual traditions, and human capital before
the mid-nineteenth-century onset of Western imperialism in China. Most
interestingly, the lower Yangzi region, which the California school cham-
pioned as having comparable living standards and developments to England
or the Netherlands during the eighteenth century, also hosted the treaty port
of Shanghai, which was to become the leading financial and industrial city of
the nineteenth–twentieth centuries. In this regard, the Shanghai “economic
miracle” of the twentieth century can be viewed as as much an import from
the West as attributable to its own superior native roots.23
This overseas intellectual inward turn towards a China “on her own” forms an
interesting contrast to Chinese historiography within mainland China, which at
least from the 1950s had long built a dogmatic version of the Marxist narrative that
posits a unilinear, universal – ironically highly Eurocentric – version of the law of
societal evolution, progressing from the lowest stage of slavery in ancient times,
to feudalism, capitalism, socialism, and communism. In this historiography, with
the nineteenth- and twentieth-century Chinese economy and society character-
ized as semicolonial and semifeudal, the encroachment of Western imperialism –
although a national humiliation – represents progress, or an assault on
a backward “feudalist“ Qing China by a more advanced and productive capitalist
system.24 In this regard, the paradigms of modernization and Marxism converge.
Very curiously, one offshoot of this mainland Marxist historiography is
a lively debate on the so-called “Sprouts of Capitalism,” inspired by Mao
Zedong’s claim in 1939: “The development of China’s commodity economy
within her feudal society has spawned the sprouts of capitalism. Even without
the influence of Western capitalism, China will gradually develop into a capitalist
22
Cohen, Discovering History in China.
23
See Ma, “The Rise of a Financial Revolution,” for the Shanghai miracle.
24
Xu Dixin 许涤新 and Wu Chengming 吴承明, 中国资本主义发展史，第一卷中
国资本主义的萌芽 (A History of Capitalist Development in China, Vol. 1, Sprouts of
Chinese Capitalism) (Beijing, Renmin chubanshe, 1993).
27

debin ma
society on its own.”25 Generations of Chinese scholarship labored to discover

those so-called traces or “sprouts” of capitalist production relations in traditional
Chinese agriculture, handicrafts, and commerce. Despite coming from different
or opposite ideological and institutional backgrounds, this emphasis on indigen-
ous sources of Chinese development saw an interesting confluence with the
California school. Alternatively put, just as China might have progressed “natur-
ally” to capitalism without the onslaught of Western imperialism, the Industrial
Revolution might have equally eluded the West or England had they not
stumbled up coal deposits or the discovery of the New World.
The idea that eighteenth-century China might already have been on the cusp
of, or a natural progression toward, modern capitalism or industrial revolution
does not sit well with the lackluster Chinese aggregate economic record of the
nineteenth to twentieth centuries, as partly revealed in Figure 1.1. While it is
beyond the scope of this chapter to evaluate the entire literature on pre-
nineteenth-century China’s initial conditions, I argue that our interpretation of
the patterns of economic change or the lack thereof in modern China needs to be
placed in the context of a rapidly industrializing West and Japan and needs to be
sought in the scarcity of intellectual resources rather than of natural resources.
Below we turn to our framework on the importance of intellectual resources.
Ideology and Economic Changes: A New Framework

Although they were an ethnic minority from China’s seminomadic northeast-
ern frontier, the ascendancy of Manchu rule under the Qing actually marked
the culmination of a millennia-long evolution and maturing of a highly cen-
tralized, unitary Chinese political regime governed by an absolutist emperor at
the top of the power pyramid. Aided by a formal bureaucracy recruited
through a highly structured nationwide (national–metropolitan–provincial)
civil service examination rooted in Confucian classics, imperial China could
implement a system of direct administrative rule (junxian 郡县) with mostly
educated officials assigned to directly govern over a thousand counties nation-
wide over her vast territories on a two- to five-year rotating basis. Hence the
combination of direct administrative rule with the legitimacy of imperial
personnel appointment became a potent instrument of political control and
rule that no or few other traditional regimes had mastered.26 In this regard,
25
Xu and Wu, 中国资本主义发展史, p. 4.
26
Two caveats were important to this system. First, however much impersonality and
neutrality characterized China’s imperial regime, they were frequently compromised by
the emperor’s personal rule, and his personal entourage of eunuchs, consort, and other
inner court staff. Second, the highly centralized personnel appointment coexisted with
28

Chinese rulers managed to break free from the constraints of feudal and local
autonomous institutions that had characterized Europe, and possibly much of
the rest of the world for that matter. Beyond the borders of empire, where they
could not implement direct rule, the Qing, like previous dynasties, constructed
a China-centered international order through the so-called tributary-states
trade system. This system engulfed neighboring small states or territories in
East and Southeast Asia as near protectorates that would pose no major
military or political threat. Hence absolutism Chinese-style curtailed interstate
competition and weakened independent vested interests, civil society, and
autonomous political and social groups, all to serve the purpose of minimizing
any potential threat to the throne from below.27
In this system, imperial or political legitimacy hinged on the state’s capacity to
suppress internal dissent and external challenge. Although military superiority or
repressive capacity remained paramount to political legitimacy, they were
insufficient by themselves and were only resorted to under extreme cases of
threat. Hence crucial to this legitimacy is the development of a consistent ruling
ideology or belief system that would be – in our case – Confucianism or neo-
Confucianism that rationalized Chinese-style absolutist imperial rule. To estab-
lish the monopoly of this ideology over the interpretation of political and social
events, systematic control and manipulation of information or alternative ideolo-
gies were essential to the system as characterized by the widespread incidence of
literary inquisitions and intellectual persecutions.28 Although this system was
effective in perpetuating imperial rule and the status quo, it stifled the possibilities
of endogenous development of ideological and institutional transformation from
within and from below, leading to what Jin Guangtao and Liu Qingfeng termed
the super-stable structure of the Chinese empire before the mid-nineteenth
century.29 Changes or even revolutions could happen in this system. But without
corresponding intellectual development in new political and institutional ideolo-
gies, this just meant that violent revolution or dynastic overthrow (as happened
a very decentralized delegation of economic resources at the local level. See D. Ma,
“Political Institutions and Long-Run Economic Trajectory: Some Lessons from Two
Millennia of Chinese Civilization,” in M. Aoki, T. Kuran, and G. Roland (eds.),
Institutions and Comparative Economic Development (Basingstoke, Palgrave Macmillan, 2012),
pp. 78–98.
27
See Ma, “Political Institutions and Long-Run Economic Trajectory”; Ma and Rubin,
“The Paradox of Power.”
28
Wang Fansen 权力, Jin Guantao 金观涛, and Liu Qingfeng 刘青峰, 兴盛与危机，论
中国社会的超稳定结构 (The Cycle of Growth and Decline: On the Ultrastable
Structure of Chinese Society) (Beijing, Falu chubanshe, 2011). Ge Zhaoguang, “十八
世纪中国的盛世危机.”
29
Jin and Liu, 兴盛.
29

debin ma
very often in Chinese history) were part of dynastic changes that harbored no
fundamental institutional change.
The link between the stagnation of ideology and the limits to institutional
change is echoed by another intellectual historian. Wang Fansen contends that
the contours of China’s traditional intellectual resource endowments define the
boundaries of political and social changes and the horizon over which Chinese
intellectuals could hover for solutions. The most outstanding examples are the
late Ming–early Qing Confucian scholars such as Huang Zhongxi 黄宗羲 and
Gu Yanwu 顾炎武, who were resurrected in the late nineteenth century as
China’s enlightened sages for their penetrating insights and attacks on the ills of
the Ming’s absolutist and repressive regime. However, when it came to pro-
posals for reform, the seventeenth-century Huang and Gu could only conceive
of a passionate call to return to Chinese antiquity and the classics. These attitudes
are, as Wang argues, typical of Chinese elite intellectuals of the time in general.30
By the mid-nineteenth century, however, the resources of a vast and
unexplored intellectual horizon had opened up, but they were from an
alien Western tradition, written in Western languages that few, if any,
Chinese intellectuals mastered or even heard of at the time. But it was not
merely the scarcity of foreign-language talents or the suitability of
Western ideology to Chinese reality that posed the problem. Rather it
was something far more basic and fundamental: the willingness and
capacity to recognize, first, cognitive dissonance within China’s existing
system, and, following that, the need for a paradigmatic change. As
eloquently stated by historian of thought Ge Zhaoguang, it was precisely
the glory of Qing’s so-called triumphant eighteenth century that sowed
the seeds of her failed response to Western challenge in the mid-
nineteenth century. The Qing rulers’ success in the control of a vastly
expanding and diverse empire increasingly through repression of intellec-
tual dissent and unity of ideology left her with a unique sense of grandeur
and confidence that meant she could only treat the outside world with
nothing but sheer ignorance and arrogance. In that regard, Qianlong’s
reply to Lord Macartney in 1792 is quite gracious by that account. This is
echoed by Wang Fansen: until a major political and social-economic crisis
could be keenly perceived, the Chinese intellectual soil – thoroughly
soaked in the deep cultural self-confidence of the traditional literati –
was simply not fertile enough for the transplant of new intellectual
resources.31
30
Wang, 戊戌. 31
Ge, 十八世纪; Wang, 戊戌.
30

In comparison, Tokugawa Japan’s relatively decentralized feudal system,

peripheral geopolitical position in the China-centered world order, and long
tradition of absorbing foreign (mainly Chinese) culture and ideologies engen-
dered a different cultural attitude towards Western impact and ideology
during the mid-nineteenth century. For example, catalytic events such as
the 1842 Sino-British Opium War did not propel immediate institutional or
political changes in Qing China, but did send a powerful warning of the
impending Western threat to the elites of neighboring Tokugawa Japan.
Indeed, China’s defeat in the 1840s and subsequent failures may have so
alarmed Tokugawa Japan that in a couple of decades that it turned China
from once being a model to be admired to being a lesson to be avoided.32
But effecting paradigmatic changes carries risks and uncertainties as they upset
the traditional ideological equilibrium. They are more likely to take place
through other successful examples or the arrival of new ideas or paradigms
that provide a new and consistent framework to explain the cognitive disson-
ance. In this regard, China’s defeat in the 1894–1896 Sino-Japanese War provoked
greater reaction for Chinese reform than the initial Western imperialism. The
Meiji success with Western models as a newly westernizing Asian nation in
response to the same Western challenge that Qing China had faced now began
to stir up cognitive dissonance in the giant neighboring Qing.
In this sense, institutional change requires time, experiments (or sometimes
historical accidents), and a feedback loop between events, experiments, and
ideas. We arrive at a model explaining why reform had the twists and turns with
their particular temporal and regional patterns as described above in this chapter.
Our framework is consistent with some of the theoretical and historical works of
Timur Kuran, Joel Mokyr, Murat Iyigun, and Jared Rubin, which emphasizes the
importance of ideas and the multiple equilibria of divergent paradigms, as well as
cumulative progress, or sudden surges forward or sliding backward. Our
emphasis on ideology and ideological change pushes us beyond just the para-
digm of institutions as emphasized by the advocates of institutionalism.33
32
Wang, 戊戌.
33
For models of ideology and ideological change, see M. Iyigun and J. Rubin, “The
Ideological Roots of Institutional Change” (2017 working paper). T. Kuran, Private
Truths, Public Lies: The Social Consequences of Preference Falsification (Cambridge, MA,
Harvard University Press, 1997). For the importance of ideas and feedback loops, see
J. Mokyr, The Gifts of Athena: Historical Origins of the Knowledge Economy (Princeton, NJ,
Princeton University Press, 2002). For an overwhelming emphasis on the role of
institutions, see D. Acemoglu, S. Johnson, and J. Robinson, “Institutions as the
Fundamental Cause of Long-Run Economic Growth,” in P. Aghion and S. Durlauf
(eds.), Handbook of Economic Growth (Amsterdam, Elsevier, 2005), pp. 385–472.
31

debin ma
Western Impact and Chinese Response Again

Within our new framework, we return to the paradigm of Western
impact and Chinese response with an important twist: the impact of
the West is critically important, but only to a degree and in the
manner in which the Chinese (or Japanese) manage to respond. We
emphasize the role of Chinese agency in utilizing, adapting, and
eventually redesigning new rules and institutions to propel economic
change.
In the nineteenth and twentieth centuries, Western imperialism did
not fully subjugate China, but manifested itself through the acquisition
of trading rights, leased territories, and treaty ports, as well as extra-
territoriality and spheres of interest. In the early twentieth century,
when central control was weakened, Western treaty ports expanded
rapidly at the expense of Chinese sovereignty. The expansion of these
privileges and extraterritorialities in an era of political chaos and
national disintegration turned out to be a blessing in disguise for two
reasons.
First, some of these “privileges” happened to coincide with necessary
conditions for growth, namely the maintenance of peace and public order,
the security of property rights and contract enforcement, freedom from
arbitrary taxation or official exaction, and the right to transparent rules and
predictable jurisprudence. Second and more importantly, as shown later,
foreign “privileges” in the treaty ports were often taken advantage of by
Chinese businesses and residents.
Here, we examine the evolution of two key consequences of Western
intervention. First is Shanghai, China’s largest treaty port, particularly its
International Settlement, which resulted from a merger of all Western (and
later Japanese but excepting the French) concessions operating with
a governance structure akin to a European type of self-governing incorpor-
ated urban community. Shanghai’s foreign residents organized a Municipal
Council with members elected by an association of taxpaying Western, later
Japanese and eventually (in 1928) Chinese, residents. The council operated
according to the rule of law vested in its own mini-constitution, levying taxes
and fees; running its own prison, police, and volunteer army; and providing
public goods such as roads, utilities, and port facilities. The power and
territory of the International Settlement greatly expanded in the wake of
32

the 1911 Qing collapse, with full territorial jurisdiction over all its residents,
including the Chinese.34
The second key institution – China Maritime Customs – had similar origins
and a similar trajectory. The low fixed tariffs imposed by Western powers
were initially collected by Chinese customs officials, but increasingly over-
seen by foreign consuls who set up the China Maritime Customs in Shanghai
in 1854. Although nominally an imperial Chinese organization, the Customs
gained autonomy with Britons dominating its senior staff. Its staff gradually
came to include large numbers of Westerners and later Japanese, with the
promotion of Chinese nationals into senior positions only starting in 1929.35
With over 20,000 staff in forty main customs houses across China, the
Customs rapidly emerged as China’s most stable and efficient centralized
hierarchic bureaucracy even as China herself descended into political disinte-
gration. Following the Qing collapse in 1911, the Maritime Customs effect-
ively took over the collection of the Customs charges and the distribution of
the net revenue.
The insertion of these colonial institutions unintentionally ruptured China’s
long-standing imperial monopoly of power. The resulting gaps in imperial
control created room for newly emergent Chinese business interests, many
with deep links to foreign businesses, to create new networks of power and
wealth that intensified growing pressure for formal political and institutional
changes, especially following China’s defeat in the 1894–1896 Sino-Japanese
War. However, the institutional changes arising from Western influence could
only have nationwide effects after they had triggered intellectual and political
responses from Chinese elites, which is where we turn to now.
Liang Qichao 梁启超, China’s foremost intellectual and reformer of the era,
succinctly summarized the Chinese response. Writing in 1923, Liang surmised
that changes or reforms would have never got off the ground until the Chinese
were willing to recognize and acknowledge that there were problems to begin
with. This partly began during the period from 1842 to 1894 when the Qing saw
firsthand the power of Western military equipment used to suppress the Taiping
Rebellion (1851–1864). This opened the door to the introduction of Western
(particularly military) technology and machines, or, in Liang’s phrase, material
things (qiwu 器物).
34
See Ma, “Economic Growth”; Ma, “The Rise of a Financial Revolution”; I. Jackson,
Shaping Modern Shanghai: Colonialism in China’s Global City (Cambridge, Cambridge
35
H. van der Ven, Breaking with the Past: The Maritime Customs Service and the Global Origins
of Modernity in China (New York, Columbia University Press, 2014).
33

debin ma
China’s 1894–1896 naval defeat by Japan shattered her confidence in her own
systems, terminated this initial stage, and inaugurated a second stage of introdu-
cing new “institutions” – (zhidu 制度). This process, which continued for two
decades until roughly 1917–1918, brought the comprehensive importation of
Western-style government and law, along with modern corporate enterprises
and financial institutions. The third stage followed the recognition that institu-
tional transfers could not succeed without understanding and absorbing their
underlying cultural and ideological foundations as reform demands changes in
both spirit and form. The result was the New Culture movement, which
brought the massive introduction of Western culture and ideology (wenhua
文化), even leading to the radical abrogation of Confucian ideology.36
Building on the insights of Liang, Yang Nianqun added a geographic
dimension to Liang’s temporal stages of reform, matching each of these
three reform sequences with distinctive regional political elites and with
Ming–Qing schools of Confucianism, centered in the provinces of Hunan,
Guangdong, and Jiangsu–Zhejiang, corresponding to China’s middle Yangzi,
Lingnan, and lower Yangzi macro-regions respectively.37 Yang’s regional
study drew inspiration from Paul Cohen’s early work on late Qing reform,
which emphasized the regional dimension of Western influence and high-
lighted the major cultural divide separating China’s littoral and hinterland.
More importantly, both Liang and Yang point to China’s own intellectual
endeavor in absorbing Western impact, as Cohen would put it, on her own
terms. These intellectual and ideological responses by the intellectual and
political elites paralleled practical efforts on the part of new economic actors
and networks who took advantage of Western privileges to push the bound-
aries beyond the traditional structure.
Patterns of Economic Change: From Machines

to Institutions to Ideology
The Age of “Machines”
The mid-nineteenth century, marked by the encroachment of Western
imperialism on Chinese shores, started off disastrously for the Qing dynasty.
36
Liang Qichao 梁启超, “五十年中国进化概论” (A Summary of Fifty Years of Chinese
Evolution), in 饮冰室文集点校第五集 (Selected Articles of the Yinbinshe Collection,
Vol. 5) (Kunming, Yunnan Jiaoyu chubanshe), pp. 3247–52.
37
Yang Nianqun 杨念群, 儒学地域化的近代形态：三大知识群体互动的比较研究
(The Modern Form of Regional Schools of Confucius: A Comparative Research into the
Interaction of Three Intellectual Groups) (Beijing, Sanlian chubanshe, 1997).
34

Qing’s defeats in the Opium War (1842) and other military battles, and even the
devastating Taiping Rebellion, which saw a calamitous cumulative loss of
60 million to 80 million lives, did not shake the elites’ faith in traditional
ideology and institutions.38 Political elites from China’s agrarian mid-Yangzi
heartland of Hunan and Anhui, such as Zeng Guofan 曾国藩 and Li
Hongzhang 李鸿章, rose to national prominence following their success in
mobilizing fiscal and military resources from their home provinces to eventu-
ally suppress the Taipings. Closely aligned with the official ruling ideology, the
Tongzhi Restoration (1861–1875) – which Mary Wright famously called the last
stand of Chinese conservatism – engineered a remarkable economic recovery
through the revitalization of traditional institutions: the reinstatement of
Confucian orthodoxy, the restoration of the national civil service examination
(largely interrupted during the Taiping Rebellion), and temporary exemption
from land taxes to lure cultivators to resettle war-torn agricultural regions.39
As a natural extension of the Tongzhi Restoration, the Self-Strengthening
movement (1860–1894) initiated programs that aimed to expand Chinese
military strength by developing a small number of Western-style, capital-
intensive enterprises financed or sponsored by the state and directed by
prestigious officials or merchants with official connections. Although these
enterprises, which included arsenals, factories, and shipyards, were fraught
with inefficiency and corruption, they did record modest achievements. (See
the chapter by Chi-kong Lai in this volume.)
Nonetheless, the overall ideological orientation during this period remained
backward-looking. In contrast to the concurrent Meiji reform in Japan, there
were no reforms that touched the fundamentals of the traditional regime: there
was no introduction of a modern constitution or commercial law and no reform
in the currency system, modern banks and modern infrastructure such as
railroads were expressly prohibited, and steamships were limited to the Yangzi
and other major rivers. Nonetheless, the ground on which the traditional
structure had rested were shifting with the insertion of Western imperialism.
Foreign Trade in the Age of Machines

Although the treaty port system accelerated the arrival of new technologies
and institutions, industrialization lagged far behind the opportunities opened
38
For the controversies among Qing officials’ interpretation of the Opium War defeat,
see Mao Haijian 茂海建, 天朝的崩溃鸦片战争再研究 (The Fall of the Celestial
Empire: A Re-examination of the Opium War) (Beijing, Sanlian shudian, 2013).
39
M. Wright, The Last Stand of Chinese Conservatism: The T’ung-Chih Restoration, 1862–1874
(Stanford, CA, Stanford University Press, 1962).
35

debin ma
up by the inflow of trade and technology. Attempts by Chinese and European

entrepreneurs to take advantage of opportunities linked to new technologies
and trade arrangements reveal the presence of powerful obstacles to innov-
ation within China’s late Qing economy. These are most clearly visible in the
obstacles confronting private efforts to introduce new technologies and
business structures.40
These obstacles existed precisely because the legality of traditional Chinese
commercial and business activities rested on patronage rather than universal
property rights. Often commercial guilds and organizations acquired mon-
opoly privileges in exchange for paying a fixed tax quota to the government,
and hence some form of shelter from arbitrary exactions. This institutional
arrangement, according to Eiichi Motono, formed the structural foundation
of traditional Chinese merchant groups and networks under official
patronage.41
However, the Western presence established a new source of power
and authority in China. The “unequal treaties” granted every foreigner
the right to trade and own property in the treaty ports, most import-
antly subject to Western legal systems. The imposition of “free trade”
and extraterritorial privileges turned out to have unintended conse-
quences. The treaty port system imposed by Western “free-trade”
imperialism set Chinese trade tariffs at a modest 3 or 5 percent during
this period. To shelter their trade from the arbitrary native transit taxes
known as lijin 釐金, the Westerners insisted on a flat transit tax when
their goods moved through China’s interior. These transit levies were
assessed and, after 1911, collected by the foreign-controlled China
Maritime Customs.42
Chinese merchants soon found opportunities to benefit from these
“Western” privileges through false registration of their produce as destined
for export (rather than domestic use), outfitting their ships (or junks) with
Western flags, and investing their capital in foreign-owned businesses or
simply registering their businesses as owned by foreign nominees.43 More
critically, what Motono referred to as English-speaking Chinese merchants
began to develop their own commercial networks outside the traditional
commercial groups, who had relied on the payment of native transit taxes as
40
See Brandt, Ma, and Rawski, “From Divergence to Convergence.”
41
Motono, Conflict and Cooperation. Also see the chapter by Chi-kong Lai in this volume.
42
See the chapter by Kaske and Lin in this volume.
43
See Motono, Conflict and Cooperation.
36

a means to acquire monopoly privileges. Motono identified the late 1880s –

a timing that corresponds well with the first surge of Chinese firms, as
revealed in Figure 1.2 – as the key turning point when traditional Chinese
commercial organization began to crumble.44 Motono argues convincingly
that these merchant–official nexuses forged through tax-farming arrange-
ments formed the structural foundation of traditional Chinese merchant
groups and networks. The Western presence, although small in relation to
China’s national economy, constituted a new source of power and authority
in China that drew away Chinese merchants towards new commercial
groups formed outside the tax-farming-based system. This eroded the imper-
ial system’s long-standing monopoly over political power, just like “a great
dike may be breached by tiny termites.”45 He sees the transit pass system as
a vehicle for eroding the authority of mercantile guilds, fracturing traditional
group solidarity among members of particular trades, and enhancing the
property rights and security available to Chinese businesses.
By co-operating with, utilizing, or embracing Western extraterritorial privil-
eges, mostly in the treaty ports, these forces formed pressure points that eroded
the power base of traditional vested interests. Or, as more effectively put by
Motono, it was the Chinese themselves who promoted a Western impact on
China.46 But being confined largely within the treaty port zones and the
Western sphere of influence, this “Western impact” generated a “dual-track”
system which often placed Chinese business interests in an unfavorable pos-
ition. They caused backlashes and pushbacks and stirred up calls for formal
political and institutional changes within China to level the playing field.
These trickles and leakages turned into torrents following China’s defeat in
the Sino-Japanese War of 1894–1895 and the signing of the Treaty of
Shimonoseki, which granted foreigners the right to establish factories in the
treaty ports. Previously, foreigners only had the “privileges” to engage in
trade and finance, but not manufacture. This new arrangement imposed by
the Treaty of Shimonoseki opened the floodgates of foreign direct invest-
ment and pushed the Qing to accept and grant much broader property rights
and protection to all Chinese, not just in business but also in manufacturing.
This forms the essence of the 1903–1911 late Qing legal reform in commercial
law and the promotion of chambers of commerce.
44
See Motono, Conflict and Cooperation, pp. 166–70.
45
See Motono, Conflict and Cooperation, p. 169.
46
Motono, Conflict and Cooperation, p. 167.
37

debin ma
The Age of Institutions and China’s Turning Point

China’s defeat in the Sino-Japanese War of 1894–1896 by Japan, a nation long
regarded as a student rather than an equal, inflicted a profound mental shock
on Chinese elites and the public at large. It marked the end of the Self-
Strengthening movement and led to a sudden surge of interest in the
Japanese experience. Huang Zunxian 黄遵宪, the Qing ambassador to and
keen observer of Meiji Japan, wrote a landmark study of Japan’s transform-
ation in 1887. Despite being delivered to Zeng Guofan and Li Hongzhong and
circulated privately, Huang’s work drew little attention even when it was
printed in 1895. But China’s stunning defeat in 1896 turned this book into an
instant best seller, as lamented by one gentry that had we all paid attention to
Huang’s book earlier, China would have been spared over 200 million silver
taels of war reparations – something equivalent to five times the Qing state’s
annual peacetime revenue – extracted by Japan following her 1896 military
victory.47
Intellectual historian Ge Zhaoguang marks 1896 as China’s key intellectual
turning point. He shows that during the 300 years before 1894, Japan trans-
lated 129 Chinese works while the Chinese translation of Japanese works
amounted to a mere twelve. This trend reversed in the decade after 1896 with
958 Chinese translations of Japanese works but only sixteen translations in the
other linguistic direction.48 Indeed, when Japan became the first Asian society
to translate Western materials, the shared vocabulary of Chinese characters
quickly installed Japan as the natural intermediary in transmitting Western
culture to China, especially because of the inflow of massive numbers of
Chinese students and the outflow of Japanese advisers and teachers to
China.49
China’s naval defeat directly triggered the Hundred Days reform in 1898
backed by the young Guangxu Emperor (r. 1875–1908). Although centered in
Hunan – the heartland of Self-Strengthening bureaucrats – the reform’s
intellectual leaders came from Guangdong-based elites, such as Huang
Zungxian, Kang Youwei 康有为, and Liang Qichao, who had prior exposure
to Western influence. Although the reform was quickly crushed by conser-
vatives surrounding the emperor’s aunt, Dowager Empress Cixi, the agenda
of the failed Hundred Days reform formed the core of the Qing constitutional
movement of 1903–1911, modeled directly on Japan’s Meiji reforms.
47
Jin and Liu, 开放中的变迁, p. 66. 48 Ge, 中国思想史, p. 478.
49
M. Jansen, “Japan and the Chinese Revolution of 1911,” in J.K. Fairbank and K.C. Liu
(eds.), The Cambridge History of China, vol. 11, Late Ch’ing, 1800–1911, part 2 (Cambridge,
Cambridge University Press, 1980). Wang, 戊戌.
38

Beyond military victories, Meiji Japan offered a remarkable example of

a nation with similar (or humbler) cultural heritage which had, however,
managed to implement a comprehensive and thorough reform agenda when
confronted with a common Western threat. Through the successful adoption
of the gold standard in 1897 and the recovery of extraterritoriality in 1899,
Japan’s westernizing reform introduced Western institutions but actually
kept Western capital and influence at bay.50
China’s new reform effort, heavily influenced by Meiji, was comprehen-
sive and ambitious. It aimed to prepare China for a constitutional monarchy
by drafting a formal constitution that would establish national, provincial,
and local parliaments. Military modernization was high on the reform
agenda. Administrative reforms sought to modernize public finance and
adopt a national budget. The reform initiative gave birth to new Ministries
of Education, Trade, and Agriculture and encouraged the founding of local
chambers of commerce. Policy initiatives aimed at currency reform, the
establishment of modern banks, and the expansion of railroads and other
public infrastructure. The abolition of the millennia-old civil service examin-
ation in 1905 opened the door to a modern school system, giving birth to what
are today China’s best-known universities.
Although the resulting political decentralization may have inadvertently
hastened the Qing collapse, the rise of the new republic opened the door to
a massive Chinese experimentation with new ideologies and institutions
from the West. This indeed began China’s age of modernity, a genuine
experiment with constitutionalism, the rise of regional cosmopolitanism,
and the introduction of modern company law and of such new business
techniques as double-entry bookkeeping.51
Growing local autonomy fostered by political decentralization encouraged
the growth and maturing of civil society and, more importantly, boosted
political and ideological competition across different provinces and towns as
well as treaty ports. The comprehensive data presented in Chapter 11 of this
volume by James Kai-sing Kung show the turn-of-the-century breakpoint in
education, missionaries, and numbers of treaty ports, which partly became
the institutional foundation underpinning the rise of modern industry, bank-
ing, public finance, and monetary regime as discussed above in this chapter.
50
Y. Yoda, The Foundations of Japan’s Modernization: A Comparison with China’s Path towards
Modernization (trans. Kurt W. Radtke) (Leiden, Brill, 1996).
51
F. Dikötter, The Age of Openness: China before Mao (Hong Kong, Hong Kong University
Press, 2010).
39

debin ma
The Rise of Financial Revolution in the Age of Institutions

The temporal retreat of central imperial power opened new possibilities for
the rise of a quasi-political structure that rested on the institutional nexus of
Western treaty ports (most notably Shanghai) and the Maritime Customs
service.52 Both these institutions were intimately connected with Western
institutions or Western imperialism in China, which initially only served to
protect the limited Western or foreign business interests in the context of
extraterritorial privileges. However, in the wake of 1903 constitutional reform
and the Qing’s collapse in 1911, this mechanism began to be transferred to
China domestically.
The financial revolution that emerged during the politically chaotic 1910s
and 1920s illustrates the interplay between Western impact and Chinese
response. Shanghai’s free-trade extraterritorial status had long attracted
what later became some of the world’s premier Western banking institutions,
such as the Hong Kong and Shanghai Banking Corporation (HSBC) and
Chartered Bank. But modern Chinese banks, which only started near the
end of the nineteenth century, also found their homes there.53 Even trad-
itional family-owned native banks increasingly chose to locate inside the
Western-controlled Settlement. The relocation of the Shanghai Native
Bankers’ Association from the Chinese part of the city to within the
Settlement in 1917 and the founding of an association of modern Chinese
banks in 1918 marked the rise of Shanghai’s Chinese banking community as
the leading force in China’s macro-economic, monetary, and financial regu-
lation within a largely self-regulated free banking framework.
Initially, when the Chinese government borrowed from the foreign mar-
ket, its obligation to repay could be partly enforced through the coercive
power of Western gunboats and through the important intermediary institu-
tion of the China Maritime Customs, as the agency was relatively insulated
from the threat of the Chinese imperial government. When, in 1911, the
Maritime Customs directly took over the collection and remittance of the
Customs revenue, it opened an account with HSBC, which eventually
became the custodian bank of that portion of Customs revenue pledged as
security for the service of the government’s foreign debt.54
52
Ma, “The Rise of a Financial Revolution.”
53
In particular, the jurisdictional autonomy of the Settlement which sheltered the Bank of
China’s Shanghai branch from the Beijing government’s ruinous fiscal demand in 1916
highlights the value of the independence of the Settlement. See the chapter by Li and
Yan in this volume.
54
Ma, “The Rise of a Financial Revolution.”
40

This institutional mechanism for external borrowing was soon transferred

to become the cornerstone for a domestic market for Chinese government
bonds through the intermediary of large Chinese public banks. China’s
domestic public debt originated in 1914 with the new Republican government
in Beijing setting up an independent committee composed of Chinese and
Western bankers and Maritime Customs officials. This mechanism ensured
that tax revenue earmarked for debt repayment was remitted to a special
revenue account set up in the Western banks and later in Chinese banks
headquartered in Shanghai’s International Settlement. Maritime Customs
revenue formed the most secure form of central revenue for a weakened
Beijing government and therefore the most reliable collateral for servicing
the government’s foreign debt. With government bonds serving as part of
reserves for banknotes, modern banks injected new sources of money into
the macro-economy. The new viability of domestic public debt spawned
a vibrant secondary market in the 1920s and 1930s.
This key nexus offered an unusual but credible commitment for the
security of bondholders’ property rights and the repayment of government
obligations, which laid the institutional foundation for a financial revolution.
In a sharp departure from the traditional political regime that left credibility
at the mercy or benevolence of a strong and stable state, management of the
state’s obligations to bondholders now rested with an institutional mechan-
ism that had grown autonomous from the center. It was that particular
mechanism that allowed Chinese bankers and bondholders to place some
constraint on the power of the government with regard to public finance, and
by doing so it enabled the Chinese government to tap into the private wealth
of Chinese citizens for borrowing without coercion.
The Age of Culture and China’s Fateful Ideological Turn

In 1923, barely four years after the May 4th mass movement in 1919 kicked off
China’s new age of cultural movement, Liang Qichao wrote with exuberance
that Chinese understanding of Western ideology and culture had gradually
but irreversibly progressed since the days of Self-Strengthening. It was only in
1876 that a remark by Guo Songtao 郭嵩焘, China’s ambassador to Britain,
that the new Western “barbarians” confronting China then – unlike previous
Asian threats – also had 2,000 years of civilization, caused a huge uproar and
was roundly condemned as near blasphemy. Under Self-Strengthening, des-
pite the recognition of Western military and technological superiority,
Western knowledge came to China only in drips through a limited number
41

debin ma
of indirect translations of Western works commissioned at various govern-

mental arsenals.
This changed rapidly following China’s military loss to Japan. Liang
quipped that the first generation of advocates of Western learning were
great classic Confucian scholars – himself included – who knew nothing of
any Western languages. The same was true of the next generation of advo-
cates of Western learning, most of whom were Chinese students returned
from study in Japan. By the 1920s, however, a third generation of young
scholars, many returning from studies in Europe and North America, finally
took over the baton of Western learning and in certain cases carried it to the
extreme, such that, as Liang claimed, Karl Marx was now vying for equal
status with Confucius.
Liang’s exuberant banter about China’s intellectual transformation could
not conceal the complex and wrenching emotions incurred in the struggle to
introduce an alien culture and ideology to China as well as Japan. Indeed, in
both countries the first step to legitimizing the pursuit of Western learning
started with a step backward: going back to the Chinese classics.
Meiji Japan, for example, called upon the orthodox Zhuxi 朱熹 neo-
Confucian principles to legitimize the ouster of the Tokugawa shogunate
and the reinstatement of the Meiji imperium under a centralized bureaucratic
system. In similar fashion, the rediscovery of China’s long-marginalized anti-
mainstream Wang Yangming 王阳明 school of Ming dynasty Confucianism
inspired generations of revolutionaries in Meiji Japan, China, and East Asia in
general. The most striking or audacious was Kang Youwei’s so-called redis-
covery and restoration of “authentic” Confucian classics, which he claimed
had been distorted and contaminated during the Han dynasty. These “genu-
ine” – most likely fictitious – classics, Kang claimed, already contained the
seeds of ideas that would support and legitimize the Chinese reform agenda
in the 1890s.55
But by the 1920s, China’s intellectual mainstream had moved far beyond
this phase of atavistic nostalgia. Indeed, after the 1911 Qing collapse, the new
Republican era saw no equivalent of the Ming loyalists like Gu Yanwu and
Huang Zhongxi who had refused to serve the new Qing ruler despite their
unrelenting assault on the ills of Ming absolutism. Similarly, there were few
who retained the faith in Qing orthodoxy that leaders like Zeng Guofan and
Li Hongzhang once had just a few decades earlier. China’s Republican era
55
Yan Shaodang 严绍璗, 日本中国学史 (Sinology in Japan) (Beijing, Xueshu chubanshe,
2009), pp. 98–134.
42

ushered in a new generation that embraced Western learning and ideology

with few nostalgic backward glances. But China’s intellectual turn toward the
West soon developed in unexpected directions.
Political reality following the Qing’s collapse in 1911 revealed harsh and
dark sides marked by the rise of warlordism, factionalism, civil and military
strife, and fiscal bankruptcy, leading to a rising sense of national disintegra-
tion and international humiliation. As initial euphoria turned to despair, the
Beijing government largely abandoned parliamentary experiments in 1923
soon after Liang enthusiastically wrote about the extent and durability of
cultural change. Ravaged by World War I, Western imperialism and liberal
ideology went into retreat and the world saw a turn from “Woodrow Wilson
towards Vladimir Lenin.”56 Liang’s suggestion that Marx might rival the
standing of Confucius in the eyes of Chinese elites suddenly acquired omin-
ous weight as the rise of a domestic Communist movement and the
Guomindang’s engagement first with the Soviet Union and then with Nazi
Germany marked an increasing turn toward authoritarianism.
The founding of a national capital in Nanjing returned China’s political
center to the heart of its wealthiest region of the lower Yangzi, which had
long been distrusted by the Qing court and suspected of anti-Manchu senti-
ments. But by the 1920s, the lower Yangzi had developed new elites associ-
ated with the Shanghai-based treaty port system, symbolized by the powerful
Song family, which hailed originally from Guangdong but were educated in
Christian mission schools and American colleges. Another was the
Nationalist leader Chiang Kai-shek, who married one of the glamorous
Song sisters. Chiang, a Ningpo native with strong connections to
Shanghai’s financial elites and underworld leaders, graduated from
a military academy in Japan, where he developed a lifelong admiration for
Wang Yangming. Although Chiang promoted modern innovations, he was
raised on the traditional Confucian classics and became a faithful admirer of
Zeng Guofan, the Taiping nemesis and upright architect of the Self-
Strengthening movement. Under Chiang, the Nanjing regime was increas-
ingly authoritarian but charged with a full ten-year national modernization
agenda until it was interrupted by the 1937 Japanese invasion.
56
Luo Zhitian 罗志田, “‘六个月乐观’的幻灭：五四前夕士人心态与政治” (The
Disillusion of “Six Months of Optimism”: The Mood and Politics of Intellectuals on the
Eve of the May Fourth Movement), 历史研究 (Historical Research) 4 (2006), 105–91.
43

debin ma
The End of the Financial Revolution or Coming Full Circle?

The unexpected cultural turn may have turned China’s modernization path
to her own authoritarian or absolutist roots. The paradox of this combination
came into full view in the continuing phase of financial revolution in the
Nanjing era. In its earlier days, the newly established Nanjing-based
Nationalist government in 1927, especially under the new pro-business
finance minister Song Ziwen 宋子文 – the American-educated brother of
the powerful Song sisters – respected the institutional mechanism for public
debt set up under the previous Beiyang regime. It succeeded in taking over
the core institutional element of financial revolution through the establish-
ment of an independent Sinking Fund Commission headed by representa-
tives of the Shanghai banking community and government officials. They
constituted a powerful independent check on the government’s promise to
repay.
The domestic debt issued in 1927–1931 nearly doubled the total for the
entire Beiyang era from 1912 to 1926. The ratio of domestic to foreign public
debt, which had stood at one to seven during the Beiyang era, rose sharply to
six to four during the decade following the 1927 creation of the Nanjing
government. In this way, the nascent Nanjing government created favorable
conditions for diffusing the fruits of financial revolution in the form of new
monetary and financial institutions and instruments unimaginable within the
treaty port framework. An important consequence of the banks’ increased
holding of securitized government bonds was the rapid increase in the banks’
capacity to expand their issue of bank notes as China’s banking regulations
allowed modern banks to use securities – mostly government bonds – to
serve as 40 percent of the reserves needed to issue notes. This serves as the
critical anchor for China’s financial revolution as it expanded the balance
sheet of private banks and enhanced the rise of credible banknote issues, as
illustrated in Figure 1.5.57
However, once the Nanjing regime consolidated its power, its leaders
moved to re-establish a more authoritarian system that would rein in the
treaty port autonomy, particularly in Shanghai, China’s commercial, finan-
cial, and industrial capital. The nationalization of major Chinese banks and
the establishment of a fiat currency in 1935 and 1936 respectively – without,
however, the simultaneous establishment of an internal form of checks and
balances and along with the waning of Western imperialism and its
57
See Ma, “The Rise of a Financial Revolution”; also see the chapter by Li and Yan in this
volume.
44

associated fiscal–financial nexus – set the stage for the rise of hyperinflation in
the 1940s.58 The resurgence of a fiat-money-based hyperinflation in the 1940s
marked another coming full circle to the eras of China’s historical hyperin-
flation in the Song and Yuan dynasties whose issuance of paper money
exploded under wartime pressures.
Japan’s full-scale invasion in 1937 and China’s drive for resource mobiliza-
tion pushed toward the rise of a wartime command economy.59 The ascend-
ancy of Communist rule and ideology followed decades of gradual but
increasing radicalization of the modernization ideology. When the
Communists took power, they were armed with another ideology borrowed
from the Soviet Union: Marxism or Communism then established as the new
ruling orthodoxy grafted onto China’s authoritarian or totalitarian roots.60
Indeed, China’s new Communist leader, Mao Zedong, who spent his forma-
tive years in his revolution-enlightened native Hunan, became a modernizer,
but remained a great admirer of China’s historic founders of often brutal
absolutist dynasties, Qin Shihuang 秦始皇 and Zhu Yuanzhang 朱元璋.
Mao drew as much inspiration from Chinese classics as from Marxist theory.
Most ironically, just as Confucianism was extended or rediscovered to
legitimize westernization and modernization in the early twentieth century,
so Communism, the most radical ideological import from the West, was now
deployed to legitimize a newly invigorated and technologically enhanced
version of Chinese absolutism.
Conclusion
This chapter has built explicit links between intellectual and ideological cycles
and phases of economic change in China. By incorporating ideology into the
narrative of modern China’s economic history, it has emphasized the import-
ance of the Western impact within a framework that also assigns a central
role to the Chinese response to and adaptation of Western ideology and
practice. This new narrative adds missing elements to the Great Divergence
debate, which has considered initial conditions mainly in the form of institu-
tional structures and natural-resource endowments. By highlighting the
importance of intellectual resources, this approach exposes the limitations
of the Great Divergence thesis, which has tended to project late development
58
See Ma, “The Rise of a Financial Revolution”; also see the chapters by Li and Yan and by
Kaske and Lin in this volume.
59
See the chapter by Bian in this volume. 60 See the chapter by Xu in this volume.
45

debin ma
experience onto historical initial conditions or institutions and to ignore the

dependence or reliance of East Asian industrialization on borrowed institu-
tions and ideology. It is the greatest irony that both the Marxist discussion of
the sprouts of capitalism and the Great Divergence inquiry into why China
failed to generate an early industrial revolution are founded on the implicit
assumption of a unilinear or Eurocentric version of historical evolution or
inevitability (of industrial revolution).
The problem of ideas looms particularly large in China’s historical context of
absolutist institutions. In the absence or weakness of a domestic “voice,” the
“exit” option often remained the sole constraint to the centralized monopoly of
power in imperial China. In this regard, leakages in the form of physical capital
or intellectual resources from this monopoly rule carried particular weight and
significance. Utilization and absorption of these leakages proceeded on two
levels – in a bottom-up process whereby rational economic agents took advan-
tage of Western extraterritorial privileges and among upper-echelon political and
intellectual elites who eventually modified ideology to legitimatize moderniza-
tion policy at the national and political level. It was only when these two levels
connected or met that China saw the largest of transformations.
This process worked not only in the nineteenth and twentieth centuries,
but also in the post-1978 reform era. There, ideology in the top echelon and
experiments at the bottom often interacted with tension.61 Ultimately, it was
a rapid convergence of ideology to meet the global standards and rules after
China joined the WTO in 2001 that saw China’s most phenomenal trans-
formation. The regime at the center was more willing to compromise rather
than exercise arbitrary and unchecked power. However, three decades of
unprecedented growth led to a resurgence in confidence, in particular in
China’s own ideology and historical legacy. But if such resurging confidence,
as the eighteenth-century Qing once had, leads to a revival of Qing-style
closure to information flow and restriction of independent thinking, will all
the reforms and revolutions be another repeat of history? Let us hope not.
Further Reading
Brandt, L., D. Ma, and T. Rawski, “From Divergence to Convergence: Re-evaluating the
History behind China’s Economic Boom,” Journal of Economic Literature 52.1 (2014), 45–123.
Cohen, P., Discovering History in China: American Historical Writing on the Recent Chinese Past
(New York, Columbia University Press, 1984).
61
See the chapters by Naughton and by Brandt and Rawski in this volume for the reform
period.
46

Fukao, K., D. Ma, and T. Yuan, “Real GDP in Pre-war East Asia: A 1934–36 Benchmark
Purchasing Power Parity Comparison with the U.S.,” Review of Income and Wealth 53.3
(2007), 503–37.
Ge Zhaoguang 葛兆光中国思想史第二卷七世纪至十九世纪中国的知识，思想与信
仰 (History of Chinese Thought, Vol. 3, Chinese Knowledge, Thought, and Beliefs
between the Seventh and Nineteenth Centuries) (Shanghai, Fudan daxue chubanshe, 2001).
Ge Zhaoguang “葛兆光十八世纪中国的盛世危机” (Crisis in China’s Glorious
Eighteenth Century), Feb. 19, 2019, www.ftchinese.com/author/%E8%91%9B%E5%
85%86%E5%85%89.
Jin Guantao 金观涛, and Liu Qingfeng 刘青峰, 兴盛与危机，论中国社会的超稳定结
构 (The Cycle of Growth and Decline: On the Ultrastable Structure of Chinese Society)
(Beijing, Falu chubanshe, 2011).
Jin Guantao 金观涛 and Liu Qingfeng 刘青峰, 开放中的变迁，再论中国社会超稳定
结构 (The Transformation of Chinese Society (1840–1956): The Fate of Its Ultrastable
Structure in Modern Times) (Beijing, Falu chubanshe, 2011).
Liang Qichao 梁启超, “梁启超五十年中国进化概论” (A Summary of Fifty Years of
Chinese Evolution), in 饮冰室文集点校第五集 (Selected Articles of the Yinbinshe
Collection, Vol. 5) (Kunming, Yunnan Jiaoyu chubanshe, 1923), pp. 3247–52.
Ma, D., “Economic Growth in the Lower Yangzi Region of China in 1911–1937: A Quantitative
and Historical Perspective,” Journal of Economic History, 68.2 (2008), 385–92.
Ma, D., “The Rise of a Financial Revolution in Republican China in 1900–1937: A Survey
and New Interpretation,” Australian Economic History Review 59.3 (2019), 242–62.
Ma, D., and J. Rubin, “The Paradox of Power: Principal–Agent Problems and
Administrative Capacity in Imperial China (and Other Absolutist Regimes),” Journal of
Comparative Economics 47.2 (2019), 277–94.
Maddison, A., Chinese Economic Performance in the Long Run, 2nd ed., rev. and updated
(Paris, Development Centre of the Organisation for Economic Co-operation and
Development, 2007).
Mokyr, J., The Gifts of Athena: Historical Origins of the Knowledge Economy (Princeton, NJ:
Princeton University Press, 2002).
Motono, E., Conflict and Cooperation in Sino-British Business, 1860–1911: The Impact of the Pro-
British Commercial Network in Shanghai (New York, St. Martin’s Press, 2000).
Rawski, T.G., Economic Growth in Prewar China (Berkeley, University of California Press, 1989).
Xu, Dixin 许涤新 and Wu Chengming 吴承明,中国资本主义发展史，第一卷中国
资本主义的萌芽 (A History of Capitalist Development in China, Vol. 1, Sprouts of
Chinese Capitalism) (Beijing, Renmin chubanshe, 1993).
Wang Fansen 王汎森, “戊戌前后思想资源的变化：以日本因素为例” (Changes in
Intellectual Resources before and after the Hundred Days Reform), 二十一世纪 (Twenty-
First Century) 45 (February 1998), 47–54.
Wang Fansen 王汎森, 权力的毛细血管作用，清代的思想、学术与心态 (The
Penetrating Role of Power, Ideas, Academics, and Moods in Qing) (Beijing, Beijing
daxue chubanshe, 2015).
Yan Shaodang 严绍璗, 日本中国学史 (Sinology in Japan) (Beijing, Xueshu chubanshe,
2009).
47

2
Economic Transition in the Nineteenth
Century
william t. rowe
The Great Qing Empire (1644–1912) was the most populous political entity
that had yet existed on the landmass that we now refer to as “China,” and its
economy was possibly one of the most developed. But by the first several
decades of the nineteenth century, the Jiaqing and early Daoguang reigns,
there had emerged a general consensus among elites both in and out of
government that the empire was facing a multifaceted and potentially cata-
strophic crisis of the economy, polity, and society. By this point the Qing had
already begun to be significantly incorporated into the early modern world
economy, although it had not yet experienced, as it very shortly would,
military conflict with the West and the invasion by Western agents of
economic and cultural change that would follow in its wake.
Articulating a Chinese establishment view of Qing economic history, the
senior historian Dai Yi invoked the image of a “camel’s hump” (tuofeng 驼峰):
recovery from devastation in the late sixteenth century and the seventeenth,
a period of great prosperity during the Yongzheng and Qianlong reigns
(1723–1795), and finally a period of collapse into poverty and semicolonialism.
Dai stressed the Opium War (1837–1842) as a benchmark in this downward
spiral, leaving more or less unanswered the question of what happened
during the early decades of the nineteenth century.1
Persons in the mid-Qing would not have found Dai’s analogy entirely
improbable. In a poem of 1758, the Qianlong Emperor first articulated the
formula chiying baotai 持盈保泰, or “hold onto the surplus and preserve
the abundance.” This would become a theme in his pronouncements over
the several decades to follow. He invoked this notion, not coincidentally, in
I wish to thank Debin Ma, Thomas Rawski, and Joshua Rowe for helpful comments on an
earlier draft of this chapter.
1
Dai Yi 戴逸, “在清代經濟宏觀趨勢與總體評價學術研討會上的發言” (Speech to
the Conference on Macro-economic Trends and Overall Appraisal of the Qing Dynasty),
清史研究 (Studies in Qing History) 2008.3, 14.
48

Economic Transition in the Nineteenth Century
the process of the conquest and incorporation of Xinjiang, probably his

greatest military adventure, and it reflected the emperor’s pride that he had
expanded imperial territory to a greater degree than had any of his pre-Qing
or Qing predecessors. At the same time, he had overseen a seemingly
unprecedentedly flourishing economy and the fiscal security of the imperial
state.
His formula, however, also expressed a new and gnawing anxiety that this
pinnacle of achievement would face constant pressures of erosion. This
mounting imperial fear, it has been argued, lay behind the increasingly
authoritarian and repressive policies of the later Qianlong reign, up to and
including the regime of the arriviste strongman Heshen. Nor did this mount-
ing fear end with the passing of Qianlong and his henchman in 1799. A parallel
trope of late Qianlong rhetoric, shoucheng 守成 (preserve the achievements of
Qing imperial forebears), was inherited by the succeeding Jiaqing Emperor,
and routinely invoked, first specifically to dismantle Heshen’s apparatus of
personal corruption, and thereafter to legitimate a range of policies to combat
decay of the fragile “complete prosperity” (quansheng 全盛) of the empire at
its mid-eighteenth-century peak.2 This anxiety about decline – political,
economic, civilizational – was by no means confined to the court. As Susan
Mann Jones and Phillip A. Kuhn wrote, in the classic English-language study
of this era, “The widespread feeling that 1775–80 was a turning point in [Qing]
history – a turning downward – pervades the political and social commentar-
ies written by officials and scholars of the early nineteenth century.”3
Gao Wangling has argued convincingly that the prosperous age of the
eighteenth century was in no small measure the result of insightful govern-
ment policies favorable to economic development.4 By contrast, such matters
as the Jiaqing Emperor’s seemingly weak-kneed failure to thoroughly purge
officialdom of the corrupt underlings installed by Heshen, and the Daoguang
Emperor’s arbitrary and unsteady pursuit of the Opium War, have contrib-
uted to the conventional view of the early nineteenth century as an era of
2
Gao Xiang 高翔, “從持盈保泰到高壓統治：論乾隆中期政治轉變” (From
“Preserve the Abundance” to Repressive Governance: Political Transformation in the
Mid-Qianlong Reign), 清史研究 (Studies in Qing History) 1991.3, 8–13; Guan Wenfa 關
文發, 嘉慶帝 (The Jiaqing Emperor) (Changqun, Jilin wenshi chubanshe, 1993), pp.
59–114.
3
S.M. Jones and P.A. Kuhn, “Dynastic Decline and the Roots of Rebellion,” in J.
K. Fairbank (ed.), The Cambridge History of China, vol. 10, Late Ch’ing, 1800–1911, part 1
(New York, Cambridge University Press, 1978), p. 161.
4
Gao Wangling 高王凌, 十八世紀中國的經濟發展和政府政策 (Government Policy
and Eighteenth-Century Chinese Economic Development) (Beijing, Zhongguo shehui
kexue chubanshe, 1995).
49

william t. rowe
weak governance. Recent scholarship, however, emphasizing such factors as

campaigns for court and official frugality, rehabilitation of upright ministers
cashiered under Heshen, regularization of the practices of such institutions as
the Grand Council and the Imperial Household Department, a somewhat
relaxed tolerance for literati dissent, and an advertised willingness to listen
and respond to popular grievances, has contributed to a more favorable
assessment of the Jiaqing court, at least, and a new credence assigned to the
genuineness of a purported “Jiaqing restoration” (xianyu weixin 咸與維新).5
Ecological Decay
A substantial body of research has concurred in the view that the Qing
empire – including most regions of its Chinese-dominated core (neidi
内地) – suffered acute ecological deterioration in the half-century straddling
the assumption of power by the Jiaqing Emperor. Floods and droughts
increased in frequency and severity, harvests were less abundant, rural
refugees were more ubiquitous, and the need for relief distribution became
more urgent. Short-term climate change may have been one factor. Studying
Lingnan (southeast China), for example, Robert Marks has found that the
annual average temperature fell by one degree centigrade over the course of
the early nineteenth century: declining in the 1810s, stabilizing in the 1820s,
falling again over the 1830s, and “bottoming out” in the 1840s. Frosts were
first reported in late 1808, with a virtually unprecedented two to three inches
of snowfall in 1809. This sudden wave of cooling “shocked harvest yields,”
leading to continuing regional dearth over the decades to follow.6
But although natural causes played some part in ecological decay, the
major factor was clearly human activity, both negligent and aggressive. In
1801, the Hai river basin of Zhili province experienced one of its greatest
floods in recorded history, affecting 122 counties and creating hundreds of
thousands of refugees. Between 1801 and 1802, the price of staple grain in the
area nearly tripled, and many farmers were forced to sell off their land at
greatly depressed prices. In 1813–1814, the same region experienced a major
5
See, for example, Zhang Yufen 張玉芬, “論嘉慶初年的咸與維新” (On the
“Comprehensive Reform” of the Early Jiaqing Reign), 清史研究 (Studies in Qing
History) 1992.4, 49–54; and D. McMahon, “Dynastic Decline, Heshen, and the
Ideology of the Xianyu Reforms,” Tsing Hua Journal of Chinese Studies, new series 38.2
(June 2008), 231–55; W. Wang, White Lotus Rebels and South China Pirates: Crisis and Reform
in the Qing Empire (Cambridge, MA, Harvard University Press, 2014).
6
R.B. Marks, Tigers, Rice, Silk, and Silt: Environment and Economy in Late Imperial South
China (Cambridge, Cambridge University Press, 1998), pp. 217–18.
50

drought-induced famine, with rainfall reported at only slightly more than half
of that in normal years; again, grain prices more than doubled, and thousands
of farmers abandoned their land to become refugees. Another major flood
came in 1822–1823. Thereafter, throughout the Daoguang reign, flood disas-
ters in the Hai river basin became a virtually annual event.
The principal reason for the flooding was the progressive silting of the Hai
and its tributaries. As Lillian Li concludes, although silting of the Hai had
become an increasing problem since the Liao dynasty (947–1125), disastrous
consequences had really only become evident in the mid-Qing. Li identifies
a “reign cycle” of government attention to dredging the river system, with lax
maintenance in the latter years of each Qing ruler contributing to cata-
strophic flooding and hence heightened responsiveness at the outset of the
succeeding reign. In terms of both maintenance and relief delivery, however,
she also notes greater diligence and effectiveness in the Jiaqing versus the
Daoguang reign, leading her to identify the Daoguang era as the “turning
point” in the Hai system’s ecological decline.7
But if the problem in the Hai river basin was primarily that of state neglect,
an equally great – perhaps yet greater – problem was uncontrolled (mostly
private) reclamation of agricultural land. As the Qing court became increas-
ingly cognizant of its growing population and the need to assure a steady food
supply, in the late Kangxi reign it began to encourage massive development
of new arable. This went well beyond the resettlement of Sichuan and other
regions that had been devastated during the dynastic transition, to include the
“opening” (kaiken 开垦) of highlands, marshes, riverbanks, lakeshores, sea-
coasts, and virtually every scrap of historically unfarmed land both in the
interior and on the frontiers. Sometimes government “encouragement of
agriculture” (quannong 劝农) got a bit out of hand, as in the Yongzhang reign,
when false reporting of reclaimed land and the forced introduction of
agriculture to unsustainable farmland demanded retrenchment in subse-
quent years. But overall the process of land development continued through
the end of the eighteenth century and into the nineteenth.8
The region in which the unintended effects of excess reclamation were
revealed most clearly, and with greatest human and economic consequence,
7
L.M. Li, Fighting Famine in North China: State, Market, and Environmental Decline, 1690s–
1990s (Stanford: Stanford University Press, 2007), pp. 31, 250–66.
8
The classic work is Peng Yuxin 彭雨新, 清代土地開墾史 (A History of Land
Development in the Qing Period) (Beijing, Nongye chubanshe, 1990). On the
Yongzheng campaigns, see W.T. Rowe, Saving the World: Chen Hongmou and Elite
Consciousness in Eighteenth-Century China (Stanford: Stanford University Press, 2001),
pp. 56–68.
51

william t. rowe
was the middle Yangzi, especially that sector of Hubei province’s central
plain lying like a slice of pie between the valleys of the Yangzi and Han rivers
northwest of its apex at the river confluence of Wuhan. This region is laced
with small tributaries and constitutes, in Pierre-Étienne Will’s apt term, a vast
“interior delta.”9 Extending this region a little more broadly, we might
include also the area of Hunan’s Dongting lake and Xiang river valley to
the south, and the Han river highlands to the northwest, all of which
participate in a coherent and interdependent ecological system.
This entire, highly productive, region existed throughout the early and
mid-Qing in what Will describes as a “fragile equilibrium” of land and water.
Spring floodwaters descended on the region from the mountains of the west
and northwest, eventually to be led to the sea via the lower Yangzi river. The
relatively lesser capacity of the lower Yangzi outlet, however, dictated that
these waters be held for a period of time in the great floodwater receptacles of
the middle Yangzi lakes, rivers, and floodplains. Initially this abundance of
water was a great asset to the region, allowing the absorbtion of rapidly
growing population into subsistence agriculture, and also entrepreneurial
investment in polder lands for commercial rice production for downriver
export to the markets of the lower Yangzi throughout the “prosperous age”
of the eighteenth century. Already by that time, however, perceptive officials
such as Hunan governors Yang Xifu in the 1740s and Chen Hongmou in the
1750s had begun to be concerned about the problem of constricting the
region’s natural capacity to contain seasonal floodwaters, and the threat
this posed to the entire region’s hydraulic security. They consequently
began to promulgate restrictions on unchecked lakeshore and riverbank
encroachment, but the pressures of both population growth and market
opportunity rendered these proscriptions largely ineffective.10
Compounding this tension was the ever more intensive settlement of
highlands, most dramatically along the Han river in northwest Hubei and
adjacent areas of Sichuan and Shaanxi. Large swaths of mountainside that had
effectively been virgin forest under the Ming had been legally opened to
farming, initially during the sixteenth century and then with ever greater
intensity during the eighteenth.11 This process was catalyzed in part by the
9
P.-É. Will, “Un cycle hydraulique en Chine: La province du Hubei du XVIe au XIXe
siècles,” Bulletin de l’École Française d’Extreme Orient 68 (1980), 261–87. See especially the
map on p. 287.
10
P.C. Perdue, Exhausting the Earth: State and Peasant in Hunan, 1500–1850 (Cambridge, MA,
Harvard University Press, 1987), pp. 192–3, 221, 228.
11
E.S. Rawski, “Agricultural Development in the Han River Highlands,” Ch’ing-shih wen-
t’i 3.4 (1975), 63–4.
52

dissemination to the area of highland-adaptable New World crops, especially

maize and sweet potatoes, but indigenous crops, including staples such as
beans and commercial products such as tung, also played a role in highland
reclamation. Perhaps the majority of this extensive agriculture up through
the eighteenth century was slash-and-burn shifting cultivation, practiced by
a group that came to be known throughout central and south China as
pengmin 棚民 (“shed people”), an emerging legal category of households
with an incipient ethnic identity. In the Han valley, as in highland areas of
Jiangxi, Jiangnan, Lingnan, and elsewhere, shed people carved out their own
ecological and occupational niche. They coexisted tenuously with econom-
ically favored residents of the plains and river valleys, providing the latter
with charcoal and other specialty products but competing often violently
over land, water, educational opportunities, and other scarce resources.12
Beginning in the third quarter of the eighteenth century, Qing officials
launched periodic drives to “register” this population, which meant effect-
ively curbing their transience and settling them on the land.13 This policy was
abetted considerably by terracing of the highlands for wet-rice cultivation in
the decades before and after 1800.14 By this time localities in the Han river
valley had also developed a wide range of specialized exports, including
pears, walnuts, chestnuts, melons, bark paper, fungus, gypsum, saltpeter,
and badger and fox furs.15
Both shifting and sedentary agriculture in the highlands contributed to the
region’s ecological deterioration. In the highlands themselves, soil fertility
declined in successive harvests after initial clearance, even as the areas were
asked to absorb continuing waves of immigration and were characterized by
the same deeply exploitative forms of land tenure (multiple lords to a field,
indentured permanent tenancy, heavy up-front rent deposits) that character-
ized most areas newly developed for agriculture during the Qing’s “prosper-
ous age.” According to the classic study by Suzuki Chusei, these conditions,
12
S. Averill, “The Shed People and the Opening of the Yangzi Highlands,” Modern China
9.1 (January 1983), 84–126; A. Osborne, “The Local Politics of Land Reclamation in the
Lower Yangzi Highlands,” Late Imperial China 15.1 (June 1994), 1–46; S.-T. Leong,
Migration and Ethnicity in Chinese History: Hakkas, Pengmin, and Their Neighbors
(Stanford, Stanford University Press, 1997).
13
Liu Min 劉敏, “論清代棚民的戶籍問題” (The Problem of Registering the Shed
People in the Qing), 中國社會經濟史研究 (Studies in Chinese Socio-economic
History) 1983.1, 17–29.
14
Rawski, “Agricultural Development in the Han River Highlands,” 65–6.
15
T.-J. Liu, Trade on the Han River and Its Impact on Economic Development, c. 1800–1911
(Nankang, Institute of Economics, 1980), pp. 7–10.
53

william t. rowe
compounded by fiscal oppression, were major contributors to the outbreak

of the White Lotus rebellion in this area in the 1790s.16
It was not only highlands and marshes that were enclosed for rice produc-
tion. Large stretches of the Jianghan plain had originally been cordoned off
for horse pasture by the Qing conquerers, under the control of the banner
garrison at Jingzhou. Between the late seventeenth and the late nineteenth
centuries, most of this land had been progressively ceded to Han farmers for
the construction of polders, with the horse population declining from over
13,000 to a mere 2,000. The encroachment onto this potential floodplain was
a major contributor to the increasingly serious flooding of the Jianghan
“interior delta” after the late eighteenth century.17
In the broader middle Yangzi region, declining water retention in
deforested highlands and accelerating river and lake silting from hillside
topsoil runoff simply aggravated critical problems of hydraulic manage-
ment. The precarious condition of the entire region was suggested as early
as 1763, when Hunan governor Chen Hongmou ordered the demolition of
the illegal Yanglinzhai dike in Xiangyin county, which resulted in the
unanticipated collapse of four other major dikes protecting Yanglinzhai,
and the dislocation of scores of households.18 But it was the 1788 bursting
of the Wancheng dike, on the Yangzi in western Hubei, that incontestably
revealed to all the calamitous state of the interior delta, leaving the major
prefectural seat of Jingzhou under three meters of water and the down-
stream Wuhan conurbation itself disastrously flooded. Less devastating
floods followed on a nearly annual basis. When the noted hydraulic expert
Wang Zhiyi was posted to the Huguang governor-generalship in 1807, he
conducted a thorough analysis of the region, deducing that the entire area
bounded by the Yangzi and the Han had been reduced to a basin, more
often inundated than arable. He undertook vast and costly repair works,
which had a positive impact on the hydraulic security of the region for the
remainder of the Jiaqing reign. But the entire Yangzi–Han–Xiang area
began again to suffer major flooding with the accession of the Daoguang
Emperor in 1820. Particularly devastating instances of these nearly annual
16
Suzuki Chusei 鈴木中正, 清朝中時研究 (A Study of the Mid-Qing Period)
(Toyohashi, Aichi University Research Institute on International Problems, 1952).
17
Y. Gao, “The Retreat of the Horses: The Manchus, Land Reclamation, and Local
Ecology in the Jianghan Plain (ca. 1700s–1850s),” in Liu (ed.), Environmental History in
East Asia: Interdisciplinary Perspectives (London, Routledge, 2014), pp. 100–25.
18
Perdue, Exhausting the Earth, p. 228.
54

events occurred in 1831 and in the closing years of Daoguang’s troubled

reign in 1848–1849.19
The most alarming and symbolic evidence of ecological decay in the mid-
Qing involved the progressive deterioration and eventual abandonment by
the administration of the Grand Canal, an ancient semi-natural, semi-
manmade waterway dating back in part to the reign of the first emperor,
Qin Shihuang.20 In the Ming and Qing the principal governmental use of the
Canal was for the annual northern shipment of central China rice, the so-
called “Grain Tribute,” relied upon (in the Qing) to feed the large populations
of bannermen in the capital and along the northern frontiers. The viability of
these shipments required scrupulous dredging not only of the canal bed itself,
but also of the Yellow River and other north China waterways that flowed
into the canal. When this was not faithfully done, as was the case in the
late eighteenth and early nineteenth centuries, the canal silted up and
northbound rice shipments were endangered and reduced. This problem
had been increasingly evident during the Jiaqing reign, prompting reform
proposals of various kinds, including the replacement of the grain
tribute system as a whole by a new structure of agricultural colonies (tuntian)
that would ideally serve to make bannerman populations self-sufficient in
grain.21
Anxiety suddenly intensified into genuine panic in 1803, when the rupture
of the Hengjialou dike in Henan spilled large quantities of silt-laden water
into the canal. Northbound grain tribute boats were halted at Zhangqiu,
Shandong, by the silt buildup, and many failed altogether to make it through.
The Jiaqing Emperor opened the “pathways of words” (yanlu), calling upon
officials to submit suggestions on how to rehabilitate the canal and the tribute
19
Will, “Un cycle hydraulique en Chine,” pp. 282–5. See also T.-J. Liu, “Dike Construction
in Ching-chou,” Papers on China 23 (1970), 1–28; and W.T. Rowe, Hankow: Conflict and
Community in a Chinese City, 1796–1895 (Stanford, Stanford University Press, 1989), pp.
147–8.
20
Hoshi Ayao 星斌夫, 大運河：中國の漕運 (The Grand Canal: Grain Tribute in China)
(Tokyo: Kondō shuppansha, 1971); H.C. Hinton, The Grain Tribute System of China, 1845–1911
(Cambridge, MA, Harvard East Asia Series, 1956); J.K. Leonard, Controlling from Afar: The
Daoguang Emperor’s Management of the Grand Canal Crisis, 1824–1826 (Ann Arbor, Center for
Chinese Studies, 1996); Jones and Kuhn, “Dynastic Decline and the Roots of Rebellion,”
119–28; Zhang Yan 張岩, “包世臣與近代前夜的 ‘海運南漕’ 改革” (Bao Shichen and
Seaborne Grain Tribute Reform), 近代史研究 (Studies in Modern History) 2000.1, 129–53;
Ni Yuping 倪玉平, “道光六年糙糧海運的幾個問題” (Some Problems Regarding the
Sea Transport of Grain Tribute in 1825), 清史研究 (Studies in Qing History) 2002.3, 70–5.
21
C. He and Y. Wei, eds., 皇朝經世文編 (Collected Writings on Statecraft from the
Present Dynasty), juan 47, 48. See also W.T. Rowe, Speaking of Profit: Bao Shichen and
Reform in Nineteenth-Century China (Cambridge, MA, Harvard University Asia Center,
2018), Chapter 4.
55

william t. rowe
system. Among the responses was the first call for abandoning the Grand
Canal altogether, in favor of coastal shipping of tribute grain from Shanghai
to the north. No major action was taken, however, and throughout the
remainder of Jiaqing’s reign grain tribute problems visibly worsened.
The succeeding Daoguang Emperor, a committed ditherer on most issues,
was forced into decisive action when the Gaojiayan dike at Qingjiang in northern
Jiangsu ruptured in 1824, allowing only very few grain tribute boats to get
through to the capital. The emperor convened an emergency conference of
high officials, but their deliberations were inconclusive and their suggestions
timid. One man, however, Board of Revenue president Yinghe (1771–1839),
presented a bold plan to shift the following year’s shipment from the canal to
the coastal route, adding that the shipment should be conducted not by govern-
ment personnel but instead by commercial contractors. With the posting to
Jiangsu of a new corps of crisis managers, including provincial governor Tao
Zhu, provincial treasurer He Changling, and both men’s activist adviser Wei
Yuan, Yinghe’s plan was in fact implemented for the 1825 shipment.
Inasmuch as that year’s tribute grain shipment was successfully completed,
by sea, the project might be considered a success, and Wei Yuan in particular
lauded it as a revolutionary and long-overdue innovation.22 But the success
was a qualified one. There were significant unanticipated cost overruns.23
There were incidences of violent unrest by laid-off canal workers that
continued to escalate in intensity and notoriety over the following quarter-
century.24 And although Yinghe had specified that the sea route was to be
a temporary measure only, with the cost savings applied largely to the repair
of the Grand Canal, in fact those repairs were never effectively carried out,
and use of the canal for tribute shipments was gradually abandoned
altogether.25 Indeed, grain tribute receipts by the capital region continued
to steadily decline over the remainder of the Daoguang reign until, with the
interruption of shipments in kind by the Taiping wars, most of the imposi-
tions were commuted to cash in 1857–1858.26
22
He and Wei, 皇朝經世文編, juan 48:70–1.
23
Ni Yuping 倪玉平, 清朝嘉道財政與社會 (Government Finance and Society in the
Qing Jiaqing and Daoguang Reigns) (Beijing, Commercial Press, 2013), pp. 244–76.
24
Hoshi, 大運河, pp. 233–6.
25
Hinton, The Grain Tribute System, p. 28; Jones and Kuhn, “Dynastic Decline and the
Roots of Rebellion,” 122.
26
Xia Nai 夏鼐, “太平天國前後長江各省之田賦問題” (Problems of Land Tax Collection
in the Yangzi Provinces before and after the Taiping Rebellion), 清華學報 (Tsinghua
University Studies) 10.2 (April 1935), 419–25; W.T. Rowe, “Hu Lin-i’s Reform of the Grain
Tribute System in Hupeh, 1855–58,” Ch’ing-shih wen-t’i 4.10 (December 1983), 33–86.
56

Mid-Qing impact on the natural environment worked in other ways as

well. Meng Zhang has recently shown that demand for timber in the building
construction boom of eighteenth-century Jiangnan spawned a long-distance
trade in timber from a progressively widening range of upriver provinces,
including Jiangxi, Hubei, Hunan, Sichuan, and Guizhou. She argues, how-
ever, that wholesale deforestation of these regions was ameliorated by self-
imposed controls on harvesting and efforts at replanting on the part of
commercial interests concerned about the lucrative trade’s long-term
sustainabililty.27
Similar restraint was not apparent in other newly booming trades. As the
prosperous age progressed, heightened demand for luxury goods spurred
“resource rushes” in both frontier and overseas production areas, along with
the development of new trade networks involving merchants and carriers of
all backgrounds, including Europeans and Americans. Sable pelts from
Siberia (via Kiakhta); sea otter furs from Hokkaido, Siberia, and North
America (via Canton); steppe mushrooms from Mongolia; pearls from Jilin;
jade from Xinjiang and Burma; ginseng from Manchuria; mother-of-pearl
from Sulu; sandalwood from Timor; birds’ nests from Borneo; sea cucumbers
from Southeast Asia, Australia, and Fiji; and many other goods flowed into
the empire. Demand was spurred initially by the court, but progressively
spread to wider circles of wealthy consumers. By the early 1800s, most of
these trades had entered a phase of rapid decline: Jilin, for example, produced
2,890 pearls in 1795 but only 895 in 1815, while 17,446 sea otter pelts entered
Canton in 1806, as compared to a mere 329 in 1831. Various factors played
a role in these declines, but the most ubiquitous feature was the exhaus-
tion of resource supply due to overexploitation. In other words, the mid-
Qing’s very prosperity contributed directly to environmental decay and
commercial contraction as the eighteenth century gave way to the
nineteenth.28
Demographic Crisis?
Contributing substantially to ecological deterioration and in turn aggravated
by it was the mid-Qing’s growing population pressure. All parties at the time,
and virtually all scholars since, have recognized the immense increase in the
27
Meng Zhang, Timber and Forestry in Qing China: Sustaining the Market (Seattle:
University of Washington Press, 2021).
28
J. Schlesinger, A World Trimmed with Fur: Wild Things, Pristine Places, and the Natural
Fringes of Qing Rule (Stanford: Stanford University Press, 2017), pp. 50–4, 73, 133.
57

william t. rowe
empire’s population, and have attributed this in part to prolonged experience

of peace and stable governance. But had this growth reached a stage of
profound crisis by the turn of the nineteenth century and the accession of
the Jiaqing Emperor? Put another way, had the Qing empire, with its
expanding borders but deteriorating ecology, reached or perhaps begun to
exceed its maximum carrying capacity for human beings by this time? This
question has provoked spirited disagreement, down to the present day.
If not the first to sound the alarm, at least the most frequently cited, was
the Changzhou scholar Hong Liangji (1746–1809). In an essay of 1793 – five
years before Thomas Malthus, at the other end of the Eurasian continent,
published his yet more famous Essay on the Principle of Population – Hong
argued that “today’s population is five times as large as that of thirty years
ago, ten times as large as that of sixty years ago, and not less than twenty
times as large as that of one hundred years ago.” By contrast, during that
same span the amount of available farmland “has only doubled, or, at the
most, increased three to five times.” What could be done about this?
Government could encourage agricultural improvement and the reclamation
of new arable, lower agrarian taxes, maintain granary stocks, and
distribute famine relief; Hong acknowledged that the Qing had done most
of these things relatively well, but this was not enough. Under these condi-
tions, Heaven will necessarily intervene to reduce the population via
“flood, drought, and plague,” but that too will likely be insufficient to
resolve the problem of food supply. This was a catastrophically bleak por-
trayal of the unforeseen negative consequences of generations of apparent
prosperity.29
Among modern scholars, Hong found his greatest support in the work of
the demographic historian Ping-ti Ho. Though hardly accepting Hong’s wild
speculations about the pace of Qing population growth, Ho’s more careful
research did show the population more than doubling (from
around 150 million to more than 300 million) over the course of the eight-
eenth century. And, while not necessarily accepting Hong’s dire predictions
of demographic crisis, Ho did conclude that in the Qing “the optimum
condition (the point at which ‘a population produces maximum economic
welfare’) at the technological level of the time, was reached between 1750 and
1775.” Barring the introduction of new technology, and especially given
continuing rapid population growth, as it entered the nineteenth century
29
L. Hong, “Yiyan,” in Hong, Juanshi geji (trans. K.-C. Liu), in W.T. de Bary and
R. Lufrano (eds.), Sources of Chinese Tradition, vol. 2, 2nd ed. (New York, Columbia
58

the Qing’s relationship between population and land was doomed to become
increasingly adverse.30 Relying largely on Ho’s population figures but adding
his own calculations of cultivated acreage (expanding nearly threefold between
1400 and 1770) and grain yields per acre (increasing by nearly 50 percent over
this same period), the agricultural economist Dwight Perkins helped explain
how the Qing had avoided a demographic crisis during the “prosperous age,”
but left unchallenged Ho’s conclusion of a dramatically worsening man–land
ratio as extensive growth approached its limits in the nineteenth century.31
Ho wrote in the late 1950s and Perkins in the late 1960s, and since that time
there has been a largely positive reappraisal of the performance of the
empire’s early modern economy. If the locus classicus for those who see
a crisis of provisioning emerging around 1800 is Hong Liangji, revisionists
might find their inspiration in Hong Liangji’s younger contemporary Bao
Shichen (1775–1855). Bao, to my knowledge, never quoted Hong directly,
though they traveled in the same scholarly circles in Changzhou and shared
some of the same patronage networks, but he seems clearly to be referring to
him when he derides “petty Confucians” (xiaoru 小儒) for making dire
predictions without the benefit of either detailed field observation or math-
ematical precision, both of which he claims to provide. Bao emphatically
denied that continuing population growth was a liability, arguing that there
was more than ample space in the existing agrarian regime for further
extensive and (especially) intensive growth, when directed by scholars, offi-
cials, and landowners who knew what they were doing. More people simply
meant more workers, and more productive workers at that.32
Bao Shichen’s most influential modern heir is unquestionably the prolific
economic historian Li Bozhong. Focusing narrowly on the empire’s most
developed economic region, Jiangnan, Li argues that early and mid-Qing
population growth is somewhat overstated. By his calculations, between 1680
and 1850 Jiangnan’s population rose from around 20 million to something
over 36 million. This seemingly impressive growth, however, was merely half
of the growth rate registered by the empire as a whole, and in fact some
20 percent slower than the growth rate of Jiangnan itself over the centuries of
Ming stability. This relatively modest growth continued into the early
nineteenth century: whereas the empire’s total population grew 39 percent
over the years from 1789 to 1838, that of Jiangsu rose only 32 percent and of
30
P.-T. Ho, Studies in the Population of China, 1368–1953 (Cambridge, MA, Harvard
University Press, 1959), pp. 270, 277–8.
31
D. Perkins, Agricultural Development in China, 1368–1968 (Chicago, Aldine, 1969), pp. 16–17.
32
Rowe, Speaking of Profit, Chapter 3.
59

william t. rowe
Zhejiang only 31 percent, and in the core Jiangnan prefecture of Jiaxing

growth was only 21 percent. Li attributes this restrained growth to self-
conscious “modern birth control techniques,” including infanticide, late
marriage, abortion, and abstinence.33
Li concedes that by the late eighteenth century the introduction of new
arable had been pretty much exhausted in Jiangnan, but, guided in part by the
writings of Bao, he argues that greater productivity was being achieved as late
as the Taiping wars of the 1850s. The adoption of double-cropping rice and
winter dry-field crops increased productivity of the land, while the develop-
ment of off-farm employment, notably in the textile trades and especially of
female workers, enhanced labor productivity. More controversially, Li
argues that the shrinking size of family farms to an “optimal” size actually
aided production by encouraging more efficient intensified labor use. He
concludes, “There was a shortage, not a surplus, of labor even in mid-Qing
Jiangnan.”34
Ho Ping-ti’s former student James Lee has developed a model of Qing
population behavior that incorporates elements both of his mentor’s argu-
ments and of those of Li Bozhong. Contra Malthus, who depicted China as an
egregious example of blind human reproduction, Lee stresses the continuing
availability in the Chinese demographic repertoire of both “preventive” and
“positive” checks on population growth, including infanticide; these checks
were applied by households with varying rigor, depending on their percep-
tion of food supply and economic opportunity. In the early and mid-Qing
“prosperous age” these checks were relaxed, and population grew relatively
rapidly. Agreeing with Li Bozhong that Jiangnan probably experienced less
dramatic growth than other areas, Lee extends this to argue that the form of
“modern population growth” in the Qing empire was nearly the opposite of
that at the other end of the Eurasian continent: whereas in Europe this
growth was evidenced most in already densely populated areas, in China
the major loci of growth were peripheral areas of pioneering settlement, such
as the southwest and internal highlands. In the early nineteenth century these
areas “filled up,” resulting in decreased immigration, increased application of
preventive and positive checks, and a slowing of empire-wide population
growth. (In one local population he has studied in detail, Lee finds male
33
Li Bozhong 李伯重, “控制增長以保福容福褣: 清代前中期江南的人口行為”
(Restricting Growth in Order to Preserve Wealth: Population Behavior in Early and
Mid-Qing Jiangnan), 新史學 (New Historiography) 5.3, 1994.9, 25–71.
34
B. Li, Agricultural Development in Jiangnan, 1620–1850 (New York, St. Martin’s Press, 1998),
p. 147.
60

population growth of 1.2 percent per year for 1774 to 1804, slowing to
0.2 percent per year from 1805 to 1873.) This suggests significant demographic
pressure, but hardly the Malthusian crisis predicted by Hong Liangji.35
Other scholarship has argued that the Qing dealt relatively successfully
with population pressure by restructuring or intensifying production.
Focusing on state policy, Gao Wangling points out that the throne was
alert to a looming population crisis already by the early eighteenth century,
and had responded with a multifaceted program not only of intensive and
extensive agricultural improvement, but also of deliberate economic diversi-
fication, through its nurturance of nonagricultural sectors such as handicrafts
and mining. (By implication, if there was a downturn in the economy after
the Qianlong reign, in Gao’s logic this would have been the result not of
factors of resources or labor, but rather of the progressive failure of official
initiative and morale.)36
Most recently, Richard von Glahn has martialed wage and price data to
argue – as Bao Shichen would have predicted – that more intensive and
efficient allocation of labor allowed the Qing to avert the worst effects of
population pressure. Prior to the Taiping catastrophe, for example, there was
no telltale spike in male mortality. Von Glahn finds a secular rise in the price
of grain over the course of the eighteenth century, but this was accompanied
by comparable rises in the price of cotton, silk, and other consumer goods,
leading him to conclude that the inflation was due not to “the stress of
overpopulation” but instead to a “substantial growth of the total money
supply.”37 Into the nineteenth century, he again finds no evident long-term
correlation of prices and wages with population growth. After a spike in the
1810s, grain prices indeed trended downward over the 1820s and 1830s, while
real wages (measured in grain) remained fairly static throughout the entire
1800–1850 period. In sum, “The evidence does not suggest that China had
reached the point of a Malthusian demographic crisis” any time prior to the
mid-century rebellions.38
35
J. Lee and C. Campbell, Fate and Fortune in Rural China: Social Organization and
Population Behavior in Liaoning, 1774–1873 (Cambridge, Cambridge University Press,
1997); J. Lee and F. Wang, One Quarter of Humanity: Malthusian Mythology and Chinese
Realities (Cambridge, MA, Harvard University Press, 1999).
36
Gao Wangling 高王凌, “關於康乾盛世的幾個問題” (Some Problems Concerning
the Prosperous Age from Kangxi to Qianlong), 清史研究通訊 (Bulletin of Qing
Historical Studies) 1990.4, 21–6.
37
R. von Glahn, The Economic History of China: From Antiquity to the Nineteenth Century
(Cambridge, Cambridge University Press, 2016), pp. 329–30.
38
Von Glahn, The Economic History of China, 363–4.
61

william t. rowe
The issue of population pressure is also reflected in what has become

a contentious debate on standards of living. Raised most provocatively by
Kenneth Pomeranz in his study of the “Great Divergence” in economic
development at opposite ends of the Eurasian continent, the argument that
wealthier areas of Qing China in the eighteenth century enjoyed living
standards at least on a par with those in more favored areas of Europe
seems to have originated in the work of the Chinese scholar Fang Xing. In
a pioneering 1996 article, Fang calculated the percentages that early Qing
peasant households in Jiangnan spent on food, shelter, clothing, and fuel. On
the basis of his estimate that such households spent only 55 percent of their
annual budget on grain and 76 percent on food overall, he concluded that
most lived in relative “comfort” (wenbaoxing 温饱型), comparing well to
other rural areas not only in China but also elsewhere in the world.39
Commenting on Fang’s conclusion, Pomeranz pointed out that contempor-
aneous English peasant households spent roughly the same portion of their
budget on food, but added that Fang “almost certainly undercounts non-
grain consumption” on items like family rituals, jewelry, and entertainment,
suggesting that food expenditures in Jiangnan were considerably lower than
those in England.40 He argues as well that consumption of “everyday luxur-
ies” such as tea and sugar was higher in eighteenth-century Jiangnan than in
England. In my view we can discount tea, a domestic product in China and
a costly import in England, but sugar makes a potentially useful surrogate for
discretionary spending. Pomeranz finds that Chinese sugar consumption as
late as 1800, while lower than that in England (the world leader), remained
notably higher than that in continental Europe.41 In terms of nutrition,
Pomeranz estimates the Qing empire-wide average calorie consumption
during the eighteenth century to be 2,651 per day, a figure which “compares
well” with that of Britain, and was “quite far above” estimates for Europe as
a whole.42 Life expectancies in the Qing empire, though there was substantial
diversity by region, were overall “quite comparable” to those in England.43
On the basis of Fang’s, Pomeranz’s, and independent calculations, the most
recent scholarship tends to concur that standards of living in the more
developed areas of the Qing during the “prosperous age” were higher than
39
Fang Xing 方行, “清代江南農民的消費” (Peasant Expenditures in Qing Jiangnan),
中國經濟史研究 (Studies in Chinese Economic History) 1996.3, 93–4.
40
K. Pomeranz, The Great Divergence: China, Europe, and the Making of the Modern World
Economy (Princeton, Princeton University Press, 2000), p. 137.
41
Pomeranz, The Great Divergence, pp. 118–23. 42 Pomeranz, The Great Divergence, p. 39.
43
Pomeranz, The Great Divergence, pp. 36–7.
62

they had ever been, higher than those in the West, and “unsurpassed by any
other contemporary society.”44
If this consensus view holds true, what happened to standards of living as
the prosperous age faded into the age of crisis in the late eighteenth and early
nineteenth centuries? In Jiangnan, population grew steadily and rapidly, from
172 million in 1750 to 254 million a hundred years later, when the Taiping wars
precipitated a sudden and dramatic decline (to 105 million in 1880). Beginning
in the early nineteenth century, this growth was accompanied not by further
expansion in cultivated acreage, but rather by a modest decline.45 Combined
with partible inheritance and relatively free alienability of land, the result was
considerable shrinkage of the average family farm size, estimated by Li
Bozhong as nine mu in the mid-nineteenth century, down from fifteen mu
two centuries earlier.46
The Jiaqing reign also witnessed a sudden spike in rice prices. Data
collected by Yeh-chien Wang for Wuchang prefecture (Hubei), a modestly
prosperous area of central China, show rice prices holding remarkably stable
at fifty-three grams of silver per hectal from 1769 through 1790; by 1801 they
had soared to 89.6 grams of silver, remaining in that range through the late
1830s.47 Annual data collected by Tanaka Issei for Xiaoshan county (Zhejiang)
indicate some radical short-term vacillations, but basically echo the Wuchang
pattern: a price of 2.3 taels per shi in 1781, remaining relatively stable for
a decade (2.2 taels in 1791), then leaping to 2.85 taels in 1800.48 A similar fin de
siècle spike occurred throughout the empire. Kishimoto Mio reports an
aggregate empire-wide rice price of 1.75 taels per shi in 1760, inflating to 2.2
taels in 1800.49 And rice prices directly mattered to life experiences. As in
other preindustrial economies, they had nearly immediate consequences for
fertility and mortality rates; in the Qing, “As prices rose, so did female
infanticide, resulting in lower recorded female births.”50
44
Huang Jingbin 黃敬斌, 民生與家計：清初至民國時期江南居民的消費 (Popular
Livelihoods and Household Budgets: Jiangnan Residents’ Expenditures from Early
Qing to the Republican Era) (Shanghai, Fudan University Press, 2009), p. 330; von
Glahn, The Economic History of China, p. 350.
45
Huang, 民生與家計, pp. 14–15. 46 Li, Agricultural Development, p. 138.
47
Yeh-chien Wang, “Spatial and Temporal Patterns of Grain Prices in China, 1740–1910,”
unpublished paper.
48
Tanaka’s data cited in Sui-wai Cheung, The Price of Rice: Market Integration in Eighteenth-
Century China (Bellingham, WA, Center for East Asian Studies, 2008), p. 153.
49
Kishimoto Mio 岸木美緖, “清朝中期經濟政策の基調” (Preliminary Investigation of
Mid-Qing Economic Policy), Chikaki ni arite 11 (1987), 18.
50
Lee and Wang, One Quarter of Humanity, p. 111.
63

william t. rowe
What were the implications for household consumption patterns? In his

pioneering study, Fang Xing found that while the share of overall food
expenditures within household budgets rose from 76 percent in the early
Qing to 83 percent around 1880, expenditures on non-grain foodstuffs com-
prised a much greater percentage of this over time, rising from 21 percent to
29 percent; Fang interprets this as evidence of rising, not falling, rural
standards of living in Jiangnan well into the late nineteenth century.51
A more detailed analysis has been undertaken by Huang Jingbin. Huang
asks whether the passage from the mid-eighteenth to the mid-nineteenth
century in Jiangnan indeed saw the progressive immiseration of rural house-
holds. He notes that literati writings from throughout the era consistently
complain about peasant “extravagance” (shechi 奢侈), but observes that this
complaint almost always was directed at ritual and festival spending, not at
daily consumption patterns. In terms of the latter, Huang calculates that
household consumption of staples such as grain and salt remained relatively
constant over the course of this period (12.5 dan for grain, 37.5 catties for salt),
while consumption of all other foodstuffs (vegetables, meat, eggs, oil, sugar,
wine) declined by an average of 20 percent. Consumption of sugar, for
example, declined from 27.5 to 20 catties per year. As percentages of house-
hold expenditure, grain rose from 39.7 percent in the mid-eighteenth century
to 45 percent in the mid-nineteenth, while expenses on other foodstuffs fell,
from 11.2 percent to 9.5 percent, and those on clothing, medical care, and
education (everything, that is, other than housing rent, which rose from
3.8 percent to 5.8 percent) shrank as well. There was, then, Huang concedes,
a fall-off in standards of living, but not so great as to qualify as immiseration,
or to negate Fang Xing’s conclusion that Jiangnan rural households remained
relatively “comfortable.”52 He adds, provocatively, that after the 1850s the
combination of population decline due to the Taiping war and regional
development of foreign trade brought surviving Jiangnan households a half-
century of renewed prosperity comparable to or greater than that of the
eighteenth century.53
We have less information on living standards in regions outside the
disproportionately prosperous Jiangnan. Based on wage data for nonfarm
workers in Beijing, Debin Ma and his colleagues argue that real income in fact
declined steadily over the late eighteenth and early nineteenth centuries,
51
Fang, “清代江南農民的消費”, 97. Pomeranz, The Great Divergence, p. 247, reports
Fang’s findings on this score with some skepticism.
52
Huang, 民生與家計, pp. 103, 307–8, 325–31.
53
Huang, 民生與家計, p. 327.
64

bottomng out below subsistence level in the 1830–1860 period.54 Another recent
study by Wu Xiaozhen looks at the area in and around Hengzhou prefecture in
southern Hunan. This was long a region of middling economic security, but
during the late eighteenth and early nineteenth centuries the region’s exports of
both rice and tea expanded dramatically, in what appears a clear pattern of
economic development. This was accompanied, however, by the first systematic
cultivation of maize, sweet potatoes, and other New World crops, which,
though highly unfavored by local tastes, displaced rice as the chief dietary staple.
Wu argues that this must be interpreted as a clear decline in living standards,
ironically accompanying and in some measure due to economic growth.55
Scattered evidence points to demographic consequences of a presumed
decline in living standards in the Jiaqing and Daoguang reigns. Studying lineage
genealogies from Tongcheng, a prosperous county in the periphery of the lower
Yangzi region, Ted Telford has found male life expectancy at birth to have
declined from 39.6 in the mid-eighteenth century to 34.9 in the early nineteenth.56
In a local banner population in Liaoning, James Lee and Cameron Campbell
found the male population growing 0.2 percent per year over the 1805–
1873 period, a dramatic decline from the average of 1.2 percent per year in the
preceding 1774–1804 era, the result primarily of lowered birth rates, resulting
from rigorous exercise of birth control by people observing a significant decline
in life opportunites.57 This decline in birth rates was undoubtedly even more
pronounced among women, leading to an increasingly skewed sex ratio, and – in
a culture in which marriage and reproduction were a moral imperative –
a growing incidence of bachelorhood. In his Tongcheng sample, for example,
Telford found that the share of unmarried males among those born in the last
quarter of the eighteenth century was 31.6 percent, a number which rose to
38.6 percent for those born in the first two decades of the nineteenth century,
46.4 percent for those born in the 1820s and 1830s, and 52 percent for those born in
the 1840s.58 Similarly, in their Liaoning sample, Lee and Campbell found
54
R.C. Allen, J.-P. Bassino, D. Ma, C. Moll-Murata, and J.L. van Zanden, “Wages, Prices,
and Living Standards in China, 1738–1925, in Comparison with Europe, Japan, and
India,” Economic History Review 64 (2011), Figs. 5, 6.
55
Wu Xiaozhen 吳小珍, “糧食生產供求變動與地方的經濟：以湘南為中心” (Shifts
in Food Supply and Qing Local Economies: The Case of Southern Hunan), 清史研究
(Studies in Qing History) 2012.3, 45–57.
56
T.A. Telford, “Patching the Holes in Chinese Genealogies: Mortality in the Lineage
Population of Tongcheng County, 1300–1800,” Late Imperial China 11.2 (December
1990), 133.
57
Lee and Campbell, Fate and Fortune in Rural China, p. 47.
58
T.A. Telford, “Family and State in Qing China: Marriage in the Tongcheng Lineages,
1650–1880,” in Institute of Modern History (ed.), Family Process and Political Process in
Modern Chinese History (Taipei, Academia Sinica, 1992), pp. 921–42.
65

william t. rowe
a declining marriage rate among males of sixteen to thirty-five sui from eighty-
eight per thousand in the pre-1805 period to seventy-five per thousand after
that year.59 Matthew Sommer has documented in dramatic detail some of the
consequences of this skewed sex ratio and distorted marriage market, including
changing cultural norms that effectively sanctioned such heterodox practices as
polyandry, wife pimping, and wife selling.60
Economic Slowdown?
If the early nineteenth century was “a watershed that marked the beginning
of a prolonged period of intensified Malthusian pressure,”61 and China’s ratio
of population to cultivated land was increasingly adverse, was the overall
economy assisting by providing nonfarm employment opportunities for
surplus agricultural labor?
In a masterful article, the economic historian Peng Zeyi traced the impres-
sive development of the Qing economy in the first half of the dynasty, but
then characterized the half-century from 1784 to 1839 as one of “stagnation” or
even “contraction.”62 In Peng’s calculations, cultivated acreage empire-wide
shrank 1.6 percent over the thirty years between 1784 and 1813, even while
population continued to grow rapidly. State-owned manufacturing enter-
prises declined. Annual capital budgets (yongyin 用银) of the porcelain
factories at Jingdezhen fell from 8,000 taels in 1740, to 7,000 taels in 1799, to
a mere 2,500 taels in 1812. Official textile mills in Nanjing and elsewhere, and
the government mints at Beijing, also experienced significant slowdowns.
Much of the contraction of state-owned enterprises, of course, represented
loss of market share to the private sector, but Peng points out that private
enterprises likewise declined during this era. The salt industry of Sichuan
shrank dramatically from 1780 to 1820, and the copper and other mines of the
southwest saw their production fall roughly 1 percent per year between 1783
and 1797. Urban handicrafts, being more dispersed, are harder to measure,
but Peng saw contraction there as well. In what might be his most debatable
argument, he sampled some seventy-nine artisan and merchant guilds from
fifteen major commercial cities (including Hankou, Suzhou, Chongqing,
59
60
M.H. Sommer, Polyandry and Wife Selling in Qing Dynasty China (Berkeley, University of
California Press, 2015). The reference to Bao Shichen is on pp. 337–8.
61
62
Peng Zeyi 彭澤益, “清代前期手工業的發展” (The Development of Handicrafts in
the Early Qing), 中國史研究 (Studies in Chinese History) 1981.1, 43–60.
66

Changsha, and Foshan), and found their dates of formal organization to

overwhelmingly concentrate in this era (four in 1671–1715, sixteen in 1716–
1785, and fifty-nine in 1786–1839). Whereas others might see this as evidence of
commercial efflorescence, Peng holds a negative view of guilds as exclusion-
ary defensive organizations, and hence finds support in this for his contention
that interregional commerce was also under severe pressure during the turn-
of-the-century decades. In this era, he concludes, “The independent develop-
ment of [urban] commercial capital proceeded in inverse proportion to the
development of the overall economy.”63
Is Peng correct in his view of this era as one of generalized economic
contraction? Let us consider one facet of the economy which might plausibly
be seen as a proxy for the overall economy, and historically the early modern
Chinese economy’s driving sector: the rice trade down the Yangzi from
Wuhan to Nanjing and Suzhou. There is an active historiography on this
question. Few dispute the baseline figure provided by Han-sheng Chuan and
Richard Kraus of between 8 million and 15 million dan of rice following this
route in the 1730s.64 A recent, careful study of grain prices throughout rice
consumption areas of south China suggests that as late as 1795 market price
cointegration (suggestive of arbitrage via vigorous inter-market exchange)
remained somewhat higher than that in Western Europe and only very
slightly behind that of England.65
The question is what happened after that, and why. Based on stele records
of the Hunan provincial club in Nanjing, Nakamura Jihei concluded that the
downriver rice trade had decreased considerably by 1800.66 Studying
a prefectural gazetteer of Yongzhou, Hunan, Kitamura Hironao found that
rice exports to the lower Yangzi from the single rice-producing county of
Qiyang, amounting to over 100,000 dan annually around 1800, had ceased
altogether by the time of the gazetteer’s publication in the 1830s; Kitamura
thus argued for a sudden cessation of the Yangzi rice trade as a whole at the
end of the Qianlong reign.67 Chuan and Kraus themselves were less categor-
ical, though they too observed a major shrinkage of the trade. Disputing
63
Peng, “清代前期手工業的發展,” 52.
64
Han-sheng Chuan and R.A. Kraus, Mid-Qing Rice Markets and Trade: An Essay in Price
History (Cambridge, MA, Harvard University East Asian Monographs, 1975), p. 77.
65
C.H. Shiue and W. Keller, “Markets in China and Europe on the Eve of the Industrial
Revolution,” American Economic Review 97.4 (September 2007), 1189–1216.
66
Nakamura Jihei 中村治兵衛, “清代湖廣米流通の一面” (One Aspect of the
Huguang Rice Trade in the Qing Era), 社會經濟史學 (Studies in Socio-economic
History) 18.3 (1952), 53–65.
67
Kitamura Hironao 北村敬直, “清代の商品市場について” (On Mid-Qing Commodity
Markets), 經濟學雜誌 (Journal of Economics) 28.3–4 (1953), 5, 8.
67

william t. rowe
Dwight Perkins’s claim that the downriver rice trade had totally disappeared
by c. 1930, they instead argued for a long-term decline to some 2.7 million dan
in the early twentieth century, roughly one quarter of the volume at the
trade’s eighteenth-century peak.68 A recent study by Sui-wai Cheung finds
two boom periods of downriver rice shipments in 1724–1727 and 1732–1755,
with a subsequent modest revival in 1777–1786. Though reports in 1788
indicate that middle Yangzi rice was “unmarketable” in Jiangnan at that
time, the trade actually revived once more in 1795 and remained active at
least through the remainder of the Jiaqing reign, though the trend thereafter
was downward.69 Surveying this historiography, Richard von Glahn con-
cludes that the overall evidence “points to a decline in the scale of inter-
regional trade in grain beginning in the 1790s.”70
Why these vicissitudes in the trade, and what does the decline signify?
Cheung argues that the key factor at all times was demand in the lower
Yangzi for rice from upriver regions – its oscillation between rice deficit and
rice surplus. Although agriculture in Jiangnan was increasingly commercial-
ized, and its population heavily engaged in handicrafts and other non-grain-
producing work, Cheung follows Li Bozhong in seeing the region as by no
means “overpopulated,” and in normal years self-sufficient in grain. Thus
what lay behind the occasional spikes in downriver rice imports in both the
eighteenth and the nineteenth centuries was weather-induced short-term
harvest shortfalls.71 Chuan and Kraus, by contrast, look for answers in the
exporting rather than the importing region.72 They reason that the decline in
the middle to lower Yangzi rice trade logically reflected either a decline in
central China’s interregional commerce overall, or else a substitution of other
commodities for rice; they indicate doubt that the latter was the case, but, as
we shall see in a minute, I think they may be wrong. A secular decline in the
downriver rice trade, moreover, might reflect a general decommercialization
of middle Yangzi (and upper Yangzi) agriculture – an option I believe to be
patently not the case. Alternatively, it might reflect a diversion of commer-
cialized rice to local consumption on the part of the production regions’ own
increasingly non-grain-producing urban and rural populations. As indicated
by the growing incidence of collective protests in extraction centers of
Hunan, against the rising local rice prices resulting from extra-regional
68
Chuan and Kraus, Mid-Qing Rice Markets, p. 77.
69
Cheung, The Price of Rice, pp. 97–102. 70 Von Glahn, Economic History of China, p. 331.
71
Cheung, The Price of Rice, pp. 113, 135–6.
72
Chuan and Kraus, Mid-Qing Rice Markets, p. 78.
68

exports – an early example being that in Xiangtan in 1819 – this is much more
likely to have been the case.73
What does this say about the state of long-distance domestic commerce
overall, as the Qing entered the nineteenth century? By most accounts, it was
flourishing. Wu Chengming conservatively estimates the overall value of
traded commodities in the empire around 1840 to be some 387,624,000 taels of
silver, among which roughly 163 million was in rice, 107 million in cotton and
cotton cloth, 59 million in salt, 32 million in tea, and 27 million in silk and silk
cloth. With the possible exception of rice, these levels were all at the highest
they had ever been.74 A classic study by Peng Yuxin showed the interregional
trade passing through the central China entrepôt of Hankou in the nine-
teenth century to link all provinces of the empire, comprise scores of
commodities, and reach annual values of hundreds of millions of taels.75
Let us look briefly at the regional economy I know best, that of the middle
Yangzi. This was the region that would have been most seriously impacted
by a decline in the downriver rice trade. But at least two commodities
advanced to offset any such losses in regional exports: tea and cotton.
Three middle Yangzi counties – Hunan’s Anhua and Linxiang and Hubei’s
Puqi – generally produced more than half of the region’s tea output, but
almost all parts of the region adopted some tea production as market condi-
tions warranted. At least by the Ming era, teas from these counties circulated
within the region itself, but due to their perceived inferiority to products of
other regions they did not enjoy much of a domestic interregional market.
The foreign market was a different story. It was middle Yangzi tea that had
supplied the famous tea-for-horse exchange with the Mongols during the
Ming and early Qing, shipped up the Han river and overland to the great
border market of Zhangjiakou. With the Kangxi Emperor’s privatization of
this trade in the 1710s it increased dramatically. Exports to Asiatic Russia had
also followed this route since the twelfth century, after 1727 exchanged at the
treaty market of Kiakhta. But the growing population of Siberia brought an
73
P.C. Perdue, “Insiders and Outsiders: The Xiangtan Riot of 1819 and Collective Action in
Hunan,” Modern China 12.2 (April 1986), 166–201. See also Shigeta Atsushi 重田德,
“清朝における湖南米市場の一考察” (Hunan Rice Markets in the Qing Dynasty),
東洋文化研究所紀要 (Memoirs of the Institute of Oriental Culture) 10 (1956), 427–98;
and R.B. Wong, “Grain Riots in the Qing Dynasty,” Journal of Asian Studies 41.4 (August
1982), 767–88.
74
Wu Chengming 吳承明, “論清代前期我國國內市場” (The Domestic Market in the
Early Qing), 歷史研究 (Historical Studies) 1983.1, 99.
75
Peng Yuxin 彭雨新, “抗日戰爭前漢口的洋行和買辦” (Foreign Firms and
Compradores at Hankou Prior to the Sino-Japanese War), 理論戰線 (Theoretical
Front) 1959.2, 22–9.
69

william t. rowe
estimated sixfold increase in the trade’s volume during the early nineteenth
century.76
The early development of Qing tea exports to Britain in the eighteenth
century left the middle Yangzi largely unaffected, since the inferior product
could little compete with the better teas of coastal areas. This changed
dramatically, however, with the abolition of the Canton monopoly in 1842.
The removal of legal restraints on the trade brought a sudden increase in
demand, causing the price of Fujian and Zhejiang teas to skyrocket, and
offering European buyers reason to seek out new and cheaper sources of
supply. The middle Yangzi in the 1840s thus underwent a virtual invasion by
Cantonese and other commercial entrepreneurs, proselytizing a shift to tea
cultivation. Fanning out from the established tea districts, these merchants
persuaded growers in a widening circle of production areas to convert from
other crops to tea, or, where green tea was already grown, to red tea for the
European market. At first most of this product went south to Guangzhou,
then gradually shifted to Shanghai, and, with the opening of the middle
Yangzi itself to foreign trade in 1861, to Hankou. Tea thus became the
region’s chief foreign export, and likely for a period its most important
commercial crop. In Anhua county, by the 1860s, it constituted the major
source of livelihood for an estimated 90 percent of the population.77
After tea came cotton. In arguing for the rapid relative decline of China’s
economy in the late eighteenth and early nineteenth centuries, Kenneth
Pomeranz argues that empire-wide cotton cultivation and processing
increased only very slightly, if at all, in the century following 1750. In the
lower Yangzi, there was probably a slight decline, and in north China the
drop-off was “considerable.” This leaves the middle Yangzi provinces, which,
in Pomeranz’s words, while they “did increase their cotton cultivation after
1750 . . . never became very large producers.”78 On this point he is simply
wrong.
Cotton had been cultivated extensively in northern portions of the middle
Yangzi since the sixteenth century, with the region’s output ranking only
behind that of the lower Yangzi. While much of this crop went for household
consumption, by the eighteenth century the majority was taken by handicraft
weavers producing for the market, and the concentration of such activity in
76
M.I. Sladkovsii, Economic Relations between Russia and China (Jerusalem, 1966), p. 71.
77
Shigeta Atsushi 重田德, “清末における湖南茶的新展開” (The New Development
of Hunan Tea Production in the Late Qing), in Shigeta, 清代社會經濟史研究 (Studies
in Qing Socio-economic History) (Tokyo, Iwanami Shoten, 1975), pp. 207–38.
78
Pomeranz, The Great Divergence, pp. 139–41.
70

centers such as Wuhan and Changde led to the rise of a brisk intra-regional
trade. By the nineteenth century, a considerable portion of middle Yangzi
cotton also found its way into the interregional trade, especially to Sichuan,
which took hundreds of thousands of bales per year.79
The middle Yangzi, notably the Hubei county of Xianning, by the late
Ming had also acquired an empire-wide reputation for its cotton cloth. By the
eighteenth century, cotton textiles spun and woven by urban, rural, and
especially suburban households were sent to many parts of the empire. The
areas around Wuhan, Changde, Yiyang, and especially Shashi and Jingzhou
prefecture in western Hubei were major suppliers to this intra-regional trade,
though production was also dispersed well beyond these centers. Shashi, with
its “Thirteen Guilds” management combine, comprising merchant groups
from throughout the empire, was described by early Western reporters as the
greatest weaving center in all of China. Much of the product went upriver, to
Sichuan, Yunnan, and Guizhou, and downriver to Jiangxi.80
In sum, if the rice trade downriver on the lower Yangzi declined in the
early nineteenth century, as most scholars now agree, and even if trade on
that artery declined as a whole (domestic customs revenues at Suzhou are
reported to have dropped from 583,000 taels in 1791 to 391,000 taels forty years
later81), this does not necessarily indicate a contraction of the overall Qing
commercial economy as the empire entered the nineteenth century. What it
more likely indicates, as G. William Skinner has argued, is that there was
a diversification of commodities and trade routes as the Qing economy
fleshed out beyond one or two staples.82
Maritime Trade
China’s maritime trade grew steadily over the course of the “prosperous
age,” and continued to do so throughout the period under discussion here.
Following the Kangxi Emperor’s lifting of the “sea ban” (haijin 海禁) and
establishment of the Maritime Customs (haiguan 海关) in 1683–1684, except
for a brief hiatus under the “second sea ban” of 1716–1728, the Qing adminis-
tration generally encouraged the development of East Asian intra-regional
79
A. Feuerwerker, “Handicraft and Manufactured Cotton Textiles in China, 1871–1930,”
Journal of Economic History 30.2 (1970), 340.
80
Wu, “論清代前期我國國內市場,” 104–6; Lyon Chamber of Commerce, La mission
lyonnaise d’exploration commerciale en Chine 1895–97 (Lyon, A. Rey, 1898), pp. 2.280–1.
81
Von Glahn, Economic History of China, p. 371, citing figures from Wu Chengming.
82
G.W. Skinner, “Regional Urbanization in Nineteenth-Century China,” in Skinner (ed.),
The City in Late Imperial China (Stanford, Stanford University Press, 1977), p. 713 note.
71

william t. rowe
maritime trade throughout this era. Qing trade with Southeast Asia con-
tinued to expand at least through the 1840s, with dozens of ports, large and
small, along the southeast coast hosting customs office branches and partici-
pating in that trade.83 Despite the Tokugawa shogunate’s imposition of
restrictions on the export of monetary metals to China in 1685, trade with
Japan remained active as well. In the single year 1830, some 222 oceangoing
ships called at Shanghai, Ningbo, and Suzhou, having stopped previously at
various ports in Southeast Asia and Japan.84
The largest single import commodity within this regional trading network
was rice from Siam. Whereas under the Ming Siamese exports to China
comprised primarily luxury goods such as fine timber, with the lifting of the
sea ban rice – a staple – became the backbone of the trade. Imports of rice into
Guangdong and other grain-deficient southeastern provinces were repeat-
edly encouraged by Qing regional officials whenever shortages became
acute, throughout the late eighteenth and early nineteenth centuries.
During the heyday of the trade in 1809–1833, rice was Siam’s largest export,
and the Qing empire its largest customer. With Qing imperial policy toward
Chinese merchants personally going abroad gradually easing over the eight-
eenth century, it was those merchants who dominated the trade; Viraphol
reports that during the first few decades of the nineteenth century, Chinese
vessels carried around eighty-five tons of rice into the empire per year, nearly
three times the volume of their closest competitor, the British. But the
number of European ships plying the China coast was growing rapidly. In
1837, over 200 Western ships called at Guangzhou, nearly twenty times the
number that had called there a century earlier, in 1732.85
It was, of course, the Sino-British trade that came to dominate this
exchange, growing progressively over the late eighteenth century. This
growth accelerated with the end of the European wars in 1815, again with
the introduction of the steamship into East Asian waters in 1830, and yet again
with the abolition of the East India Company’s monopoly in 1834. The driving
force behind this expansion was the British demand for Chinese tea. Tea
exports from the Qing to England rose from virtually nil in the final years of
the sea ban, to 15 million pounds in 1785, and double this amount around 1830.
83
Huang Guosheng 黃國盛, 鴉片戰爭前的東南四省海關 (Maritime Customs in
China’s Four Southeastern Provinces Prior to the Opium War) (Fuzhou, Fujian renmin
chubanshe, 2000), pp. 178–9.
84
G. Zhao, The Qing Opening to the Ocean: Chinese Maritime Policies, 1684–1757 (Honolulu,
University of Hawaii Press, 2013), pp. 135–6, 171–2.
85
S. Viraphol, Tribute and Profit: Sino-Siamese Trade, 1652–1853 (Cambridge, MA, Harvard
University Press, 1977), pp. 109–20, 180.
72

Total Sino-British trade – imports and exports – grew in value from 23 million
dollars in 1817 to nearly 37 milliion in 1833. The imports to China included
various luxury goods (pepper, ivory, shark fins, etc.) as well as the trade’s
staple, South Asian raw cotton, which was overtaken by opium in 1823. Qing
imports of this last item grew ninefold between the turn of the century and
the outbreak of the Opium War.86
If there was, as some have suggested, a contraction of Qing interregional
domestic trade at the same time that the foreign maritime trade continued its
rapid growth, is the story of the early nineteenth century one of transition
from a balanced redistributive economy to an extraction economy in service
to an external metropolis? The best data we have suggest that this was hardly
the case. According to figures generated by Wu Chengming, as late as 1840
Qing maritime exports of tea (some 605,000 dan) remained less than a third of
tea produced for the domestic market (more than 2,000,000 dan), while raw
cotton imports remained less than a quarter of that grown and marketed
domestically.87 The Qing empire’s economy was simply too large and its
markets too developed to be reduced to dependency.
The Daoguang Depression

The reign of the Daoguang Emperor (1821–1850), and especially the quarter-
century following 1825 or 1826, have become familiar to historians as the era
of the “Daoguang depression” (Daoguang xiaotiao 道光萧条). In some recent
Chinese scholarship, indeed, this phenomenon is paired with the so-called
“Kangxi depression” of the decades around 1700, essentially bookending the
intervening “prosperous age” of the eighteenth century, and signifying the
completion of an economic cycle.88
While there is much debate about the nature and causes of the depression,
there is general agreement regarding three major trends. First is the deflation
of commodity prices (in silver). We generally accept that prices underwent
a prolonged benign inflation over the century following the Kangxi
Emperor’s declared opening of the Qing to foreign trade in 1682, followed
by perhaps forty years (1780–1820) during which the price rise continued at
a markedly slower pace. Then, in Yeh-chien Wang’s words, “The second
quarter . . . of the nineteenth century was generally a period of severe
86
M. Greenberg, British Trade and the Opening of China, 1800–42 (Cambridge, Cambridge
University Press, 1951), pp. 3, 76–89, 103, 217, 221.
87
Wu, “論清代前期我國國內市場,” 99.
88
Ni, 清朝嘉道財政與社會, p. 1.
73

william t. rowe
deflation. The movement of prices lost its vigor at the beginning of the
century and remained virtually at a standstill for more than a decade; then it
took a sharp turn downward until 1850.” Wang very ambitiously constructs
a price index of all commodities throughout the empire. Setting the year 1682
as 100, he finds a steady rise to 300 in 1815, followed by a dramatic decline to
150 by 1850.89
Second, scholars agree that during this same period there was a sudden
shift in the Qing’s balance of silver flows, from large annual inflows to
massive outflows. Table 2.1, originally by Richard von Glahn, offers the
following table of this dramatic turnaround:90
A study by Yu Jieqiong found a net inflow of 74.7 million silver dollars over
the years from 1801 to 1826, compared to a net outflow of 133.7 million in 1827–
1849.91
The third component of the Daoguang depression about which scholars
agree is that silver became much more expensive in terms of copper (yingui
qianjian 银贵钱贱). The Qing inherited from its predecessor a bimetallic cur-
rency system, based on uncoined silver specie, the unit of account for which was
the “tael” (liang 两), and statutory coin (zhiqian 制钱) made of copper alloy, the
unit of the latter being the “cash” (wen 文). Whereas a string of 1,000 copper cash
(chuan 串) carried a par value of one silver tael, the disastrous experiences of the
late Ming had convinced the Qing that efforts to maintain a 1,000:1 exchange rate
by administrative fiat were counterproductive. Instead, it settled for the goal of
Table 2.1 Net flow of silver to

China, 1818–1850
1818–1820 + 9.89 million pesos

1821–1825 + 21,01 million pesos
1826–1830 – 12.96 million pesos
1831–1835 – 19.81 million pesos
1836–1840 – 29.49 million pesos
1840–1845 – 51.33 million pesos
1846–1850 – 30.59 million pesos
89
Yeh-chien Wang, “The Secular Trend of Prices during the Ch’ing Period (1644–1911),”
Journal of the Institute of Chinese Studies of the Chinese University of Hong Kong (1972), 361.
See also R.H. Myers, The Chinese Economy: Past and Present (Belmont, CA, Wadsworth,
1980), pp. 71–4.
90
Von Glahn, Economic History of China, p. 367. The “peso” is the Spanish Carolus peso,
known after 1800 in China as the “yuan.”
91
Cited in Wang, “The Secular Trend of Prices,” 365.
74

maintaining “stability” in the money market, over both time and space. In this
effort it enjoyed only mixed success. In the late seventeenth century, copper–
silver exchange rates of well over 1,000:1 were not at all unusual, but, from the
beginning of the eighteenth century, copper coin became increasingly “expen-
sive,” and rates of 700:1 or 800:1 were the norm.92
In the Daoguang reign this trend was dramatically reversed. Local situations
varied considerably, of course, but we might consider as representative the
series of exchange rates discovered for Shaanxi province by Hans Ulrich Vogel.
For the entire eighteenth century, Vogel tells us, the exchange rate in Shaanxi
remained well below 1,000:1, but in the Daoguang reign the value of copper
cash in terms of silver taels rapidly declined: 1,370:1 in 1831, 1,480:1 in 1842, and as
high as 1,800:1 in 1846.93 This increasingly skewed exchange rate also implies that
commodity prices in silver were ever more divorced from prices in copper.
Following a price series for some dozen retail items in a store in Ningjin county,
Zhili, Yeh-chien Wang shows prices in copper cash remaining relatively stable
over the first half of the nineteenth century, even as prices of the same items in
silver fell by more than half between 1815 and 1849.94
These, then, are the areas of general agreement. The relationship among
these trends, and patterns of causation, have occasioned spirited debate, both
at the time and today. It was very likely none other than Bao Shichen who, in
1820, first tied the silver shortage directly to the opium trade.95 Bao calculated
that in contemporary Suzhou, opium users spent altogether between
3 million and 4 million taels per year on their habit, and on this basis estimated
an empire-wide annual outlay of no less than 100 million taels.
Unsurprisingly, since opium was nearly wholly a foreign import, some
80 million taels of silver per year were exported to pay for the drug. Bao
did not identify this specifically as a net unfavorable balance of trade,
but this was done at least as early as 1833 in a memorial of censor Sun
92
The classic work is Peng Xinwei 彭信威, 中國貨幣史 (Chinese Monetary History)
(Shanghai, Renmin chubanshe, 1965; first published 1954). See also Chen Zhaonan 陳昭
南, 雍正乾隆年間的銀錢比價變動 (Changing Relative Prices of Silver and Copper in
the Yongzheng and Qianlong Reigns) (Taipei, Zhongguo xueshu ju, 1966);
Kuroda Akinobu 黑田明伸, “清代銀錢二貨制的構造とその崩壞” (The Structure
and Collapse of Silver–Copper Bimetallism in the Qing), 社會經濟史學 (Studies in
Socio-economic History) 57.2 (1992), 93–125; R. von Glahn, Fountain of Fortune: Money
and Monetary Policy in China, 1000–1700 (Berkeley, University of California Press, 1996);
W.T. Rowe, “Provincial Monetary Practice in Eighteenth-Century China,” in
C. Moll-Murata, J. Song, and H.U. Vogel (eds.), Chinese Handicraft Regulations of the
Eighteenth Century (Munich, Iudicium, 2005), pp. 347–71.
93
H.U. Vogel, “Chinese Central Monetary Policy, 1644–1800,” Late Imperial China 8.2
(1987), 27.
94
Wang, “The Secular Trend of Prices,” 355. 95 Rowe, Speaking of Profit, pp. 151–5.
75

william t. rowe
Lanzhi.96 Bao argued that exports of silver for opium caused a severe short-
age of capital in domestic markets, but again did not refer to the problem of
bimetallic exchange rates. Nevertheless, he argued that the problems caused
by opium imports, economic and otherwise, were so severe that it was worth
the risk of war to close them down altogether.
In a powerful article of 1983, Peng Zeyi greatly elaborated on Bao Shichen’s
analysis, sharing with him the identification of opium imports as the principal
cause of the Daoguang depression. Peng noted that the outflow of silver in the
four years between 1843 and 1846 alone equaled close to 10 percent of the total
amount of silver in circulation in the Qing domestic economy. This prompted
hoarding behavior, decreasing investment in agricultural and handicraft produc-
tion, especially in the empire’s most developed areas – Jiangnan, Hunan, and
Sichuan. A crisis of credit in the commercial economy threated the acceptability
of widely circulated banknotes. Prices paid to producers fell – by as much as
50 percent for commercial rice producers. Tax burdens on cultivators (normally
assessed in silver and collected in the depreciated copper coin) rose,97 as did rents
charged by taxpaying landlords, and many farmers lost land title and leaseholds.
Resistance movements became more frequent and severe. The first sign of relief
came in the 1850s, when increased tea and silk exports helped correct the balance
of trade and the Qing’s hemorrhage of silver slowed.98
More recent scholarship, however, has called into question the role of
opium in precipitating the Daoguang depression. He Liping, for example,
calculates that the total outflow of silver due to opium imports over the
entire depression era lay between 3.6 percent and 6.7 percent of the empire’s
total silver supply. While this was not insignificant, to be sure, He argues that
it was not sufficient in itself to cause either the skewed monetary exchange
rate or the commodity price deflation that characterized the period.99
96
Wang Hongbin 王宏斌, “林則徐關於銀貴錢賤的認識與困惑” (Lin Zexu on the
Problem of Expensive Silver and Cheap Copper Coin), 史學月刊 (History Monthly)
2006.9, 35–41.
97
Yeh-chien Wang found conversion rates used in land tax collection to have risen from
880:1 in Jianding in 1759 to 1,360:1 in Xiaoshan in 1800, and to 2,600:1 in Ningbo in 1852.
Y.-C. Wang, Land Taxation in Imperial China, 1750–1911 (Cambridge, MA, Harvard
University Press, 1973), p. 61.
98
Peng Zeyi 彭澤益, “鴉片戰後十年間銀貴錢賤波動下中國經濟與階級關係” (The
Impact of Expensive Silver and Cheap Copper Coins on the Chinese Economy and
Class Relations in the Post-Opium War Decade), in Peng Zeyi, 十九世紀後半期的中
國財政與經濟 (Government Finance and the Economy in Late Nineteenth-Century
China) (Beijing, Renmin chubanshe, 1983), pp. 24–71.
99
He Liping 賀力平, “鴉片貿易與白銀外流關係之再檢討” (A Re-examination of the
Relationship between the Opium Trade and Silver Outflow), 社會科學戰線 (Social
Science Front) 2007.1, 63–80.
76

Man-houng Lin argues that the annual outflow of silver in the years from
1808 to 1856 averaged roughly 8 million taels, an amount equivalent to one-
quarter of the Qing’s annual land tax revenue and 6 percent of the govern-
ment’s annual revenues as a whole. She argues, like He Liping, that this
amount, though large, was not large enough to constitute the determining
factor in the depression. She agrees that a silver shortage was largely to
blame, but rather a short-term “global silver shortage” brought about by
declining production at the source, the New World mines.100 Richard von
Glahn agrees with Lin that opium was not the major problem, but disagrees
on the causal role of global silver supplies. Instead, he finds the roots of the
depression in the overall contraction of the domestic economy. He cites as an
indicator of this the slowdown in domestic interregional trade (a factor about
which, as noted above, I am not necessarily in agreement), and points out
that, unlike the domestic trade, the empire’s foreign trade – its silk and tea
exports – continued to grow steadily throughout the depression era. The
outflow of silver, then, was for von Glahn “a consequence, not a cause, of an
economic decline that traced back to the final decades of the eighteenth
century.”101
Government Finance
Nothing more clearly reveals the economic tribulations of the Jiaqing and
Daoguang reigns than the vicissitudes of government finance. As Gao Xiang
has pointed out, when the Qianlong Emperor boasted that he had presided
over a “prosperous age,” it was less the condition of the overall economy
than that of the central government treasury to which he was referring.
Although there were several stores of wealth at the court’s disposal, both the
most important and the readiest measure of the Qing’s fiscal health was the
silver accounts of the Board of Revenue treasury (hubu yinku 户部银库). By
this measure, the Qianlong Emperor’s boast was hardly an idle one. Standing
at 33,950,000 taels at the outset of his reign in 1736, treasury reserves had risen
by about 25 percent through the late 1750s, when the imperial adventures in
what became Xinjiang essentially ate up this surplus. On the rise again after
1760, due in large measure to active campaigns for sales of gentry degrees and
100
M.-H. Lin, China Upside Down: Currency, Society, and Ideologies, 1808–1856 (Cambridge,
MA, Harvard University Asia Series, 2006), pp. 22, 133.
101
Von Glahn, Economic History of China, p. 374. See also R. von Glahn, “Foreign Silver
Coins in the Market Culture of Nineteenth-Century China,” International Journal of
Asian Studies 4.1 (2007), 61–2.
77

william t. rowe
official ranks (juanna 捐纳), treasury reserves grew substantially over the
following three decades, reaching an all-time high of 83,408,014 taels in 1778.
Corruption and waste in the final years of Qianlong’s reign reduced this
surplus somewhat, but in the final year of Gaozong’s formal reign, 1795, the
level of reserves remained at a quite comfortable 69,391,990 taels.102 Over the
first decade of the Jiaqing reign, according to the standard account,
the treasury surplus was effectively “wiped out.”103 This is an exaggeration,
to be sure, but there is little question that the severe depletion of treasury
funds in this era was real, and legitimate reason to view this as a turning point
in the long-term decline of Qing fiscal capacity.
Treasury reserves underwent a sudden dramatic decline during the final
years of the retired emperor’s life and his successor’s first years on the throne.
Reserves were 19,185,592 taels in 1798, less than a third of what they had been
three years before, and less than a quarter of their level of a decade earlier.104
By 1801, they were 16,930,000 taels – less than they had been at any time since
the late seventeenth century. But this would also prove to be a low point for
the Jiaqing reign. Over the first decade of the nineteenth century, treasury
reserves gradually recovered to the 30-million-tael range, and they remained
relatively stable over the 1810s, 1820s, and 1830s. Beginning in 1840, at the
height of the Opium War, they plummeted again, reaching a new low of
9,933,790 taels in 1843.105
The decline in treasury reserves in the Jiaqing and Daoguang reigns
reflected a highly unsteady balance of current accounts, and a reluctant
abrogation of the long-standing Qing policy of liangru weichu 量入为出 –
“calculate expenditures on the basis of revenues.” The treasury enjoyed
a surplus of revenues over expenditures of nearly 12 million taels as late as
1791, but in the first year of Jiaqing’s nominal rule in 1796 this had turned
around to a deficit of nearly 13 million, and the following year it peaked at
negative 28,807,266. In 1798 the annual deficit declined to around 9 million,
reflecting a sudden rise in revenues of between 6 million and 16 million. With
some occasional ventures into positive territory, annual deficits continued at
102
Gao, “從持盈保泰到高壓統治,” 9; Ni, 清朝嘉道財政與社會, pp. 83–4.
103
Jones and Kuhn, “Dynastic Decline and the Roots of Rebellion,” 144.
104
Peng, “鴉片戰後十年間銀貴錢賤波動下中國經濟與階級關係,” 51; Ni, 清朝嘉道
財政與社會, p. 169.
105
Shi Zhihong 史志宏, 清代戶部銀庫收支和庫存統計 (Statistics on Income, Outlays,
and Current Balances in the Board of Revenue Treasury during the Qing Era) (Fuzhou,
Fujian renmin chubanshe, 2009), p. 104. See also E. Kaske, “Fund-Raising Wars: Office
Selling and Interprovincial Finance in Nineteenth-Century China,” Harvard Journal of
Asiatic Studies 71.1 (June 2011), 91; and Ni, 清朝嘉道財政與社會, pp. 169–70.
78

levels between 1 million and 5 million until the death of the Daoguang
Emperor at mid-century.106
What so rapidly tipped the balance between income and expenditure was the
enormous cost of suppressing the White Lotus Rebellion, coinciding with the first
ten years of the Jiaqing Emperor’s reign. Military expenses had always comprised
the largest share of central government expenditures; Wei Yuan estimated them
at an average of 17 million taels per year during the Qianlong reign (far greater
than the next largest budgetary items: yanglian 养廉 pay supplements to civil and
military officials, at 4.3 million, and Yellow River maintenance costs, at
3.8 million).107 But the cost of suppression of the White Lotus was out of all
proportion to this – due in no small part to real and fraudulent payments for
civilian militiamen to fight for the Qing cause.108 The sudden rise in negative
balance of accounts in 1797 was due to a dramatic jump in annual treasury
expenditures, from 18 million to 35 million taels, mostly spent on the war. The
total cost of defeating the rebellion has been estimated at between 120 and
200 million taels, several times the reserves of the treasury in even its best years.109
Turning to the revenue side, Yeh-chien Wang long ago established that
there was a substantial growth in total central government revenues over
the second half of the Qing era, from an annual income of just under
74 million taels in 1753 to 292 million in 1908. Most of this came from new
exploitation of commercial taxation, which rose tenfold over the later Qing,
from accounting for 26.5 percent of total revenues to 64.9 percent. But
though the land tax (diding 地丁) thus comprised an ever smaller percentage
of revenues, it too rose in absolute terms, nearly tripling from 37,817,000 in
1753 to 103,400,000 in 1908.110
Wang’s long-term figures, however, mask the fact that for most of the
Jiaqing and Daoguang reigns – before the landmark shift to reliance on
commercial taxes in the 1850s – government revenues were in decline.
There were wide fluctuations from year to year, but the trend of decline is
suggested by the figures in Table 2.2 for revenue collected in silver at the
Board of Revenue treasury:111
106
Ni, 清朝嘉道財政與社會, pp. 165–8, 380.
107
Ni, 清朝嘉道財政與社會, p. 79.
108
Y.C. Dai, “Civilians Go into Battle: Hired Militias in the White Lotus War, 1796–1805,”
Asia Major, third series 22.2 (2009), 145–78.
109
Jones and Kuhn, “Dynastic Decline and the Roots of Rebellion,” 144; Dai, “Civilians Go
into Battle,” 163; Kaske, “Fund-Raising Wars,” 91.
110
Wang, Land Taxation, 72.
111
Shi, 清代戶部銀庫收支和庫存統計, pp. 197–8, 227–8; Ni, 清朝嘉道財政與社會,
pp. 166–7.
79

william t. rowe
Table 2.2 Silver revenue

collected at Board of
Revenue treasury, 1802–
1850
1802 11,496,754 taels

1811 9,448,666 taels
1821 7, 630,389 taels
1831 7,010,524 taels
1841 6,796,038 taels
1850 7,748, 585 taels
Within this total revenue, the land tax, the single largest component,
continued to increase modestly at least until the 1810s, and remained rela-
tively stable for several years thereafter. At least by the 1840s, however, most
provinces began to report massive shortfalls in land tax and grain tribute
collections relative to assessments. One of the most egregious was Hubei,
which in 1842 actually collected only 528,486 taels – merely 46 percent of its
assessment – and by 1850 only 334,179. Similar shortfalls occurred in the grain
tribute, the government’s second-largest contributor to total revenue in these
years. Official response in both cases was reform, not by increasingly effective
collection, but by lowering assessments.112
The third-largest contributor to Qing central revenues was the salt admin-
istration, which in the mid-eighteenth century yielded just under 9 million
taels, amounting to some 11.8 percent of central government revenue.113 Salt,
however, was among the first revenue sources to exhibit marked decline:
Peng Zeyi calculates a shrinkage of salt revenues by 13.3 percent over the
years from 1753 to 1812, even while land tax revenues increased by 10.9 percent
and total revenues by 6.3 percent.114 The basic problem was smuggling of
“private salt,” which had already become noticeable at the height of the
prosperous age, and which significantly worsened into the early nineteenth
century.
Most observers agreed that smuggling was a direct result of the inefficient
structure of the existing salt administration. The empire was divided into
several large salt districts for salt production and distribution. Within each
district, a limited number of enfeoffed “official-merchants” held franchises to
112
Xia, “太平天國前後長江各省之田賦問題,” 420. 113 Wang, Land Taxation, p. 72.
114
Peng, “鴉片戰後十年間銀貴錢賤波動下中國經濟與階級關係,” 51.
80

purchase from the producers and distribute very large annual quotas of salt,
via specified routes and to specified local markets. Successive price markups
along the way made the eventual price to consumers in many cases unaf-
fordably high. But there were alternative sources of salt: that privately
produced outside the state system, that illegally brought into one salt district
from a competing neighboring district, and, not least, that carried over quota
by government merchants themselves. All of these increasingly undersold
the legal “government salt,” and reduced the revenue that salt produced.
In 1832–1833 Liang-Jiang governor-general Tao Zhu undertook
a fundamental reform of operations in the Liang-Huai salt district, of which
he was collaterally commissioner. Liang-Huai was the most lucrative of the
several geographic districts into which the imperial salt administration was
divided. It was tasked with distributing the rich product of the salt fields in the
Yangzhou hinterlands throughout two subdistricts: Huainan, encompassing
portions of northern Jiangsu and most of Jiangxi, Hubei, and Hunan, and
Huaibei, covering most of Anhui and Henan. Tao’s reform lifted most of the
restrictions on salt distribution in Huaibei: (1) transport and sales were no
longer restricted to the 200-odd enfranchised merchants, but declared open to
any enterprising merchant whose integrity and financial soundness could be
ascertained, (2) to facilitate entry into the trade of these often smaller
merchants, units of salt for sale were drastically reduced in size to more
manageable job lots, which might be carried alongside commodities other
than salt; and (3) routes and marketplaces within districts were no longer for
the most part prescribed, and state-mandated brokers and other middlemen
were eliminated. Most contemporaries agreed that Tao’s reforms were an
unmitigated success, and they were soon imitated in some of the empire’s
other salt districts.115 By 1908, central salt revenues amounted to roughly 45
milllion taels, nearly five times what they had been in the mid-eighteenth
century.116 This was celebrated by contemporaries as evidence that fiscal and
economic reform could indeed be achieved, even in this age of manifest crisis.
115
Saeki Tomi 佐佰富, “清代道光朝における淮南鹽政的改革” (The Reform of the
Huainan Salt Administration during the Qing Daoguang Reign), in Saeki, 中國史研究
(Studies in Chinese History) (Kyoto, Tōyōshi kenkyūkai, 1969), 2.621–66; Saeki, 清代鹽
政の研究 (A Study of the Qing Salt Administration) (Kyoto, Tōyōshi kenkyūkai, 1956);
T.A. Metzger, “T’ao Chu’s Reform of the Huaipei Salt Monopoly (1831–1833),” Papers on
China 16 (1962), 1–39; Ni, 清朝嘉道財政與社會, pp. 276–311; Rowe, Speaking of Profit,
Chapter 5. For follow-up reforms affecting the Huainan portion of the Liang-Huai
district, see W.T. Rowe, Hankow: Commerce and Society in a Chinese City, 1796–1889
(Stanford: Stanford University Press, 1984).
116
Wang, Land Taxation, p. 72.
81

william t. rowe
In the shorter term, faced with growing imbalances between revenues and
expenditures, the Jiaqing court and its successors increasingly relied upon
improvised and irregular generators of income. Notable among these were
campaigns to solicit “contributions,” in exchange for granting of gentry
degrees or brevet (less often substantive) official rank. Sales of degrees and
ranks (juanna) had been utilized in the prosperous age itself to finance such
things as the Yongzheng Emperor’s massive buildup of reserves in the ever-
normal granary (changping cang 常平仓) system, and Qianlong’s prosecution
of the Jinchuan campaigns. But in the Jiaqing reign they became critical to the
regime’s very survival. The 1798 Sichuan and Hubei Reconstuction
Campaign (Chuan Chu shanhou juanshu 川楚善后捐输), enacted to finance
the protracted White Lotus war, attracted nearly 11 million purchasers and
netted 30 million taels, in Elisabeth Kaske’s words “the highest sum ever
reached in a single contribution campaign.” In Zhejiang province alone it
brought in just under 2 million taels, or approximately 90 percent of the
province’s annual land tax quota. Such ad hoc revenue-raising mechanisms,
along with proliferating non-statutory surtaxes, would become increasingly
central to government finance over the remainder of the nineteenth
century.117
What Happened in the Early Nineteenth Century?

Economic historians have been fairly unanimous in identifying the years
around 1800 as a turning point in Chinese economic history. Writing in the
Cambridge History of China in 2002, for example, Ramon Myers and Yeh-chien
Wang argued that it was then that “negative externalities,” such as environ-
mental decay, joined with “market failure,” monetary deflation, rising
unemployment, a growing tax burden, a cooling climate, and rising social
tensions to turn what had been until recently a prosperous economy into
a desperately struggling one.118 More recently, Richard von Glahn has like-
wise pinpointed the year 1800 as marking the transition from an early modern
era of economic “maturation” to one of wrenching “restructuring.” Though
skeptical about the claim of an impending Malthusian trap, he is inclined to
117
Kaske, “Fund-Raising Wars,” 91; L. Zhang, “Legacy of Success: Office Purchase and
State–Elite Relations in Qing China,” Harvard Journal of Asiatic Studies 73.2 (2013), 260,
270; Xu Daling 許大齢, 清代捐納制度 (The System of Sales of Degrees and Ranks in
the Qing Period) (Taipei, Wenhai chubanshe, 1977).
118
R.H. Myers and Y.-C. Wang, “Economic Developments, 1644–1800,” in W.J. Peterson
(ed.), The Cambridge History of China, vol. 9, part 1, The Ch’ing Empire to 1800 (New York,
Cambridge University Press, 2002), pp. 564, 640–1.
82

accept arguments for a declining per capita GDP, and emphatic that this era
witnessed the onset of a general “economic depression,” marked by “static or
declining real prices and wages,” lasting at least through the great rebellions
of the mid-nineteenth century.119
Kenneth Pomeranz has famously sought to place China’s economic experi-
ence of this period in comparative perspective, identifying it as that of
a “Great Divergence” from both the Western experience and its own pre-
sumed path of development. Pomeranz argued eloquently that the Qing
during the “prosperous age” enjoyed standards of living comparable or even
superior to those of Europe, and that there seemed little in Chinese economic
thought and institutions that would forecast a future markedly worse than
that of Europe. And yet that is precisely what happened. The reason, of
course, was industrialization in Europe and its absence in China, a divergence
Pomeranz attributes not to pathology but to “conjunctures,” the essentially
serendipitous convenient availability of fossil fuel deposits in Europe and the
exploitation of the New World (using unfree African labor) to remove the
“land constraint” that typically inhibited industrialization in the Qing and
elsewhere.
On the basis of what we have seen above, I am inclined to agree with the
general thrust of Pomeranz’s arguments (though not, for the most part, his
identification of the specific reasons for Qing divergence). Certainly, we can
all agree that the Qing did not undergo anything like industrialization before
this was introduced from outside. But if the Jiaqing and Daoguang reigns did
not see anything as revolutionary as industrialization, they did witness
a continued incremental development of the proto-industrial economy that
had been emerging at least since the sixteenth century: agricultural commer-
cialization and diversification, niche-seeking product specialization, exploit-
ation of new markets, forging of highly sophisticated distribution networks,
and proliferation of handicraft manufacture. All of this required a substantial
level of innovative entrepreneurship, which early modern China was hardly
lacking.
An outstanding but not unique example was the capital-intensive salt
industry of Furong, Sichuan. Building upon the highly sophisticated capital
mobilization instruments of the lineage trust (tang 堂) and the highly refined
regime of property rights and written contracts in late imperial business
culture, Furong merchants created and re-created partnerships of great
scale and achieved considerable vertical integration of fuel supply, salt
119
Von Glahn, Economic History of China, Chapter 9.
83

william t. rowe
extraction and refinement operations, and long-distance marketing. The real

takeoff of Furong’s salt industry came only in the 1850s, when Sichuan salt
captured much of the middle Yangzi market away from the rebel-plagued
Liang-Huai saltyards at Yangzhou. But Furong was able to do this so adroitly
because the organizational and technological structure had already been built
up, largely in the Jiaqing and Daoguang reigns.120
In this chapter we have concentrated on just a few prominent features
relevant to the Jiaqing and early Daoguang economy. Ecological deterior-
ation was widespread, severe, and largely man-made, a direct consequence of
developmental choices made during the “prosperous age.” Its first critical
manifestations appeared in this era, but worsened thereafter. None of the
ecological cataclysms of the Jiaqing and Daoguang reigns remotely
approached in severity those of the century to follow – the north China
famines of 1876–1879 and 1928–1930, and the great Yangzi flood of 1931 – in
each of which millions of persons lost their lives. This was in part because the
chief state bulwark against massive dearth, the granary system, continued to
function tolerably well in the early nineteenth century, forfeiting its utility
altogether only during the civil wars of the 1850s.121
On the question of demographic crisis, we would come down somewhere
between the pessimism of Hong Liangji and the optimism of Bao Shichen:
while there might not have been a looming Malthusian crisis, and there was
still room for per acre and even perhaps per capita productivity rises, it does
seem likely that continued population growth had some, if selective, negative
impact on standards of living.
There was a notable slowdown in the Qing economy beginning in the late
eighteenth century, but this was somewhat selective. If intra-regional trade in
grain on the Yangzi declined a little, this was almost certainly offset by
corresponding growth of other trade routes and commodities. Any overall
decline in long-distance domestic trade was offset by the rise in overseas
trade – not merely with the West but with Nanyang and other intra-Asian
trade partners.122 The development of new markets brought with it, as in the
middle Yangzi region, not devolution into dependent monoculture but
rather a vigorous diversification of product base. Neither did the Qing
120
M. Zelin, The Merchants of Zigong: Industrial Entrepreneurship in Early Modern China
121
P.-É. Will and R.B. Wong, with J. Lee, Nourish the People: The State Civilian Granary
System in China, 1650–1850 (Ann Arbor, Center for Chinese Studies, 1991), p. 75.
122
See, for example, figures on domestic versus maritime customs revenues in von Glahn,
Economic History of China, p. 370.
84

economy as a whole devolve into dependency on a foreign metropolis,

notwithstanding the steady growth in volume of maritime trade. After
c. 1825, economic “depression” certainly occurred, marked by the shortage
of silver, the skewing of bimetallic exchange rates, and the deflation of
commodity prices (especially those paid to farmers), but the more fundamen-
tal causes of the depression remain under debate; by the 1840s, and the
opening of new ports for foreign trade, the effects of this depression seemed
to be easing.
A severe fiscal crisis around the turn of the nineteenth century was only
gradually, and probably never completely, relieved. The short-term answers
included stopgap measures such as sales of offices and contribution drives,
and significant but piecemeal reforms such as those undertaken in the grain
tribute and salt administrations during the Daoguang reign. The long-term
solution was a wholesale fiscal restructuring, shifting the basis of state
finances from agriculture to commerce, but this only began to be enacted
in the 1850s, under the pressure of the Taiping wars. One is tempted to agree
with Ni Yuping’s conclusion that the late Qing state – far from achieving
a condition of lasting “sustainability” – never fully recovered the degree of
initiative, energy, fiscal sufficiency, and economic policy competence that it
sacrificed in the last decades of the eighteenth century.123 The mid-Qing state
was far from perfectly clear-sighted – its lapses in vision in large measure led
to the negative consequences of the post-“prosperous age” – and the Jiaqing
and Daoguang administrations did have their own moments of decisive
action. But clearly one of the factors that underlay the Great Divergence
was the declining leadership of the imperial state.
Further Reading
Chuan, H.-S., and R.A. Kraus, Mid-Qing Rice Markets and Trade: An Essay in Price History
(Cambridge, MA, Harvard University East Asian Monographs, 1975).
Gao Wangling 高王凌, 十八世紀中國的經濟發展和政府政策 (Government Policy
and Eighteenth-Century Chinese Economic Development) (Beijing, Zhongguo
shehui kexue chubanshe, 1995).
Huang Guosheng 黃國盛, 鴉片戰爭前的東南四省海關 (Maritime Customs in China’s
Four Southeastern Provinces Prior to the Opium War) (Fuzhou, Fujian renmin
chubanshe, 2000).
123
Ni, 清朝嘉道財政與社會, p. 377. See also the chapter by Elisabeth Kaske in the
present volume.
85

william t. rowe
Huang Jingbin 黃敬斌, 民生與家計：清初至民國時期江南居民的消費 (Popular

Livelihoods and Household Budgets: Jiangnan Residents’ Expenditures from Early
Qing to the Republican Era) (Shanghai, Fudan University Press, 2009).
Jones, S.M., and P.A. Kuhn, “Dynastic Decline and the Roots of Rebellion,” in J.K. Fairbank
(ed.), The Cambridge History of China, vol. 10, Late Ch’ing, 1800–1911, part 1 (New York,
Cambridge University Press, 1978).
Kuroda Akinobu 黑田明伸, “清代銀錢二貨制的構造とその崩壞” (The Structure
and Collapse of Silver–Copper Bimetallism in the Qing), 社會經濟史學 (Studies in
Socio-economic History) 57.2 (1992), 93–125.
Lee, J., and F. Wang, One Quarter of Humanity: Malthusian Mythology and Chinese Realities
(Cambridge, MA, Harvard University Press, 1999).
Li, Bozhong, Agricultural Development in Jiangnan, 1620–1850 (New York, St. Martin’s Press,
1998).
Ni Yuping 倪玉平, 清朝嘉道財政與社會 (Government Finance and Society in the Qing
Jiaqing and Daoguang Reigns) (Beijing, Commercial Press, 2013).
Peng Zeyi 彭澤益, “清代前期手工業的發展” (The Development of Handicrafts in the
Early Qing), 中國史研究 (Studies in Chinese History) 1981.1 43–60.
Perdue, P., Exhausting the Earth: State and Peasant in Hunan, 1500–1850 (Cambridge, MA,
Harvard University Press, 1987).
Pomeranz, K., The Great Divergence: China, Europe, and the Making of the Modern World
Economy (Princeton, Princeton University Press, 2000).
Rowe, W.T., Speaking of Profit: Bao Shichen and Reform in Nineteenth-Century China
(Cambridge, MA, Harvard University Asia Center, 2018).
Saeki Tomi 佐佰富, 代鹽政の研究 (A Study of the Qing Salt Administration) (Kyoto,
Tōyōshi kenkyūkai, 1956).
Suzuki Chusei, 鈴木中正, 清朝中時研究 (A Study of the Mid-Qing Period) (Toyohashi,
Aichi University Research Institute on International Problems, 1952).
Von Glahn, R., The Economic History of China: From Antiquity to the Nineteenth Century
(Cambridge, Cambridge University Press, 2016).
Wu Chengming 吳承明, “論清代前期我國國內市場” (The Domestic Market in the
Early Qing), 歷史研究 (Historical Studies) 1983.1, 96–106.
86

3
Agriculture
For over two millennia, China has sustained the largest single human society
on the planet through the development of one of the most sophisticated
agrarian systems in history. Even until quite recently, agriculture occupied
a central place in the Chinese economy, commanding a dominant 60 to
70 percent of the total economy. Agricultural institutions define the Chinese
economic system and agricultural production drove long-run economic
change or growth in China. Agriculture was at the center of the Great
Divergence debate. Agricultural harvest successes or failures sometimes
spelled the rise and fall of dynasties throughout history. Moving to the
modern era, Chinese agriculture became the scapegoat for China’s modern-
ization failure and was regarded as the incubator for Communist revolution.
However, given its overriding importance, research on modern Chinese
agriculture has been surprisingly understudied for the last few decades.
This chapter provides a selective survey of some of the key themes in
modern Chinese agriculture during the nineteenth and twentieth centuries. It
provides a critical survey on the general trends in agricultural output during
this period and existing interpretations. The deteriorating land–labor ratio
over time has easily given rise to a Malthusian type of interpretation of
Chinese agriculture in the form of the so-called “high-level equilibrium
trap” hypothesis proposed by Mark Elvin and later the “involution” thesis
according to Philip Huang. Both posit a largely pessimistic vision of a long-
run decline in agricultural productivity and output per capita in the face of
resource constraints and overpopulation over time. However, the more
optimistic vision, as recently championed by the California school, posits
that Chinese agricultural expansion achieved efficiency from gains in the
use of better fertilizers, rationalization of resource use, agricultural
We wish to thank Matthew Noellert and Thomas Rawski for very useful comments.
87

intensification, and cash crop cultivation. This technical bias induced by the
lower Yangzi’s relative factor endowment, combined with the expansion in
regional trade and geographic division of labor, constituted what they viewed
as Smithian growth.1 Our survey goes on to reconstruct a model that recon-
ciles these relatively pessimistic and optimistic assessments of the Chinese
mode of agricultural production. Based on that, it argues that traditional
Chinese institutions have shown remarkable resilience to adapt to the intro-
duction of new crops, heightened commercialization, and technological and
organizational innovation.
Long-Run Trends in Agricultural Output

Labor-intensive cultivation has been a long-standing feature of Chinese and
East Asian agriculture for millennia, marked by small-scale family farms, very
high labor input, and land productivity. China’s first large-scale survey, by the
team led by John Lossing Buck of Nanjing University in the 1920s and 1930s,
revealed a mean farm size of a mere 3.76 acres compared with 14.28, 63.18, and
156.85 acres for the Netherlands, the UK, and the US respectively. Farm sizes
in China’s northern wheat region were slightly larger than in the southern
rice region, standing at 5.63 and 3.14 acres respectively.2 John Buck’s study also
mapped out the main crop mix across China’s twenty-two provinces in the
1930s, as shown in Map 3.1.
The most typical form of this labor-intensive cultivation (精耕细作) is the
highly developed rice culture in the lower Yangzi region, where the so-called
“ten mu per family” (one mu = 0.16 acre) came into firm shape at the beginning of
the nineteenth century.3 Intensive agriculture has given rise to very high land
productivity, probably matching some of the highest in the most advanced
agricultural regions employing modern technology even in the early twentieth
1
The articles of debate can be found in the May 2002 (61.2) issue of the Journal of Asian
Studies. For the earlier debate between Philip Huang, Thomas Rawski, and Ramon
Meyers around the 1990s on whether or not there were improvements in agricultural
productivity and living standards in Chinese agriculture for the early twentieth century,
see Chapter 6 of P. Richardson, Economic Change in China, c. 1800–1950 (Cambridge,
2
J.L. Buck, Land Utilization in China (Nanking, University of Nanking Press, 1937), p. 268.
For per capita acreage between the eighteenth and early twentieth centuries, see
Shi Zhihong, 史志宏, 清代农业的发展和不发展 (1661–1911 年) (Development or
Non-development of Qing Agriculture: 1611–1911) (Beijing, Shehui kexue wenxian chu-
banshe, 2017), pp. 141–3.
3
Li Bozhong 李伯重, 多视角看江南经济史 (1250–1850) (An Economic History of
Jiangnan in Multiple Perspectives (1250–1850)) (Beijing, Sanlian shudian, 2003), pp.
241–88.
88

Map 3.1 Agricultural areas of China
Source: P. Richardson, Economic Change in China, c. 1800–1950 (Cambridge, Cambridge University Press, 1990), p. 71
century. However, labor productivity would be far lower in that regard.4 Here,
we turn to examine long-term agricultural output trends under China’s land
cultivation system for the nineteenth and twentieth centuries.
Long-Term Agricultural Decline: Fact or Fiction?

Figure 3.1 shows an upward trend in long-term crop yields between the
sixteenth century and the turn of the nineteenth century, both in the
Jiangnan region and nationally. However, after about 1800, all series except
Perkins’s (1969) show a sharp downturn. Indeed, the estimates by Wu (1985,
1998) show that the decline in crop yield for the few decades after the mid-
eighteenth century actually wiped out all the gains accumulated during the
previous centuries. Figure 3.2 shows this decline in yields that accounted for
a corresponding decline in various per capita grain output estimates from the
early nineteenth century, and ultimately per capita GDP estimates in the face
of rising population.5 Again, with the exception of Perkins’s estimates, these
declines continued after the mid-nineteenth century even when there occurred
catastrophic population loss during the 1860s Taiping Rebellion and the gradual
modernization of Chinese agriculture from the late nineteenth century.
How well do these estimates of declining land yields and per capita grain
output stand scrutiny? One thing to bear in mind is that the crop yield data for
the 1930s and 1950s derived from large-scale or nationwide statistical surveys are
of much better quality than estimates for earlier periods. Therefore it is striking
to see that all these studies cited above, except Perkins (1969),6 show crop yield
and per capita grain output estimates for the peak period of the Qing much
higher than for the 1930s and 1950s, leading to a possible suspicion of sample
selection bias in terms of crop yield data for the pre-1930s period. As we will
show later, 1930s levels of per capita grain output were high even compared
with the Communist period or the early reform era of the 1980s. It is also
possible that the peak value for grain output per capita for Qing China has been
overestimated, and that its actual decline during and after the nineteenth
century was more gradual and slower than estimated by Shi and Guo.
4
See J.L. Buck, Land Utilization in China (Nanking, University of Nanking, 1937), p. 226;
D.H. Perkins, Agricultural Development in China, 1368–1968 (Chicago, Aldine Publishing,
1969), Chapter 2. Also see R.C. Allen, “Agricultural Productivity and Rural Incomes in
England and the Yangtze Delta, c. 1620–c. 1820,” Economic History Review 62.3 (August
2009), pp. 525–50.
5
For long-term GDP estimates, see S. Broadberry, H. Guan, and D.D. Li, “China, Europe
and the Great Divergence: A Study in Historical National Accounting, 980–1850,” Journal
of Economic History 78.4 (2018), 955–1000.
6
These in-text citations are given in full in Figure 3.1.
90

Wu (1985, 1998)
Rice (1950s, Shanghai)

Chao et al. (1995) & Chao (2001)
Shi (2015) Li (2003, 2010)
400
Perkins (1969)
Guo (2001)
Yield (Raw grain, kg per mu)
200
Rice (1950s)
1930s
100 1950s
50
1450 1550 1650 1750 1850 1950
Year
Figure 3.1 China’s historical crop yield
Notes: (1) Except for rice yield in the 1950s, crop yield per mu in this figure refers to yield
averaged by agricultural acreage, while sown acreage, equal to multiplying agricultural
acreage by the multiple-crop index, is more used in China’s current statistics.
(2) Li’s (2003; 2010) figures are for the Yangzi delta only, while all other series are national
estimations.
(3) Chao et al. (1995) estimates average land yield from 1750 to 1900, and Chao (2001)
estimates land yield during the Song dynasty and links it to 1750.
(4) For comparison, estimations based on statistical surveys in the 1930s and 1950s are
labeled specifically. Yield in the 1950s is the average of yields from 1952 to 1957, and the
estimation method is described in Peng (2015).
Sources: D.H. Perkins, Agricultural Development in China, 1368–1968 (Chicago, Aldine Publishing,
1969); Wu Hui 吴慧, 中国历代粮食亩产研究 (Study of Grain Yield per Mu in Chinese
History) (Beijing, Zhongguo nongye chubanshe, 1985); Wu Hui, “历史上粮食商品率商品
量测估—以宋明清为例” (Estimates of Ratio and Quantity of Food Commercialization:
Song, Ming, and Qing Case Study), 中国经济史研究 (Research in Chinese Economic
History) 1998.4, 16–36; Kang Chao 赵冈 et al., 清代粮食亩产量研究 (Study of Grain Yield
per Mu in the Qing Dynasty) (Beijing, Zhongguo nongye chubanshe, 1995); Kang Chao, 农业
经济史论集：产权、人口与农业生产 (Collection of Papers on Agricultural Economic
History: Property Rights, Population, and Production) (Beijing, Zhongguo nongye
chubanshe, 2001), p. 173; Guo Songyi 郭松义, “明清时期的粮食生产与农民生活水平”
(Food Production and Living Standard of Peasants in the Ming and Qing Dynasties), 中国社
会科学院历史研究所学刊 (Journal of the CASS Institute of History) 2001.1, 373–96; Li
Bozhong 李伯重, 多视角看江南经济史 (1250–1850) (An Economic History of Jiangnan in
Multiple Perspectives (1250–1850)) (Beijing, Sanlian shudian, 2003), pp. 83, 128, 326–7; Li
Bozhong, 中国的早期近代经济—1820年代华亭 – 娄县地区 GDP 研究 (An Early
Modern Economy in China: A Study of the GDP of Huating–Lou Area, 1823–1829) (Beijing,
Zhonghua shuju, 2010), pp. 395–8; Shi Zhihong 史志宏, “清代农业生产指标的估计”
(Estimates of Indicators of Agricultural Production in the Qing Dynasty), 中国经济史研究
(Research in Chinese Economic History) 2015.3, 5–30; Z. Shi, Agricultural Development in Qing
China: A Quantitative Study, 1661–1911 (Leiden, Brill, 2017); Peng Kaixiang 彭凯翔, “人口增长
下的粮食生产与经济发展—由史志宏研究员的清代农业产出测算谈起” (Grain
Production and Development under the Growth of Population: A Discussion from Shi
Zhihong’s Estimation of Agriculture Production in the Qing Dynasty) 中国经济史研究
(Research in Chinese Economic History) 7 (2015), 38–49

750 Wu (1985, 1998)

Shi (2015)
700 Guo (2001)
Per capita grain (Raw product, kg) Perkins (1969)
650
600
550
500
450
400 1930s, rural

consumption
350
1950s
300
250
1450 1550 1650 1750 1850 1950
Year
Figure 3.2 Per capita output of raw grain, 1500–1950

Notes: for the discussion of 1930s rural grain consumption, see later section on consistency
checks. Sources for other estimates are the same as for Figure 3.1.
Take the example of the estimates by Chao et al. (1995), which provides the
most systematic evidence of this decline based on the Qing primary sources.
Chao et al. calculated a long-term annual decrease of 0.2 percent between 1750
and 1900, as shown in Figure 3.1. Their nationwide average yield between 1750
and 1900 of 2.3 dan (about 177 kilograms) per mu is slightly lower than the land-
tax-based yield estimate made by Wu (1985; 1998) but somewhat higher than
that made by Shi Zhihong for the first half of the nineteenth century. But
Chao et al.’s estimate of land yield was derived by doubling the rental share
rate recorded from the land tax account book by assuming a constant 50/50
share of rent, a practice also followed by Wu (1985; 1998), Shi (2017), Guo
(2001), and others. This reliance on land rent data, due to their abundance and
low variation in land rent, is understandable, but the assumption of a constant
ratio of land rent to yield is highly problematic for the entire period of study.
Indeed, given the increasing trend of urban-based landlords, who were becom-
ing increasingly detached from their land, and the increasing switch from
sharecropping to fixed land rents, tenants’ control over their land strengthened
over time with landlords’ share of the output possibly going down and
correspondingly reducing their investment in land improvement.7 Hence the
assumption of a constant ratio could lead to a downward bias in crop yield.
7
Gao Wangling 高王凌, “关于近代粮食亩产量的估算问题—清代粮食亩产量研究
>读后” (On the Estimation of Modern Grain Yield per Mu: Review of Study on Grain
92

Agriculture
Table 3.1 Mean and distribution of agricultural yields (shi per mu) in Qing
China
Region Period obs mean median sd min max
Northern China 1644–1795 (early Qing) 264 0.71 0.55 0.62 0.02 3.40
Northern China 1796–1850 (mid-Qing) 56 0.70 0.58 0.65 0.06 3.25
Northern China 1851–1911 (late Qing) 137 0.64 0.50 0.57 0.06 4.00
Southern China 1644–1795 (early Qing) 801 2.72 2.38 1.62 0.26 18.00
Southern China 1796–1850 (mid-Qing) 555 3.00 2.65 1.70 0.40 13.66
Southern China 1851–1911 (late Qing) 473 3.03 2.67 1.65 0.38 12.50
Note. According to Shi, Agricultural Development in Qing China, northern China

includes 东北 the Northeast, 甘肃 Gansu, 河南 Henan, 山东 Shandong, 山西
Shanxi, 陕西 Shaanxi, and 直隶 Zhili. Southern China includes 安徽 Anhui, 福建
Fujian, 广东 Guangdong, 广西 Guangxi, 贵州 Guizhou, 湖北 Hubei, 湖南 Hunan,
江苏 Jiangsu, 江西 Jiangxi, 四川 Sichuan, 云南 Yunnan, and 浙江 Zhejiang.
Here we will conduct some cross-checks and tests by looking at nearly

3,000 cases of yield data collected from local gazettes by Shi.8 One advantage
of these yield data is that they are mostly for “public land” (gongtian 公田),
with much simpler and more stable rental contractual relationships than
those used in land rent account books. These records also include data from
places like Huizhou, where the rental record books and land transaction
books showed a declining yield in late Qing given by other studies. Table 3.1
shows that the mean and distribution of land yields across the three different
periods of Qing are quite close. Actually, land yields in mid- and late Qing
were slightly higher in southern China. The standard deviation of land yields
in northern China shrank because of the increase in sample size. As shown
later, the New World crops, which diffused rapidly in the late Qing, were not
recorded, so the actual distribution of land yields in late Qing could tilt more
towards the right. All in all, these micro-level data do not support declining
land yields in mid- and late Qing.9
Yield per Mu in the Qing Dynasty”), 中国经济史研究 (Research in Chinese Economic

History) 2000.2, 156–60.
8
Shi Zhihong, Agricultural Development in Qing China: A Quantitative Study, 1661–1911
(Leiden, Brill, 2017).
9
To justify their findings, Chao et al., Wu, and others have listed factors such as soil
degradation, climate change, and environmental damage caused by the introduction
of maize into China. Kang Chao 赵冈 et al., 清代粮食亩产量研究 (Study of Grain
Yield per Mu in the Qing Dynasty) (Beijing, Zhongguo nongye chubanshe, 1995);
Wu Hui 吴慧, 中国历代粮食亩产研究 (Study of Grain Yield per Mu in Chinese
History) (Beijing, Zhongguo nongye chubanshe, 1985). For a different interpretation
93

Figure 3.3 Distribution of Qing crop yields based on local gazettes

Source: Shi, Agricultural Development in Qing China.
Consistency Checks from the Consumption and Demand Side

We now turn to some cross-checks from consumption and market prices. For
example, Wu himself estimates that historically (mostly in the Ming and Qing
eras) the farming and non-farming population consumed, on average, 17.5 kilo-
grams and 16.4 kilograms of processed rice per person per month, respectively.
With about 70 percent of the total population engaged in farming, this amounts
to an aggregated average consumption per capita of 17.2 kilograms per month
(0.575 kilogram per day for thirty days per month), or 206.5 kilograms per year.
Adding non-staple foods, seeds, fodder, etc., which, as Wu suggests, constituted
a 30 percent share of total grain output, we can derive a per capita grain
consumption of 295 kilograms per year.10 This level is consistent with the
307 kilograms of processed grain consumption per capita for rural households
for the 1930s, but higher than the national average in the 1950s,
on the environmental damage caused by the introduction of maize into China, see
Wang Baoning 王保宁 and Zhu Guangyong 朱光涌, “从抵制到接受: 清代浙江的
玉米种植” (From Resistance to Acceptance: Maize in Zhejiang Province in the
Qing Dynasty)”, 中国历史地理论丛 (Journal of Chinese Historical Geography),
2019.1, 108–17. We will examine environmental issues below.
10
Wu, 中国历代粮食亩产研究, 64–8.
94

Agriculture
which was about 297 kilograms of raw grain or 208 kilograms of processed
grain.11 Note that the 1930s level of per capita grain consumption, based on
a large-scale survey, is not exceeded until the late 1980s. Furthermore, the
reported per capita consumption for meat and eggs was sixteen kilograms
per year in the 1930s, similar to the average consumption level in 1978.
Similar studies also show that daily calorie and protein intake reached
about 3,000 to 4,000 calories and 100 grams during the 1930s.12 All these
consumption-side estimates match the 1930s per capita grain output levels
better than they do those from before the nineteenth century, as shown in
Figure 3.2.
If there was a drastic decline in per capita grain output and consump-
tion in the late nineteenth century, there would have also been
a corresponding decline in real wages as well as large shifts in relative
prices between wages (prices of labor-intensive goods) and grain prices.
However, based on the limited regional wages and prices, we do not
observe such a decline for the set of relative prices. Figure 3.4 presents
a set of wages for agricultural, handicraft and industrial laborers in Beijing
and other regions relative to grain price. Although scattered and some-
what volatile, none of them show a persistent downward trend. Data
elsewhere on relative prices also reveal that rural and urban wages kept
pace with other items of daily living such as cotton cloth, fuel, vegetables,
meat, and housing.13
11
Peng, “人口增长下的粮食生产与经济发展.”
12
Data for the PRC period come from 中国农业统计资料汇编 1949–2004 (Statistics on
Chinese Agriculture, 1949–2004) (Beijing, Zhongguo tongji chubanshe, 2005). Data for
the 1930s come from 农情报告 1937–38 (Crop Reporting in China, 1937–38) (Nanjing,
National Agricultural Research Bureau, Ministry of Industry); Qiao Qiming 乔启明, 中
国农村社会经济学 (Social Economics in Rural China) (Shanghai, Shangwu
Yinshuguan, 1945), pp. 410–18.
13
R.C. Allen, C. Moll-Murata, D. Ma, and J. van Zanden, “Wages, Prices, and Living
Standards in China 1738–1925: In Comparison with Europe, Japan, and India,” Economic
History Review 64.S1 (2011), 8–38. Peng Kaixiang 彭凯翔, 清代以来的粮价: 历史学的解释
与再解释 (Rice Prices since the Qing Dynasty: A Historical Interpretation and
Reinterpretation) (Shanghai, Shanghai renmin chubanshe, 2006), pp. 32–41. Peng Kaixiang
彭凯翔, 从交易到市场—传统中国民间经济脉络试探 (From Dealings to the Market:
A Discussion of the Private Economy of Traditional China) (Hangzhou, Zhejiang daxue
chubanshe, 2015), pp. 90–101, 302–12. Zhang Li 张丽, 非平衡化与不平衡: 从无锡近代农
村经济发展看中国近代农村经济的转型 (Unbalanced and Imbalance: The
Transformation of China’s Modern Rural Economy from the Perspective of Wuxi’s
Modern Rural Economic Development) (Beijing, Zhonghua Shuju, 2010), pp. 16–18, 327–
45. T.G. Rawski, Economic Growth in Prewar China (Berkeley, University of California Press,
1989). L. Brandt, Commercialization and Agricultural Development: Central and Eastern China,
1870–1937 (Cambridge, Cambridge University Press, 1989).
95

300
Relative wage index 250
200
150
100
50
0
1800 1820 1840 1860 1880 1900 1920 1940
Year
Relative farm wage, Beijing, 1901 = 100 Relative carpenter wage, Beijing, 1901 = 100
Relative farm wage, Yanshan, 1914 = 100 Relative farm wage, Jing Country, 1912 = 100
Relative carpenter wage, Jing Country, Relative coalminer wage, Kailuan, 1929 = 100
1912 = 100
Figure 3.4 Wage index relative to grain price in Beijing and Hebei province
Notes: Most data are indexed with different base years. Kailuan prices are for flour and
other series are for wheat.
Sources: Beijing: Peng Kaixiang 彭凯翔, 从交易到市场—传统中国民间经济脉络试探
(From Dealings to the Market: A Discussion of the Private Economy of Traditional China)
(Hangzhou, Zhejiang daxue chubanshe, 2015), pp. 95–7, 304–5.
Jing county: 景县志 (Jing County Annals) (1932), 中国地方志集成·河北府县志辑
Collection of Chinese Local Chronicles, Hebei Section, 50 (Shanghai, Shanghai shudian
chubanshe, 2006), vol. 6, pp. 4–6. See Peng, 从交易到市场, pp. 84–5, 194–5, for details
about silver–copper exchange in conversion.
Yanshan: John Lossing Buck 卜凯, 河北盐山县一百五十农家之经济及社会调查
(Economic and Social Survey of 150 Farmers in Yanshan County, Hebei Province)
(Nanjing, Jinling daxue nonglinke, 1929), pp. 160–1.
Kailuan: Nankai daxue jingji yanjiusuo jingjishi yanjiushi 南开大学经济研究所经济史
研究室, 旧中国开滦煤矿的工资制度和包工制度 (Wage System and Labor Contract
System of Kailuan Coal Mine before the PRC) (Tianjin, Tianjin renmin chubanshe, 1983),
p. 123. The series is based on real wages of coal miners.
In short, the case for a Malthusian trap or decline for Chinese agriculture in
the nineteenth and twentieth centuries seems far from being established
based on existing studies of land yield and per capita grain output. While
this chapter does not reconstruct new estimates, it may be fair to say that
while Perkins (1969) might have given a “subsistence-level” estimate for per
capita output, other scholars’ estimates have probably reflected the product-
ive capacity of China’s arable land. The two groups capture the lower and the
upper bounds of grain output respectively.
96

Agriculture
Explaining Per Capita Output Stability

If our survey here casts doubt on a long-term decline, how do we explain the
relative stability throughout this period, and how did Chinese agriculture
keep pace with population increase given the land constraints of the nine-
teenth and twentieth centuries? Per capita grain output is derived from
multiplying yield by total acreage divided by total population. As is well
known, historical data on population and acreage are heavily contested.
Although it may be difficult to ever arrive at precise historical estimates,
there may well be various mitigating factors to population and acreage data
that prevented a Malthusian fall in per capita grain output.
On total land acreage, Ho observes that the arable land was substantially
underreported in official historical documents.14 For this reason, Shi uses the
modern-era land data published in 1952 by the National Bureau of Statistics of
China as a reference point to go back and re-estimate aggregate arable land
areas for the Qing and modern periods.15 He concludes that there were
probably 1.2 billion mu, or 50 percent, more arable land than was officially
recorded in the late Daoguang reign (1821–1850), and that agricultural acreage
increased from 0.83 billion mu to 1.46 billion between early seventeenth-
century Ming and the late nineteenth century, an increase of 76 percent.16 Shi
goes further to say that as the 1952 land data were underassessed as well, his
estimates for the Qing period should be on the conservative side.17
Shi argues that by the mid-Qing, cultivation peaked and reached its satur-
ation point in inland core agricultural regions. The pace of increase slackened
between the early nineteenth century and 1911 in China proper, but growth in
southwestern China remained at a high level. Moreover, the opening of new
frontiers in northeastern China, Mongolia, and Xinjiang raised the total
acreage by another 0.3 billion mu.18 It is likely that agricultural acreage
14
Ping-ti Ho 何炳棣, 中国古今土地数字的考释与评价 (Interpretation and Evaluation of
Ancient and Modern Land Figures in China) (Beijing, Zhongguo shehui kexue chubanshe,
1988).
15
Shi, Agricultural Development in Qing China.
16
Shi, 清代农业的发展和不发展, pp. 130–5.
17
In fact, until the late twentieth century, a general practice was to state “1 billion mu” to
describe the aggregate arable land amount of China. But, systematic census and
satellite-based assessment put the actual total above 2 billion mu, catching everyone
by surprise.
18
Cao estimated that the total population increased from 383 million to 436 million during
1820–1911, with population in northeast China rising from 2.49 million to 18 million,
which is about 30 percent of the total population increase. See Cao Shuji 曹树基, 中国
人口史 (Population History of China), vol. 5, 清时期 (Qing Dynasty) (Shanghai, Fudan
daxue chubanshe, 2001); updated estimations can be found in Chapter 8 by Cao Shuji in
our volume I.
97

increases largely slowed or stagnated in the early twentieth century given the
civil strife and natural disasters.19 As Xin finds, there were substantial unculti-
vated areas around the Three Gorges on the Yangzi river even by the time of
the late nineteenth century. Cultivatable land was not totally exhausted until
the 1930s.20 There is also the possibility of exploration of hilly or marginal land,
more efficient utilization of agricultural land, innovation in agricultural tech-
nology and crop varieties, and so on, a point we will elaborate on later.21
The second factor is obviously population. The eighteenth-century popula-
tion explosion inevitably led to a dramatic decline in the land–labor ratio in
agriculture and a potential food crisis. However, the nineteenth century saw
a dramatic downturn in the fourteen-year-long Taiping Rebellion (1851–1865) that
led to a population loss of 70 million, mostly in China’s most densely settled and
developed area of the lower and middle Yangzi. This was followed by another
population loss exceeding 10 million in the Muslim rebellion in northwestern
China (1862–1873).22 Whether or not these disasters can be characterized as
Malthusian is another question, but they certainly leveled the population to
that of the early nineteenth century and temporarily relieved land constraints.
Another notable cause of population loss is natural disaster. Figure 3.4 shows
that the period under study had one of the highest frequencies of natural disasters
documented in recent studies. Although the increasing frequency of disaster may
well be due to better records during this period, it may also be closer to reality.
The most prominent “Ding-wu Disaster” (丁戊奇荒) of 1876–1879 and 1928–1930
led to population losses, respectively, of as much as 20 million and 10 million in
northern China and nationwide. Figure 3.5 shows that among all the natural
disasters, drought and flood were more frequent and caused possibly more
human loss compared with others such as pests, earthquakes, typhoons, or the
cold. It is important to note that apart from these large-scale disasters, most
disasters are fairly localized and of limited duration. Hence their damage may be
limited. We have evidence from Huizhou land rental data showing that
most disasters in the second half of the nineteenth century remained
19
Yan Zhongping 严中平 et al., 中国近代经济史统计资料选辑 (A Selection of
Statistical Data of the Modern Economic History of China) (Beijing, Zhongguo shehui
kexue wenxian chubanshe, 2012), p. 238.
20
Xin Deyong 辛德勇, 历史的空间与空间的历史 (The Space of History and the
History of Space) (Beijing, Beijing shifan daxue chubanshe, 2005), pp. 36–7.
21
Han Maoli 韩茂莉, 中国历史农业地理 (Historical Agricultural Geography of China)
(Beijing, Beijing daxue chubanshe, 2012), pp. 5–10; J.L. Buck (ed.), Land Utilization in
China: Statistics (Nanjing, Jinling daxue chubanshe, 1937), pp. 40–4.
22
Cao, 中国人口史, vol. 5, 清时期, pp. 455–689.
98

Agriculture
Northwest China famine, 1928–30

10000 North China famine, 1876–79
Calamity
Drought/flood
Henan famine, 1942–43
Other
Death toll (1,000)
1000
Death toll (1,000)

2500
5000
100 7500
10000
12500
10
1850 1875 1900 1925 1950
Year
Figure 3.5 Major disasters with a death toll above 1,000 (1840–1949)
Source: Xia Mingfang 夏明方, 民国时期自然灾害与乡村社会 (Natural Disasters and
Rural Society during the Republic of China) (Beijing, Zhonghua shuju, 2000), pp. 395–403.
localized.23 Also, by the early twentieth century, as Hou Yangfang shows, public
health, railroads, and, to a lesser extent, the rise of relief organizations provided
much more effective relief efforts and possibly prevented much larger population
loss compared with earlier periods.24
Models of Chinese Agriculture: High-Level

Equilibrium Trap, Involution, or Industrious
Revolution?
What are the micro-factors in terms of the choice of technology, combination of
factor use, and household production and management decisions that underlie
23
See Jiang Taixin 江太新 and Su Jinyu 苏金玉, “论清代徽州地区的亩产” (On the
Yield per Mu of the Huizhou Area in the Qing Dynasty), 中国经济史研究 (Research in
Chinese Economic History) 3 (1993), 36–61. It provides rent collection records of forty-
seven pieces of land from 1883 to 1908 to show the effect of disasters. However,
correlation between most of the forty-seven areas was very weak, which shows that
the effect of disasters was localized even within one region.
24
Hou Yangfang 侯杨方, 中国人口史 (Population History of China) (1910–1953), vol. 6
(Shanghai, Fudan daxue chubanshe, 2001), pp. 587–610.
99

the long-run macro trends of Chinese agricultural output and productivity? The
deteriorating land–labor ratio over time easily gave rise to a Malthusian-type
interpretation of Chinese agriculture in the form of the hypothesis of the so-called
“high-level equilibrium trap” proposed by Mark Elvin and later the “involution”
thesis according to Philip Huang. Both posit a largely pessimistic vision of a long-
run decline in agricultural productivity and output per capita in the face of
resource constraints and overpopulation over time. However, the more optimis-
tic vision recently championed by the California school posits that Chinese
agricultural expansion, particularly in the highly developed lower Yangzi area,
proceeded in a different technological and institutional trajectory from the well-
known British or Western European model. In agriculture, efficiency came from
gains in the use of better fertilizers, the rationalization of resource use, agricultural
intensification, and cash crop cultivation. This technical bias, induced by the lower
Yangzi’s relative factor endowment, and combined with the expansion of regional
trade and the geographic division of labor, constituted what has been viewed as
Smithian growth. To a certain degree, the California school thesis is a variation of
the Boserupian classification (according to Ester Boserup) of long-term economic
growth which viewed resource constraints more optimistically as a stimulus to
technical change and intensification.25 Here, by expanding and developing an
analytical framework originally derived from Mark Elvin and Philip Huang, but
more rigorously formulated by Kang Chao, we show that a more rigorous
Boserupian framework of innovation can reconcile the two opposing sides of
the debates and reveal the resilience of Chinese agriculture beyond the
Malthusian trap.
From a Malthusian to a Boserupian Framework

In Figure 3.6(a), we show a production function where total output (TP)
reaches the highest average output (AP) at point A given the fixed land
constraint. In a Malthusian model, whenever AP exceeds average subsistence
cost (indicated by the forty-five-degree line), it will induce a population rise to
25
It is regrettable that the recent heated debate between the “involutionists” and the
California school has taken few cues from the extensive induced-innovation literature.
The induced-innovation theory, a more rigorous formulation of the Boserupian thesis,
was statistically tested in seminal works by Hayami Yujiro and Vernon Ruttan in the
1980s to explain the successful economic growth of modern Japan under severe factor-
endowment constraints. It clearly revealed the insufficiency of the simplistic Malthusian
framework, which ignored the potentials of factor-biased technological progress, factor
substitution, and the expansion of trade based on comparative advantage that would
prevent the fall in marginal productivity of labor and release the factor endowment
constraints. See Y. Hayami and V. Ruttan, Agricultural Development: An International
Perspective (Baltimore, Johns Hopkins University Press, 1985).
100

Agriculture
Subsistence cost
Production
Production
D TP′
MP
C TP MP′
B
A D
A C AP′
AP
B Ave. subsistence cost
Labor force Labor force

Figure 3.6 Malthusian and Boserupian models
eat up the surplus, leading to new equilibrium AP at point B, a point that

approximates the so-called high-level equilibrium trap condition. As men-
tioned by Kang Chao, given inequality, the actual Malthusian equilibrium
point tilts towards the left of B.
In the case of modern agricultural improvement, the Malthusian trap or
equilibrium is relieved through technical innovation of the labor-saving and
(land) resource-using type (as in the case of the British or American type of
agriculture, which uses inanimate power sources and mechanization). This
can be indicated by a general upward shift of the production function that
raises both marginal product (MP) and AP over time. Clearly, Chinese or East
Asian agriculture under severe land constraint did not take that direction.
Instead, we can show a Boserupian type of innovation in the adoption of
labor-using technology and institutions that shift TP to TPʹ in Figure 3.6(a). In
the Boserupian model, when population reaches point C, people can opt out
of a new production function in the form of TPʹ. TPʹ is not a general upward
shift of production function or an increase in total factor productivity across
the board, but it does increase total output and absorb more labor with the
much larger population size.26 By moving AP from A to D, the Boserupian
innovation temporarily releases the system from the Malthusian threat
through a combination of intensive cultivation and a greater degree of the
division of labor and commercialization in response to rising population
density.
In Figure 3.6(b), we illustrate the changes using marginal and average product
curves. The shift from TP to TPʹ leads to a shift towards APʹ and MPʹ. If the
Boserupian innovation is continuous, it could lead to multiple or constant shifts of
TP to TPʹ, which can trace out a more general production function that is the
envelope of TP and TPʹ. This envelope expands the production frontier and
26
See K. Chao, Man and Land in Chinese History: An Economic Analysis (Stanford, Stanford
101

improves efficiency even though APʹ at point D may be lower than AP at point
A (although this is theoretically ambiguous). As you can see, APʹ and MPʹ at point
D represent a drastic improvement over AP and MP had TP not shifted to TPʹ,
which represents the dire scenario of the Malthusian trap or involution under a
much larger population size and a deteriorating land–labor ratio.
The Boserupian innovation leading to the movement from MP to MPʹ is
realized through several channels: the greater application of irrigation, fertil-
izer, and intensive cultivation, which raised land productivity of a single crop;
the intensification of multi-cropping and crop rotation; the introduction of
more labor-intensive, profitable cash crops; and finally the shift to agricultural
and handicraft sideline production which relies less on the use of land. This is
sometimes alternatively called Smithian growth.
By the nineteenth century, the potential for raising land yield on a single
crop was largely exhausted within China’s long-established traditional agri-
cultural regions. Although there are slight differences in the application of
draft animals and fertilizer, various studies show that yield based on single
cropping shows small variation with farm size by the 1930s.27 Based on the
study by John Buck, Table 3.2b shows that smaller farm size tended to have
a slightly higher double-cropping index and higher yield, but the difference is
not overwhelming. On the other hand, the increasing share in the cultivation
of cash crops such as peanuts, tobacco, and American cotton due to height-
ened commercialization, as shown later, could generate higher revenue,
flatten the marginal revenue, or push it further outward.
We can see in Table 3.2 that average labor productivity seems to be rising
with farm size. In other studies in northern China and the lower Yangzi, we
also see labor productivity and net income increase with farm size before it
reaches a certain scale. These all seem to indicate that the impact of techno-
logical adjustment in the labor-intensive direction is limited, and support the
involution thesis proposed by Philip Huang.28 However, this argument needs
to be qualified on several theoretical and empirical grounds. First, the
efficiency argument based on average product rather than marginal product
27
These new findings based on the micro-level data used by John Buck can be seen in
H. Hoken and Q. Su, “An Analysis on the Inverse Relationship between Yield and Farm
Size in Rural China in the 1930s,” in H. Hu, F. Zhong, and C.G. Turvey (eds.), Chinese
Agriculture in the 1930s: Investigations into John Lossing Buck’s Rediscovered “Land Utilization
in China” Microdata, https://doi.org/10.1007/978-3-030-12688-9.
28
P. Huang, The Peasant Economy and Social Change in North China (Stanford, Stanford
University Press, 1985). Cao Xingsui 曹幸穗, 旧中国苏南农家经济研究 (Study of the
Economy of Farmers in Southern Jiangsu before the PRC) (Beijing, Zhongyang bianyi
chubanshe, 1996), pp. 86–130.
102

Agriculture
Table 3.2 Farm size and productivity

Medium- Very
Small Medium large Large large
farms farms farms farms farms
Average farm area in acres 1.43 2.84 4.92 7.17 13.02

Crop acres per man-equivalent 1.5 2.1 2.6 3.2 4
Index of double cropping 153 151 149 147 143
Index of crop yield 100 99 100 98 100
Composite land yield index* 129.74 101.43 95.51 91.11 82.21
Man-equivalent per farm 1.2 1.7 2.3 2.8 3.7
Production of grain equivalent in 828 1168 1448 1679 2073
kilograms per man-equivalent
Percentage of net income from 21 14 11 10 9
other than farm sources
Number of idle months per able- 1.6 1.7 1.8 1.8 1.8
bodied man
Notes: Except for the composite land yield index, all other data are from John Lossing
Buck, Land Utilization in China: Statistics (Nanking, University of Nanking, 1937),
pp. 267–88, 295. The composite land yield index was compiled by Hao Hu and Minjie
Yu, who recalculated it with slightly different criteria based on the Buck data. See Hao
Hu and Minjie Yu, “The Relationship between Farm Size and Land Productivity in
Early Twentieth-Century China,” in Hao Hu, Funing Zhong, and Calum G. Turvey
(eds.), Chinese Agriculture in the 1930s: Investigations into John Lossing Buck’s Rediscovered
“Land Utilization in China” Microdata, https://doi.org/10.1007/978-3-030-12688-9.
may exaggerate the role of farm size given that large farm size could also be
associated with other larger fixed investment or outlays. In fact, with an
active market in land and labor (including short-term labor), higher labor
productivity in large farms may be attributable to greater management skills,
better access to capital, other productive inputs, and, most importantly, the
hiring of laborers from outside the farm. Larger farms may also benefit from
lower transaction costs in marketing and other activities and higher potential
for crop diversification than smaller farms. All these are more likely to push
small-scale farmers into sideline production or off-farm work (particularly in
areas where farms are located near urban areas or newly industrialized areas).
This is also revealed by the higher percentage of income derived from other
farm sources from smaller farmers in Table 3.2.29
29
D. Benjamin and L. Brandt, “Markets, Discrimination, and the Economic Contribution
of Women in China: Historical Evidence,” Economic Development and Cultural Change
44.1 (October 1995), 63–104.
103

Agricultural Sideline Production

We now turn to examine the issue of agricultural sideline production, a long-
standing feature of the Chinese agrarian system which combines main agri-
cultural production with household subsidiary and handicraft activities. The
increasing role of household subsidiary and handicraft production has long
been held to be the cornerstone of the “involution” thesis that posits
overpopulation pushing agricultural households increasingly into lower-
productivity activities and ultimately into poverty. However, a critical prob-
lem with this argument is that it may overlook the important fact that
subsidiary or sideline production may be closely related to high agricultural
seasonality due to China’s monsoon climate.30 Handicraft industries such as
cotton spinning or weaving, developed since China’s Ming and Qing periods,
require little capital and are easily tailored to household production by
women and children during the agricultural slack season. Hence the argu-
ment for involution based simply on lower or diminishing returns to labor in
cotton cultivation or textile handicrafts relative to grain or staple production is
insufficient. While more systematic evidence is needed, we illustrate the
impact of agricultural seasonality by showing the striking inverse relationship
between daily agricultural wages and the percentage of idle time throughout
a year in northern China in Figure 3.7.
We now develop a new theoretical framework to interpret the impact of
seasonality. Figure 3.8 shows that in the absence of cash crop cultivation or
handicrafts, the marginal revenue of labor is MR1 during the agricultural season
but drops drastically to MR2. The total number of work days will be deter-
mined by the intersection of marginal cost (MC) of labor and MR2. Now we
introduce cash crops or sideline activities such as cotton cultivation. Given that
returns on cotton cultivation were lower than on grain cultivation, as indicated
in the literature, marginal returns on combined grain and sideline activities
would be MR2 and MR2ʹ, which are both lower than MR1 and MR1ʹ respect-
ively. However, the total income based on MR2 and MR2ʹ may not necessarily
be lower than that from MR1 and MR1ʹ for two reasons. First, the total number
of workdays has now been extended in the scenario of combined agricultural
and sideline activities. Second, MR2ʹ, which takes account of the introduction
of sideline activities, is higher than MR1ʹ during the agricultural slack season.
30
For an extended interpretation of this issue, see Lu Feng 卢锋, “我国传统农业生产结
构特征” (Structural Characteristics of Traditional Agricultural Production in China), in
半周期改革现象：我国粮棉流通改革和食物安全研究 (The Phenomenon of
Semi-cycle Reform: A Study of Grain and Cotton Circulation Reform and Food
Safety in China) (Beijing, Beijing daxue chubanshe, 2004), pp. 437–44.
104

Agriculture
35 160
30 140
Wage rates (copper cash)

Percentage of idle time (%)
120
25
100
20
80
15
60
Percentage of idle time
10
Wage rates 40
5 20
0 0
1 2 3 4 5 6 7 8 9 10 11 12
Month
Figure 3.7 Percentage of idle time (left axis) and wage rates in units of copper cash (right axis)
Source notes: percentage of idle time of farmers from Buck, Land Utilization in China,
p. 296; wage rates (1807–1858) from Sydney Gamble, “Daily Wages of Unskilled Chinese
Laborers 1807–1902,” Far Eastern Quarterly 3.1 (November 1943), 41–73
Moreover, if the desire to keep small farms as families’ livelihood reduces the
marginal disutility of work, this will push MC towards MCʹ, further extending
workdays annually.31
We can now see the insufficiency of the involution argument of Philip
Huang which hinges on the simple fact that both MR1ʹ and MR2ʹ are lower
than MR1, but ignores the fact that MR1 is only applicable during the
agricultural harvest season. The new system of mixed cultivation and sideline
production extends the number of workdays and at the same time raises the
marginal returns during the agricultural slack season. Over time, this com-
bined effect may increasingly dominate the income from pure agricultural
work given that increasing commercialization, improving transport, and
rising prices for cash crops raise the returns on handicraft and sideline
production. Rather than immiseration or exploitation, as the “involution”
thesis may imply, total annual household income may increase.32 All of these
go to show that so-called “hidden employment” or “surplus labor” was far
31
However, data do not seem to support this hypothesis, as Table 3.2 reveals that the
number of idle months varies little with farm size. Data show only that among the very
small farms – the bottom 1 percent of farms – the number of idle months reaches as high
as 2.3. Buck, Land Utilization in China, p. 307.
32
There is no definitive resolution on the empirical evidence supporting either side of the
debate. See Li Bozhong 李伯重, 江南农业的发展 1620–1850 (Agricultural
Development in Jiangnan, 1620–1850) (Shanghai, Shanghai guji chubanshe, 2007);
105

Production
MR1
MR1′
MC
MC′
Wage rate
MR2′
MR2
Working days
Figure 3.8 Agricultural sideline production
less severe and has been increasingly resolved through the combined agro-
handicraft eco-system in China. It is ironic that from a comparative perspec-
tive what is classified here as “involution” or the “high-level equilibrium trap”
in the Chinese context bears a striking resemblance to what other scholars
describe as “industrious revolution” for preindustrial Europe and Japan.33
One revealing “natural experiment” on the importance of sideline activity
can be seen in the case of the Kaixian Gong village 开弦弓村, known as Jiang
village, in the lower Yangzi area known once for its highly developed silk
production. In 1956, in an attempt to modernize, radical governmental
policies largely eliminated its sideline production and shifted labor and land
to grain production. With massive investment in irrigation and fertilizer,
along with enhanced multi-cropping, land productivity and total agricultural
output increased by nearly 60 percent in 1956 compared with 1936. However,
it turned out that this just barely made up for the value of sideline production
that had been sacrificed in the process. The value of sideline production,
which had once accounted for 45 percent of total income in 1936, was reduced
P. Huang, The Peasant Family and Rural Development in the Yangzi Delta, 1350–1988
(Stanford, Stanford University Press, 1990); K. Pomeranz, The Great Divergence: China,
Europe, and the Making of the Modern World Economy (Princeton, Princeton University
Press, 2000). Also see the chapter by Kenneth Pomeranz in Volume I of The Cambridge
Economic History of China.
33
For “industrious revolution,” see J. de Vries, The Industrious Revolution: Consumer
Behavior and the Household Economy, 1650 to the Present (Cambridge, Cambridge
University Press, 2008); A. Hayami, Japan’s Industrious Revolution: Economic and Social
Transformations in the Early Modern Period (Tokyo, Springer, 2015). For seasonality and
sideline occupation in the US, England, and Japan, see K.L. Sokoloff and D. Dollar,
“Agricultural Seasonality and the Organization of Manufacturing in Early Industrial
Economies: The Contrast between England and the United States,” Journal of Economic
History 57.2 (1997), 288–321; O. Saito and M. Takashima, “Estimating the Shares of
Secondary- and Tertiary-Sector Outputs in the Age of Early Modern Growth: The
Case of Japan, 1600–1874,” European Review of Economic History 20.3 (2016), 368–86.
106

Agriculture
to only 20 percent (despite the much higher price of agricultural sideline

goods in 1956).34
What this Communist era “natural experiment” reveals is the resilience of
traditional Chinese rural household production to deal with the seasonality
problem. Labor reallocation across seasonal cycles, coupled with a very
active labor market, particularly for short-term hire, promoted commercial-
ization and industrialization. Indeed, given the relatively high labor return
during the agricultural harvest season, incentives to migrate to urban centers
for full-time nonagricultural work may be dampened. As shown in Figure 3.8,
if urban wage rates are still below MR1 or MR1ʹ, rural laborers may not find it
attractive enough to move to full-time occupations in the city. This explains
the viability of rural-based industry even in the face of rapid
industrialization.35 Overlooking or grossly misunderstanding this important
mechanism in the Chinese agricultural–industrial ecosystem led to disastrous
outcomes in the Communist era’s collectivization movement. Indeed, the
revival of rural-based industrialization, such as the township and village
enterprise, provided the engine of Chinese economic growth during the
reform era of the 1980s and 1990s.36
New Elements in Chinese Agriculture

In comparison with the sweeping transformations in modern industry,
finance, and other sectors, changes in the agricultural sector remained far
more limited. However, some changes did happen in both crops and the
spatial structure of cultivation. The opening of treaty ports, the rise of
agricultural exports more closely linked to international markets, and the
shift of grain tribute from the Grand Canal to the sea route from the mid-
nineteenth century, along with the construction of railroads from the twen-
tieth century, are among the main factors effecting change. Institutional and
political changes also led to agricultural improvement, investment in techno-
logical innovation, and organizational changes, particularly in marketing,
34
Fei Xiaotong 费孝通, 江村经济—中国农民的生活 (Economy in Jiang Village:
Peasant Life in China) (Shanghai, Shangwu yinshuguan, 2001), pp. 258–69.
35
Xiao Bucai 萧步才, “江苏省江阴县手工织布业调查资料: 上” (Investigation Data of
Handwoven Cloth in Jiangyin, Jiangsu Province: I)”, 学术月刊 (Academic Monthly)
1958.1, 87–9.
36
See Chapter 18 on the Great Leap Famine by James Kung below for the disasters of
agricultural collectivization. For the importance of 1980s rural industrialization, see the
chapter by Barry Naughton and also B. Li, Agricultural Development in Jiangnan, 1620–1850
(Basingstoke, Macmillan, 1998).
107

processing, and the production of agricultural commodities for export and

industrial production.
Acceleration in the Diffusion of New Crops

One of the longest-lasting transformations in Chinese agriculture from the
sixteenth century onward is the introduction and diffusion of New World
crops such as maize and sweet potatoes. While their introduction fundamen-
tally changed the structure of grain consumption in China, the nationwide
diffusion of these crops achieved consequential proportions and nationwide
significance only by the late nineteenth and early twentieth centuries. Even in
the eighteenth century, despite governmental encouragement, most crops
were diffused through the migration of people. By the nineteenth century,
maize and sweet potatoes were largely restricted to the southern regions
where they initially came in, with maize concentrated in the high plateau of
the southwest and sweet potatoes in the hills of the southeast.37 But by the
1910s, things had fundamentally changed as both crops, which had few
demands regarding irrigation and temperature, expanded throughout the
whole of northern China, with the cold-resistant corn spreading further into
northeast China and Inner Mongolia following the new wave of Chinese
migration. By that point, with northern China rising to become the center of
cultivation, the share of these crops in total output only trailed rice and wheat
at the national level. While the share of maize and sweet potatoes was only
about 2 percent in the early nineteenth century, it reached 5 percent and
10 percent in 1914 and 1930 respectively (converted using a 4:1 ratio between
sweet potatoes and raw grain).38
It turns out that the process of diffusion of New World crops is far more
complicated and multifaceted. Often the diffusion started with migration,
which itself is driven by population increase.39 Then, once brought to wider
areas of China, these New World crops would take time to take root and
37
Cao Shuji 曹树基, “清代玉米番薯分布的地理特征” (Geographical Characteristics
of the Distribution of Corn and Sweet Potato in the Qing Dynasty)”, 历史地理研究
(Historical Geography Study) 1990.2, 287–303.
38
Li Xinsheng 李昕升 and Wang Siming 王思明, “清至民国美洲作物生产指标估计”
(An Estimate of the Production of New World Crops in China from the Qing Dynasty
to the Republican Era), 清史研究 (Qing History Journal) 2017.3, 126–39.
39
This contradicts a recent study that argues that the introduction and diffusion of New
World crops accounted for the population explosion during the eighteenth and early
nineteenth centuries. See S. Chen and J. Kung, “Of Maize and Men: The Effect of a New
World Crop on Population and Economic Growth in China,” Journal of Economic
Growth 21.1 (2016), 71–99.
108

Agriculture
thrive after local adaptation and competition with existing indigenous

crops.40 This process reinforced an interactive rather than a one-way causal
relationship between the diffusion of New World crops and population
expansion. Indeed, crop diffusion often followed rather than preceded (or
caused) population expansion. This explains why the spread of New World
crops continued and even accelerated nationwide after population pressure
in China significantly decreased due to the massive population loss and the
opening of China’s northeastern frontier after the mid-nineteenth century.
Migration, warfare, and new transport infrastructure all enhanced the move-
ment of population and reinforced the diffusion of crops.
Besides corn and sweet potatoes, the diffusion of other New World Crops
such as Irish potatoes also accelerated during this period. Although introduced to
China relatively late as they appeared in only scattered records in the eighteenth
century, the importance of Irish potatoes in China varies greatly by location.41 As
a highly cold-resistant crop, Irish potatoes reached much higher-elevation plat-
eaus than can be reached by maize and sweet potatoes and opened up new
frontiers of cultivation on China’s barren northwestern frontier. In the 1930s, Irish
potatoes had a share in total staples consumption ranging from 10 to 20 percent
among rural households in Shanxi, Suiyuan, Chahar, Qinghai, Gansu, Ningxia,
and other northwestern provinces.42 For these provinces, the importance of Irish
potatoes as a staple food exceeded that of maize and sweet potatoes. Other
notable crops, such as peppers, tomatoes, pumpkins, apples, and other foreign
vegetables, as well as American types of peanuts for oil production, also came
into significance after the nineteenth century.
Agricultural Commercialization
Another important driving force behind agricultural development during this
period is the heightened degree of commercialization. The broad quantitative
profile is captured by the well-known comprehensive study of Wu Chengming
and Xu Dixin. Figure 3.9 shows that both the absolute values and the shares of
almost all crops produced for the market rose between 1840 and 1920 (in 1894
40
Han, 中国历史农业地理, pp. 529–56. Wang Baoning “王保宁, 花生与番薯：民国
年间山东低山丘陵区的耕作制度” (Peanuts and Sweet Potatoes: The Farming
System in the Low Hill Area of Shandong in the Republic of China), 中国农史
(Agricultural History of China) 2012.3, 54–68.
41
Ping-ti Ho 何炳棣, 美洲作物的引进、传播及其对中国粮食生产的影响
(Introduction and Dissemination of American Crops and Their Impact on China’s
Grain Production, I I I), 世界农业 (World Agriculture), 1979.6, 25–31. Han, 中国历史
农业地理, pp. 661–4.
42
农情报告, 1937–1938.
109

Figure 3.9 Trends of agricultural commercialization

Source: Xu Dixin 许涤新 and Wu Chengming 吴承明, 中国资本主义发展史 (History
of Capitalist Development in China) (Beijing, Renmin chubanshe, 1990), vol. 2, pp. 279–
304, 966–95
constant prices). The figure also reveals different patterns, with cocoons produced
for domestic reeling production and cotton increasingly sold to mechanized
manufacturers. But both silk and cotton in the end were becoming more
commercialized. On the other hand, tea and soybeans increased their share of
commercialization by being directly sold to the international market. Meanwhile,
self-consumption and the domestic market remained predominant for grain.43
There is variation in the performance of agricultural commodities under
enhanced commercialization and international trade. Chinese tea, which
43
Another estimate by Shi Zhihong also shows that the share of cash or economic crops
slowly increased between the eighteenth and early twentieth centuries, and that it had
reached 20 percent by the nineteenth century. See Shi, 清代农业的发展和不发展,
p. 127.
110

Agriculture
used to be dominant in the global market, faced severe competition from

India and Japan and began to decline from the end of the nineteenth century,
and only attained half its previous level by the 1930s, leading to
a corresponding reduction in tea production and tea cultivation acreage.44
The result for cotton production was mixed, with the share sold to modern
spinning factories increasing but the share for home production decreasing.
Meanwhile, cocoon production kept up with the increased demand for raw-
silk exports and domestic demand until the Great Depression and the rise of
artificial silk in the 1940s.45
Increased commercialization led to the rise of new cash commodities,
most notably soybeans and bristles from northeastern China; hides, leather,
wool, and eggs from the middle and western parts of China; and wood oil
from southern China. Using the comparative advantage of low labor cost,
eastern coastal provinces began to turn out miscellaneous textiles, such as
trimmings, straw cloth, and straw braid.46 Clearly commercialization
brought rural households genuine economic benefits, and in some cases
even wealth.47 They also introduced market risk and in some cases global
risk, as marked by the 1930s Great Depression.48 The most extreme case is the
cultivation and commercialization of opium, a product over which China
once lost a war with Britain in the 1840s. Ironically, through domestic
cultivation and import substitution, Chinese opium finally drove out foreign
opium and became the number one cash crop by the late nineteenth century
(see Figure 3.9). Obviously, opium generated huge profits for cultivators as
well as warlords, and in many cases led to serious social and economic
problems.
44
Xu Daofu 许道夫, 中国近代农业生产及贸易统计资料 (Statistics on Agricultural
Production and Trade in Modern China) (Shanghai, Shanghai renmin chubanshe, 1983),
p. 257.
45
Zhang, 非平衡化与不平衡, pp. 29–35.
46
See Wang Jingyu 汪敬虞, 中国近代经济史 1895–1927 (China’s Modern Economic
History 1895–1927) (Beijing, Renmin chubanshe, 2012), pp. 879–970; Xu, 中国近代农业
生产及贸易统计资料, pp. 160–336.
47
See R.H. Myers, The Chinese Peasant Economy: Agricultural Development in Hopei and
Shantung, 1890–1949 (Cambridge, MA, Harvard University Press, 1970); Brandt,
Commercialization and Agricultural Development; Rawski, Economic Growth in Prewar
China; David Faure, The Rural Economy of Pre-Liberation China: Trade Expansion and
Peasant Livelihood in Jiangsu and Guangdong, 1870–1937 (Oxford, Oxford University Press,
1989).
48
Xia Mingfang 夏明方, 民国时期自然灾害与乡村社会 (Natural Disasters and Rural
Society during the Republic of China) (Beijing, Zhonghua shuju, 2000), pp. 171–9.
111

Agricultural Improvement and Modernization

Although Chinese agriculture remained largely traditional, technological and
organizational modernization did proceed prominently in several key agri-
cultural crops or commodities often connected with international factors and
in some cases under the direct sponsorship and management of industrial
enterprises. A more significant development is the gradual establishment of
national infrastructure to support agriculture. Although long an agrarian
state where agriculture is at the center of economic activity, traditional
imperial China had little formal infrastructure to support agriculture.
Agricultural administration was loosely attached to the board of finance
and construction. The late Qing reform (1903–1911) set up a special agricul-
tural department (农务司) under the New Ministry of Agriculture, Industry,
and Commerce to implement agricultural improvement and policies such as
organizing and managing newly established agricultural schools and experi-
ments. The local level also saw the official establishment and sponsorship of
schools, experimental stations, and publications for research and education.
The founding of the new Nationalist government heralded a new level of
agricultural-improvement policies and infrastructure, as seen in the establish-
ment of a national-level research institute, the National Agricultural Research
Bureau (中央农业实验所), along with prominent private institutions such
as the well-known Nanjing University Agricultural Department. China now
had a truly national-level agricultural survey, research, and improvement
infrastructure and policies.49
The best-known case of agricultural improvement through multinational
corporations is tobacco cultivation. Although introduced to China from the
New World during the sixteenth century and widely diffused by the end of
the Qing, domestically grown tobacco leaves were not suited to the modern
production of cigarettes. In the early twentieth century, the giant British
American Tobacco Company began to introduce American tobacco.
However, unlike the spontaneous introduction of centuries earlier, the
British American Tobacco Company sent a team of experts to survey nearly
100 counties in fourteen provinces in order to determine the most suitable
location of cultivation before 1914. Once they had started, the company
aggressively adopted such measures and policies as distributing free seeds,
49
A case in point is the compilation of agricultural manuals, which go back two millennia
in China. With intense interest in sericultural improvement, the publication of new
sericulture manuals from the end of the nineteenth century reached more than 140
types, which exceeded the total of the past millennia. Hua Degong 华德公, 中国蚕桑
书录 (Bibliography of Chinese Sericulture Books) (Beijing, Nongye chubanshe, 1990).
112

Agriculture
demonstrating new ways of planting, designing tobacco drying rooms, and

adopting a high-price repurchase program. These aggressive direct purchase
and distribution programs outcompeted and supplanted traditional distribu-
tion networks, such as guilds. Given the taxation revenue from tobacco
cultivation, the local government was also taking a more proactive
attitude.50 This “corporate” type of agricultural sponsorship achieved great
success, later imitated by large Chinese tobacco companies such as the
Nanyang Brothers. Through successful import substitution, Chinese-
produced tobacco rapidly expanded, taking a large share of the domestic
market and even gaining a share of the international market, leading to the
rise of specialized areas of tobacco cultivation in China.51
The introduction of American cotton followed a very different pattern as
the huge size of the domestic cotton market and production made it impos-
sible for any single firm or corporation to take on the task. The main
improvement needed was that Chinese short-staple cotton was unsuitable
for machine spinning. Here the government stepped in first through the
initiative of the Hu-Guang governor Zhang Zhidong and other local officials,
later joined by the newly founded Ministry of Agriculture, Industry, and
Commerce to distribute the seeds of American cotton in the late Qing. By the
Republican period, the initiative had been taken up by both the central and
local governments, by cotton manufacturing associations, and by individual
enterprises. More importantly, efforts had broadened to include such activ-
ities as the establishment of experimental farms, developing improved seed
varieties, organizing associations, and conducting scientific research, some-
times in collaboration with the newly risen universities. Although overall
achievements were limited due to a host of technical and environmental
factors and the share of long-staple cotton remained relatively low, the
cultivation of American cotton did expand rapidly. Statistics for 1934 show
that American cotton took a 50 percent share of the total cultivated acreage of
cotton and 52 percent of total cotton production.52
50
Li Gengwu 李耕五, 许昌烤烟发展史话 (History of the Development of Tobacco in
Xuchang) (Xuchang, Xuchang kaoyan fazhan shihua bianji weiyuanhui, 1992), pp. 18–25.
51
S. Cochran, Big Business in China: Sino-Foreign Rivalry in the Cigarette Industry, 1890–1930
(Cambridge, MA, Harvard University Press, 1980); Zhang Youyi 章有义, 中国近代农
业史资料 (Historical Data of Modern Chinese Agriculture), vol. 2 (Beijing, Shenghuo
dushu xinzhi sanlian shudian, 1957), pp. 225–6.
52
See Yan Zhongping 严中平, 中国棉纺织史稿 (History of Chinese Cotton Textiles)
(Beijing, Kexue chubanshe, 1955), pp. 323–40; Kang Chao 赵冈 and Jessica C.Y. Chao
陈钟毅, 中国棉纺织史 (The Development of Cotton Textile Production in China)
(Beijing, Zhongguo nongye chubanshe, 1997), pp. 40–3; Shen sung-chiao 沈松侨, “經濟
113

The diffusion pattern of sugar beet falls between those of American

tobacco and cotton. Originally sugar was only produced from sugarcane
grown in southern China. In 1906, the governor-general of Fengtian, Zhao
Erxun, set up experimental farms to introduce sugar beet following the
advice of Japanese experts and later the continued efforts of the Japanese
Southern Manchurian Company. Upon success, the Southern Manchurian
Company set up a sugar manufacturing company to advance seeds, loans,
and fertilizer to farmers with guaranteed purchase of the autumn harvest.
This model was followed by Chinese manufacturers, such as the Jinan Puyi
sugar factory in Shandong province. Like American tobacco, this corporate
plus rural-household model led to the rise of concentrated production areas
of sugar beet.53
One of the most fascinating developments was in sericulture, which
produces cocoons for the reeling of raw silk for both export and the domestic
market. Unlike Chinese tea exports, Chinese raw-silk exports remained
competitive until about the 1910s; from then on Japanese silk began to take
over and was dominant by the 1920s and 1930s. Japanese success was built on
nearly five decades of sericultural improvement and research in the breeding
of better silkworm varieties. Its major breakthrough came in the discovery
and development of the so-called first generation (F1) hybrid variety based on
the systematic application of Mendelian genetic principles. In just a decade,
diffusion of the F1 variety rose to nearly 100 percent between 1913 and 1920.
The Japanese success was founded on a combination of university-based
research, conglomerate enterprise promotion of better varieties, and distri-
bution arrangements with sericultural farmers.
The Japanese success awoke the world’s long-time leader based in the lower
Yangzi. After the founding of the new Nationalist government in Nanjing,
Jiangsu province, in 1927, the Chinese silk reeling center migrated from
Shanghai to the city of Wuxi in Jiangsu province. Under the leadership of
a giant silk reeling conglomerate, the Yongtai Company, there were efforts to
push for sericultural improvement and technological diffusion in the 1930s.
With government support and through the establishment of silkworm rearing
co-operatives, cocoons were sold directly to reeling factories. In 1932, the
作物與近代河南農村經濟 (1906–1937)—以棉花與菸草為中心” (Cash Crops and

the Modern Rural Economy of Henan Province, 1906–1937: Focusing on Cotton and
Tobacco), 近代中国农村经济史论文集 (Collection of Papers on the Rural Economic
History of Modern China) (Taipei, Zhongyang yanjiuyuan jindaishi yanjiusuo, 1989),
pp. 347–9.
53
Han, 中国历史农业地理, p. 743. Zhang, 中国近代农业史资料, vol. 2, pp. 161, 171.
114

Agriculture
Yongtai Reeling Company began to set up long-term exclusive contracts with

farmers or co-operatives in the lower Yangzi, a system very similar to the
subcontractual direct-purchase system pioneered by large silk reelers in Japan.
Starting in 1932, the provincial governments of both Jiangsu and Zhejiang began
to take a direct role in promoting sericultural improvement and technological
diffusion by designating model districts across the region, and set up a national-
level sericultural research and improvement organization.
The outcome was a 1930s lower Yangzi catch-up with Japan that was
nothing short of remarkable. Figure 3.10 plots the lower Yangzi diffusion
curve for the scientifically improved variety, mostly of the F1 types. For
Jiangsu province, the share of scientifically produced silkworm eggs increased
from 5 percent to almost 100 percent within only five to seven years.
Diffusion lagged somewhat in Zhejiang, but the overall rate of diffusion in
the lower Yangzi in the 1930s was comparable to the Japanese diffusion of the
F1 hybrids in the 1910s and 1920s.54
In comparison with this remarkable, though limited, success, agricultural
development in terms of mechanization remained low, except in northeastern
China, which had farms relatively larger in size. By 1949, there were only 401
tractors. Mechanized water pumps gained some ground in the highly commer-
cialized lower Yangzi. In Wuxi, the acreage irrigated by mechanized water
pumps reached more than 60 percent. However, less than 1 percent of the
acreage was irrigated mechanically in the entirety of Jiangsu and Zhejiang
provinces combined.55 This is understandable given the very labor-intensive
nature of Chinese agriculture; modernization went more in the direction of
biological or chemical improvement rather than in the mechanical direction.56
Land Institutions, Distribution, and Rural Credit

Land Institutions
Despite the nominal claim that all land belongs to the emperor, imperial China
was under a more or less de facto regime of private property rights with a vibrant
54
D. Ma, “Why Japan, Not China, Was the First to Develop in East Asia: Lessons from
Sericulture 1850–1937,” Economic Development and Cultural Change 52.2 (January 2004),
369–94. Wang Xiang 王翔, 中国近代手工业史稿 (History of the Modern Chinese
Handicraft Industry) (Shanghai, Shanghai renmin chubanshe, 2012), p. 380.
55
Wang Fangzhong 王方中, “旧中国农业中使用机器的若干情况” (Some
Information about the Use of Machinery in Agriculture in Traditional China), 中国
近代经济史论文选 (Essays on the Modern Economic History of China), vol. 2
(Shanghai, Shanghai renmin chubanshe, 1985), pp. 843–6.
56
See Hayami and Ruttan, Agricultural Development; Ma, “Why Japan, Not China.”
115

100% 100%
80% 80%
Jiangsu (A)
60% 60% Jiangsu (S)
Zhejiang (A)
40% 40% Zhejiang (S)

Japan (A)
20% 20%
0% 0%
1913 1916 1919 1922 1925 1928 1931 1934 1937
Figure 3.10 Diffusion of scientifically produced silkworm varieties in Japan and the
lower Yangzi. A = autumn crop; S = summer crop
Source: D. Ma, “Why Japan, Not China, Was the First to Develop in East Asia: Lessons from
Sericulture 1850–1937,” Economic Development and Cultural Change 52.2 (January 2004), 369–94
land market regulated by local customs and rules. Because of the multiple claims
to a single piece of property, land rights took on many different forms or layers in
commercial transactions.57 These multiple and sometimes vaguely defined land
rights systems have been viewed as the cause of agricultural stagnation in
traditional China. One of the most cited cases is the pervasive tenancy system,
particularly sharecropping – often viewed as inefficient, reducing the incentives
of cultivators. Such a view has been well challenged both theoretically and on
empirical grounds. Theoretical work based on risk diversification and monitor-
ing costs reveals the rationality of sharecropping not just in China but around the
world.58 Indeed, the Chinese case reveals precisely that it was the most devel-
oped regions, such as the lower Yangzi, that also had the most-developed
tenancy system. Furthermore, from the nineteenth century onward, the import-
ance of sharecropping declined drastically relative to the fixed rent system.
57
For a nuanced discussion of Chinese traditional land rights, see the chapter by
Kishimoto Mio in Volume I. For a more recent contribution see Long Denggao 龙登
高, 中国传统地权制度及其变迁 (The Evolution of Traditional Institutions of Land
Property Rights in China) (Beijing, Zhongguo shehui kexue chubanshe, 2018); Cao Shuji
曹树基 and Liu Shigu 刘诗古, 传统中国地权结构及其演变 (The Structure and
Evolution of Traditional Land Property Rights in China) (Shanghai, Shanghai jiaotong
58
S.N.S. Cheung, The Theory of Share Tenancy, with Special Application to Asian Agriculture
and the First Phase of Taiwan Land Reform (Chicago, The University of Chicago Press,
1969); A. Braverman and J. Stiglitz, “Sharecropping and the Interlinking of Agrarian
Markets,” American Economic Review 72.4 (1982), 695–715; K. Otsuka, H. Chuma, and
Y. Hayami, “Land and Labor Contracts in Agrarian Economies: Theories and Facts,”
Journal of Economic Literature, 30.4 (1992), 1965–2018.
116

Agriculture
Indeed, there is a large literature that argues that Chinese tenants often had far
stronger tenancy rights, as indicated by the widespread existence of perpetual
tenancy or soil rights. These rights were relatively secure and could even be
transacted or used as collateral in the market, as regulated by well-developed
customs and rules. These greatly resolved the incentive problems and reduced
the transaction costs associated with the tenancy system.59 Table 3.3 reveals that
areas in central and southern China that had a much higher rate of tenancy were
among the most productive and commercialized regions in China, possibly with
stronger and more sophisticated traditions of customary law and enforcement
mechanisms.60
There is a fundamental problem of land property rights regulated by customs
rather than by formal rules, courts, and legal procedures. Although county
magistrates and imperial government did conduct the equivalent of civil or
commercial litigation, its aim focused heavily on social and political stability
rather than legal justice.61 Clearly, customary rules carry strong elements of
moral economy. The most important example is the practice of zhaojia 找价,
a custom that allows the seller of the land to reclaim additional compensation
from the buyer after the purchase has been finalized, especially when the land
value increases subsequent to the transaction. These claims could occur on
a regular basis many years after the purchase. While there is no denying that
this understandably leads to disputes, opportunistic behavior, and uncertainty
over investment and property, various new archives from after the nineteenth
century show that the zhaojia practice become regularized through customs and
evolved into a more predictable system of installment payments by the buyer.62
All in all, new studies increasingly recognize the rationality and efficiency
attained by what had long been thought “backward” traditional practices in
a society without strong formal legal rules and enforcement.63 While there is
some validity in the claim that the Chinese legal institutions and practice
might have led to divergence in long-term agricultural trajectories between
China and England,64 it is difficult to argue the case for general institutional
59
Cao and Liu, 传统中国地权结构及其演变, pp. 18–34, 63–99.
60
Z. Chen, K. Peng, and W. Yuan, “Usury, Market Power and Poverty Traps: A Study of
Rural Credit in 1930s’ China,” Frontiers of Economics in China 3 (2018), 369–96.
61
D. Ma, “Law and Economy in Traditional China: A ‘Legal Origin’ Perspective on the
Great Divergence,” in Ma, Law and Long-Term Economic Change: A Eurasian Perspective
62
Peng, 从交易到市场, p. 227–9.
63
Long, 中国传统地权制度及其变迁; Cao and Liu, 传统中国地权结构及其
演变.
64
T. Zhang, The Laws and Economics of Confucianism: Kinship and Property in Preindustrial
China and England (Cambridge, Cambridge University Press, 2017).
117

Table 3.3 Tenancy rates in the 1930s

Share of tenants and part-tenants in
total farm households
Share of rental land in
Province total cultivated acreage (1934) NARB (1931–1934) Buck (1929–1933)
South
Jiangsu 42.33 59.60 62.15
Zhejiang 51.31 77.80 65.46
Anhui 52.64 65.00 62.12
Jiangxi 45.10 72.80 71.07
Hubei 27.89 69.00 60.31
Hunan 47.80 73.40 60.70
Fujian 39.33 73.20 74.30
Guangdong 76.95 81.00 81.55
Guangxi 21.20 67.00 50.35
Sichuan –– 75.40 55.07
Yunnan –– 63.80 46.67
Guizhou –– 64.20 55.22
North
Hebei 12.89 32.60 19.26
Shandong 12.63 30.80 21.79
Henan 27.27 44.00 31.10
Shanxi –– 38.20 34.55
Shaanxi 16.64 46.80 30.65
Chahar 10.20 61.80 ––
Suiyuan 8.75 45.60 8.45
Gansu –– 41.20 36.54
Ningxia –– 39.00 3.00
Qinghai –– 41.00 53.73
Liaoning –– –– 64.10
Average 30.73 54.25 50.80
Notes. The ratio of rented crop land comes from Guomin zhengfu zhujichu tongjiju
国民政府主计处统计局, 中国土地问题之统计分析 (Statistical Analysis of Land
Problems in China) (Nanjing, Zhengzhong shuju, 1946), p. 63. The percentages for
farmers who were tenants or part-owners are taken from two sources. “NARB” cites
农情报告汇编 (Crop Reporting in China, 1934) (Nanjing, National Agricultural
Research Bureau, Ministry of Industry, 1936), p. 62. “Buck” cites the averages of
farm and agricultural surveys in Buck, Land Utilization in China, pp. 57–9.
failure in the traditional property rights regime as far as short-term efficiency

is concerned. Indeed, the explosive growth and recovery in agricultural
output in the early 1980s following the adoption of the so-called
118

Agriculture
Agricultural Household Responsibility and the resolution of the collective

commune reveals again the resilience of traditional Chinese agrarian land
institutions.
Distribution and Credit

One of the long-standing debates around Chinese agriculture is the question
of distribution and exploitation. The thesis mostly advanced by the Marxist
framework advocated that the prevalence of small-scale farming and vulner-
ability to risk exposes peasants to land deprivation and exploitation through
usury. The actual empirical evidence of any of these is unclear. An important
study by Loren Brandt and Barbara Sands on land and income distribution in
early twentieth-century China does not reveal very high inequality in land
distribution. Their calculations give a Gini coefficient for land distribution of
0.55–0.6, which is high but not extreme. But their income Gini for rural
households stood at 0.4–0.45. Using village-level data, they reveal that the
discrepancy between income and land distribution was derived from two
major sources. First is that factor markets in land and labor have served to
equalize returns on agricultural production through the renting of surplus
land and the hiring of surplus labor. Second, with commercialization, a third
of net village income was actually derived from nonagricultural labor,
sometimes from out-migrants working in newly arisen industrial centers.65
Their studies of moderate inequality in Chinese rural areas have been
confirmed by a host of new studies.66 Indeed, according to Kang Chao’s
study based on much more scattered and controversial evidence, there was
no long-term increase in land inequality in the millennium between the Song
and Republican China.67
Turning to the question of usury, new studies reveal that the story is far
from a simple case of exploitation by the rich and powerful with monopoly
power. Surveys show that in the 1930s most regions in rural China had an
annual interest rate in the range of 30 percent. Interest rates were far higher in
the more impoverished regions than in highly commercialized areas. But this
had less to do with exploitation and monopoly power than with higher
transaction costs incurred in the size of loans often in relatively more
65
L. Brandt and B. Sands, “Beyond Malthus and Ricardo: Economic Growth, Land
Concentration, and Income Distribution in Early Twentieth-Century Rural China,”
Journal of Economic History 50.4 (1990), 807–27.
66
Guan Yongqiang 近代中国的收入分配：一个定量的研究 (Income Distribution in
Modern China: A Quantitative Study) (Beijing, Renmin chubanshe, 2012), Chapter 3.
67
Chao Gang 赵冈, 中国传统农村的地权分配 (Land Distribution in Traditional Rural
China) (Beijing, Xinxing chubanshe, 2006).
119

impoverished areas.68 This is even true of relatively low-risk loans with land
collateral. What is more accurate is that it reflected a general phenomenon of
relative capital scarcity in the Chinese economy. It was overall capital
constraint that clearly acted against capital accumulation in Chinese agricul-
ture and tilted it further in the direction of labor-intensivity.
In Table 3.4 we reproduce a recent study that uses the county data from the
survey conducted by the Department of Agricultural Economics at the
University of Nanking under the supervision of Buck from 1929 to 1933 to
test different hypotheses. The interest rate is regressed to the wealth-
inequality or Gini coefficient; the percentage of creditors from the “exploiting
class” (class); the percentage of owner-farmers (ownerp); proxies of per capita
income, including the crop output per capita (GrainPC) and daily wage
(Wage); and proxies of market and monetization development level, includ-
ing the level of waterage in local agricultural trade (waterage) and the
proportion of rent payment in money (mrent). We also control for the total
percentage of lending by formal financial institutions (finst) and per capita
acreage of arable land (land).
Regression results show that the Gini coefficient has no robust positive
effect on the interest rate, while the coefficient class alone is insignificant. The
percentage of owner-farmers, ownerp, is significantly positive at the 5 percent
level; that is, a higher percentage of farmers tilling their own land means
higher interest rates, rejecting the hypothesis of class exploitation and sup-
porting the other hypotheses that tenancy reflects more efficient and func-
tioning markets and better-developed institutions for the rural credit market.
Meanwhile, the income effect of GrainPC and Wage, and the effect of the
monetization development level represented by mrent, are negative, as
expected, and significant at the 5 percent level.
Conclusion
Ultimately, agricultural modernization and rising agricultural productivity
will lead to a shrinking of the agricultural sector and a shift of resources to
other sectors. Until about three decades ago, Chinese agriculture and agri-
cultural population weighed heavily on the Chinese economy and were at
the center of debate about long-term economic growth or stagnation. Much
68
Chen, Peng, and Yuan, “Usury, Market Power and Poverty Traps.” Peng Kaixiang 彭凯
翔, Chen Zhiwu 陈志武, and Yuan Weipeng 袁为鹏, “近代中国农村借贷市场的机
制” (Mechanisms of Rural Credit Markets in Modern China), 经济研究 (Economic
Research Journal), 2008.5, 147–59.
120

Agriculture
Table 3.4 Explaining interest rates across regions

(1) (2)
Variable Ordinary least squares Two-stage least squares
Gini 20.283** −26.979

(10.019) (33.071)
class 0.050 0.051
(0.044) (0.065)
finst −0.173* 0.182
(0.097) (0.246)
ownerp 0.058** 0.133**
(0.028) (0.067)
GrainPC −0.330*** −0.356***
(0.083) (0.110)
land 2.040** 1.895
(0.882) (1.289)
Wage −11.250** −12.234**
(4.783) (5.743)
mrent −0.060** −0.116***
(0.025) (0.036)
waterage −3.725* −3.441
(1.999) (2.443)
C 35.090*** 36.870***
(5.738) (11.506)
adj. R2 0.450 0.247
Obs. 118 118
Notes. dependent variable: average interest rate; Gini: Gini coefficient;

class: weight of exploiting class in credit; ownerp: percentage of
owner-farmers; finst: weight of financial institutions in credit;
Wage: daily wage; GrainPC: crop output per capita; land: per capita
arable land; waterage: waterage in local agricultural trade; mrent:
proportion of rent payment in money. Astersisks ***, **, *
represent significance levels of 1 percent, 5 percent, and 10 percent
respectively. For more details, see Zhiwu Chen, Kaixiang Peng, and
Weipeng Yuan, “Usury, Market Power and Poverty Traps: A Study
of Rural Credit in 1930s’ China,” Frontiers of Economics in China 3
(2018), 369–96.
more research needs to be done. This survey, though highly selective, does
not support the case of a rapid or general decline in overall agricultural
performance or per capita grain output during the period under study. It
also shows that while Chinese agriculture remained largely traditional, there
were important and significant new developments towards modernization.
121

Most importantly, despite being saddled with too many ideological labels and
conceptual frameworks, both the technology and the institutions of Chinese
agriculture displayed remarkable resilience, not only in sustaining the liveli-
hood of a quarter of the world’s population, but also in adapting to new
modernization challenges. Hence it is time to recognize and understand the
logic of Chinese agriculture on her own terms through the changing times,
and a thorough understanding of it will be critical to the success of any
genuine reform.
Further Reading
Brandt, L., and B. Sands, “Beyond Malthus and Ricardo: Economic Growth, Land
Concentration, and Income Distribution in Early Twentieth-Century China,” Journal
of Economic History 50 (December 1990), pp. 807–27.
Buck, J.L., Land Utilization in China: Statistics (Nanking, University of Nanking, 1937).
Cao Shuji 曹树基 and Liu Shigu 刘诗古, 传统中国地权结构及其演变 (The Structure
and Evolution of Traditional Land Property Rights in China) (Shanghai, Shanghai
jiaotong daxue chubanshe, 2014).
Cao Xingsui 曹幸穗 and Wang Siming 王思明 (eds.), 中国农业通史: 近代卷 (General
History of Chinese Agriculture: Modern Times) (Beijing, Zhongguo nongye
chubanshe, 2020).
Chao, K., Man and Land in Chinese History: An Economic Analysis (Stanford, Stanford
Chen, Z., K. Peng, and W. Yuan, “Usury, Market Power and Poverty Traps: A Study of
Rural Credit in 1930s’ China,” Frontiers of Economics in China 3 (2018), 369–96.
Han Maoli 韩茂莉, 中国历史农业地理 (Beijing, Beijing daxue chubanshe, 2012).
Huang, P., The Peasant Economy and Social Change in North China (Stanford: Stanford
Li, B., Agricultural Development in Jiangnan, 1620–1850 (Basingstoke, Macmillan, 1998).
Li Bozhong 李伯重, 中国的早期近代经济—1820 年代华亭 – 娄县地区 GDP 研究
(Beijing, Zhonghua shuju, 2010).
Min Zongdian 闵宗殿 (ed.), 中国农业通史: 明清卷 (General History of Chinese
Agriculture: Ming and Qing Dynasties) (Beijing, Zhongguo nongye chubanshe, 2020).
Ma, D., “Why Japan, Not China, Was the First to Develop in East Asia: Lessons from
Sericulture 1850–1937,” Economic Development and Cultural Change 52.2 (January 2004),
369–94.
Peng Kaixiang 彭凯翔, 清代以来的粮价: 历史学的解释与再解释 (The Rice Price
since the Qing Dynasty: A Historical Interpretation and Reinterpretation) (Shanghai,
Shanghai renmin chubanshe, 2006), pp. 32–41.
Peng Kaixiang 彭凯翔, “人口增长下的粮食生产与经济发展—由史志宏研究员的清
代农业产出测算谈起” (Grain Production and Development under the Growth of
Population: A Discussion from Shi Zhihong’s Estimation about Agricultural Production
122

Agriculture
in the Qing Dynasty), 中国经济史研究 (Research in Chinese Economic History) 7

(2015), 38–49.
Perkins, D.H., Agricultural Development in China, 1368–1968 (Chicago, Aldine Publishing,
1969).
Richardson, P., Economic Change in China, c. 1800–1950 (Cambridge, Cambridge University
Press, 1990).
Shi, Z., Agricultural Development in Qing China: A Quantitative Study, 1661–1911 (Leiden, Brill,
2018).
123

4
Handicraft and Modern Industries
l i n d a g r o v e a n d t ō r u k u b o
Industrialization in China has followed complicated paths over the last

century and a half. China, like Russia, Germany, and Japan, followed in the
footsteps of the pioneering industrial nations. For the first pioneering gener-
ation, industrialization developed indigenously, building on preindustrial
handicraft traditions, inventing new technologies using water and steam
power, and creating new corporate management systems. The new tech-
nologies of steamships, railroads, the telegraph, and the telephone trans-
formed transportation and communication networks. Private entrepreneurs
played central roles in the development of the new industrial systems, aided
by protective tariffs and other state measures designed to promote industrial
and commercial development.
For the second generation of industrializing nations, there was no need to
invent technologies since they could be purchased and transplanted. In
China, as in Japan and other later developing countries, the state took the
lead in pushing for change, formulating industrialization policies and provid-
ing investment funds for new industries. The early push for industrialization
came during a period of political crises. The late imperial Chinese state
suffered repeated shocks, beginning with the loss of the Opium War and
the subsequent forced opening of Chinese territory to foreign trade, followed
by the fourteen-year struggle against the Taiping and other rebellious groups
that resulted in the deaths of a reported 20 million people and widespread
destruction in the Yangzi valley, China’s most developed region. It is not
surprising that early industrialization efforts focused on military technology.
As Chinese industry developed in the late nineteenth and early twentieth
centuries, industrialization moved beyond military defense to light industry.
Factories were established to process goods for export and produce goods to
substitute for imports that had flooded the markets following the opening of
the treaty ports. By the 1920s and 1930s, there were large and small factories
124

not only in the treaty port cities, but in many smaller cities and towns. Some
produced for export markets, others for domestic consumption.
Some early Chinese industrial enterprises match conventional images of
modern industry – large-scale enterprises in purpose-built factory buildings
with mechanized production processes, staffed by disciplined workers, and
following Western corporate models. But alongside those modern factories,
we find many other industrial and semi-industrial forms that mixed the use of
mechanical and human power, relied on skills learned through training in
traditional modes of production, located production in small workshops and
homes, and drew on indigenous commercial networks and business prac-
tices. Minami and Makino’s compilation of long-term Asian statistics, which
provides the best estimate of the balance between modern industry and
handicraft production in important industries, shows that in 1933 over 90 per-
cent of output in the food, beverages and tobacco sector was produced by
handicraft means, over 60 percent of textile goods, apparel and leather came
from handicraft, and 53.9 percent of goods in the chemical, metal, and
machinery sector also came from handicraft workshops (Figure 4.1).1
China’s industrialization experience followed multiple paths. The first path
was a government-led path based on imported technology and investment of
relatively large supplies of capital to create modern factories, most commonly in
heavy industry. A second path, developed by indigenous Chinese capital, was
modeled on Western enterprise, but was based on intensive use of labor in light
industrial production. A third path, also created by indigenous Chinese capital,
grew out of handicraft traditions and combined human labor and mechanized or
semi-mechanized production processes to produce light industrial goods.
This chapter examines the “new” – i.e. the modern industrial and semi-
industrial systems that developed in the late nineteenth and the twentieth
centuries – but also what happened to the “old”; that is, the systems that linked
agricultural production with handicraft processing to provide food, clothing, and
other goods consumed by China’s vast rural and urban populations. We will
look at China’s largest traditional handicraft industries, silk and cotton, which
were transformed beginning in the late nineteenth century. The first of these,
silk, was directly linked to an explosion in the export trade while the second,
cotton, is an example of the transformations involved in import-substitution
industrialization. We will add to the consideration of these two well-studied
examples brief consideration of industries that produced soap, matches,
1
Minami Ryoshin 南亮進 and Makino Fumio 牧野文夫 (eds.), アジア長期経済統計 3 中
国 (Long-Term Asian Statistics, vol. 3, China) (Tokyo, Tōyō Keizai Shinpōsha, 2014), p. 111.
125

linda grove and tō ru kubo
Million yuan
0 2,000 4,000 6,000 8,000
Food, beverages and tobacco 565 5,713
Textile, apparel and

leather industries 1,098 1,813
Factories
Handicraft
Chemicals, metal, machinery and others 501585
Figure 4.1 Industrial production in mainland China, 1933

Source: Minami Ryoshin 南亮進 and Makino Fumio 牧野文夫 (eds.), アジア長期経済
統計 3 中国 (Long-Term Asian Statistics, vol. 3, China) (Tokyo, Tōyō Keizai Shinpōsha,
2014), p. 111
cosmetics, and other new products that entered the Chinese market in the
twentieth century. For a detailed analysis of the role of the state in building
heavy industries, including mining and shipping, and the creation of an ideology
supporting the development state, readers should refer to Chapter 6 by Morris
L. Bian in this volume.
Foreign models, as well as foreign investment, played major roles in
China’s industrialization. The foreign models came from the pioneering
industrialized states of Britain and the United States, but also – and most
importantly – from one of the first successful late-developing states, Japan.
Shortly after the end of the Sino-Japanese War, Sino-Japanese relations took
a positive turn: thousands of Chinese students were sent to Japan to study in
universities and technical and vocational schools, and hundreds of Japanese
scholars, teachers, and technicians were invited to work in China as advisers
to national and provincial government-sponsored projects.2 In later years,
2
D. Reynolds, China 1898–1912: The Xinzheng Revolution and Japan (Cambridge, MA, Council on
East Asian Relations, Harvard, 1993); Jiang Pei 江沛, 留日学生、东游官绅与直隶省的近
代进程 1900–1928 (Overseas Students and Official Missions and Modern Development in
Zhili Province, 1900–1928), 史学月刊 5 (2005), 56–66; Ōsato Hiroaki 大里浩秋 and Son
An Suk 孫安石 (eds.), 中国人日本留学史研究の現段階 (The Current State of Research
on the History of Chinese Students in Japan) (Tokyo, Ochanomizu shobō, 2009).
126

Japanese companies became major foreign investors in Chinese light indus-

try, and during the war years took over Chinese firms in the occupied areas of
China. The legacy of Japanese involvement in light industry and in the
development of heavy industry in Manchuria during the colonial period
helped to create a base for the rapid development of Chinese manufacturing
after 1949.
While foreign models and technology are an important part of the story of
Chinese industrialization, the largest role was played by the thousands of
Chinese entrepreneurs who established factories and workshops throughout
the nation. For a long period after 1950, these entrepreneurs received little
attention; however, since the launching of the reform policies in the early
1980s, there has been renewed interest in the contributions of China’s early
entrepreneurs to modern economic development.
The final section of this chapter will look at the wartime and postwar
reorganization of Chinese industry that laid the base for the nationalization of
industry in the early 1950s and offer an evaluation of modern industrial
growth and its impact on the development of the Chinese economy in the
first half of the twentieth century.
Industrialization from Abroad

Modern mechanized industry was introduced to China in the mid-nineteenth
century. China’s experience differed from that of the early-industrializing
nations, including Britain and the United States, where industrialization
began with light industry, primarily textiles. One of the distinctive character-
istics of the Chinese case – a characteristic it shares with Japan and some other
late-developing countries – is that mechanization was first introduced in
defense-related industries, including munitions and shipbuilding. The first
mechanized factories were transplants from the West, importing machinery,
raw materials, and Western engineers who supervised production. By the
early twentieth century, China began to import machinery from Japan, and
Chinese engineers who had studied in Europe, the United States, and Japan
began to replace foreign engineers, assuming leadership roles in modern
enterprises.
In the 1860s the Qing government, involved in a long internal war to
suppress the Taiping Rebellion, decided to create a modern defense industry
to ensure that in the future China could provision its military and build
modern ships to control shipping in its own waters. The government estab-
lished three state-owned enterprises, the Jiangnan Arsenal in Shanghai (1865),
127

the Mawei Shipbuilding Yard in Fuzhou (1866), and the Tianjin Arsenal
(1867). The Jiangnan Arsenal was the most successful of these enterprises:
by the early 1890s the Jiangnan Arsenal had 662 machines and employed 2,913
workers and had developed into a major shipbuilding company that is still
active in the twenty-first century.3 The Mawei Shipbuilding Yard and the
Tianjin Arsenal had less happy fates. The Mawei Shipbuilding Yard was
partially destroyed during the 1884 Sino-French war, and the Tianjin
Arsenal was destroyed by the troops of the Eight-Power Allied Expedition
Force that occupied north China during the Boxer Rebellion.
Ship repair and servicing facilities established by foreign companies in the
treaty ports after the 1840s were a second source of the modern defense
industries. Foreign companies established docking facilities in Hong Kong,
Guangzhou, Shanghai, and other port cities to repair and service commercial
ships; some of those establishments later developed shipbuilding capacity.
For example, two companies established in Shanghai, the Boyd Company
(1862) and S.C. Farnham (1863), by the 1880s had the capacity to construct
2,000-ton ships. Each of the firms had divisions for forging parts and other
metalwork and employed several hundred Chinese workers.4
These factories used imported machinery and imported raw materials, and
Western engineers provided technical direction. Maritime Customs statistics
show a steady increase in machinery imports and raw materials for the
factories after 1880, when such goods were first included in the statistical
records (Figure 4.2). Scrap metal occupied a large share of the imports,
including scrap metal from European factories, as well as used horseshoes
from the cities of Glasgow and London.5
Technology Transfer
Acquiring technology from overseas was the first step in building a modern
industrial base. However, for sustained growth the receiving country needed to
develop capacity to educate engineers and skilled workers to take over manage-
ment of the transplanted factories. In the late nineteenth century, there was only
limited progress in China in transferring technological knowledge and skills. In
3
Shanghai Shehui Kexueyuan Jingji Yanjiusu 上海社会科学院经济研究所 (ed.), 江南
造船厂厂史 (A Factory History of the Jiangnan Shipbuilding Company) (Shanghai,
Shanghai Shehui kexueyuan chubanshe, 1983).
4
Sun Yutang 孙毓棠 (ed.), 中国近代工业史资料 (Materials on the History of Chinese
Modern Industry), 2 vols., part 1 (Beijing, Kexue chubanshe, 1957).
5
China Imperial Maritime Customs, Returns of Trade and Trade Reports for the Year of 1891
(Shanghai, 1892).
128

1,000 HK tael
16,000
Machinery
14,000
Iron and iron
products
12,000 Steel and
steel products
10,000 Tin plate
8,000
6,000
4,000
2,000
0
1886
87
89
90
91
92
93
94
95
96
97
98
99
1900
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
88
Figure 4.2 Imports of machinery, iron, and steel, 1886–1915

Source: China Imperial Maritime Customs, Returns of Trade and Trade Reports, 1887–1910,
1911–1915
the early factories, including both defense-related and light-industrial operations,

foreign engineers supervised work. The Qing government was aware of the
problem and launched a major project to develop human capital by sending
Chinese students to study in Europe and America; however, only about 4–5
percent of the first generation of returnees took positions in manufacturing.6
Overall, Chinese modern manufacturing in the late nineteenth century was
almost completely dependent on European and American machinery and engin-
eering skills.
While the formal government programs had little impact on developing
higher-level engineering skills, the state-run defense factories did serve as
incubators for building a cadre of skilled workers who played a major role in
the development of the Chinese machine-making industry. Workers who
had mastered skills through employment in the defense factories were
6
Xu Dingxin 徐鼎新, 中国近代企业的科技力量与科技效应 (Impact and Scientific–
Technical Levels in China’s Modern Enterprises) (Shanghai, Shanghai Shehui kexue
chubanshe, 1995).
129

responsible for the establishment of most of the early twentieth-century

machine workshops and factories.
Some of the earliest examples of this spin-off phenomenon are Chinese-
owned machine-making enterprises, including Da Long in Shanghai and the
machine-making workshops of Tianjin’s Santiaoshi district.7 Da Long began
repairing ships and later began to make parts for modern textile enterprises;
the Tianjin machine shops produced parts for modern industry as well as iron
gear looms for rural industrial districts. In the mid-Yangzi city of Hanyang,
the Zhou Heng Shun machine-making factory, which developed out of
a handicraft metalworking shop, grew into a mechanized factory that could
produce steam engines and pumps for mining operations.8
Late Qing Technical Education Efforts

Following China’s defeat in the Sino-Japanese War (1894–1895), ever-broadening
groups came to share a sense of crisis about the survival of China. Zheng
Guanying’s popular book Warnings to a Prosperous Age raised the specter of
a “commercial war” with the West, inspiring efforts to “save the country
through enterprise.” During the first decade of the twentieth century, the
national government and some provincial governments established training
schools for technicians and scholarship programs to send talented students
overseas. Sheng Xuanhuai, one of the leading officials promoting modern
industry, played a central role in developing engineering education, establishing
the Beiyang Xuetang (forerunner of Tianjin University) in 1895 and the Nanyang
Gongxue (forerunner of Shanghai Jiaotong University) in 1896.
Technical training entered a new stage under the New Government
policies, instituted by Yuan Shikai after he was appointed governor-general
of Zhili province in the wake of the Boxer Rebellion. Yuan’s industrial
promotion policies were based on practices of Meiji Japan. He hired
Japanese advisers and recruited Japanese teachers to staff the Zhili
7
Shanghai shi Gongshang Xingzheng Guanliju 上海市工商行政管理局 (eds.), 上海民
族橡胶工业 (Shanghai National-Capital Rubber Industry) (Beijing, Zhonghua shuju,
1979); Shanghai Shehui Kexueyuan Jingji Yanjiusuo 上海社会科学院经济研究所 (ed.),
大隆机器厂的产生、发展和改造 (The Birth, Development, and Reform of the
Dalong Machine Making Company) (Shanghai, Shanghai renmin chubanshe, 1980);
Nankai daxue lishixi 南开大学历史系 (ed.), 天津三条石早期工业资料调查 (A
Survey of Materials on Early Industry in Tianjin’s Santiaoshi), unpublished manuscript,
1958; Gail Hershatter, “Flying Hammers, Walking Chisels: The Workers of Santiaoshi,”
Modern China 9 (October 1983), 387–419.
8
Yan Peng 严鹏, 战略性工业化的曲折展开: 中国机器工业的演化 1900–1957 (The
Torturous Development of Strategic Industrialization: The Evolution of China’s
Machine-Making Industry, 1900–1957) (Shanghai, Shanghai renmin chubanshe, 2015),
pp. 39–40.
130

Gongyiju, which promoted light industries. The Zhili Gongyiju established

a training school, a demonstration factory, and a product exhibition hall. Six
hundred student apprentices studied weaving, dyeing, textile design, soap-
making, carpentry, ceramics, matches, and drafting. Some of the students
were sent to Japan to study at technical schools, and some served apprentice-
ships in Japanese factories.9
The Role of Japanese Technology Transfer

While earlier studies have focused on technology transfer from Britain,
Europe, and the United States, Japanese technology also played a significant
role in Chinese industrial development, particularly in the silk and cotton
industries. Many of the leaders of the postwar textile industry in China,
including Zhu Xianfang, head of the Nantong Textile Institute, and Zhang
Fangzuo, head of the Textile Research Institute, had studied in Japan.
From the 1920s on, Japanese textile technology came to be used widely in
China, with the introduction of the Minorikawa multi-thread reeling
machine and the Toyoda automatic loom. Japanese technology also had
a major impact on the handicraft weaving industry, with the import and
spread of the iron gear loom. The iron gear or treadle loom, invented in Japan
in the late 1880s and introduced to China by the Zhili Gongyiju, functioned
like a power loom, but with “power” supplied by the weaver pumping the
treadles. The loom had been used in rural workshops in Japan that did not
have access to electricity and came to be widely used in weaving workshops
all over China into the 1950s.10 Japanese technology played a role in the
development of other light industries, including matches and rubber shoes;
overseas Chinese entrepreneurs from Japan included the founders of the
Nanyang Brothers Tobacco Company, the chief rival of the British American
Tobacco Company.11
A third major transfer from Japan in the 1920s and 1930s was on-the-job
training (OJT) regimes for technicians and skilled workers, which spread
from the Japanese-owned cotton mills in China to their Chinese-owned
9
L. Grove, “Technology Transfer, Imitation and Local Production: The Soap Industry in
Early Twentieth Century Tianjin,” in K. Furuta and L. Grove, Imitation, Counterfeiting
and the Quality of Goods in Modern Asian History (Singapore, Springer, 2017), pp. 161–82;
D. Reynolds and C. Reynolds, East Meets West: Chinese Discover the Modern World in
Japan, 1854–1898 (Ann Arbor, Association for Asian Studies, 2014).
10
L. Grove, A Chinese Economic Revolution: Rural Entrepreneurship in the Twentieth Century
(Lanham, Rowman & Littlefield Publishers Inc., 2006).
11
S. Cochran, Big Business in China: Sino-Foreign Rivalry in the Cigarette Industry, 1890–1930
131

rivals.12 In the case of the silk industry, there had been little technological
innovation following the introduction of European technology in the late
nineteenth century; this began to change in the early twentieth century,
when silk filatures started to provide training for their workers.
Privately Funded Technology Education

Technical education was not exclusively the work of public institutions.
Entrepreneurs also sponsored specialized training schools. In 1912 the well-
known textile entrepreneur Zhang Jian set up a Textile Training Institute in
Nantong, the home base of his Da Sheng cotton mills, and by the latter half of
the 1930s two-thirds of the textile engineers working in Chinese mills were
graduates of his school.13
Export-Oriented Industrialization
The second path to industrialization was modeled on Western experience in
producing light industrial goods, with factories combining imported machin-
ery with intensive labor. Sugihara Kaoru has argued that this labor-intensive
industrialization model was characteristic of industrial production in both
China and Japan.14 Private capital played the central role in light industrial-
ization. We have divided our consideration into two parts, first looking at
export-oriented production, which developed relatively early, and then con-
sidering industries which supplied substitutes for imports in domestic
markets.
Export-oriented production was not a new phenomenon in the nineteenth
century. Chinese craftsmen had engaged in export-oriented production for
centuries. Chinese ceramics are found in archaeological sites throughout
Southeast Asia and the Middle East, and reports of Chinese silk can be
found in Roman records. In more recent centuries, eighteenth-century crafts-
men produced a variety of goods for export. Among the goods in the Canton
trade were Chinese specialties, including silk, tea, and ceramics, but also
made-to-order furniture; ceramic dinner and tea sets with company coats of
arms; genre paintings; glass paintings, including reproductions of Peale’s
12
Tomizawa Yoshia 富澤芳亜, 近代中国の工業教育と紡績技術者養成 (Technical
Education in Modern China and the Development of Textile Engineers), 経済史研究
20 (2017), 47–96.
13
Xu, 中国近代企业的科技力量与科技效应, 6.
14
K. Sugihara, “Labour-Intensive Industrialization in Global History: An Interpretation of
East Asian Experiences,” in G. Austin and K. Sugihara (eds.), Labour-Intensive
Industrialization in Global History (London and New York, Routledge, 2013), pp. 20–64.
132

famous portrait of George Washington; and much else. What was new in the
late nineteenth century was the use of machines, imported from the West,
that changed manufacturing processes.
In this section on export-oriented industries, we will first review the
introduction of new technologies and their impact on the silk industry and
then briefly consider other export-oriented industries, including those that
processed agricultural raw materials, as well as the development of new
export industries that drew on the abundant supply of rural surplus labor,
organized in new ways and serving new markets.
The Movement for Modern Filatures

The Chinese began producing and trading silk fabrics several thousand years
ago, and silk was already an important product in the Canton export trade
before the opening of the treaty ports in the mid-nineteenth century. The
Yangzi delta area had developed as the center of Chinese silk production
during the Song and Yuan dynasties. During the Ming dynasty, silk produc-
tion flourished in areas around Lake Tai, and by the sixteenth and seven-
teenth centuries, Huzhou prefecture, south of the lake, had established
a reputation for its high-quality Tsatlee silk. After the opening of the treaty
ports, demand for Chinese raw silk exploded and silk became the most
important commodity in China’s export trade. A silkworm epidemic in
Europe contributed to rising demand, but equally important was a more
than fivefold increase in the per capita consumption of silk goods in the
United States, Great Britain, Germany, and France.15
The boom in the market for Chinese raw silk did not last. In China raw silk
was produced in rural households using handicraft methods, and the quality
of the silk varied depending on the skills of individual producers. At the
height of the boom even inferior raw silk could be sold for a good price, but
the mixing of good and inferior qualities damaged the reputation of Chinese
silk.16 As questions were raised about quality, the price of raw silk traded in
Shanghai fell: the average price in 1874–1883 was 30 percent lower than ten
years earlier.17
15
G. Federico, An Economic History of the Silk Industry (Cambridge, Cambridge University
Press, 1997).
16
L. Li, China’s Silk Trade: Traditional Industry in the Modern World, 1842–1937 (Cambridge,
Council on East Asian Studies, 1981).
17
Suzuki Tomoo 鈴木智夫, 洋務運動の研究 (Research on the Self-Strengthening
Movement) (Tokyo, Kyūko Shoin, 1992), p. 322.
133

Falling silk prices were the impetus for foreign trading firms to introduce
mechanized silk filatures to improve the quality of silk. The American trading
firm Russell & Co. set up the first modern silk filature in 1878 and a leading
Chinese silk merchant, Huang Zongxian, established a filature in 1880. By the
early twentieth century, 80 to 90 percent of the output of the mechanized silk
filatures was destined for the export market.18
Although the factories referred to above are usually cited as the first
modern silk filatures, there had been efforts to set up mechanized filatures
twenty years earlier. The British trading firm Jardine, Matheson had sup-
ported the establishment of a mechanized filature in Shanghai in 1861.
However, difficulties in obtaining cocoons, together with a shortage of
skilled labor, made the factory financially unstable, and it closed in 1869.19
Other efforts to set up mechanized filatures in the 1870s also ran into
problems: local merchants and members of the elite who were worried that
mechanized filatures would destroy the existing handicraft industry, impov-
erishing thousands of rural producers, organized protests. As a result, when
mechanized filatures were established in the early 1880s, the Shanghai intend-
ant ordered the factories to close. Representatives of foreign governments
entered the dispute, eventually reaching an agreement with the Liangjiang
governor-general that allowed the mechanized filatures to operate; however,
the factories were assessed a new tax and limits were placed on the number of
factories.20
At about the same time that these struggles were taking place in the
Shanghai region, an overseas Chinese merchant from Vietnam, Chen
Qiyuan, set up a mechanized filature in Nanhai county, Guangdong; other
merchants in the region soon followed his example. Chen’s factory used
treadle-style reeling equipment rather than steam-powered machines. Chen’s
efforts, like the early efforts to establish modern filatures in Shanghai, met
with protests. As the number of silk reeling workshops increased, there was
a shortage of cocoons, and traditional silk-goods merchants began to protest;
in 1881 protesters attacked the factory. Worried about social stability, the
local government ordered the factories to temporarily close.
18
Xu Xinwu 徐新吾, 中国近代缫丝工业史 (A History of Chinese Raw Silk
Manufacture) (Shanghai, Shanghai renmin chubanshe, 1990).
19
Kubo Tōru 久保亨, Kajima Jun 加島潤, and Kigoshi Yoshinori 木越義則, 統計でみ
る中国近現代経済史 (Economic History of Modern China: Based on Statistical Data)
(Tokyo, Tōkyō daigaku shuppankai, 2016), p. 33.
20
Suzuki, 洋務運動の研究, pp. 324–8.
134

250
Shanghai
200 Wuxi
Other Yangzi delta
150 Guangdong
Sichuan
100
50
0
1890 90 1902 12 20 26 30 36
Figure 4.3 Silk reeling filatures
Source: Kubo Tōru 久保亨, Kajima Jun 加島潤 and Kigoshi Yoshinori 木越義則 (eds.),
統計でみる中国近現代経済史 (Economic History of Modern China: Based on
Statistical Data) (Tokyo, Tōkyō daigaku shuppankai, 2016), p. 36
While some have argued that the opposition to the introduction of

modern silk filatures reflected the conservative nature of Chinese society,
we should not overemphasize the protests, since from the 1890s to the 1920s
modern silk filatures were established in many parts of the country, with
large concentrations in Wuxi in Jiangsu province and in Sichuan province, as
well as in Shanghai and Guangdong. The number of mechanized filatures
grew from eighty-seven in 1894 to 115 in 1902, 171 in 1910, and 274 in 1918
(Figure 4.3).21 Patterns of production varied from region to region. Soda has
argued that three factors contributed to the success of the Shanghai modern
filatures: first, a steady supply of women workers from northern Jiangsu
province; second, the establishment of a system of cocoon brokers which
stabilized the market for raw materials; and third, the establishment of
a system of rental factories, which reduced start-up costs.22 In Wuxi, which
was a relative newcomer to sericulture, there was no local tradition of silk
hand-reeling; rural producers specialized in raising silkworms and selling the
cocoons to modern filatures.23 In Shunde there were various styles of
21
Xu, 中国近代缫丝, 142.
22
Soda Saburo 曽田三郎, 中国近代製糸業史の研究 (Research on China’s Modern
Silk Industry) (Tokyo, Kyūko Shoin, 1994).
23
L.S. Bell, One Industry, Two Chinas: Silk Filatures and Peasant-Family Production in Wuxi,
1865–1937 (Stanford, Stanford University Press, 1999).
135

management; at the peak, some 150,000 women worked in filatures, and the
area became well known for its independent women workers who refused to
marry.24
In the mid-nineteenth century, the advent of silkworm disease in Europe
presented new market opportunities, but Japan also rose as a strong competi-
tor, soon surpassing China as an exporter of raw silk.25 To understand the
competition, we need to consider the technologies in different production
centers. Shanghai’s mechanized filatures used Italian technology that separ-
ated the boiling and reeling processes, directly reeling the silk. While this
method was not as efficient as other reeling technologies, it produced the
high-quality raw silk that was in great demand in the French textile center of
Lyon for the production of luxury fabrics. In contrast, Japanese filatures used
“Suwa-style” equipment in which the boiling of cocoons and reeling were
completed in one process. The Suwa-style process produced a consistent
quality of medium-grade raw silk at a low cost; such raw silk was in high
demand in the United States, where it was well adapted as weft for power-
loom weaving.26 The differential in production costs gave Japan an advan-
tage, and the gap in exports between Japan and China steadily increased. As
for the raw silk produced in Guangdong, most was of lower quality and was
sold in the domestic market or exported to the United States.
One of the keys to Japanese success was a government-sponsored effort to
improve sericulture techniques, including the establishment of special
schools to train instructors in improved techniques. By the late 1890s local
gentry and county magistrates in the silk regions of Jiangnan also began to set
up schools, the most successful of which were the Hangzhou School for
Sericulture and the Hushuguan Sericulture School for Girls; many of the
graduates of these schools were sent to Japan for further studies, including Fei
Dasheng, who was the elder sister of the anthropologist Fei Xiaotong and
played a central role in modernizing the Wuxi silk industry.27
With the beginning of the Depression in 1929, demand for luxury products
like silk sharply declined, and the Chinese silk industry suffered a major blow.
This downturn coincided with the growing popularity of rayon, which
served as a cheap substitute for silk. As the recovery from the Depression
began, there was a shift in the location of the Chinese mechanized silk
24
Lü Xuehai 吕学海, 顺德丝业调查报告 (Research Report on the Shunde Silk
Industry), manuscript copy, Nankai University Library, 1938.
25
D. Ma, “The Modern Silk Road: The Global Raw-Silk Market, 1850–1930,” Journal of
Economic History 56.2 (1996), 330–55.
26
S. Soda, 中国近代製糸業. 27 Bell, One Industry.
136

industry. Jiangnan production shifted from Shanghai, where capital shortage

had resulted in instability, to Wuxi, which was closer to the cocoon-
producing regions.28 The shift to Wuxi was related to the development,
with provincial government encouragement, of infrastructure to support
sericulture, including sericulture schools, agricultural extension, and manda-
tory use of inspected silkworm eggs. Reforms of the silk industry were
promoted by the Wuxi Cocoon Merchant Guild, whose members included
most of the larger silk firms.
Scholars have offered different explanations for what was once viewed as
Japan’s success and China’s failure in the global silk markets of the 1920s and
1930s. Some praised Japan’s initiatives at quality control and lamented China’s
failure to adopt similar policies until later in the 1930s; others argued that the
commercial systems involved in Chinese cocoon supply jacked up prices and
reduced profits; while others put the blame on the slow response of the
Chinese state.29 Debin Ma, focusing on the period up to 1900, argued that
technological adaptation was more successful in Japan and Guangdong than
in the Yangzi delta region and that development of short-term capital
markets was equally important.30
A careful look at the records shows that Chinese silk exports continued to
rise throughout the 1920s and that the number of silk filatures continued to
increase. So rather than Japanese success and Chinese failure, we might see
Japan as very successful, and China also as successful, but to a lesser degree.
The major differences were directly related to the markets each country
targeted: Japan focused on the booming market for middle-quality silk in the
US, while China targeted the European market for high-quality silk, which
stagnated in the 1920s. Moreover, in the 1930s the silk industries of both
countries entered a new stage of growth. One sign of this was an effort by
China’s largest silk company, Yong Tai, to directly enter the New York silk
market.31 Another sign of China’s relative success was the response of
28
Okumura Satoshi 奥村哲, 中国の資本主義と社会主義：近現代史像の再構成
(Capitalism and Socialism in China: Reformulating Images of Modern and
Contemporary History) (Tokyo, Sakurai shoten, 2004).
29
Li, Chinese Silk Industry; R. Eng, Economic Imperialism in China: Silk Production and
Exports (Berkeley, University of California Press, 1986); Federico, Economic History.
K. Furuta, “Peasant, Market Town, and Handicraft Technology,” in A. Hayami and
Y. Tsubouchi (eds.), Economic and Demographic Development in Rice Producing Societies:
Some Aspects of East Asian Economic History, 1500–1900 (Proceedings of the 10th
International Economic History Congress, Leuven, 1989).
30
D. Ma, “Between Cottage and Factory: The Evolution of Chinese and Japanese
Silk-Reeling Industries in the Latter Half of the Nineteenth Century,” Journal of the
Asia Pacific Economy 10.2 (2005), 195–213.
31
Okumura, 中国の資本主義.
137

Japanese silk companies to challenges from their Chinese rivals: in the early
war years, Japanese silk companies lobbied occupation authorities to limit
production in China’s modern silk factories.
Mechanized Silk Filatures and the Handicraft Silk Industry

The silk complex – the raising of silkworms, reeling raw silk, and silk
weaving – was for centuries one of China’s major handicraft industries.
Rural households cultivated mulberry trees, raised silkworms, and used
hand machines to reel the raw silk. Some households also wove silk fabrics,
while others sold raw silk to merchants who sold it or put it out to weavers
working under their direction. The development of mechanized silk filatures
pulled apart the silk complex, creating two separate and distinct markets: the
first market was for raw silk processed in mechanized filatures, and was, at
least at first, primarily for the export market. The other market was for raw
silk produced in rural homes or semi-industrial handicraft workshops using
hand- or foot-powered reeling equipment; this market came to primarily serve
domestic customers. From the 1870s to 1929 we see a slow, but steady,
increase in China’s production of raw silk, and a slow, but steady, move
from home processing toward processing of cocoons in modern filatures or in
small workshops using hand- or foot-powered machines (Figure 4.4). The
percentage of silk processed in mechanized filatures advanced from 13.1 percent
in 1881–1885, to 25 percent by the first decade of the twentieth century, to
60 percent in 1936.32 For producers of raw silk, the changes came over a long
period of time, and as late as the mid-1930s, 40 percent of China’s raw silk was
still produced by handicraft methods. When Lü Xuehai investigated the
Shunde silk industry in Guangdong in 1937, he found that households involved
in sericulture moved back and forth between selling cocoons to filatures and
processing by hand at home, adjusting their activities to market prices.
New Export-Oriented Handicraft Industries

Silk was the most important semi-processed export commodity in China in
the late nineteenth and early twentieth centuries. However, following the
opening of the treaty ports, foreign merchants identified other potential
export opportunities: some linked farmers to foreign markets, and others
made use of the large supply of underemployed female labor in the country-
side. An example of the first was the export of dried-egg products. The egg
32
Wang Xiang 王翔, 中国近代手工业史稿 (A History of China’s Modern Handicraft
Industry) (Shanghai, Shanghai renmin chubanshe, 2012), p. 295.
138

ton
12,000
Filature
10,000
Handicraft
8,000
6,000
4,000
2,000
0
80
00
85
90
95
05
10
15
20
25
30
35
40
18
19
Figure 4.4 Silk exports by production source

Source: L. Hsiao, China’s Foreign Trade Statistics, 1864–1949 (Cambridge, MA, Harvard
University Press, 1974), pp. 33, 36 for filature, pp. 31, 32, 34, 35 for handicraft
export business, which was started by a German merchant in Hankou in 1887,

flourished during the First World War when European demand for egg
products soared.33 Hairnets, straw hats, and embroidery are other examples
of new export items that made use of rural labor. Many of these new rural
handicrafts were introduced by Western missionaries, who sought ways to
provide remunerative labor for the rural poor. In some cases, the new
processing industries used local raw materials, for example the straw used
for weaving braid for hats, but in other cases the raw materials were
imported. When missionaries first promoted hairnet weaving as a rural
industry, human hair was imported from Europe. Later, Chinese hairnet
producers switched to use Chinese hair. In the case of the lace and embroi-
dery industries, the designs came from overseas, with rural women working
in their homes under putting-out arrangements.34
33
Yoshida Tateichiro 吉田建一郎, “戦間期中国における鶏卵：鶏卵加工品輸出と
養鶏業” (The Chinese Poultry Industry and Export of Egg Yolk and Processed Egg
Products during the Interwar Years), 東洋学報 (Journal of East Asia) 86.4 (2005), 503–34.
34
S. Mann, “Women’s Work in the Ningbo Area, 1900–1936,” in T. Rawski and L. Li (eds.),
Chinese History in Economic Perspective (Berkeley, University of California Press, 1992),
139

By the 1930s lists of Chinese exports came to include products of modern

factories as well. Japanese-owned cotton mills in China began to pioneer
markets in Southeast Asia, and in 1935 a Chinese Industrial Goods Export
Association was established to promote the export of manufactured goods.35
Import-Substitution Industrialization
When nineteenth-century British traders lobbied their government to force
open Chinese markets, the motivation was access to one of the world’s
largest markets. However, the market never reached the levels foreign
manufacturers envisioned, in large part because of the strong push-back
from Chinese entrepreneurs, who developed import-substitution industries.
The largest of the import-substitution industries produced consumer goods
like cotton textiles, flour, cigarettes, beer, matches, and rubber shoes, as well
as construction materials, including cement.
Many of the products manufactured in China’s factories and workshops
were new “exotic products”; first introduced from the West, consumption
spread out from the treaty port cities into smaller towns in the interior.36
Some of the new products replaced existing Chinese consumer items: for
example, Western-style candles used for lighting quickly replaced traditional
Chinese candles, just as modern laundry and toilet soaps replaced indigenous
cleansing agents. While some of the products – for example cotton yarn and
cement – were produced in modern factories using imported technology,
others were produced in small factories and workshops, using a mixture of
handicraft and mechanized production processes. In the early years, treaty
restrictions limited the use of protective tariffs and Chinese manufacturers
struggled against cheap imports. After China regained tariff autonomy in
1930, domestic producers were protected by higher tariffs and came to
dominate the domestic market for many consumer goods, as seen in
Table 4.1.
Production processes and product chains in China’s modern consumer-
goods industries linked modern factories with rural handicrafts. The
pp. 243–70; Peng Zeyi 彭泽益 (ed.), 中国近代手工业史资料 (Materials on the

Chinese Modern Handicraft Industry) (Beijing, Sanlian shudian, 1962).
35
Kubo Tōru 久保亨, 戦間期中国〈自立への模索〉：関税通貨政策と経済発展
(China’s Quest for Sovereignty in the Interwar Period: Tariff Policy and Economic
Development) (Tokyo, Tōkyō daigaku shuppankai, 1999).
36
F. Dikötter, Exotic Commodities: Modern Objects and Everyday Life in China (New York,
Columbia University Press, 2007).
140

Table 4.1 Market share of domestic products, 1920–1936 (%)

Cotton yarn Cotton pieces Cement
1920 68.9 19.4 59.9

1930 102.3 55.3 84.6
1936 102.3 86.5 92.0
Source: Kubo, Kajima, and Kigoshi, 統計でみる, pp. 23, 24, 52
best-known example of this phenomenon was use of cotton yarn from urban
spinning mills as raw material in rural weaving.
The handicraft cotton industry was one of the first of China’s trad-
itional industries to feel the impact of the opening of foreign trade in the
mid-nineteenth century. British manufacturers had hoped to find a large
market in China for cotton piece goods, but soon discovered that while
there was a large market for machine yarn, it was more difficult to sell
cotton fabrics. Machine yarns made steady inroads in the Chinese market,
and by the end of the nineteenth century, imports supplied about 40 per-
cent of the total Chinese consumption of yarn.
Seeing the high demand for machine yarns, European and American
trading firms were inspired to establish spinning mills in China. In the
1860s and 1870s two British trading firms in Shanghai drafted a plan to set
up a modern spinning and weaving mill to produce fabrics in imitation of
Chinese “native cloth.”37 Chinese cotton fabric merchants, who con-
trolled the trade in native cloth, opposed the plan. The Qing govern-
ment, concerned about the fate of rural weaving households whose
livelihood depended on income from weaving and worried about allow-
ing foreigners to establish factories within its territory, blocked the plan.
Following the failure of the foreign investors to receive approval for
a modern cotton mill, a consortium of Chinese merchants, including the
Guangdong comprador Zheng Guanying (1842–1922), gained support from
the leading official Li Hongzhang (1823–1901) for a similar plan. The company,
based on a collaboration between private capital and the state, was estab-
lished with the explicit purpose of providing substitutes for imports to
improve China’s unfavorable balance of trade. At that time, cotton cloth
accounted for 30 percent of China’s total imports.
37
Suzuki, 洋務運動, 328–9.
141

It was not long before more cotton mills were established in Shanghai. The
Treaty of Shimonoseki, ending the Sino-Japanese War, for the first time
explicitly allowed foreign capital to establish factories in China, and this
shift touched off a factory-founding rush, first by European and American
investors, and then by Chinese entrepreneurs.38
Production in cotton spinning mills grew rapidly in the 1910s and the 1920s,
and in cotton weaving mills in the 1920s and 1930s: by the 1930s, both sectors
were supplying most of the domestic demand (Table 4.2). Japanese-owned
and Chinese-owned firms played the central roles in this rapid expansion,
while many of the early European and American firms failed. Among the
European and American firms, the only successful venture was the Ewo (Yi
He) spinning mill founded by Jardine, Matheson, the leading British trading
company.
What accounted for the Western investors’ bad track record? Nakai has
argued that many of the foreign firms invested very little of their own capital,
leaving them dependent on Chinese merchants for the supply of raw cotton
and the sale of finished goods, resulting in high capital and distribution costs.
Beginning in the 1910s, some of the firms declared bankruptcy and others sold
out to Japanese or Chinese buyers.
A second question is related to the strong role of Japanese capital in the
Chinese cotton industry. What inspired Japanese investment and why was it
so successful? As background, we should note that Japanese cotton compan-
ies had been important players in the Chinese market from the late nine-
teenth century. As the Japanese cotton industry developed in the 1880s and
1890s, production capacity exceeded demand in the Japanese domestic mar-
ket. Firms had to export if they wanted to continue to grow, and Japanese
firms had built up capital holdings, so they had money to invest. At the same
time, stricter enforcement of the 5 percent import tariff rate was making it
more difficult for Japanese goods to compete in the Chinese market. This
combination of factors led Japanese cotton firms to invest in China. The
Japanese investments were made with adequate capital backing, overcoming
the problems that had faced Western firms.39
Let us now turn to the fate of the Chinese-owned cotton mills. How were
they able to survive in a situation where their capital and technological resources
38
Kubo, Kajima, and Kigoshi, 統計でみる中国近現代経済史, p. 25; Nakai Hideki 中井
英基, 張謇と中国近代企業 (Zhang Jian and the Modern Chinese Enterprise)
(Sapporo, Hokkaidō daiguku toshokankokai, 1996).
39
Takamura Naosuke 高村直助, 近代日本綿業と中国 (The Modern Japanese Cotton
Industry and China) (Tokyo, Tōkyō daigaku shuppankai, 1982).
142

Table 4.2 Production and trade of machine-made cotton yarn and cotton
pieces
Production of Import of Production of Import of Export of
Year cotton yarn cotton yarn cotton pieces cotton pieces cotton pieces
(10,000 tonnes) (10,000 tonnes) (million m2) (million m2) (million m2)
1880 － 0.9 － 451.6 －

1890 0.4 6.5 5.2 518.2 －
1900 6.1 9.0 42.3 531.6 －
1910 8.6 13.8 50.5 566.5 －
1920 16.8 8.0 160.4 670.5 2.2
1930 44.0 1.0 565.3 500.7 43.9
1936 39.7 0.5 1,203.1 196.5 8.1
1950 43.7 － 2,520.0 － 27.7
Source: Kubo, Kajima, and Kigoshi, 統計でみる, pp. 19–20
were less than those of their Japanese rivals? The rivalry between Japanese and
Chinese firms played out in different ways in different parts of the country. In
northern China (Tianjin and Qingdao), where Japanese military and political
influence was particularly strong, Japanese firms held the leading position. In
Shanghai, competition was even; sometimes Japanese firms moved ahead, and
sometimes Chinese firms seemed to be in a more favorable position. Chinese
firms dominated in the rest of the country. Mori has noted that Japanese and
Chinese firms targeted different markets, with Chinese spinning firms producing
lower-count yarns, while Japanese firms specialized in higher-count yarns.40 This
division was not absolute and was reversed in some cases. When we look at
profitability, we can see that the profits of some Chinese-owned mills – for
example Yong An and Shen Xin in Shanghai, Hua Xin in Ji Xian (Henan), and Jin
Hua in Yuci (Shanxi) – equaled those of the Japanese-owned mills. The Chinese-
owned mills in Shanghai benefited from superior management, good engineers,
and access to relatively cheap imported cotton. The Chinese-owned firms in the
interior of north China and in the Zhejiang–Jiangsu region were close to cotton-
growing regions which supplied cheaper cotton; located in areas where rural
weaving was highly developed, there was a ready market for their output.41
40
Mori Tokihiko 森時彦, 中国近代綿業史の研究 (A Study of the History of the
Modern Chinese Cotton Industry) (Kyoto, Kyōto daigaku shuppankai, 2001).
41
Kubo Tōru 久保亨, 戦間期中国の綿業と企業経営 (Cotton Industry and
Management in Wartime China) (Tokyo, Kyūko Shoin, 2005); Juanjuan Peng, The
Yudahua Business Group in China’s Early Industrializaton (Lanham, Lexington Books, 2020).
143

The domestic cotton industry benefited from favorable conditions

between 1910 and 1930. During World War I, imports sharply declined,
creating what Marie-Claire Bergère described as the “golden age of Chinese
capital.” Other factors that have been cited as contributing to the growth of
the cotton industry include government policies to protect native industries,
developments in the Chinese machine-making industry, and co-operative
efforts by government and nongovernment groups to promote a shift to
longer-staple American cotton which was more suitable for machine
spinning.42
Modern Cotton Mills and Competition with Handicraft

One of the central debates in Chinese economic history has been the question
of what happened to traditional hand-loom weaving when faced with com-
petition from modern machine production. The traditional system of cloth
production was embedded in the rural household economy, which linked
agricultural production of fiber with hand spinning and weaving. In China
the spinning and weaving parts of this production nexus were usually joined
together. While some households sold homespun yarn in the periodic mar-
kets in Jiangnan, the trade was limited, and merchants were only able to gain
control over the production process through putting-out arrangements after
machine yarns became available.43
Access to machine yarn produced change in two directions: first, cotton
weaving spread to regions of the country which did not grow cotton and had
previously imported finished fabrics from the Jiangnan region. Second, hand-
loom weavers in many of the older cotton-producing regions began to
substitute machine-spun yarn for homespun.44 The spread of the use of
machine-spun yarns was uneven, progressing rapidly in some regions,
much more slowly in others. By the early 1920s hand spinning had virtually
disappeared in the old cotton centers, and by 1936 roughly 75 percent of the
yarn consumed in China came from modern spinning mills. A study of
Jiangsu province by Ma Junya and Tim Wright provides a detailed examin-
ation of regional differences. They argue that weaving declined in the
42
M.-C. Bergère, The Golden Age of the Chinese Bourgeoisie (Cambridge, Cambridge
University Press, 1989). Kubo, 戦間期中国〈自立〉.
43
Nishijima Sadao 西嶋定生, “The Formation of the Early Chinese Cotton Industry,”
translated in L. Grove and C. Daniels (eds.), State and Society in China (Tokyo,
University of Tokyo Press, 1984), pp. 17–77.
44
L. Grove, “Rural Manufacture in China’s Cotton Industry, 1890–1990,” in D. Farnie and
D. Jeremy (eds.), The Fibre That Changed the World: The Cotton Industry in International
Perspective, 1600–1990s (Oxford, Oxford University Press, 2004), pp. 431–59.
144

Songjiang–Taicang region, which had been China’s leading cotton-weaving

center since the sixteenth century. In the Songjiang–Taicang region, labor
moved from the countryside into Shanghai. At the same time, handicraft
weaving using machine yarn flourished in regions just north of the Yangzi
river, while in the far north of the province, rural households that had
abandoned weaving before machine yarn became available devoted them-
selves to farming.45
To understand what happened to hand weaving we need to briefly explore
changes in technology. In 1840 home weavers used various types of wooden
looms that produced “native cloth,” a narrow, relatively thick fabric made
with low-count cotton yarn. This competed in the market with much wider
“foreign cloth” that was machine woven using higher-count yarns. By the late
nineteenth century, hand weavers were beginning to use the “flying shuttle,”
which produced wider fabrics and doubled output per day. The next step was
the introduction of the iron gear loom, which again doubled the daily volume
of cloth in comparison with a wooden loom with a flying shuttle.
These improvements in technology led to the establishment of new
industrial districts that produced large volumes of cloth, competing with
imported goods and the output of mechanized weaving factories. The best
known of these districts were in north China – Gaoyang and Baodi in Hebei
province and Wei Xian in Shandong. They represented a new mode of
production that combined weaving on the iron gear loom with finishing
and dyeing in workshops and factories that used imported chemical dyes and
mechanized finishing equipment. The flexibility of the technology allowed
for production of a wide variety of stripes, checks, and patterned fabrics.
Wholesale merchants in the industrial districts managed production as well
as sales through their nationwide networks of sales branches that closely
linked production to market demand.46 By the 1920s and 1930s, the iron gear
loom had spread from north China to small workshops and factories in the
Jiangnan region. Xu Xinwu estimated that by 1930 there were 1,500 small
weaving factories in the Shanghai region alone, producing 3 million bolts of
cloth a year.47
45
J. Ma and T. Wright, “Industrialization and Handicraft Cloth: The Jiangsu Peasant
Economy in the Late Nineteenth and Early Twentieth Centuries,” Modern Asian Studies
44.6 (October 2010), 1337–72.
46
Grove, Economic Revolution.
47
Xu Xinwu 徐新吾 (ed.), 江南土布史 (A History of Native Cloth in Jiangnan)
(Shanghai, Shanghai Shehui kexue chubanshe, 1992).
145

The small weaving factories that flourished in the years before the war are
prime examples of what Peng Nansheng has termed semi-industrialization,
a style of operation that combined features of handicraft and modern indus-
try: work was done in workshop or factory settings, used a mixture of
machine and human labor, and had close links to the market.48 Semi-
industrial workshops and small factories were located in small towns or
villages. For weavers employed in the workshops and factories, farming
became a sideline activity, with weaving providing the main source of
income. Technologies were mixed: for example, on the eve of the anti-
Japanese War there were 8,000 looms in operation in Jiangsu’s Changshu
county, including eighty-four Toyoda power looms, 300 iron gear looms that
had been motorized, 2,000 iron gear looms powered by treadles, and 4,000 to
5,000 flying-shuttle looms.49
Merchant capital played a central role in organizing the weaving industry,
linking producers with markets. The production style was sharply distin-
guished from that of the modern mechanized factories that produced large
volumes of standard products, which were sold through independent trading
firms. Small-scale producers kept a sharp eye on the market, producing
a changing array of goods, adjusting production to market demand. The
development of modern infrastructure in the twentieth century – telegraphy
for communicating with distant branch offices, the postal service, motorized
vehicles and trains for dispatching goods – facilitated the new styles of
operation.
Other New Import-Substitution Industries

As new consumer products were introduced from overseas, Chinese entre-
preneurs studied production and sales techniques and set up firms to manu-
facture substitutes. The list of such industries and products is long; here we
will look at the production of matches and soap as two examples of light
industry that developed to serve domestic markets. In both cases, technology
transfer from Japan played a crucial role. Soap and matches were introduced
to the Chinese market by Western firms in the treaty ports. Both products
had also been introduced to the Japanese market in the nineteenth century,
and by the 1870s Japanese firms were producing domestic substitutes; by the
early twentieth century, Japanese producers began to market their products
48
Peng Nansheng 彭南生, 半工业化: 近代中国乡村手工业的发展与社会变迁
(Semi-industrialization: The Development of Rural Handicraft Industry in Modern
China and Social Change) (Beijing, Zhonghua shuju, 2007).
49
Xu, 江南土布.
146

in China and other parts of Asia. This process of imitation was repeated in
China, but rather than learning from the Western originators of the products,
Chinese entrepreneurs turned to Japanese imitators for models and technical
advice.50
Our first example is the production of matches. Matches were invented in
Europe in the early nineteenth century and entered the Chinese market in the
1860s. It was not long before Chinese entrepreneurs began to make imita-
tions, and by the late 1920s, 180 workshops and factories were producing
matches for Chinese consumers. By the 1930s there were match factories in
almost every province of China: some were small workshops that used hand
labor; others were modern factories with specialized equipment for both
production and packaging. It was quite common for factories to combine the
use of mechanized production with hand labor by women and children to
produce and pack the matchboxes.51
Matches are a good example of a new “exotic” product that quickly spread
among Chinese consumers to become an essential part of daily life; they are
also a good example of the market-oriented pattern of production of light-
industrial goods and of Japanese influence on the development of a Chinese
industry. The first Chinese-owned factory was established by a returnee
student from Japan, and through the early 1920s many of the Chinese
factories employed Japanese technicians who supervised production and
trained Chinese staff.52 Most match factories, including the smallest operators
with only a few employees, employed sales personnel who managed market-
ing, directly linking production with sales.53
Soap, like matches, was first introduced to the Chinese market through the
treaty ports in the nineteenth century, and use spread quickly, pushed by
early twentieth-century campaigns for a more hygienic lifestyle which was
promoted as one of the central features of modernity. The first Chinese-
owned soap factory was established in Tianjin in 1903 by a patriotic business-
man who was an early promoter of the “national products” movement,
which urged Chinese to produce and consume Chinese-made products.
The Tianjin Soap Company was a spin-off based on technology introduced
50
Furuta and Grove, Imitation, Counterfeiting and the Quality of Goods in Modern Asian
History.
51
Liu Dajun 刘大钧, 中国工业调查报告 (Report on a Survey of Chinese Industry) (1937;
reprint by Fujian Jiaoyu chubanshe, 2010).
52
K.Y. Chan, “Playing with ‘Alien Fire’ (Yanghuo): Matches in Late Nineteenth- and Early
Twentieth-Century China,” in Furuta and Grove, Imitation, Counterfeiting and the
Quality of Goods in Modern Asian History, pp. 203–23.
53
Liu, 中国工业.
147

from Japan by the Zhili Gongyiju. The company employed a graduate of the
Tokyo Higher Industrial School as its chief engineer and used Japanese and
German machinery. The company quickly became the model for other
aspiring entrepreneurs. By the 1930s there were at least thirty-five soap
factories in the city of Tianjin, serving the north China market.54
Soap factories were developed independently in other cities, most of which
followed the same “spin-off” pattern of development. The technology in use
in the soap industry, like that in match production, was relatively easy to
master; factories used a mixture of methods and levels of mechanization, and
in most factories at least some of the processes – most commonly packaging –
was done by hand.
Both in the West and in Japan some firms successfully created dominant
brands that sold in national and international markets; well-known foreign
brands like Leverhulme’s Sunlight, Unilever’s Lux, and Kao from Japan’s
largest soap manufacturer, were aggressively advertised in Chinese periodic-
als in the 1920s and 1930s. In contrast, none of the Chinese-owned soap or
match companies in the pre-World War I I era were able to create national
brands, leaving space in the market for dozens of competitors, each selling in
local or regional markets.
The small workshops and factories in the match and soap industries are
representative of small-scale production in light industry in the prewar period.
These enterprises shared certain common characteristics which can be seen in
contemporary small-scale industry in China in the post-1980 reform era. In terms
of scale, firms were small and organized following the principles of traditional
Chinese business partnerships. Production processes combined the use of
machines with human labor. Products were produced in small lots, and
a single firm often produced small lots of different variations of the same generic
product. Production was closely linked to assessments of market demand, and
producers had their own sales staff which managed liaison with wholesale and
retail customers. Technologies were simple and flexible, allowing for quick
adjustment of product characteristics to meet market demand.
Company Founders and Management Styles

Most of the first generation of industrial entrepreneurs were merchants with
experience in foreign trade, including some who had worked for foreign
54
L. Grove, “Technology Transfer, Imitation and Local Production: The Soap Industry in
Early Twentieth-Century Tianjin,” in Furuta and Grove, Imitation, Counterfeiting and the
Quality of Goods in Modern Asian History, pp. 161–82.
148

companies.55 Zheng Guanying, who played a central role in the establish-

ment of the first modern cotton mill, was a well-known comprador merchant
and public intellectual, and Liu Hongsheng, who established the Da
Zhonghua Match Company and the Shanghai Cement Company, got his
start as an agent selling coal from the British Kailuan Mines. Overseas
Chinese entrepreneurs also made major contributions to China’s industrial-
ization. Guo Le, the head of Yong An, China’s second-largest cotton spinning
and weaving corporation, came from an Overseas Chinese family from
Australia; the brothers Jian Zhaonan and Jian Yujie, who founded the largest
Chinese-owned tobacco company, the Nanyang Brothers Tobacco
Company, and Yu Zhiqing, founder of the largest rubber company, were
Overseas Chinese from Japan; Chen Qiyuan, who established the first mech-
anized silk filature in Guangdong, was from a family that had business
enterprises in Vietnam and had worked in Vietnam for almost twenty
years. Most of the men who founded the mechanized silk filatures in
Shanghai, and later in Wuxi, also began as merchants, as did Yan Yutang,
the first manager of the Da Long Machine Workshop, who began his career
working for one of the foreign-capital ship repair companies.
The Rong brothers, founders of the Mao Xin and Fu Xin flour mills, as well
as China’s largest textile conglomerate, Shen Xin, were unusual in that they
got their start as managers of a native bank (qianzhuang). Another group of
entrepreneurs began from official backgrounds. Zhang Jian, the Nantong
entrepreneur who founded the Da Sheng cotton mill; Xie Nanming, who
established the Yong Tai filature in Wuxi; and Zhou Xuexi, founder of the Qi
Xin Cement Company in Tangshan, all belong to this category. However,
there were only a few entrepreneurs who began their careers as bureaucrats,
and most of the investors in their companies were merchants. Overall,
merchants played the central role in the rush to found modern industries in
the late nineteenth and early twentieth centuries.
There was a distinct shift in the kinds of men who came to head Chinese
industrial companies in the 1920s and 1930s. The new generation of enterprise
leaders were men who had graduated from universities with degrees in
engineering or business management. Our list of such industrial leaders
includes many with overseas degrees. Among the industrial leaders were
graduates from Japanese institutions, including the Tokyo Higher School of
Technology, Kyoto University, and the Kiryu Higher School of Technology;
from MIT and the University of Illinois in the United States; from the Berlin
55
Kubo, 戦間期中国の綿業, 235–58.
149

Technical University in Germany; and from the Jiangnan Arsenal School in

China. Among the foreign-university graduates we can identify several who
were the sons of first-generation entrepreneurs, sent by their fathers to
acquire technological or business skills.
Just as the second-generation managers tended to have professional skills,
management styles also changed: many of the founders organized their firms
following the partnership forms (hegu 合股) that were common in Chinese
commerce.56 As firms grew larger, there was a tendency to move toward
a Western-style limited-liability corporate form; however, there is no ques-
tion that the principles of traditional partnership forms continued to have
a major impact on Chinese business management. This was particularly true
in the early years when Western-style business firms were rare, and investors
were worried about the risks; moreover, alternative investment opportun-
ities, including moneylending, offered much higher interest rates.57 As
a result, firms like the Shen Xin cotton spinning and weaving company
began operations with much borrowed capital. To recruit investors, some
firms, including the Shanghai Merchant Steamship Company and Da Sheng,
guaranteed investors a fixed rate of return (guanli 官利), which they began to
pay even before the enterprise began operations.58
As we move from the 1920s to the 1930s, leading industrial firms began to
find ways around this problem. For example, the Yong An textile company,
strongly committed to the benefits of the limited-liability form of organiza-
tion, found ways to limit the level of guaranteed return. Thus, by the 1930s,
firms that had begun as partnerships expanded their operations, increased
capital, and reorganized as limited-liability companies. There were, of course,
firms that failed: for example, despite the expansion of its scale of operations,
Shen Xin hesitated and by the mid-1930s had piled up huge debts.
Nevertheless, the overall trend was for a shift from the traditional partnership
form to the limited-liability company.
Our explanation of business practices among industrial firms would not be
complete without some reference to the question of family involvement in
management. Just as family management plays a major role in the twenty-
first century in many Chinese firms in Taiwan and Southeast Asia, so families
56
Negishi Tadashi 根岸佶, 商事に関する慣行調査報告書：合股の研究 (A Report
on the Study of Commercial Customs: A Study of Chinese-Style Business Partnerships)
(Tokyo, Tōa Kenkyūsho, 1943).
57
Zhu Yingui 朱荫贵, 引进与变革: 近代中国企业官利制度分析 (Introduction and
Change: An Analysis of the System of “Guaranteed Interest” in Modern Chinese
Enterprises), 近代史研究 4 (2001), 145–67.
58
Zhu, 引进与变革, Introduction; Nakai, 張謇と中国近代企業.
150

were at the core of some of China’s most successful 1930s firms. Firms in
which the family was a central organizing principle included the cotton
textile firms of the Rong and Guo families, the silk firms of the Xie family,
and the Nanyang brothers’ tobacco company. Keeping management within
a family was an effective way to guarantee a family’s control. However, when
we compare the Chinese family firm to family firms in Japan, we also note
some significant differences. First, the man taking control of a company in
the second generation was not always the eldest son. While this practice
reflected China’s system of equal inheritance among sons, in contrast to
Japan’s stress on the rights of the eldest son, it also gave the Chinese-style
family firm flexibility in choosing appropriate successors.
Entrepreneurs in Small-Scale Firms

Entrepreneurial practices among small-scale, semi-industrial firms also
experienced major change in the early twentieth century. Most of the early
proprietors of semi-industrial firms were merchants who moved from trading
into control over production, primarily through putting-out systems.
Partnerships, which were usually drawn up for limited periods of time,
were the most common form of organization. The capital demands of semi-
industrial firms were much lower than those of modern industrial enter-
prises, and since most of the investment was in circulating capital used to buy
raw materials, it was possible to form partnerships for two to three years. At
the end of the contract, partners decided whether to continue the arrange-
ment or to seek new business partners.
Many of the managers of modern industrial firms got a start through
working for foreign firms in the treaty ports. The entrepreneurs of semi-
industrial firms were more likely to have started as commercial apprentices
before breaking off to set up their own independent firms. In the Gaoyang
industrial district another source of entrepreneurial talent was men with
technical skills, gained through study at modern vocational schools in
China and Japan.
We know much less about entrepreneurship in semi-industrial firms than
we do about the capitalist entrepreneurs who created large industrial firms.
Thanks to the survival of company archives, business historians have been
able to analyze the management practices of large modern firms.59 Accounts
of small-scale firms, on the other hand, depend on oral history and the
59
E.Köll, From Cotton Mill to Business Empire: The Emergence of Regional Enterprises in Modern
China (Cambridge, MA, Harvard University Press, 2003).
151

recollections of former employees, and provide less information on the day-

to-day operations of enterprises.
After 1950, large firms were taken over by the state and became the basis of
state-owned enterprises. In some cases, their former owners and managers
were kept on as technical advisers. Small factories were also taken over by the
state in the 1950s. In industrial districts like Gaoyang and Weifang, the small
firms were mechanized, and technically skilled former owners were kept on
as managers while many of the men who had worked in the wholesale trade
were absorbed into state-owned marketing companies. In the early 1980s,
when the reform policies allowed private enterprise to flourish, the sons and
grandsons of former entrepreneurs were among the first to set up factories
and trading companies.60
Industry during the War Years

Beginning during the war years there was a major shift in investment from
light industry to heavy and chemical industries. This shift was evident in areas
under the national government, as well as in Japanese-occupied areas. During
the war years, China was split between five different administrative zones: (1)
areas under the control of the national government with its wartime capital
in Chongqing; (2) Manchukuo, which was set up as a puppet state under
Japan in 1932; (3) Japanese-occupied north China under a puppet provisional
government of the Republic of China (after 1940 under the North China
Political Affairs Commission); (4) central and south China under the
reformed government of the Republic of China; and (5) Communist-
controlled anti-Japanese base areas in the northwest and behind the
Japanese military lines in rural areas of northwest, north, central, and south
China.
In the 1920s and 1930s, the national government had worked hard to build
infrastructure that integrated the various regions of the country. The division
during the war years brought great economic losses, not only through loss of
plants and equipment, but also through the cutting of production chains that
linked the regions of China. During the war years, administrators in each of
the zones drafted industrial policies which they implemented with varying
degrees of success. These state-led development initiatives in each of the
60
D. Wank, Commodifying Communism: Business, Trust and Politics in a Chinese City
(Cambridge, Cambridge University Press, 1999); Grove, Economic Revolution.
152

zones shaped investment decisions and laid a base for the reorganization of
industry in the late 1940s and early 1950s.
Wartime Developments in Areas under the National

Government
As tensions with Japan rose in the early 1930s, the national government began
planning for war, creating a National Defense Planning Commission in 1932,
known after 1935 as the National Resources Commission. The commission
brought together China’s leading economists, scientists, and engineers to
draw up plans for a “national defense economy” to meet the Japanese
military challenge. Overall strategy called for a retreat to the interior, transfer
of existing industrial resources from the coastal regions, and creation of new
defense-related industries under a system of state capitalism.61
The wartime national government retreated to a region where modern
industry was poorly developed. In a 1933 national survey of factories regis-
tered under the factory law, the provinces of Sichuan, Hunan, Guangxi, and
Shaanxi, which were under national government control during the war
years, had only 2.6 percent of the total of 2,435 factories, accounting for
3.1 percent of total invested capital and 2.2 percent of workers. During the
war the government invested in defense-related industries, including power
generation, mining, chemical industries, fuel refining, and munitions. At the
same time, it strove to provide a minimal level of consumer goods to
a population that had been swollen by an influx of refugees from Japanese-
occupied areas. Much of the investment in light industry was left to private
investors, and by the 1940s the government sought to strengthen its control
over private industry.
As we can see from Figure 4.5, the index numbers for industrial production
in national government-controlled Sichuan and Yunnan provinces rose stead-
ily from 1938 to 1942, followed by a sudden decline beginning in 1943. The
early war years saw much new investment, as well as the transfer of factory
equipment from Japanese-occupied areas. Private investors also created some
new factories. However, as the war progressed, Japanese blockades made it
more difficult to acquire the necessary machinery and raw materials, and
some factories were forced to reduce or suspend production.
Production in national government-controlled areas made up 8 percent of
the national total of mining and industry in 1938–1940, rising to about 10 percent
61
W. Kirby, “The Chinese War Economy,” in J. Hsiung and S. Levine (eds.), China’s Bitter
Victory: War with Japan 1937–1945 (Armonk and London, M.E. Sharpe, Inc., 1992), pp.
185–212.
153

160.0
140.0
120.0
100.0
80.0 Free China

All China
60.0
40.0
20.0
0.0
1938 1939 1940 1941 1942 1943 1944 1945
Figure 4.5 Industrial production index, 1938–1945
Source: Kubo Tōru 久保亨, 20 世紀中国経済史論 (Economic history of Twentieth-
Century China)（Tokyo, Kyūko Shoin, 2020), p. 135
during the 1941–1944 period. Thus, while the 10 percent figure for 1941–1944 may
still seem to be very small, it represented more than double the prewar level.
When we look at the figures by sector (Figure 4.6) we see an increase in the
number of firms in defense-related sectors, including metallurgy, machine
making, electronics, and chemicals. Taking into consideration the munitions
industry, which was not included in the statistical calculations, we estimate
that two-thirds of the new factories were in heavy industries.62 Moreover,
since capital investment in chemical and heavy industries tends to be much
higher than investment in other lines of production, this may lead to an
overestimate of the weight of heavy industry in total productive capacity.
The earliest new factories in the interior were established in 1938, with most of
the investment coming from private capital. Between 1939 and 1941, new state-
owned firms took the lead: some were under the National Resources
Commission, and others were established by provincial governments. After
1942, new investment was largely from private investors. A comparison of the
total invested capital with records from before the war shows that state invest-
ment represented about 20 percent of total capital investment in 1936, but was up
to 40 percent in 1944. These figures also exclude the munitions industry.
62
We should note that light-industrial matches and soap were included in the chemical
industry.
154

140
Others
Food
120
Fiber
Chemical
100 Electronic
Metal
Machine
Million yuan
80
Refinery
60
40
20
0
37 38 39 40 41 42 43 44
Year
Figure 4.6 Capital of newly built mills, classified by industry
Source: Kubo, 20 世紀中国経済史, p. 138
Developing industry in the interior under wartime conditions was not easy.
Difficulties in acquiring raw materials and fuel were a major block to sustained
development. This was particularly a problem for privately owned light
industry, especially after 1942. In the case of the cotton industry, transportation
costs rose, and there was a shortage of electricity and machine parts, as well as
a shortage of skilled labor. After 1943 there was a sharp drop in the supply of
raw cotton, which led to a dramatic decline in the production of cotton yarn.
Peasants, unhappy with the government’s low procurement price for raw
cotton, sometimes sold cotton to mills in Japanese-occupied areas. As the
controls stiffened, textile entrepreneurs lost the will to go on.63
63
Kubo Tōru 久保亨, 20 世紀中国経済史論 (Economic History of Twentieth-Century
China) (Tokyo, Kyūko Shoin, 2020), p. 166.
155

Wartime Developments in Japanese-Occupied Areas

Japanese occupation forces strove to build economic systems to sustain the
war effort in China, as well as to contribute to the economic development of
the Japanese empire. Economic planning began in Manchukuo in 1932 and
continued in north and central China after those areas were occupied by the
Japanese Army. Comprehensive plans were drawn up independently for each
occupied area, laying out investment strategies that highlighted the contribu-
tions each of the areas was expected to make to the development of the
Japanese empire.
Manchukuo
Japanese planners targeted Manchukuo for heavy industrial development,
including development of iron and steel, as well as expansion of mining of
coal, magnesium, and other metals. In August 1936, as efforts were being
made to prepare for war, the Army Ministry issued guidelines for develop-
ment, and the Kantō Army, the military unit responsible for control of
Manchukuo, together with the South Manchurian Railway Company,
a statutory company that had been created to manage railroads and eco-
nomic development in the region, drew up a five-year plan for industrial
development. The plan called for what can be characterized as slow, steady
development.
When war broke out in the summer of 1937, Japanese authorities in the
home islands pushed the Manchukuo puppet government to speed up the
plan. Investment in the mining sector doubled, and in the following years
there was increased pressure to develop natural resources. Production targets
steadily rose, even though there was little evidence that the original lower
targets were attainable. As the war in China expanded and war conditions at
home worsened, it became more difficult to gain foreign exchange, shipping
capacity declined, and it was increasingly difficult to obtain from Japan or
from Germany the construction materials that were needed to implement
the policies. Few of the targets stipulated in the plans were met.
The plans for Manchukuo had put little stress on developing light indus-
tries, and by the end of the war the region was left with a distorted economic
structure, including bloated steel and mining sectors. Production of steel and
sulfur continued to rise, reaching a peak in 1943. We can get some idea of the
scope of the development by looking at the facilities that remained after
Japan’s defeat: the steel industry had an annual capacity of 2.5 million tons,
and electric generation capacity (primarily hydropower) was 1.7 million
156

kilowatts. There were three main hydropower plants, and transmission lines
linked the power plants to the cities of Anshan and Dalian.64
North China
Economic planning in the occupied regions of north China began in
November 1938 with the establishment of the North China Development
Company, a national policy company. Half of the initial capital of 350 million
yen was contributed by the Japanese government, with the remaining
investment coming from private companies. A series of subsidiary companies
were established to oversee development of railroads, power generation, and
telephone and telegraph networks. Development plans envisioned comple-
mentary relations between the region and the Japanese home economy, with
north China supplying raw materials as well as agricultural products, primar-
ily cotton, for Japanese industry. Seventy percent of the investment of the
North China Development Company was put into infrastructure, including
railroads and expansion of ports to improve shipment of natural resources.
With the support of the state sector, private firms also flocked to north China.
A new steelworks was established at Shijingshan (the forerunner of post-1949
Capitol Steel), and major Japanese makers like the Toyota automobile
company and the precision machine company Koito Seisakujo, along with
other heavy industrial firms, built facilities in occupied north China.
However, not everything went according to plan. In the early stages of the
war, Japanese policy was focused on exploiting resources to supply the
Japanese economy, a policy that weakened the economy of north China. In
response to the shortages of goods and rising inflation, in the early summer of
1940 the Japanese government revised occupation policies, drafting a new
five-year development plan which put more stress on assuring adequate
supplies of goods for the north China population. However, with the begin-
ning of the Pacific war, there were new calls for goods to supply the Japanese
home economy, and pressure on north Chinese resources increased. This
exploitation of resources exhausted north China’s economy, and shipments
of goods to Japan met only 60 to 70 percent of targets.65
Overall, the Japanese plans for economic construction in north China
ended in failure. However, some of the investments did create facilities
that remained after the war. For example, the newly constructed steel plants
64
Yamamoto Yuzo 山本有造, 「満洲国」経済史研究 (An Economic History of
“Manchuria”) (Nagoya, Nagoya daigaku shuppankai, 2003).
65
Nakamura Takafusa 中村隆英, 戦時日本の華北経済支配 (Japan’s Control of the
Wartime North China Economy) (Tokyo, Yamakawa shuppansha, 1983).
157

reached a production capacity of 756,000 tons, and an electric power grid was
constructed in Hebei and Shandong Provinces.66
Central China
Japanese government planning for the areas under the provisional govern-
ment of the Republic of China followed the same organizational strategy as
that for the occupied areas of north China. A national policy company, the
Central China Promotion Company, was given responsibility for supervision
of railroads, electric power, mining, industry, and fisheries. Subsidiaries
under this company stressed heavy industry, including the mining of iron
ore and coal, maintenance of the railroads, and economic recovery in
Shanghai and other cities.67
The central China region, including Shanghai and the Yangzi valley, had
been the center of light-industrial development in the first three decades of
the twentieth century, and Japanese plans called for the exploitation and
development of the existing light-industrial base. As we have seen earlier, silk
and cotton were the most important light industries in the central China
region.
Silk Filatures
Reforms in the silk industry were still in progress at the beginning of
the war. After Japanese forces occupied the silk-producing regions of
Zhejiang and Jiangsu, the Central China Cocoon and Silk Company
(Kachū Sanshi Kabushiki Kaisha) took control of the silk trade. Despite
the efforts of the new state policy company, the silk industry went into
a sudden and sharp decline. The major problem was a shortage of
cocoons, brought on by shifts in the agricultural economy. Rising
demand for cotton and wheat drove up prices, and land that had
been devoted to cultivating mulberry trees to feed silkworms was
converted into cotton and wheat fields. With the start of the Pacific
war, the global market was basically closed, and the Central China
Cocoon and Silk Company piled up debts and was shut down in 1943.
Meanwhile, the national government supported investment in the
Sichuan silk industry, building an infrastructural base that would
66
Ju Zhifen 居之芬 and Zhang Limin 张利民 (eds.), 日本在华北经济统制掠夺史
(History of Japanese Plunder and Economic Control in North China) (Tianjin,
Tianjin Guji chubanshe, 1997).
67
Hara Akira 原朗, 「大東亜共栄圏」の経済的実態 (Economic Realities of the
“Greater East Asia Co-prosperity Sphere”), 土地制度史学 18.3 (1976), 1–28.
158

support development of Sichuan as a silk production center in the

postwar years.68
Cotton Spinning and Weaving

The cotton industry fared somewhat better than the silk industry during the
war years. Some of the Chinese owners of cotton mills followed the national
government to the wartime capital in Chongqing, moving equipment from
their coastal plants into the interior. While Chinese industry in the coastal
provinces suffered losses, the great exception to that story of wartime loss
was the success of Japanese-owned cotton mills during the early war years.
Wartime shortages led to a sharp rise in prices and to higher profits. Japanese
cotton firms, which were already well established in the lower Yangzi region,
increased capacity when the Japanese military turned operation of confis-
cated Chinese-owned mills over to Japanese firms.69 Chinese-owned mills
located in the concession areas of Shanghai, which remained outside the
control of the Japanese military until the start of the Pacific war, also reaped
high profits, and some Chinese firms continued operations by dividing and
transferring their machinery to rural locations in the Jiangnan region.70
Textile entrepreneurs who had followed the national government to its
wartime capital also earned high profits during the war years, benefiting
from the inflation in prices for consumer goods.
Wartime Developments in the Anti-Japanese Base Areas

The anti-Japanese base areas under Communist Party control in north-
west China and behind the Japanese lines in north and central China had
been established in remote areas with no modern industrial base. During
the war years, efforts were made to mobilize handicraft producers to
provide basic commodities for the war effort. Small factories and work-
shops, using traditional technology, produced munitions for the resist-
ance forces, and women throughout the border areas were mobilized to
spin and weave to produce uniforms for the army and clothing for the
civilian population. The co-operative forms developed in the base areas
68
Kubo, Kajima, and Kigoshi, 統計でみる中国近現代経済史, p. 38.
69
Takamura, 近代日本綿業と中国.
70
Wang Zijian 王子建, “‘孤岛’时期的民族棉纺织工业” (National-Capital Cotton
Industry during Shanghai’s “Isolated Island” Period), in 中国近代经济史研究资料
(Materials on Modern Chinese Economic History), vol. 10 (Shanghai, Shanghai shehui
kexue chubanshe, 1990); Kubo, 20 世紀中国経済史論, pp. 467–9.
159

served as a model in the early 1950s for the reorganization of rural

handicraft industries.
Postwar Reorganization of Chinese Industry

At the end of the war the national government faced the formidable task of
reintegrating the economy and reviving industrial production against
a background of continued political turmoil. The uneasy wartime alliance
with the Communist Party quickly unraveled, and civil war began in the
summer of 1946. In the northeast (Manchuria), Soviet armies launched an
invasion less than a week before the end of the war and occupied most of the
major cities of the region
As we can see from the industrial production index (Figure 4.7), industrial
production had begun to recover in the first months after the end of the war,
only to plunge beginning in 1947. To understand why this happened, we need
to consider the unintended consequences of trade and financial policies.
While the war had been long and exhausting, physical destruction of indus-
trial facilities in most parts of China was relatively limited. Both in the areas
under the national government and in the Japanese-occupied areas of the
country, wartime investment in heavy industrial facilities had increased
capacity in mining, iron and steel production, and hydroelectric power
generation. Light industries, on the other hand, had suffered losses, and
factories that were still in operation were running below capacity, largely
because they could not acquire raw materials. Production of cotton goods
and raw silk, the former for domestic consumption and the latter for export
trade, required the restoration of commodity chains that linked agriculture to
industry.
During the war years, the Republic of China had joined the American-led
alliance of free economies under the Bretton Woods agreement, and postwar
trade and monetary policies followed the alliance’s guidelines – including
a system of fixed exchange rates, convertibility of currency, and support for
free trade. China seemed to be in an advantageous position to move from
a wartime economy to a peacetime economy. First, the government had built
up significant foreign-currency reserves during the war years: these funds had
been accumulated from wartime aid and profits on controlled trade and are
estimated at US$800 million to US$900 million. Added to this were the assets
that were seized from the Japanese at the end of the war. The government
thought it had adequate funds, and in February 1946 announced that it was
freeing the foreign-currency market and opening up free trade.
160

1933 = 100
140.0
120.0
100.0
80.0
60.0
40.0
20.0
0.0
12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48
19
Figure 4.7 Industrial production index, 1912–1949
Source: Kubo, 20 世紀中国経済史, 90–1
The open-market plans ended in disaster. The foreign-exchange rate was

set at a higher rate than current market exchange values, and the favorable
exchange rates led to a huge inflow of imports from the United States.
Government planners had anticipated an influx of producer goods to aid
the recovery of domestic manufacturing, but the influx of consumer goods
acted as a drag on the recovery of domestic producers. By June of 1947,
domestic manufacturing stood at only 35.1 percent of the highest prewar
level.
On the other hand, the exchange rate for export exchange had been set at
a high rate, and it served as a brake on exports, which went into sharp decline.
This imbalance produced a huge trade deficit, resulting in a reduction of
US$115 million in foreign-currency reserves within less than six months.
There was no question that the plans to open the market had been intro-
duced at too rapid a pace.71
In August 1946, a delegation of Shanghai economic leaders demanded
a fundamental reconsideration of economic policies. They argued that the
71
Kubo Tōru 久保亨, “対外経済政策の理念と決定過程” (Principles behind the
Decision-Making Process Related to Trade-Related Economic Policies), in
Himeta Mitsuyoshi 姫田光義, 戦後中国国民政府史の研究, 1945–1949 (Study of
the History of the Chinese National Government in the Postwar Period, 1945–1949)
(Tokyo, Chūo daigaku shuppanbu, 2001), pp. 235–61.
161

Sino-American trade agreement of November 1946, which had promoted

freer trade, had speeded up the inflow of American goods. Even though the
treaty terms were in theory equal, given the great gap in economic strength
between the US and China there was little question that the benefits of
reducing Chinese regulations were going to American capital. Opposition
from Chinese economic leaders led to a growing national consensus and the
national government was forced to abandon its economic liberalization
policies. On February 16, 1947, the government announced emergency regu-
lations that banned the sale of foreign exchange and reinstituted rationing for
daily necessities.
Meanwhile, plans to take over Japanese assets in the former occupied areas
were running into difficulties. Reintegration of the regional economies was
a major stumbling block. During the war years there were four separate
currency regimes: the northeast (Manchukuo) ran on Japanese yen; north
China ran on lianyinquan 聯銀券; central China ran on choubeiquan 儲備券;
and areas under the control of the national government, as well as the
Communist anti-Japanese base areas, ran on the national government’s fabi
法币. One of the first steps to achieve reintegration of the economy was
reintegration of the currencies. Conversion rates between the puppet curren-
cies and the fabi were all set at disadvantageous rates for those in the former
occupied areas, leading to an overvalued fabi. Conversion thus produced
inflation in the coastal cities. At the same time, the rates facilitated the flow of
high-quality goods produced in the coastal regions into the interior at low
prices, which had a negative impact on manufacturers in the interior. As
funds flowed from the interior into the coastal regions, the interior regions of
China suffered from a shortage of capital and fell into depression.
In the coastal regions, there were clashes over who should reap the
benefits of the assets seized from Japanese entities in the formerly occupied
areas, and efforts to adjudicate the matter ran into great difficulties. In the
north China region individual factories, power plants, and mines were
taken over by different organizations, breaking the integration that had
been created under the Japanese occupation. And in the Soviet-occupied
northeast, many industrial assets were dismantled and shipped to the
Soviet Union.
Failure of the economic liberalization policy and the difficulties of taking
over Japanese assets created financial problems for the postwar government.
Delays in the recovery of both production and circulation of goods led to
shortages in the markets and rising prices. Despite rising inflation, the
162

government continued to pile up large deficits as it struggled to acquire goods

to support the civil war, adding to inflationary pressure.
Although the national government had won victory in its war with Japan,
the Chinese economy failed to recover. The failure of the postwar liberaliza-
tion policy and the unrealistic plans for integrating formerly occupied and
unoccupied regions resulted in rising inflation and the inability of the gov-
ernment to control the economy, dealing a death blow to economic integra-
tion efforts.
Organizational Legacies
Organizations created in the postwar period to deal with assets confiscated
from Japanese owners smoothed the way for nationalization of industry in
the post-1949 period. We can see this by looking at what happened in
the cotton industry. After the war, Japanese-owned assets were seized by
the national government. The government established a state company, the
Chinese Textile Construction Company (Zhongguo Fangzhi Jianshe Gongsi
中国纺织建设公司) to operate the mills for three years, after which they
were to be sold to private investors.72 The national government collapsed
before the sale to private investors could be carried out, and the government
of the People’s Republic of China took over the company, which became the
core of a new state corporation to control the cotton industry. Similar
companies were established for other major industries, creating the base
for consolidation of industrial production under a series of state-owned
companies. Through the process of socialist transformation in the early
1950s, factories that were still in private hands were converted to public
ownership and put under state-owned industrial groups.
Much of Chinese industrial production in the early 1950s came from what
we have referred to as semi-industrial firms, which mixed machine produc-
tion and handicraft methods. State policies in the early 1950s for the socializa-
tion of handicraft industries drew on the practices of handicraft co-operatives
developed in the anti-Japanese base areas during the war years. Individual
handicraft workers were organized into co-operatives, which operated under
the supervision of state bodies that controlled the supply of raw materials and
sale of finished products. Over the years, some of the co-operatives devel-
oped into factories, and the most successful of the factories were taken over
by the state and turned into state-owned enterprises. Others continued as
72
Kubo, Kajima, and Kigoshi, 統計でみる中国近現代経済史, p. 29.
163

small-scale operations controlled by rural communes, providing low-cost

goods for local consumption.
Concluding Remarks
In the late 1980s, when Thomas Rawski outlined his approach in Economic
Growth in Prewar China, he summarized his argument in four key concepts:
“domestic, private, civilian, and competitive.”73 Arguing against Marxist-
influenced interpretations that stressed the impact of foreign imperialism,
state control, and the exploitative nature of elite manipulation of markets,
Rawski chose those four concepts to argue that key players in the prewar
modern Chinese economy were private firms, primarily owned by Chinese
capital, and that state investment, planning, and control played a relatively
minor role. Furthermore, he argued that Chinese markets were highly
competitive, and that the Chinese economy was characterized by signifi-
cant growth in both the agriculture and industrial sectors. Our understand-
ing of the prewar economy is in basic agreement with Rawski’s view,
although we have a more expansive understanding of modern industrial-
ization. We have argued for multiple paths to industrialization: a state-
centered path, a path centered on private capital, and a semi-industrial
path. Our analysis of the two basic types of private capital development –
export-oriented industrialization and import-substitution industrialization –
fits very well with Rawski’s understanding of prewar growth. Our third
path – semi-industrial development – adds an additional dimension by
including in the industrial sector the thousands of small semi-industrial
firms that mixed mechanized production with hand labor. As we, and
others, have argued, production by semi-industrial firms was strongly
market-dependent, making full use of the new forms of transportation
(railroads and scheduled bus services) and communications (telephone,
telegraph, and the postal network) to link rural and small-town production
units to regional and national markets. We see in this third path to
industrialization organizational forms and production processes that share
many similarities with the hundreds of thousands of small private firms
created in China after the launching of the economic reforms in the early
1980s. In many cases, contemporary twenty-first-century firms have been
73
T. Rawski, Economic Growth in Prewar China (Berkeley and Los Angeles, University of
California Press, 1989), p. 3.
164

created in regions that were the sites of semi-industrial development in the

prewar period and build on the entrepreneurial practices of earlier times.
Wartime developments serve as a bridge to the postwar period and early
1950s. The Japanese occupation of China and the eight-year-long war led to
heavy loss of human life and property. The war years also saw a major
restructuring of the Chinese economy in Japanese-occupied areas, as well as
in areas under the national government and in Communist base areas. We
have stressed several common themes that characterized development in all
the areas. First, state planning played a central role in the wartime economy
in each of the areas, and there was a shift from a dominance of privately
owned industry to a greater role for state planning and investment in
infrastructure, mining, and industrial enterprises. Second, to meet wartime
needs, investment in each of the areas concentrated on sectors that could
contribute to the war effort.
In the confusion of the postwar era and the difficulties resulting from the
overly hasty introduction of freer trade and capital markets, the economy
went into severe decline. It is thus difficult to measure the impact of wartime
investment in the various regions of occupied and unoccupied China on
industrial output. However, as our argument for the wartime period sug-
gests, there was significant investment in infrastructure, including mining
and power generation, in both occupied and unoccupied China, and that
investment, together with the organizational changes, created a base for
development of the state-centered economy in the 1950s.
Further Reading
Cochran, S., Big Business in China: Sino-Foreign Rivalry in the Cigarette Industry, 1890–1930
Cochran, S., and A. Hsieh, The Lius of Shanghai (Cambridge, MA, Harvard University
Press, 2013).
Faure, D., China and Capitalism: A History of Business Enterprise in Modern China
(Hong Kong, Hong Kong University Press, 2006).
Furuta, K., and L. Grove (eds.), Imitation, Counterfeiting and the Quality of Goods in Modern
Asian History (Singapore, Springer, 2017).
Grove, L., A Chinese Economic Revolution: Rural Entrepreneurship in the Twentieth Century
(Lanham, Rowman & Littlefield Publishers, Inc., 2006).
Kirby, W., “The Chinese War Economy,” in J.C. Hsiung and S.I. Levine (eds.), China’s
Bitter Victory: War with Japan 1937–1945 (Armonk and London, M.E. Sharpe, Inc., 1992),
pp. 185–212.
Köll, E., From Cotton Mill to Business Empire: The Emergence of Regional Empires in Modern
China (Cambridge, MA, Harvard University Asia Center, 2003).
165

Kubo, T., “Changing Patterns of Industrialization and Emerging States in Twentieth

Century China,” in K. Otsuka and K. Sugihara (eds.), Paths to the Emerging State in
Asia and Africa (Singapore, Springer, 2019), pp. 20–64.
Kubo Tōru 久保亨, 20 世紀中国経済史論 (Economic History of Twentieth-Century
China) (Tokyo, Kyūko Shoin, 2020).
Kubo Tōru 久保亨, Kajima Jun 加島潤, and Kigoshi Yoshinori 木越義則 (eds.), 統計で
みる中国近現代経済史 (Economic History of Modern China: Based on Statistical
Data) (Tokyo, Tōkyō daigaku shuppankai, 2016).
Peng Nansheng 彭南生, 半工业化：近代中国乡村手工业的发展与社会变迁 (Semi-
industrialization: The Development of Rural Handicraft in Modern China and Social
Change) (Beijing, Zhonghua shuju, 2007)
Peng Zeyi 彭泽益 (ed.), 中国近代手工业史资料 (Materials on the History of the
Chinese Modern Handicraft Industry), 4 vols. (Beijing, Sanlian shudian, 1957).
Rawski, T., Economic Growth in Prewar China (Berkeley, University of California Press,
1989).
Sheehan, B., Industrial Eden: A Chinese Capitalist Vision (Cambridge, MA, Harvard
Sugihara, K., “Labour-Intensive Industrialization in Global History: An Interpretation of
East Asian Experiences,” in G. Austin and K. Sugihara (eds.), Labour-Intensive
Industrialization in Global History (London and New York, Routledge, 2013), pp. 20–64.
Sun Yutang 孫毓棠 (ed.), 中国近代工业史资料 (Materials on the History of Chinese
Modern Industry), 4 vols. (Beijing, Kexue chubanshe, 1957).
Wang Xiang 王翔, 中国近代手工业史 (A History of China’s Modern Handicraft
Industry) (Shanghai, Shanghai renmin chubanshe, 2012)
Xu Dingxin 徐鼎新, 中国近代企業的科技力量与科技効応 (Technical Strength and
Technical Productivity of Modern Chinese Enterprises) (Shanghai, Shanghai kexue
xueyuan chubanshe, 1995).
Xu, Xinwu, “The Struggle of the Handicraft Cotton Industry against Machine Textiles in
China,” Modern China 14.1 (1988), 31–49.
Xu Xinwu 徐新吾, 中国近代缫丝工业史 (A History of the Modern Chinese Silk
Industry) (Shanghai, Shanghai renmin chubanshe, 1990).
166

5
The State and Enterprises in Late Qing
China
chi-kong lai
This chapter focuses on the so-called Self-Strengthening era during

the second half of the nineteenth century when the Qing empire was
expanding state involvement in industry and technology from its traditional
ideology. The origin and motivation of state involvement in the private
market during this era is different from that in the early twentieth century,
during which China began to take lessons from Meiji Japan and the ideology
of modernization (see Bian’s chapter in this volume). There were some
elements of continuity here – in remnants of “Self-Strengthening” through
the state efforts both to build up modern enterprises and develop science and
technology. This chapter explains the impact of the state sector’s emergence
that was manifested in the development and expansion of state arsenals and
modern enterprises.
There is already a considerable literature, to which Chinese, Japanese, and
Western scholars have contributed, on the role of the state in the develop-
ment of Chinese enterprises in the late nineteenth century, among which the
studies by Albert Feuerwerker, Kwang-Ching Liu, Wellington K.K. Chan,
Yen-p’ing Hao, Chi-kong Lai, Xia Dongyuan 夏东元, Zhang Guohui 张国
辉, and Long Denggao 龙登高 are a few of the most important.1 Elisabeth
1
See, for example, A. Feuerwerker, China’s Early Industrialization: Sheng Hsuan-huai (1844–
1916) and Mandarin Enterprise (Cambridge, MA, Harvard University Press, 1958);
K.-C. Liu, Anglo-American Steamship Rivalry in China (Cambridge, MA, Harvard
University Press, 1962); Liu, “Steamship Enterprise in Nineteenth-Century China,” Journal
of Asian Studies 18.4 (November 1959), 435–55; W.K.K. Chan, Merchants, Mandarins and
Modern Enterprise in Late Ch’ing China (Cambridge, MA, Harvard University Press, 1977); Y.-
P. Hao, The Comprador in Nineteenth Century China: Bridge between East and West
(Cambridge, MA, Harvard University Press, 1970); Hao, The Commercial Revolution in
Nineteenth Century China: The Rise of Sino-Western Mercantile Capitalism (Berkeley,
University of California Press, 1986); C.-K. Lai, ‘“The Qing State and Modern Enterprise:
The China Merchants’ Company, 1872–1902,” in J.K. Leonard and J.R. Watt (eds.), To
Achieve Security and Wealth: The Qing Imperial State and the Economy, 1644–1911 (Ithaca, NY,
The Cornell East Asia Series, 1993), pp. 139–56; Zhang Guohui 张国辉, 洋务运动与中国
167

chi-kong lai
Köll and Anne Reinhardt clearly show how the government utilized the
building of the railway and the shipping companies to expand its political
power and agenda.2
The Origin of Self-Strengthening Ideologies

and Policies
This section will illustrate the links between state policy and enterprises in the
cultural, institutional, and economic contexts of late Qing China. The trad-
ition of state involvement in large-scale enterprises can be traced back to
three government-regulated enterprises in the late Ming and mid-Qing
periods: the special role of the Jiangsu and Zhejiang seagoing junks in the
transport of tribute grain; state monopolies in salt, tea, and ginseng; and the
mining industries.
The key to government-sponsored enterprises was the concept of
zhaoshang 招商 – recruiting investors and managers. Under the system of
zhaoshang, officials tapped private wealth to meet the state’s needs. In return
for the chance to obtain a profit, investors had to meet the obligations placed
upon them by officials. China has had a long tradition of using the method of
“recruiting merchants” in government-sponsored projects. A major example
in Chinese economic thought was Liu Yan’s 刘晏 (718–80) policy in the mid-
Tang period. Liu Yen advanced the zhaoshang concept of recruiting mer-
chants to undertake government-regulated enterprises, such as the salt trade,
the grain transport service, and the granary system.3 Later, this legacy of
zhaoshang became one of the most effective means of government interven-
tion in the private economy. Such zhaoshang terminology appeared more and
more frequently as the years went by. The phrase zhaoshangju 招商局 was
近代企业 (The Self-Strengthening Movement and China’s Modern Enterprise) (Beijing,

Zhongguo shehui kexue chubanshe, 1979); Long Denggao 龙登高, Chang Xu 常旭, and
Xiong Jinwu 熊金武, 国之润，自疏浚始：天津航道局120年发展史 (One Hundred
Years of Development: The Tianjin Waterway Bureau) (Beijing, Qinghua daxue chu-
banshe, 2017). Long, Chang, and Xiong’s book critically evaluates the quality of state
building in the modern enterprise’s role in conservancy operations.
2
See E. Köll, Railroads and the Transformation of China (Cambridge, MA, Harvard
University Press, 2019). Anne Reinhardt also compares China with India in the steamship
navigation business. She also illustrates how the coming of Western steamships
impacted the state in the development of China’s shipping network in late Qing
China. See her Navigating Semi-colonialism: Shipping, Sovereignty, and Nation-Building in
China, 1860–1937 (Cambridge, MA, Harvard University Press, 2018).
3
See K.-C. Liu, “Statecraft and the Rise of Enterprise: The Late Ch’ing Perspective,” in
The Second Conference on Modern Chinese Economic History (Taipei, The Institute of
Economics, Academia Sinica, 1989), p. 8.
168

The State and Enterprises in Late Qing China
used in the salt trade and other enterprises before 1865,4 seven years before
the establishment of the China Merchants Steam Navigation Company
(CMSNC) under the so-called Self-Strengthening movement. Did the policy
change over time? Many Chinese officials not only studied the classics, but
also cited them as metaphors to justify their policies.
Most of the Confucian officials who endorsed ritual and ethics believed
that wealth should be shared by the public.5 The Chinese state’s legitimacy
was also usually achieved through policies for “benefiting the people.” The
rhetoric of jun 均 (equitability) and gong 公 (publicness)6 was dominant in
Confucian discourse.7 The ideal of equal allotment of land can be traced back
to the so-called “well-field system” in China’s fabled antiquity.8 Conceptually,
this notion of civic betterment may be viewed as a type of public moral
discourse. This public moral discourse focused on three sets of preferences or
priorities: (1) public good over private interest, (2) the fundamental over the
secondary (benmo 本末), and (3) “state intervention” over “laissez-faire”
economic policy.
Another way of approaching the political economy of China during the
mid-nineteenth century is to look at the statecraft debate about the state in
relation to the market. The main debate centered on the issue of whether the
government should interfere with the market and, if so, in what way. In the
Chinese tradition, there were at least two contending views of the relation-
ship between the state and the market: “state intervention” as formulated in
4
See Li Hongzhang 李鸿章, 未刊朋僚信稿 (Unpublished Letters to His Friends and
Colleagues) (Shanghai Library), Li Letters 6:29.
5
For Chinese economic thought, see D. Faure, “The Introduction of Economics in China,
1850–2010,” in V. Goossaert, J. Kiely, and J. Lagerwey (eds.), Modern Chinese Religion II,
1850–2015, vol. 1 (Leiden, Brill, 2016), pp. 65–88. Also see Faure, China and Capitalism:
A History of Business Enterprise in Modern China (Hong Kong, Hong Kong University
Press, 2006); Kwang-ching Liu, “Introduction: Orthodoxy in Chinese Society,” in Liu
(ed.), Orthodoxy in Late Imperial China (Berkeley, University of California Press, 1990),
pp. 8–12.
6
On the Chinese concept of “public” (gong), see J. Fewsmith, Party, State, and Local Elites
in Republican China: Merchant Organizations and Politics in Shanghai, 1890–1930 (Honolulu,
University of Hawaii Press, 1985), pp. 18–20; W.T. Rowe, “The Public Sphere in Modern
China,” Modern China 16.3 (July 1990), 309–29; and M.B. Rankin, “The Origins of
a Chinese Public Sphere: Local Elites and Community Affairs in the Late Imperial
Period,” Études chinoises 9.2 (Autumn 1990), 13–60.
7
According to Lien-sheng Yang, “Chinese economic history is full of terms using the
character chun [jun].” See L.-S. Yang, “Economic Aspects of Public Works in Imperial
China,” in Yang, Excursions in Sinology (Cambridge, MA, Harvard University Press, 1969),
p. 232.
8
See D.C. Lau, trans., Mencius (Hong Kong, The Chinese University Press, 1984), pp.
96–101.
169

chi-kong lai
the doctrine of Legalism,9 on the one hand, and the “laissez-faire” principle,
on the other.
From the above discussion of public moral discourse in Chinese economic
thought, we can understand that most Chinese officials tended to favor state
control in the economic sphere. There were two main reasons. The first was
the official belief in the “public good,” which is best represented in the texts of
the Confucian classics and other writings of ancient thinkers. The second
was the rhetoric of benmo, which disparaged both the merchant’s status and
mercantile activity in premodern China.
But in reality, the Chinese state alone did not have sufficient resources to
regulate the market bureaucratically. In this institutional context, how did
a centralized, premodern political system such as the Chinese bureaucracy
administer markets? The Qing state was using more remunerative methods to
recruit private merchants in the later period.10 This perspective on the state’s
positive role in the economy can also be documented by brief reference to the
seagoing junk business. Beginning in the early 1800s, ideas about statecraft
were reformulated by scholars – the most influential of whom was Wei Yuan
魏源 (1794–1857), an adviser to key government officials – in response to the
problems of the mid-Qing period, for example the need for reform in the salt
monopoly, the control of river flows, and the transport of tribute grain.
Chinese state officials could manipulate symbolic capital, such as official
titles, as an inducement for merchant participation. Officials were trying to
appeal to “norms” associated with high status and prestige. Remunerative
methods imply profit or (and) monetary rewards. In the case studies above,
the Qing state also used material rewards to recruit other types of merchants.
This combination of both normative and remunerative strategies was one of
the more successful aspects of mid-Qing state policy in the economy.
This mid-Qing legacy of zhaoshang was the key to governmental enter-
prises in late Qing China. When late Qing officials undertook to promote
modern enterprises in China, the first company was in fact popularly called
the Bureau to Recruit Investors and Managers (zhaoshangju). Later, this
legacy became one of the most effective means of government intervention
in the economy. During the mid-nineteenth century, the Qing state, in
9
The concept of state intervention is Legalist economic doctrine. To a Legalist, material
welfare far exceeds any virtues; only material welfare can centralize both the govern-
ment and the economy into a strong state, powerful enough to survive in the “modern
world.” See A. Waley, Three Ways of Thought in Ancient China (Stanford, Stanford
10
See V.J. Symons, Ch’ing Ginseng Management: Ch’ing Monopolies in Microcosm (Tempe,
Center for Asian Studies, 1981), p. 44.
170

response to the economic and commercial penetration of foreign powers,

played an active role in many new business and industrial ventures, particu-
larly those with security implications.11
The Beginning of Self-Strengthening and the

Arsenal Industry
The 1861 coup d’état to replace the eight powerful ministers, planned by Prince
Gong (also known as Yixin 弈訢, 1833–98) and the Empress Dowager Cixi
濨禧 (1835–1908), resulted in greater flexibility in government policy during
the Tongzhi 同治 (1862–74) reign and the early years of the Guangxu 光绪
(1875–1908) period.12 After various turbulent events, such as the Taiping
Rebellion (1850–64), a number of senior ministers and provincial leaders,
including Wen Xiang 文祥 (1818–76), Zeng Guofan 曾國藩 (1811–72), Li
Hongzhang 李鸿章 (1823–1901), Zuo Zongtang 左宗棠 (1812–85), and Ding
Richang 丁日昌 (1823–1882) were advocates of Self-Strengthening reforms to
safeguard China from the peril of Western incursions. These reformist
officials built arsenals and organized state-sponsored enterprises for the
steamship carrying trade, coal and iron mining, textile manufacturing, tele-
graphs, banking, and the railway service. They may be regarded as advocates
for China‘s early industrial modernization.13
Before China’s state involvement in industry and commerce, the first
priority was national defense against the rising threat of Western
imperialism.14 The greatest concern for the Qing state was the development
of China’s industrial and technological capabilities for the improvement of
military forces.15 In the 1860s and 1870s, China accelerated its pace in naval
construction; officials thus advocated domestic manufacturing of weapons,
11
For an overview of the evolution of state control economy in modern China, see
Du Xuncheng 杜恂诚, Yan Guohai 严国海, and Sun Lin 孙林, 中国近代国有经济思
想,制度与演变 (The Institution and Evolution of State-Owned Economy and Thought
in Modern China) (Shanghai, Shanghai renmin chubanshe, 2007).
12
See M.C. Wright, The Last Stand of Chinese Conservatism: The T’ung-Chih Restoration,
1862–1874 (Stanford, Stanford University Press, 1957), pp. 16–17. Also see T.-Y. Kuo and
K.-C. Liu, “Self-Strengthening: The Pursuit of Western Technology,” in J.K. Fairbank
(ed.), The Cambridge History of China, vol. 10, Late Ch’ing, 1800–1911, part 1 (Cambridge,
Cambridge University Press, 1978), pp. 491–3.
13
For an overview of the PRC scholarship in the 1980s, see J. Chen, “Recent Chinese
Historiography on the Western Affairs Movement (the Yangwu yundong, ca. 1860–
1895),” Late Imperial China, June 1986, 112–27.
14
Xia Dongyuan 夏东元, 洋务运动史 (The History of the Self-Strengthening
Movement) (Shanghai, Huadong shifan daxue chubanshe, 2010), pp. 67–8.
15
Li Hongzhang and other officials believed that the Qing lost wars because Western
armies used more advanced weapons while the Qing army was still using swords.
171

chi-kong lai
especially the building of steamships, for coastal defence.16 During this

period, Qing officials established some arsenals, such as the Anqing Arsenal
(1861),17 the Jinling Machine Manufacturing Arsenal (1865), and the Tianjin
Machine Manufacturing Bureau (1867).18 But the most significant was the
Jiangnan Arsenal, constructed in Shanghai, which contributed mightily to
China’s industrial and military modernization, developing vital fields of the
economy for the benefit of the state’s military endeavors. It became the
leading ordnance enterprise in the Far East.19 In the period from 1868 to 1876,
the arsenal produced eleven ships which were deemed to be higher in quality
than those produced in the Yokosuka Dockyard in Japan.20
The Foochow Naval Shipyard (Fuzhou Shipping Bureau, also known as
Mawei Shipping Bureau), managed by Shen Baozhen 沈葆桢 for much of its
existence and founded by Zuo Zongtang in 1866, was intended as a multi-role
facility.21 Financing a modern defense enterprise required much greater
governmental intervention than did traditional enterprises. However, the
Qing government, whose revenue represented no more than 3 percent of the
country’s gross national product, was an “unbelievably weak instrument” for
playing this role alone.22
Like the arsenal at Jiangnan, the Fuzhou shipyard’s production declined
after 1874 with financial and administrative problems taking their toll on the
availability of funds and expertise.23 The period from 1874 to 1897 saw a more
focused approach to the training of engineers and officers to manage and
maintain the fleets as well as to gain knowledge on the latest developments in
16
D. Pong, Shen Pao-Chen and China’s Modernization in the Nineteenth Century (New York,
Cambridge University Press, 1994), p. 312.
17
D. Wright, “Careers in Western Science in Nineteenth-Century China: Xu Shou and Xu
Jianyin,” Journal of the Royal Asiatic Society 5.1 (1995), 49–90.
18
Xia, 洋务运动史, pp.119, 124–5, 127–30.
19
T. Kennedy, The Arms of Kiangnan: Modernization in the Chinese Ordnance Industry, 1860–
1895 (Boulder, CO, Westview Press, 1978), pp. 146, 150–60; B.A. Elman, “Naval Warfare
and the Refraction of China’s Self-Strengthening Reforms into Scientific and
Technological Failure, 1865–1895,” Modern Asian Studies 38.2 (2004), 292; K. Biggerstaff,
The Earliest Modern Government Schools in China (Ithaca, NY, Cornell University Press,
1961), pp. 165–6.
20
Meng Yue, “Hybrid Science versus Modernity: The Practice of the Jiangnan Arsenal,
1864–1897,” East Asian Science, Technology, and Medicine 16 (1999), 16–17; Pong, Shen Pao-
Chen and China’s Modernization in the Nineteenth Century, p. 224; Biggerstaff, Government
Schools, pp. 246–7.
21
H.-C. Wang, “Transferring Western Technology into China: 1840s–1880s,” D.Phil.
thesis, Oxford University (2007).
22
Pong, Shen Pao-Chen and China’s Modernization in the Nineteenth Century, p. 199. See
Kaske and Lin’s chapter in this volume.
23
Biggerstaff, Government Schools, pp. 214–19.
172

mining, metallurgy, and weaponry.24 Students were sent to Britain, France,

and Germany to study Western science and technology so that they could
establish a generation of engineers skilled in modern warfare and equipment
design. Nevertheless, military modernization during the Self-Strengthening
movement forced the development of other modern industries, e.g. commu-
nications and textiles.
Officials and Business Enterprise in the

Self-Strengthening Era
Beyond the arsenal, the concept of fighting or competing for liquan 利权
(economic profits and sovereign rights) against Western imperial business
intrusion was beginning to take hold.25 “Merchant thinkers” and officials such
as Zheng Guanying 郑官应 (1842–1922), Ma Jianzhong 马建忠 (1845–1900),
Guo Songtao 郭嵩焘 (1818–91), and others elaborated on the importance of
the development of “self-interest” discourse for encouraging merchants’
motivation to invest in modern enterprises. On the governmental side,
powerful local officials sponsored new industries and made sure that these
efforts would advance the consolidation of their political status and the
expansion of their spheres of influence. These officials assumed successive
new roles: first supervisors, then managers, then investors, and finally, for
some, official-entrepreneurs. In assuming these new roles, official promoters
raised new ideas about liquan and its connection to the role of merchants in
the state efforts to build up modern enterprises.
Past research on the modern enterprise in the Self-Strengthening move-
ment mainly focused on organizational structure, the scope of influence and
the reasons for failure, and its role in the emergence of the bureaucratic and
national capitalists.26 Below, I will explore the influence of the relationship
between political and commercial relations on the establishment, develop-
ment, and survival of these enterprises by comparing the growth experiences
of the enterprises founded by Li Hongzhang, Zhang Zhidong 张之洞 (1837–
1909), and Yuan Shikai 袁世凱 (1859–1916) in the last fifty years of the Qing
dynasty. It helps illuminate the role played by government control in the
development of enterprises under conditions in which owner-managers
themselves were lacking legal protection.
24
Elman, “Naval Warfare,” 300.
25
See S. Halsey, Quest for Power: European Imperialism and the Making of Chinese Statecraft
(Cambridge, MA, Harvard University Press, 2015), pp. 5–11.
26
Feuerwerker, China’s Early Industrialization.
173

chi-kong lai
Li Hongzhang was a major figure in the push for the Qing regime’s
modernization effort. In the 1870s he began a twenty-five-year term as
governor-general of the capital province, Zhili, during which time he initi-
ated projects in commerce and industry. He served as a grand secretary and
superintendent of trade for the northern ports and thus was responsible for
supervising trade with the West out of treaty ports north of the Yangtze river.
During this long tenure, Li involved himself in several major modernizing
projects: an arsenal at Tianjin, a commercial steamship line, a coal mine,
a railroad, a telegraph line, a cotton mill, and a modern mint. Thus Li
Hongzhang played a double role as grand secretary and superintendent of
trade – a national-level appointment – and also a provincial-level role in
establishing his enterprises. Thus he was able to encourage and promote
modern enterprises simply through policies at the national level. It is also
worth mentioning that governors or governors-general could establish enter-
prises outside their provinces, so that the enterprises were, in a sense,
national enterprises even though they were established by provincial-level
officials. Li was able to contribute to the modernization of China’s economy
by utilizing the role of the state and in building modern enterprise through
corporations in four major westernized industries: the shipping industry, the
telegraph industry, mining (coal, steel, lead, copper, etc.), and textiles.
Under the system of so-called guandu shangban 官督商办 (official supervi-
sion with merchant management), industries were joint official–merchant
undertakings.27 This system was designed to tap into the accumulated funds
of Chinese merchants from foreign trade and ancillary services in the treaty
ports. Under this system, Li Hongzhang required merchants to invest their
own capital and to run modern enterprises at their own risk and under
government supervision. For example, officials such as Sheng Xuanhuai 盛
宣怀 supervised while comprador Tong Kingsing 唐景星 managed some of
the state’s modern enterprises. Their supervisory and managerial skills led to
discontent as financial losses and abuse of power occurred. But even though
Li’s power was not enough to protect his enterprises from the influence of
other segments of the bureaucracy, he did try to protect merchant managers
from bureaucratic interference.
The China Merchants Steam Navigation Company was the first example
of this kind of enterprise. At that time, navigation transport was controlled by
foreign companies, which enabled foreigners to earn huge profits. The
27
Feuerwerker, China’s Early Industrialization, p. 11.
174

introduction of Chinese navigation would attract Chinese merchants to use

their own national steamships and thus boost other Chinese business.
Li oversaw the founding of the China Merchants Steam Navigation
Company (CMSNC) in 1872. This groundbreaking project was to be owned
and primarily financed by merchants, thus shifting the risk of the success of
the endeavor over to private investors. In the first several years the endeavor
proved successful, becoming bigger than any of the Western steamship
companies. In 1877 the CMSNC became the first Chinese national enterprise
to acquire a foreign company. By 1877, the company owned twenty-nine
steamships. The ships were not only useful for commercial use; they carried
food relief during the famine in north China in 1876 and carried troops to
where they were needed at various times.28 The company also invested in
different business sectors.
The CMSNC was a unique, hybrid experiment, undertaken by Qing
officials and Chinese merchants to counter the inroads of Western steam
shipping in China’s coastal trade. Its purpose was to “recover China’s eco-
nomic rights [liquan].” The CMSNC was China’s first joint-stock company
(gongsi 公司), and its use of this organizational structure marked a new
departure in Chinese business practice. Yet government–merchant co-
operation in the creation and development of this enterprise followed earlier,
well-established patterns of Qing government co-operation with private
merchant groups to achieve mutually advantageous goals. In such cases,
the government recruited private human, organizational, and material
resources to launch various kinds of joint ventures that used different
approaches to “government supervision and merchant operation” (guandu
shangban). From 1872 to 1884, the CMSNC struggled to weld together old
patterns of government–business co-operation with its new joint-stock struc-
ture during a period of challenging and rapidly changing political and
economic conditions. In the end, the company failed. But the experiment
28
See C.K. Lai, “Li Hung-chang and Modern Enterprise: The China Merchants’ Company,
1872–1885,” in S.C. Chu and K.-C. Liu, Li Hung-chang and China’s Early Modernization
(Armonk, M.E. Sharpe, 1994), pp. 227, 222. Also see Chi-kong Lai 黎志刚, “轮船招商局
国有问题, 1878–1881,” Bulletin of the Institutute of Modern History, Academia Sinica 17 (1988),
15–40; Chi-kong Lai, “轮船招商局经营管理问题, 1872–1901” (Problems in the
Management of the China Merchants’ Company), Bulletin of the Institutute of Modern
History, Academia Sinica 19 (1990), 67–105. Most recently, Reinhardt, Navigating Semi-
colonialism, Chapters 2–3, explains the steamship business by discussing three major
shipping companies between 1860 and 1882. The China Navigation Company and the
Indo-China Steam Navigation Company were controlled by British companies. On the
other hand, the China Merchants Steam Navigation Company, which was sponsored by
the Qing government, became the largest shipping firm.
175

chi-kong lai
showed that, from 1872 to 1884, when the operational management of the
company was left in the hands of merchants and simultaneously received
government support in the form of annual subsidies (assigned as the official
carrier of state goods on designated routes and other temporary financial
assistance), and the exclusive right to carry the grain tribute rice to the capital
(with the imported British vessel Eaton; then later Yongqing, Fuxing, and
Liyun), it succeeded. However, the corruption of the era, including nepotism
and embezzlement, resulted in the loss of funds to other ventures which
collapsed in the turmoil of the 1880s. When government involvement in the
actual management of the firm increased from 1885 to 1911, the CMSNC
failed.
The China Merchants Company was established as an effort to compete
profitably with foreign shipping. Profits were returned to investors, but
ultimately failed to reclaim the domestic shipping industry.29 Under the
supervision of governor-general Li Hongzhang (later Yuan Shikai, then the
minister of posts and communication), new regulations came out in 1909, and
semiofficial status was only modified rather than ended by the government’s
acceptance of the new corporate structure. In September 1909, the company
held the first shareholder meeting in Shanghai and the first board of directors
was formed. It was registered with the Ministry of Agriculture, Industry, and
Commerce as a private mercantile enterprise in 1911.30 In 1912 a decision was
made in an extraordinary shareholder meeting, agreeing to lend the com-
pany’s assets to the new government of Sun Yat-sen for a mortgage loan.
With the early success of the CMSNC project, it became a model for the
founding of similar commercial ventures such as the Hubei Coal Mining
Company in 1875, the Kaiping Mining Company in 1877, the Shanghai Cotton
Textile Mill in 1878, and the Imperial Telegraph Administration in
December 1881.31
In 1877, Li Hongzhang founded the Kaiping Mining Company, which is
considered the most triumphant enterprise supported by the Qing state
because there was also a collaboration with the commercial sector to allow
private investors to fund the mine.32 Li drew on his successful experience in
forming the China Merchants Shipping Company, and he appointed Tong
29
Xia, 洋务运动史, pp. 192, 195–6.
30
Chu and Liu, Li Hung-chang and China’s Early Modernization, p. 167.
31
Lai, “Li Hung-chang and Modern Enterprise,” 218.
32
See T. Wright, Coal Mining in China’s Economy and Society 1895–1937 (Cambridge,
Cambridge University Press, 1984); E.C. Carlson, Origins and Early Development of the
Kaiping Enterprise: The Kaiping Mines, 1877–1912 (Cambridge, MA, Harvard University
Asia Center, Harvard University, 1971), pp. 1–23; Q. Liu, “Yan Fu and Kaiping Mines:
176

Kingsing to supervise the mines. As an excellent comprador businessman,

Tong was familiar with market conditions and the management practices of
new enterprises.33 And more importantly, his arrival attracted much mer-
chant investment in the new mining business. Tong Kingsing was
a professional manager, trusted by Li. With his prestige in the business
world, the Kaiping Mining Company managed to raise 7,000 of the 8,000
shares planned in its first year of operation.34 With the “introduction” of
Sheng Xuanhuai and Li Hongzhang’s “sponsorship,” Tong was able to
successfully take charge of the Kaiping Mining Company. Li Hongzhang
maintained a “confident” attitude towards his managers, did not interfere too
much in the management and operation of enterprises, and generously
provided help and political protection when enterprises faced difficulties in
funding and public opinion.35
During this period, various operations were established, including (not
a complete list) the Zichuan Lead Mine (1875), the Lanzhou Weaving Bureau
(1877), the Jingmen Coal Mine (1879), the Imperial Telegraph Administration
(1881), the Pingquan Copper Mine (1881), the Mohe Gold Mine (1889), the
Shanghai Cotton Cloth Mill (1890), the Hanyang Ironworks (1890), and the
Imperial Bank of China (1896).
This modern-enterprise model in late Qing was not a fully “westernized”
management model and the start-up scale of investment in most of these
enterprises in the 1895–1913 period was limited.36 With government supervi-
sion of funds, investment in these modern enterprises came with significant
financial risks as cases of mismanagement, financial misconduct, financial
discrepancies, and corruption occurred. As a result, merchants became
discouraged and increasingly cautious of becoming involved in state-
supervised modern enterprises.
The Meaning of Economic Liberalism in Early Modern China,” in G. Campagnolo

(ed.), Liberalism and Chinese Economic Development: Perspectives from Europe and Asia
(New York, Routledge, 2016), pp. 49–62. For Chinese scholarship, see Yun Yan 云姸,
近代开滦煤矿研究 (A Study of the Kailuan Coal Mines in Chinese Modern Times)
(Beijing, Renmin chubanshe, 2015).
33
Tang Tingshu 唐廷枢, “请开采开平煤铁並兴办铁路禀” (Self-Strengthening
Movement V I I) (Shanghai, Shanghai Renmin Publishing, 2000), pp. 119–24.
34
F. Zhao, “Kailuan Coal Mines: The Best Coal Mine Was Robbed by Japan,” Yingcai 10
(2012), n.p.
35
W.K.K. Chan, “Government, Merchants and Industry to 1911,” in J.K. Fairbank and
K.-C. Liu (eds.), The Cambridge History of China, vol. 11, Late Ch’ing, 1800–1911, part 2
(Cambridge, Cambridge University Press, 1980), p. 436.
36
Sun Yutang 孫毓棠, 中国近代工業史資料 (Information on the Modern Industrial
History of China), first series, 1840–1895, vol. 2 (Beijing, Kexue chubanshe, 1957), pp.
1166–9; A. Feurwerker, “Economic Trends in the Late Ch’ing Empire, 1870–1911,” in
Fairbank and Liu, The Cambridge History of China, vol. 2, p. 37.
177

chi-kong lai
Zhang Zhidong became the governor of Shanxi in 1882; governor-general of

Guangdong and Guangxi from 1884 to 1889; governor-general of Jiangsu, Anhui,
and Jiangxi in the lower Yangzi region during the First Sino-Japanese War; a long-
serving governor-general of Hubei and Hunan from 1889 to 1907; and a grand
councillor after 1907.37 He was the most influential reformer after Li Hongzhang.
To escape the vicious cycle of officially supervised modern enterprises, Zhang
Zhidong pursued the new scheme of guanshang heban 官商合办. Merchant–
state official joint management was introduced not only to improve merchant–
state relations but also to allow for a different business environment after a series
of setbacks. This scheme offered merchant partnerships to those who invested in
the Hubei Cotton Cloth Mill, and they were assured of annual dividends; it also
allowed for a more balanced relationship between the officials and the mer-
chants. These partnerships with merchants did not change this trend of increas-
ing bureaucratic involvement.
As early as 1881, Zhang was at great pains to duplicate the Tianjin model
(pioneered by Li Hongzhang) in Shanxi. When Zhang Zhidong established
his enterprises, he took a different approach from Li Hongzhang. He didn’t
trust businessmen who knew the economy. From 1888 to 1906, in the
management of several factories, including the Hubei Cotton Cloth Mill,
regardless of the proportion of merchant investment, Zhang insisted on
appointing his own managers to manage the enterprise. These policies
greatly hurt the enthusiasm of businessmen for investing in his business.
Steel- and ironworks were the key industries of Zhang’s Self-Strengthening
industrial policy. The Hanyeping company was not the first industrial plant
established in China,38 but it was officially established under Zhang Zhidong.
Quan Hansheng’s history of the Hanyeping Company is the classic on this
topic. He explores the running of the Hanyeping Company in three phases:
guanban 官办 (government-run), guandu shangban (official supervision and
merchant management), and shangban 商办 (merchant management).39
37
See Li Xizhu 李细珠, 张之洞与清未新政硏究 (The Study of Zhang Zhidong and Late
Qing Reform) (Beijing, Zhongguo shehui kexue chubanshe, 2015); Feng Tianyu 冯天瑜
and Chen Feng 陈锋 (eds.), 张之洞与中国近代化 (Zhang Zhidong and China’s
Modernization) (Beijing, Shehui kexue chubanshe, 2010). For English-language scholar-
ship on Zhang Zhidong, see, for example, W. Ayers, Chang Chih-tung and Educational
Reform in China (Cambridge, MA, Harvard University Press, 1971); and D.H. Bays, China
Enters the Twentieth Century: Chang Chih-tung and the Issues of a New Age, 1895–1909 (Ann
Arbor, The University of Michigan Press, 1978).
38
The first industrial plant was built in Guizhou in 1889 due to misperceptions that the
area was rich in raw materials and in order to counter foreign imports.
39
Quan Hansheng 全汉昇, 汉冶萍公司史略 (A Brief History of the Hanyeping Iron and
Coal Mining and Smelting Company, 1890–1926) (Hong Kong, The Chinese University
of Hong Kong Press, 1972). Also see Zhang Houquan 张后铨, 汉冶萍公司史 (A
178

When Zhang Zhidong founded the Hanyang Ironworks, he was supported

by the central government. Zhang lacked relevant knowledge of production
technology and scientific business management concepts. As a result, mis-
takes were made in fuel, machinery, and equipment, and in the selection of
the site of the steel plant, leading to losses. When losses occurred, Zhang
Zhidong pinned his hopes on the Qing government at first. However, due to
defeat in the Sino-Japanese War of 1894–1895, the Qing government was
unable to support his enterprises and he had to invite businessmen to
undertake them. The government became the creditor of Hanyang
Ironworks. After the failure of the guanban Hanyang Ironworks, Sheng
Xuanhuai took over, which also marked the beginning of the guandu shangban
stage of the Hanyeping Company.
A shortage of funds in the course of private-enterprise operations was
likely to cause bankruptcy, but this was not a difficult problem for these
guandu shangban enterprises – they pulled money from other companies to fill
the gap. In 1904, Sheng Xuanhuai drew more than a million taels from the
China Merchants to add to the Hanyang Ironworks.40 In 1908, he drew more
than 40,000 taels of share capital from the Hua Sheng Spinning and Weaving
Mill to add to the Hanyeping Coal and Iron Co.41 But such adjustments could
also pose a hazard to the company being drawn from. Zhang Zhidong drew
on the profits and capital of the Hubei Cotton Cloth Mill to support his other
enterprises, resulting in stagnation in its development.42
After Sheng Xuanhuai took over the Hanyang Ironworks, the first thing to
be solved was the problem of raising capital. His solution was to raise funds
through a public offering, recruit private shareholders, and hope to obtain the
support of private capital.
During the period of guandu shangban, the output and turnover of the
company increased significantly and achieved gratifying results. However,
Sheng Xuanhuai used the guandu shangban model to supervise the business
and did not recognize the rights of investors. On the surface, ownership was
subordinate to the government, but in fact it was controlled by Sheng. This
History of the Hanyeping Corporation) (Beijing, Shehui kexue wenxian, 2014); Dai Lu
代鲁, 汉冶萍公司史硏究 (A Study of the History of the Hanyeping Corporation)
(Wuhan, Wuhan daxue chubanshe, 2013). Also see T. Kennedy, “Chang Chih-tung and
the Struggle for Strategic Industrialization: The Establishment of the Hanyang Arsenal,
1884–1895,” Harvard Journal of Asiatic Studies 33 (1973), 154–82.
40
Chan, “Government, Merchants and Industry to 1911,” p. 429.
41
Tianjin archives collection, Interest-Bearing Roll of Deposits in the Shanghai Textile
Mill and Hanyeping Coal and Iron Co., New Stock Volume.
42
Chan, “Government, Merchants and Industry to 1911,” p. 429.
179

chi-kong lai
was a serious violation of the original intention of private investors to pursue

profits, not to mention promote the growth of new enterprises. This mode
greatly limited and fettered the collection of private funds, which was very
unfavorable to the capital accumulation of the Hanyeping Company. After
1908, Sheng Xuanhuai put forward the idea of shangban or merchant manage-
ment. The Hanyeping Company emerged as the result of a merger between
the Hanyang Ironworks, the Daye Iron Mine (大冶鐵礦), and the Pingxiang
Coal Mine.
The merger was an attempt to regain investor confidence and secure
funding to maintain the daily operations of the company which were essen-
tial in the production of top-quality steel. The merger allowed companies to
purchase shares, giving Hanyeping greater business opportunities with
investors’ money, e.g. to repay outstanding debts, reinvest in existing
machinery, and obtain extended loans. Past problems of poor credibility,
such as excessive borrowing and high debt accumulation, were enough to
drive away existing and/or potential investors.
In the industrial development of modern China, the combination of
bureaucrats (guan) and merchants (shang) promoted the emergence of indus-
trial development. However, in order to continue national industrial devel-
opment, it was necessary for guan and shang to go their separate ways and
reduce the interference of state power in the industrial process. Despite this
company’s failure in some respects, it became a crucial part of the Qing
government’s industrial promotion efforts, bringing China one step further
along the road to industrialization.
Railway development undertaken by late Qing officials like Zhang
Zhidong signaled a major advance in Chinese transport capabilities, as
shown in Elisabeth Köll’s recent book focusing on the Jin-Pu Railway.
Many years later, Li Hongzhang, along with Zhang Zhidong and Liu
Mingchuan, advocated for railways to be put in place. The development of
the railways took the form of partnerships with the foreign powers. The
government signed agreements with the British and the Germans to build
important lines to connect the coast to cities like Beijing. By using the
concessions, the government did not have to invest the whole of the capital
for this state building project. The Jin-Pu Railway was roughly concentrated
around the northeastern part of China, connecting Beijing and Shanghai, as
well as connecting with the countryside. However, as most countries had
significant railway networks by the 1900s, China was lagging behind.43
43
Köll, Railroads and the Transformation of China, pp. 19–50.
180

As governor of Shandong (1899–1901), as governor-general of the capital

province, Zhili (1901–7), and later as grand councillor (1908) and prime
minister, Yuan Shikai enacted broad effective reforms in areas such as
currency, banking, agriculture, industrial development, policing, education,
and others.44 Yuan Shikai did contribute positively to China through reforms
and enterprise policies. Yuan’s official, Zhou Xuexi, offered companies polit-
ical protection. The creation and development of the Qi Xin Cement C.o
启新洋灰公司 was like that of the Kaiping Mining Company, but depended
even more on the support of official funds. At the beginning of the develop-
ment of the Qi Xin Cement Co., one million silver dollars in start-up funds
was needed, and Zhou Xuexi received the strong support of Yuan Shikai
because of their affinity. This million dollars was all supported by a Tianjin
Official Silver bank loan at Yuan Shikai’s behest, and the loan period was for
ten years with annual interest of 5 percent. Yuan Shikai was involved in the
first attempt to introduce business regularization by initiating the Company
Law in 1904 under the Qing government.45 As David Faure argues, the
introduction of company law was a game changer as it provided some legal
protection for private business.46
Conclusion
The relationship between “officials” and “businessmen” was quite special
in the “Self-Strengthening” enterprises, which perhaps marked the first time
in the recent history of China that the administration had publicly supported
businesspeople in establishing and developing large modern enterprises
related to the state. In the absence of a strong private sector, the rise of
modern industry during the late Qing was mainly a governmental effort,
which yielded uncertain benefits. Thus modern-type enterprises were started
44
E. Yong, The Presidency of Yuan Shih-k’ai: Liberalism and Dictatorship in Early Republican
China (Ann Arbor, University of Michigan Press, 1977), pp. 54–5. Also see S.R. MacKinnon,
Power and Politics in Late Imperial China: Yuan Shi-kai in Beijing and Tianjin, 1901–1908
(Berkeley, University of California Press, 1980); J. Ch’en, Yuan Shih-K’ai, 2nd ed.
45
W.C. Kirby, “China Unincorporated: Company Law and Business Enterprise in
Twentieth-Century China,” Journal of Asian Studies 54.1 (February 1995), 43.
46
D. Faure, China and Capitalism: A History of Business Enterprise in Modern China
(Hong Kong, Hong Kong University Press, 2006), pp. 45–64. Also see
Zhang Zhongmin 张忠民, 艰难的变迁:近代中国公司制度硏究 (The Difficulty of
Change: A Study of the Institution of the Modern Corporation in Modern China
(Shanghai, Shanghai shehui kexue chubanshe, 2002); Gao Xinwei 高新伟, 中国近代
公司冶埋 (1872–1949) (Corporate Governance in Modern China (1872–1949) (Beijing,
Shehui kexue wenxian chubanshe, 2009).
181

chi-kong lai
in the late Qing with state encouragement, but failed to develop into
a genuine “industrial revolution.”
The Qing court does deserve credit for supporting massive infrastructure
projects that led to the expansion of shipping companies, as well as the
extension of telegraph networks and railway lines. This infrastructure
improved commercial transactions in late Qing and later Republican China.
The absence of, for example, commercial and private property laws led the
private sector to rely too heavily on official protection. The partnership also
sometimes resulted in the incompetent interference of officials, and permit-
ted corruption by entrepreneurs. Hence the introduction of a nationwide
Company Law (1904) to protect industrialized corporations marked the
beginning of a new phase as the Chinese state adapted its relationship with
private business to modernize the economy.47 In the end, the law still did not
provide sufficient protection for private property rights as nationalization of
private businesses still occurred on a regular basis.
Further Reading
Bickers, R., and L. Jackson (eds.), Treaty Ports in Modern China: Law, Land and Power
(London, Routledge, 2016).
Chan, W.K.K., Merchants, Mandarins and Modern Enterprise in Late Ch’ing China
Chen, J., “Recent Chinese Historiography on the Western Affairs Movement (the Yangwu
yundong, ca. 1860–1895),” Late Imperial China, June 1986, 112–27.
Chu, S.C., and K.-C. Liu (eds.), Li Hung-chang and China’s Early Modernization (Armonk,
M.E. Sharpe, 1994).
Faure, D., China and Capitalism: A History of Business Enterprise in Modern China
Feng, Tianyu 冯天瑜 and Chen Feng 陈锋 (eds.), 张之洞与中国近代化 (Zhang
Zhidong and China’s Modernization) (Beijing, Shehui kexue chubanshe, 2010).
Feuerwerker, A., China’s Early Industrialization: Shen Hsuan-huai (1844–1916) and Mandarin
Enterprise (Cambridge, MA, Harvard University Press, 1958).
Halsey, S.R., Quest for Power: European Imperialism and the Making of Chinese Statecraft
Hao, Y., The Commercial Revolution in Nineteen-Century China: The Rise of Sino-Western
Mercantile Capitalism (Berkeley, University of California Press, 1986).
Kirby, W.C., “China Unincorporated: Company Law and Business Enterprise in
Twentieth-Century China,” Journal of Asian Studies 54.1 (February 1995), 43–63.
Leonard, J.K., and J.R. Watt (eds.), To Achieve Security and Wealth: The Qing Imperial State
and the Economy, 1644–1911 (Ithaca, NY, Cornell University Press, 1992).
47
Kirby, “China Unincorporated,” 43.
182

Li, Xizhu 李细珠, 张之洞与清未新政硏究 (The Study of Zhang Zhidong and Late Qing
Reform) (Beijing, Zhongguo shehui kexue chubanshe, 2015).
Pong, D., Shen Pao-chen and China’s Modernization in the Nineteenth Century (New York,
Quan, Hansheng 全汉昇, 汉冶萍公司史略 (A Brief History of the Hanyeping Iron and
Coal Mining and Smelting Company, 1890–1926) (Hong Kong, The Chinese University
of Hong Kong Press, 1972).
Reinhardt, A., Navigating Semi-colonialism: Shipping, Sovereignty, and Nation-Building in
China, 1860–1937 (Cambridge, MA, Harvard University Press, 2018).
Xia Dongyuan 夏东元, 洋务运动史 (The History of the Self-Strengthening Movement)
(Shanghai, Huadong shifan daxue chubabshe, 2010).
Zhang Guohui 张国辉, 洋务运动与中国近代企业 (The Self-Strengthening Movement
and China’s Modern Enterprise) (Beijing, Zhongguo shehui kehui chubanshe, 1979).
Zhang Zhongmin 张忠民, 艰难的变迁:近代中国公司制度硏究 (The Difficulty of
Change: The Study of the Institution of the Modern Corporation in Modern China
(Shanghai, Shanghai Shehui kexue chubanshe, 2002).
183

6
State Enterprises during the First Half
of the Twentieth Century
morris l. bian
The first half of the twentieth century witnessed the rise of the state sector of
the Chinese economy. The rise of the state sector manifested in the develop-
ment and expansion of central state enterprises and regional state enterprises
and resulted from the ideology and policy of the developmental state.1 This
chapter traces the emergence and evolution of the ideology and policy of the
developmental state, describes the development and expansion of central
state enterprises and regional state enterprises, and addresses the issue of
change and continuity across the 1949 divide.
Creating the Ideology and Policy of the

Developmental State
During the first half of the twentieth century, the Nationalist elite developed
an ideology of the developmental state. In its final form, the ideology of the
developmental state consists of an emphasis on state-owned enterprise,
1
The concept of the “developmental state” was used first by Chalmers Johnson in his
study of Japanese industrial policy. According to Johnson, the Japanese developmental
state is characterized by the existence of an elite bureaucracy with sufficient scope to
take the initiative and operate effectively, the existence of a pilot organization to lead
economic development, and the perfection of market-conforming methods of state
intervention in the economy. Taking as their point of departure Johnson’s account of
the Japanese developmental state, recently a number of scholars have examined the
theory and practice of the developmental state in the context of East Asia. The concept
of the developmental state has also been fruitfully appropriated by scholars of
Republican China. See C. Johnson, MITI and the Japanese Miracle: The Growth of
Industrial Policy, 1925–1975 (Stanford, CA, Stanford University Press, 1982); Meredith Woo-
Cumings (ed.), The Developmental State (Ithaca, NY, Cornell University Press, 1999);
V. Chibber, “Building a Developmental State: The Korean Case Reconsidered,”
Politics & Society 27.3 (September 1999), 309–46; W.C. Kirby, “Engineering China: Birth
of the Developmental State, 1928–1937,” in Wen-hsin Yeh (ed.), Becoming Chinese:
Passages to Modernity and Beyond (Berkeley, University of California Press, 2000), pp.
137–60; J.B. Knight, “China as a Developmental State,” World Economy 37.10 (October
2014), 1335–47.
184

State Enterprises during the First Half of the Twentieth Century
a stress on heavy industry, a focus on national defense, and a determination

to create a planned socialist economic system. Although this ideology took its
final form during the Second Sino-Japanese War (1937–1945), core elements of
this ideology can be traced to the writings of Sun Yat-sen, the founding father
of the Chinese republic. As early as 1912, Sun stated that all major industries in
China should be owned by the state. Sun Yat-sen’s International Development of
China also laid the foundation for the Nationalist conception of heavy indus-
try, discussing the development of key and basic industries.2 Finally, the idea
of economic planning came from Sun Yat-sen as well. According to Sun,
China must develop a comprehensive plan to take advantage of the oppor-
tunity of the post-World War I period and to make use of new capital and
experienced personnel of the belligerent nations to develop large-scale
industries.3 In his 1924 lectures on the Three Doctrines of the People, Sun
continued to emphasize state-owned industries.
It should be pointed out that Sun Yat-sen was not alone in emphasizing
state ownership; leading members of the Beijing government such as Zhou
Xuexi and Zhang Jian also advocated state ownership and management of
key industries.4 As a result, despite the enactment of laws and the creation of
regulations promoting the development of private enterprise, those very
laws and regulations also stipulated state ownership and management of
key industries such as salt and kerosene as well as telecommunications.
During the 1920s, the Nationalist Party incorporated many of Sun’s views
into its platform, including those on the scope and function of state-owned
and private enterprise. Thus the 1924 Manifesto of the First Nationalist Party
Congress provided that enterprises that were monopolistic in nature and the
development of which lay beyond private means ought to be undertaken and
managed by the state.5 After the success of the Nationalist Revolution, the
party’s leadership accepted Sun’s vision of economic development and made
it part of the dominant ideology of the Chinese state. During the late 1930s
2
Sun Yat-sen 孙中山, 建国方略 (Strategy for China Reconstruction) (Shanghai,
Shangwu yinshu guan, 1927). The second part – Material Reconstruction – was translated
into English in 1921 and published in 1922 as International Development of China by G.
P. Putnam’s Sons, 2nd ed. (New York and London, The Knickerbocker Press, 1929).
3
Sun Yat-sen, Fundamentals of National Reconstruction (Taipei, Sino-American Publishing
Co. Ltd, 1953), p. 189.
4
Wu Qian 武乾, “北洋政府时期的经济法与经济体制的二元化” (Economic Law
and the Development of Two Orientations of the Economic System), 法商研究
(Study of Law and Commerce) 1 (2003), 126–32.
5
Xiao Jizong 萧继宗 (ed.), 中国国民党宣言集 (Collection of Manifestos of the Chinese
Nationalist Party), vol. 69 (Taipei, Zhongguo Guomindang zhongyang weiyuanhui
dangshi weiyuanhui, 1976), p. 91.
185

morris l. bian
and the 1940s, in response to the crisis and war, leading figures of the
Nationalist government, such as Weng Wenhao, elaborated on Sun Yat-
sen’s views. Weng Wenhao was the most important economic figure in the
Nationalist government (outside the financial sector), serving, for example, as
head of the Ministry of Economic Affairs from 1938 to 1946. In a 1938 report,
Weng divided state-owned enterprise into four broad categories: national
defense, basic manufacturing industries, important mineral products, and
power. Weng declared emphatically that these enterprises aimed to increase
the productive capacity of the nation; they were not designed to compete for
profit with the people.6 By 1942, however, Weng’s position had shifted to
emphasize state-owned enterprise even more. In that year, Weng argued
that, first, China must industrialize. Second, China must focus on heavy
industry in order to bring industrialization to fruition. Finally, the govern-
ment must rely on state-owned enterprise to lay the foundation for China’s
heavy industrial development.7 Less than a year later, Weng offered a more
concrete definition of the scope of state-owned and private enterprise. In
principle, basic industries should be owned and managed by the state. In
contrast to the large scope of state-owned enterprise in these industries, light
industries should in principle be owned and managed by private enterprise.8
As was the case with the emphasis on state-owned enterprise and heavy
industry, the focus on national defense was also the outcome of an evolution
of several decades. Fundamentally, it was the increasing Japanese threat to
China’s national security after the mid-1930s and the crisis triggered by the
war that caused the Nationalists to underline the importance of national
defense. The focus on national defense did not find explicit expression in Sun
Yat-sen’s Three Doctrines of the People. The first indication of a major shift
from the preoccupation with people’s livelihood to a focus on national
defense was the four-year plan of national reconstruction adopted by the
Fourth Nationalist Party Congress in response to the Japanese invasion of
Manchuria in November 1931. With national survival at stake, the plan
focused on national defense, used a hypothetical enemy as the target of
6 “经济部近时工作纪要” (Work Summary by the Ministry of Economic Affairs), 新经

济 (The New Economy) 1.2 (December 1, 1938), 58–62.
7
Weng Wenhao 翁文灏, 国营重工业的意义与任事同人的责任 (The Meaning of
State-Owned Heavy Industry and the Responsibility of Fellow Comrades), 资源委员
会公报 (Bulletin of the National Resources Commission) 3.2 (August 16, 1942), 75–80.
8
Weng Wenhao 翁文灏, “中国经济建设概论” (Survey of Chinese Economic
Reconstruction), in Weng Wenhao, 中国经济建设论丛 (Collection of Essays on
Chinese Economic Reconstruction) (National Resources Commission, 1943), pp. 81–2.
186

reconstruction, and defined the scope of reconstruction according to neces-

sity and viability.9
As war approached, the Fifth Nationalist Central Executive Committee in
February 1937 declared that Chinese economic reconstruction had two
objectives: to meet the needs of national defense and to improve people’s
welfare. The first objective was designed to bring to fruition the Doctrine of
Nationalism; the second objective was meant to implement the Doctrine of
People’s Livelihood.10 Subsequently, the crisis prompted by the war greatly
strengthened this national-defense orientation. In a 1938 essay, Qian
Changchao, deputy director of the National Resources Commission, stated
that all economic reconstruction should focus on national defense, and China
must take immediate steps to create a center for national defense in the
interior.11 Weng Wenhao shared Qian’s belief in the primary importance of
national defense. In a 1942 speech, Weng described how the Soviet Union and
Germany had undertaken large-scale heavy industrial reconstruction by
using five-year and four-year plans, and he pointed out that both the Soviet
Union and Germany believed that economic reconstruction was the only
solution to national rejuvenation. As long as economic reconstruction was
designed to increase national strength, it was appropriate for the whole
nation to work hard and endure hardships.12
Another integral part of the Nationalist ideology of economic develop-
ment was the determination to create a planned socialist economic system.
Should the government follow the earlier policy of the first industrialized
nations by keeping its hands off economic activities, or was the goal of
a modern industrial nation to be achieved through economic planning and
eventually through a planned socialist economic system? The Nationalist
elite clearly chose the latter path. After the Sino-Japanese War erupted, the
Provisional Nationalist Party Congress, meeting in April 1938, adopted
a Program for the War of Resistance and Nation Building, declaring that
9
国家建设初期方案案 (Resolution on National Reconstruction in the Near Future),
November 21, 1931, in 中华民国史档案资料汇编 (Collection of Archival Materials in
the History of Republican China), part 5, vol. 1, zhengzhi, no. 2, p. 337.
10
中国经济建设方案 (Resolution on Chinese Economic Reconstruction), February 19,
1937, in 中华民国史档案资料汇编, part 5, vol. 1, zhengzhi, no. 2, pp. 618–25.
11
Qian Changchao 钱昌照, “两年半创办重工业之经过与感想” (The Experience of
and Reflection on Creating Heavy Industry for the Past Two and a Half Years), 中国第
二历史档案馆档案 (Second Historial Archives, China), 28.2, 939. See also 新经济
(New Economy) 2.1 (June 16, 1939), 2–6.
12
Weng Wenhao 翁文灏, “国防经济建设之要义” (The Meaning of Economic
Reconstruction for National Defense), July 25, 1941, reprinted in 资源委员会公报 1.2
(August 16, 1941), 69–72.
187

morris l. bian
China must create a planned economic system, encourage domestic and

foreign investment, and expand wartime production.13 Writing in 1941,
Weng Wenhao distinguished among the planned economic system of the
Soviet Union; the economic systems completely controlled by the govern-
ments of Germany, Italy, and Japan; and the partially controlled economic
systems in Great Britain and the United States. Although Weng recognized
merits in all three types of economic system, he believed that none of them
was entirely consistent with China’s needs. Consequently, Weng proposed
that China eclectically adopt elements from all three of these systems. As he
put it, China ought to create a planned and controlled economic system,
under which the government should take major responsibility by making
national defense its top priority, by promoting state-owned enterprise, and by
controlling private enterprise.14
In short, the Nationalist ideology of economic development resulted in
part from Sun Yat-sen’s legacy, but, more importantly, also from the increas-
ing Japanese threat to China’s national security and the wartime crisis. The
Nationalist elite collectively developed an ideology of economic develop-
ment based on state-owned enterprise, heavy industry, national defense, and
the determination to create a planned socialist economic system. Ultimately,
this ideology led directly to the development and expansion of central and
regional state enterprise.
Developing and Expanding Central State

Enterprises in Heavy Industry
After establishing a functional central government in 1928, the Nationalist
regime began creating the institutional framework for economic planning
and the development and expansion of central state enterprises. The govern-
ment created a National Reconstruction Commission in 1929 and a National
Economic Commission in 1931. During this period, the government drew up
at least four major economic plans. The 1932 National Defense Planning
Commission was the first institutional embodiment of the new defense
orientation. Chiang Kai-shek was president of this commission and the
Military Affairs Commission to which it reported, but Weng Wenhao was
13
中国国民党抗战建国纲领 (Program for the War of Resistance and Nation Building),
April 1, 1938, in 中华民国史档案资料汇编, part 5, vol. 2, zhengzhi, no. 1, 386–9.
14
Weng Wenhao 翁文灏, “经建方向与共同责任” (The Direction of Economic
Reconstruction and Our Shared Responsibility), 中国第二历史档案馆档案 (Second
Historial Archives, China), 28.2, 314. See also 新经济 6.7 (January 1, 1942), 136–9.
188

secretary-general, and Qian Changzhao deputy secretary-general. The com-

mission was staffed by technocrats and had well over 100 members by 1934.15
In April 1935, the commission was renamed the National Resources
Commission, which marked an important change in purpose and direction.
During the next three years the National Resources Commission, under the
same leadership, changed its orientation from resource investigation and
planning to heavy industrial reconstruction. In effect, the organization was
transformed from Chiang Kai-shek’s brain trust to an organization in charge
of industrial development. Under the Three-Year Plan for Heavy Industrial
Reconstruction completed in 1936, heavy industry received the lion’s share of
investment capital. Geographically, most planned factories were to be built in
interior provinces such as Hunan and Jiangxi for fear of further Japanese
aggression.16 Twenty-one out of thirty planned factories and mines were
under construction by the time the Sino-Japanese War broke out in July 1937.
The outbreak of war and the relocation of the Nationalist government to
Chongqing led to further changes in the National Resources Commission’s
organization and activities. In March 1938 the Nationalist government placed
the commission under the jurisdiction of the newly created Ministry of
Economic Affairs, with Weng Wenhao as minister. Although Weng
Wenhao and Qian Changzhao continued to lead the organization, their
official titles were changed from secretary-general and deputy secretary-
general to director and deputy director respectively. The commission’s
function was redefined to include the creation and management of state-
owned industries and state-owned enterprises.17
The eruption of war caused serious disruptions in the implementation of
the three-year reconstruction plan and forced some of the half-finished
factories to relocate to the interior. Despite the disruptions and forced
relocation, the National Resources Commission continued to rely on the
mechanism of planning for heavy industrial reconstruction. If anything,
wartime military and economic mobilization added a greater sense of
urgency for planned and co-ordinated development. Thus, after relocating
15
国防设计委员会人事文件 (Personnel Documents of the National Defense Planning
Commission), 1934, 中国第二历史档案馆档案 (Second Historial Archives, China),
47, 115.
16
Qian Changchao 钱昌照, “两年半创办重工业之经过与感想,” 中国第二历史档案
馆档案 (Second Historial Archives, China), 28.2, 939. See also 新经济 2.1 (June 16,
1939), 2–6.
17
经济部资源委员会组织条例 (Organizational Regulations of the National Resources
Commission), August 1, 1938, in 资源委员会月刊 (National Resources Commission
monthly) 1.1 (April 1939), 63–4.
189

morris l. bian
and constructing factories in the interior between late 1938 and early 1939, the
commission developed a new Three-Year Reconstruction Plan for Heavy
Industry. The plan called for the establishment of new factories or the
expansion of existing ones in the metallurgical industry, the chemical indus-
try, the liquid-fuel industry, the machine industry, transportation and com-
munication equipment, and the leather and rubber industries.18 Although
precise figures of total government spending are unavailable, documents that
are available indicate that the government appropriated ¥329.97 million
(¥28.21 million in 1936 constant price) for heavy industrial reconstruction
during the three-year period.19
In early 1941, the National Resources Commission drafted another com-
prehensive economic plan, the Outline of a Three-Year Plan for National
Defense Industries. The Eighth Plenary Session of the Fifth Central Executive
Committee of the Nationalist Party adopted it in April 1941. According to this
plan, between 1942 and 1944 the National Resources Commission would
establish new factories or expand existing ones in the metallurgical, machine,
electrical, chemical, food, and energy industries. The total initial capital
needed was estimated at ¥816.69 million and $25.56 million, in addition to
¥259.74 million in liquid assets.20
Where did the National Resources Commission receive the investment
capital from? How did it finance heavy industrial projects? Archival
documents indicate that the commission received investment capital
from three sources: annual budget appropriation from the state treasury,
short-term loans and investment from state-run banks, and profits from
the export of mineral resources. Annual budget appropriations from the
state treasury constituted the major source of investment capital. Among
these, the single most important source of investment was an annual
budget appropriation from the state treasury. Archival sources indicate
that, after experiencing a steady rise from 1936 to 1941, the value of annual
state appropriations suffered a steep decline from 1942 to 1945. The share
of the National Resources Commission in the state budget showed
18
西南各省三年国防建设计划 (Three-Year Reconstruction Plan for National Defense
in Southwestern Provinces), 1939, 中国第二历史档案馆档案 (Second Historial
Archives, China), 28.2, 37.
19
Ziyuan weiyuanhui 资源委员会 (ed.), 复员以来资源委员会述要 (A Survey of the
Achievements of the National Resources Commission since 1945) (Nanjing, National
Resources Commission, 1948), pp. 38–9.
20
国防工业战时三年计划纲要 (Outline of the Three-Year Plan for National Defense
Industries), 中华民国史档案资料汇编, part 5, vol. 2, caizheng jingji, no. 6, 120–8.
190

a similar trend: after reaching a high of 2.8 percent in 1940, it declined

rapidly.21
Once the appropriations from the national budget were made available,
the National Resources Commission distributed them among heavy indus-
tries, such as energy, coal, petroleum, metal, iron and steel, machines,
electrical, and chemical. The commission’s own statistics reveal that between
1936 and 1945 the commission received ¥119.21 billion in investment capital
(¥71.78 million in 1936 constant price). Among the nine industries, the energy
industry received ¥50.73 billion, which accounted for 42.56 percent of the
commission’s investment capital. The industry that received the second-
largest share was the petroleum industry (15.83 percent), with a total of
¥18.87 billion.22
The allocation of investment capital made it possible for the National
Resources Commission to establish or take over roughly 130 heavy industrial
enterprises and organizations between 1936 and July 1945: nine in metallurgy,
seven in machines, five in electrical equipment, thirty-seven in chemicals,
thirty-eight in mining, twenty-seven in energy, and seven service organiza-
tions. Among them, the commission wholly owned and managed seventy-
five enterprises and organizations, partially owned and managed thirty-seven
enterprises and organizations, and invested in eighteen enterprises and
organizations. In addition, the majority were established between 1938 and
1942. Among those enterprises and organizations whose dates of establish-
ment are available, four were established in 1936, three in 1937, eleven in 1938,
nineteen in 1939, fifteen in 1940, twenty-four in 1941, sixteen in 1942, nine in
1943, ten in 1944, and one in early 1945.23 After the war ended, the commission
expanded its operations as it took over or confiscated enterprises controlled
or managed by the Japanese and their Chinese collaborators. By 1949, the
commission had 30,000 staff members and more than 600,000 workers. After
1949, some commission personnel followed the Nationalist regime to
Taiwan, but many stayed in mainland China and joined the Communist
government, including key commission leaders Weng Wenhao and Qian
Changzhao. Virtually all commission enterprises and corporations remained
state-owned enterprises or were reorganized as state-owned enterprises after
the Communist takeover.
21
Ziyuan weiyuanhui, 复员以来资源委员会述要 (1948), pp. 38–9.
22
Ziyuan weiyuanhui, 复员以来资源委员会述要 (1948), p. 40.
23
This calculation is based on “资源委员会经办事业一览表” (Tabulation of
Enterprises and Organizations of the National Resources Commission), 资源委员会
公报 9.2 (August 16, 1945), 43–51.
191

morris l. bian
Developing and Expanding Regional State

Enterprises
Regional state enterprises emerged in China toward the last decades of the
nineteenth century. Until the early 1930s, however, most of these enterprises
were individual entities with little connection or co-ordination between
them. Many were enterprises in the ordnance industry. The creation of
a national government in Nanjing in 1927, the diffusion of Sun Yat-sen’s
Three Doctrines of the People in various provinces, the national effort to
make plans for economic reconstruction, Japanese occupation of Manchuria
in 1931 and the threat it imposed to national security, and the perceived need
for Self-Strengthening by provincial authorities, all contributed to new efforts
to develop regional state enterprise in the early 1930s. In Shanxi province, for
instance, provincial authorities began to prepare for the establishment of
what would become the Northwestern Industrial Corporation in 1932. Then,
during the late 1930s and early 1940s, large enterprise corporations consisting
of numerous enterprises developed in sixteen provinces. Although the enter-
prises under these corporations varied significantly in terms of the cause of
their establishment, source of investment, authority and management struc-
ture, mode of operation, and scale and scope, they were all regional state
enterprises in terms of ownership, sources of investment, and stated object-
ives. Below I examine regional state enterprises in Guangdong, Shanxi, and
Guizhou provinces, focusing on developments during the 1930s and early
1940s.
Regional State Enterprises in Guangdong Province

In Guangdong, regional state enterprises date from Xicun Cement Plant,
which the governor of Guangdong established in 1906. The plant failed to
succeed even after its conversion to a private enterprise in 1921.24 After the
establishment of a national government in Nanjing in 1927, and echoing
national effort toward economic reconstruction, the Department of
Reconstruction in Guangdong drafted in 1929 a Program of Material
Reconstruction for Guangdong, which included a plan to build a new cement
factory at Xicun in Guangzhou. More plans for Guangdong province and for
24
Zhang Xaohui 张晓辉, 民国时期广东社会经济史 (A Social and Economic History of
Guangdong Province during the Republic Period) (Guangzhou, Guangdong renmin
chubanshe, 2005), pp. 180–1; Wu Zhengliang 卢征良 and Ke Weiming 柯伟明, “20
世纪 30 年代广东省营企业统制经营问题研究 – – 以广东士敏土厂为中心” (A
Study of Controlled Business Operation in Guangdong Province during the 1930s with
a Focus on the Guangdong Cement Plant), 民国档案 (Republic Archives) 1 (2017), 96–103.
192

Guangzhou municipality were made during the next two years. By the
summer of 1932, Chen Jitang, governor of Guangdong province, had com-
bined existing plans and programs into what became known as the Three-
Year Administrative Plan for Guangdong Province.25 Among other things,
the plan envisioned the establishment of twenty-four regional state enter-
prises with an initial capitalization of ¥97 million. In the end, the plan’s
implementation resulted in the establishment of only six regional state
enterprises: the Guangdong Cement Plant, two sugar refineries, the
Guangdong Chemical Plant, the Guangdong Textile Plant, and the
Guangdong Chemical Fertilizer Plant.26 In 1936, these seven factories were
capitalized at ¥6.5 million with a total of 3,860 employees.
After the Japanese occupation of Guangzhou in 1938, there was an increas-
ing paucity of military and civilian goods, as well as rising inflation. In
response, in February 1940 the provincial government created the
Guangdong Wartime Trade Administration for distributing goods and
administering wartime trade. The Guangdong Wartime Trade
Administration succeeded in what it did and made a profit the year it was
created. In June of 1941, the national government promulgated Regulations
Concerning Wartime Supervision and Management of Regional Trade,
which provided that provincial governments establish a corporation in
accordance with regulations concerning special limited-liability corporations.
Following this mandate, the provincial government reorganized the
Guangdong Wartime Trade Administration into the Guangdong Enterprise
Corporation (GDEC), which became operational on January 1, 1942.27
Unlike its predecessor, the GDEC’s business was not confined to trade; it
encompassed agriculture, industry, mining, trade, and transportation, for the
GDEC was designed to “promote economic reconstruction of Guangdong
25
Lian Haowu 连浩鋈, “陈济棠据粤的由来与‘广东省三年施政计划’的缘起” (How
Chen Jitang Took Over Guangdong Province and the Origins of the Three-Year
Administrative Plan for Guangdong Province), in 广东党史资料 (Material for Studying
the History of the Chinese Communist Party in Guangdong Province) 35 (2001), 378–420; A.
H.Y. Lin, “Building and Funding a Warlord Regime: The Experience of Chen Jitang in
Guangdong, 1929–1936,” Modern China 28.2 (April 2002), 177–212; Xiao Zili 肖自力, 陈济棠
(Chen Jitang) (Guangzhou, Guangdong renmin chubanshe, 2002), pp. 304–5.
26
Lian Haowu 连浩鋈, “陈济棠主粤时期 (1929–1936年) 广州地区的工业发展及其
启示” (Industrial Development and Its Implications When Guangdong Province Was
under Chen Jitang’s Control, 1926–1936), 中国社会经济史研究 (Studies in Chinese
Social and Economic History) 1 (2004), 90–9.
27
Ge Hongbo 葛洪波, “广东实业有限公司经营管理研究” (A Study of the Operation
and Management of Guangdong Industrial Limited Liability Corporation), unpublished
master’s thesis, Jinan University, 2001, 7.
193

morris l. bian
province.”28 Initially, the GDEC made the development of agriculture its

top priority because of Guangdong’s traditional lack of grain supply and
the worsening situation of grain supply after war had erupted. Still, the
GDEC began making preparations to build factories soon after its estab-
lishment. By the end of 1942, the GDEC had established a total of eight
plants in machinery, communications equipment, pharmaceutical prod-
ucts, chemical products, construction material, sugar products, and chem-
ical fertilizers (the factories were the Yuecang Machine Works, the
Yuehua Electrical Works, the Yuede Pharmaceutical Factory, the Yueli
Fertilizer Plant, the Yuexin Construction Material Plant, the Yuebei
Chemical Plant, and the Yuexing Sugar Refinery).29 These plants were
capitalized at ¥50 million at the end of 1944.30 The GDEC also developed
eight farms during the same period. Unfortunately, Japanese occupation
of northern Guangdong led to the confiscation and destruction of the
buildings and equipment of these factories.31
After the Japanese surrender in August 1945, the Guangdong provincial
authorities took over the plants run by the Japanese and their Chinese
collaborators and placed them under the supervision and management of
the Guangdong Industrial Corporation (GDIC).32 In November 1945, the
provincial government made the GDIC the leading organ for developing
the Guangdong economy by placing under its jurisdiction all factories and
farms as well as marketing and transportation operations that aimed to
increase production and improve people’s welfare.33 As of July 1948, the
GDIC had a total of eight plants under its jurisdiction. They were the
Xicun Cement Plant, the Shunde Sugar Refinery, the Guangzhou Textile
Plant, the Guangzhou Second Textile Plant, the Xicun Brewery, the
Guangzhou Ice Plant, the Meilu Jute Plant, and the Guangzhou Machine
28
广东企业公司章程 (Bylaws of Guangdong Enterprise Corporation), 1941, 广东省档
案馆档案, 19/1/13.
29
广东企业公司三十一年度业务报告 (Report by Guangdong Enterprise Corporation
on Its 1942 Business Operation), 广东省档案馆档案, 19/1/46.
30
Ge, 广东实业有限公司经营管理研究, 9.
31
Huang Juyan 黄菊艳, 抗战时期广东经济损失研究 (A Study of Economic Damage
Done to Guangdong Province during the Sino-Japanese War) (Guangzhou, Guangdong
renmin chubanshe, 2005), pp. 132–3, 141–2.
32
The Guangdong Enterprise Corporation (GDEC) was renamed the Guangdong
Industrial Corporation (GDIC) on 1 September 1943. See 广东实业公司业务报告,
1945 年 9 月至 1946 年 7 月 (Report by Guangdong Industrial Corporation on Its
Business Operations from September 1945 to July 1946), 广东省档案馆档案, 19/1/64.
33
广东省政府工作报告 (Work Report of Guangdong Provincial Government for
January through June 1946), p. 176, 广东省档案馆档案, zhengzhi/204. See also internal
correspondence of the Guangdong Industrial Corporation dated December 4, 1945,
广东省档案馆档案, 19/1/370.
194

Works.34 By early 1949, the number of GDIC factories had increased to

ten.35 Even though the number of enterprises was incredibly small, these
factories had some of the best machinery and equipment of any Chinese
enterprise, for most of their machinery and equipment had been imported
from advanced industrial countries such as Denmark, Britain, and the
United States.36
In terms of source of investment, the sole and exclusive investor in the
GDEC (GDIC after September 1, 1943) was the Guangdong provincial gov-
ernment. Correspondingly the provincial government appointed individuals
to serve as the corporation’s general manager and associate managers. At the
same time, the provincial government appointed three superintendents to
supervise the corporation’s management and operation. Finally, the provin-
cial government created a board of directors for the corporation, which
consisted of the corporation’s general manager and high-ranking officials of
the provincial government, such as the head of the Department of
Reconstruction and the Department of Finance. The superintendents and
the board of directors did not meet separately, however; they held joint
meetings. According to the corporation’s bylaws, the board of directors
ought to have met once a month. In reality, however, the board of directors
only met nine times during the war.37
In terms of corporate structure, the corporation created departments of
general affairs, planning, operation, and transportation at the time of its
establishment. Each department was subdivided into sections. In addition,
the corporation had an audit office, a secretary, an engineer, and a special
commissioner. In September 1943, the corporation abolished the Department
of Operation and reorganized the Department of Planning into the
Department of Production and the secretary into a secretariat. Another
major reorganization took place in July 1946 when the board of superintend-
ents and directors approved new Regulations on the Organizational
Establishment of the Guangdong Industrial Liability Corporation, which
stipulated the creation, under the corporation’s general manager, of divisions
of general affairs, production, business operation, and transportation; offices
34
广东实业公司沿革及业务状况 (The Evolution and Business Operations of
Guangdong Industrial Corporation), July 1948, 广东省档案馆档案, 19/1/358.
35
广东实业公司概要 (Introduction to Guangdong Industrial Corporation),
February 1949, 广东省档案馆档案, 19/1/56.
36
Ge, 广东实业有限公司经营管理研究, 13.
37
Ge, 广东实业有限公司经营管理研究, 22–3.
195

morris l. bian
of central accounting and inspection; and a secretariat.38 Further reorganiza-

tion took place in early 1947.39
Regional State Enterprises in Shanxi Province

In Shanxi province, the first regional state enterprise emerged in the 1880s,
but most of these enterprises had failed by the end of the first decade of the
twentieth century.40 The first major effort to develop modern industries
began after Yan Xishan became governor of Shanxi province in 1917.41
Between 1917 and 1930, the Shanxi provincial government established
more than twenty regional state enterprises.42 Although Yan Xishan was
defeated by Chiang Kai-shek during the Central Plains War of 1930, he was
appointed by Chiang as Shanxi’s pacification commissioner after the
Manchurian Incident in 1931. Yan Xishan was determined to consolidate
his power and contribute to national rejuvenation by developing the
Shanxi economy. In April 1932 Yan Xishan created a Planning
Commission of Shanxi Province, with himself as commissioner, and
began making a ten-year reconstruction plan for Shanxi province.
Among other things, the Ten-Year Shanxi Reconstruction Plan called for
“an increase in people’s production and the development of public
enterprise.”43
Although this document served as a blueprint for Shanxi’s development,
planning for the Northwestern Industrial Corporation (NIC) – the govern-
ment entity in charge of developing Shanxi’s industrial economy through the
creation of regional state enterprises – had started at the beginning of 1932,
several months before Yan Xishan created the Planning Commission of
Shanxi Province. Still, not until August 1933 was the NIC formally
established.44 Yan Xishan exercised direct control of the NIC in his capacity
38
广东实业有限公司组织规程 (Regulations on the Organizational Establishment of
Guandgong Industrial Corporation), 广东省档案馆档案, 6/2/1592.
39
广东实业公司业务报告, 1946 年 10 月至 1947 年 4 月 (Report by Guangdong
Industrial Corporation on Its Business Operations from October 1946 to April 1947),
广东省档案馆档案, 19/1/48.
40
Jing Zhankui 景占魁 and Kong Fanzhu 孔繁珠, 阎锡山官僚资本研究 (A Study of
Bureaucratic Capitalism under Yan Xishan) (Taiyuan, Shanxi jingji chubanshe, 1993),
pp. 40–1.
41
On how the Beijing government appointed Yan Xishan Shanxi governor, see
Jing Zhankui 景占魁, 阎锡山传 (A Biography of Yan Xishan) (Beijing, Zhongguo
shehui chubanshe, 2008), pp. 70–5.
42
Jing and Kong, 阎锡山官僚资本研究, pp. 69–72.
43
Jing Zhankui 景占魁, 阎锡山与西北实业公司 (Yan Xishan and Northwestern
Industrial Corporation) (Taiyuan, Shanxi jingji chubanshe, 1991), pp. 41–4.
44
Lian Feng 连峰, “山西地方工业化的初步尝试—以西北实业公司为例” (Early
Efforts toward Industrialization in Shanxi Province: A Case Study of the
196

as company president.45 According to NIC bylaws promulgated in

August 1935, the NIC was “designed to promote industrial development
in the northwest.” The bylaws stipulated a governance structure that
consisted of a central administration; an office of managers; a factory
administration in charge of all eleven plants; departments of mining,
technology, and research; and sections of general affairs, accounting,
operation, and evaluation/review.46 In 1936, the NIC underwent a major
reorganization. The central administration and factory administration
were abolished and replaced by company headquarters. Under the com-
pany headquarters were four division-level units: divisions of work affairs,
operation, general affairs, and accounting. Each division was subdivided
into several departments.47
The NIC expanded rapidly between 1933 and 1937. By the time the Sino-
Japanese War broke out, the NIC had thirty-three enterprises with a total of
20,648 employees. Twenty-three of them were capitalized at ¥20.514 million.
After the Japanese occupation of Taiyuan in November 1937, the Japanese
military confiscated much of the NIC’s machinery and equipment and
shipped them to Japan and other parts of occupied China in support of the
Japanese war effort.48 The NIC managed to ship a small portion of the NIC’s
machinery and equipment into Sichuan before Taiyuan fell. In the summer of
1939, the NIC reinvented itself as the New Northwestern Industrial
Corporation in Yichuan, Sichuan province, with Yan Xishan as company
president and Peng Shihong as general manager. Given its drastic reduction
of business operations, the New Northwestern Industrial Corporation had
only three division-level units: divisions of work affairs, general affairs, and
accounting. The earlier division of operation was now a department under
the division of general affairs. The period of the Sino-Japanese War was not
the best years for the NIC: it had only nine plants as of August 1945, with 3,486
employees.49
Despite wartime setbacks, the war’s conclusion brought about the NIC’s
revival and expansion (the NIC began using its old name immediately after
the war). The NIC not only took over a total of twenty-eight of its former
Northwestern Industrial Corporation), unpublished master’s thesis, Shanxi University,

2012, 20.
45
Jing, 阎锡山与西北实业公司, pp. 55–6.
46
西北实业公司章程 (Bylaws of the Northwestern Industrial Corporation), in Jing, 阎
锡山与西北实业公司, pp. 59–62.
47
Jing, 阎锡山与西北实业公司, p. 80.
48
49
197

morris l. bian
enterprises; it also confiscated more than a dozen private enterprises and

made them part of the NIC.50 At about the same time, the number of NIC
employees increased from 14,060 in 1946 to 23,421 toward the end of 1947.51
Due in part to the expansion of its scale and scope, the NIC revised its
governance structure in early 1946 to consist of company headquarters, a
secretariat, and six division-level units: divisions of industry, mining, power,
operations, general affairs, and accounting, with each division subdivided
into several departments.52 The case of Shanxi reveals that a pattern of
development relying on regional state enterprises had taken shape before
the Communist takeover of China.
Regional State Enterprises in Guizhou Province

Unlike the early establishment of the NIC in Shanxi province, the Guizhou
Enterprise Corporation (GZEC) in Guizhou province came into existence in
1939. The GZEC’s establishment was a direct response to the sustained
systemic crisis triggered by the outbreak of the Sino-Japanese War in the
context of increasing penetration of central authority into Guizhou province.
Politically, Guizhou remained under warlord control until 1935, when the
Nationalist government reorganized the Guizhou provincial army and
appointed Wu Zhongxin provincial governor. Economically, modern indus-
try was virtually nonexistent in Guizhou province as late as the late 1930s.
After the Sino-Japanese War erupted, the Nationalist government relocated
to Chongqing in Sichuan province. Many factories, organizations, and uni-
versities also relocated to southwestern provinces, including Guizhou prov-
ince. According to a Ministry of Economic Affairs official in charge of factory
relocation, 452 private firms had relocated to southwestern provinces by 1942.
As a result, southwestern provinces acquired a strategic position they had
never occupied before the war.
In December 1937, the Nationalist government appointed Wu Dingchang
(1884–1950) governor of Guizhou province. As a young man, Wu studied in
Japan’s Tokyo Business College before the 1911 Revolution. During the two
decades from 1911 to 1931, he was actively involved in politics and business,
occupying various positions such as deputy head of the Ministry of Finance
(1918) and general manager of the Salt Commercial Bank (1925). He was
a founding member of the National Defense Planning Commission. He
50
Jing, 阎锡山与西北实业公司, p. 271–6.
51
52
198

became head of the Ministry of Industries in 1935. He served in that position

until his appointment as governor of Guizhou province two years later.53
Wu Dingchang arrived in Guiyang toward the end of December 1937.
Within a year, Wu had initiated efforts to create a Guizhou enterprise
corporation. According to a resolution adopted by the Commission of
Guizhou Provincial Government in March 1939,
Guizhou province has become an important base area and a center of
transportation in the interior. We must exploit Guizhou resources in order
to effectively contribute to the mission of war of resistance and nation
building . . . However, we cannot fulfill such a heavy responsibility without
creating new organizations, centralizing human and financial resources, and
making long-term comprehensive plans.54
For the next two months, a preparatory committee drafted corporate bylaws,
identified sources of investment, and determined the scope of business
operations. The GZEC became operational on June 1, 1939.
Throughout the 1930s and 1940s, the GZEC relied heavily on the central
government and state-run financial institutions for investment capital due to
Guizhou’s economic underdevelopment. On average, central government
subsidies accounted for 58 percent of total Guizhou provincial revenue
each year from the 1936 to the 1940 fiscal year.55 Statistics compiled by
Guizhou provincial government revealed a similar situation: on average,
central government subsidies accounted for 58 percent of total Guizhou
provincial revenue each year from the 1938 to the 1941 fiscal year.56 To realize
its objective of exploiting Guizhou resources and developing Guizhou indus-
tries, the GZEC had to rely on the central government and state-run financial
institutions for investment capital.
At the time of its founding, the GZEC was capitalized at ¥6 million, and
much of that came from provincial and central governments and state-run
banks. The Guizhou provincial government invested ¥1.23 million (20.5 per-
cent). The Ministry of Economic Affairs provided ¥1.25 million (20.83 per-
cent). The Bank of China, the Bank of Communications, and the Farmers’
Bank of China between them contributed ¥3.50 million (58.33 percent).
53
Zhuang Kuming 莊焜明, “吴鼎昌与抗战时期贵州经济建设 (一九三八至一九四
四)” (Wu Dingchang and Guizhou Economic Reconstruction during the Sino-Japanese
War, 1938–1944), 近代中国 (Modern China) 133 (1999), 117–39.
54
贵州省档案馆档案, M1/374, March 1939.
55
Ding Daojian 丁道谦, 贵州地方财政概况 (A Survey of Guizhou Financial
Conditions) (Guiyang, Guizhou minyi yuekanshe, 1949), p. 23.
56
Wu Dingchang 吴鼎昌, 黔政五年 (Administrative Experience in Guizhou over the
Past Five Years) (Guiyang, Guizhou sheng zhengfu bianyin, 1943), p. 4.
199

morris l. bian
Private individuals subscribed ¥20,000 (0.34 percent). Despite the relatively

large initial capitalization, the expansion of business activities, inflation
pressure, and the lack of working capital prompted the GZEC to increase
its capital to ¥10 million in June 1940, to ¥20 million in May 1942, to
¥30 million in February 1943, and to ¥50 million in August 1947.57 The lion’s
share of investment capital continued to come from provincial and central
governments and state-run banks after each capital infusion. At the same
time, given its stated objective of exploiting Guizhou resources and develop-
ing Guizhou industries, the GZEC made the development of industrial and
mining enterprises its top investment priority.
As someone who had discussed economic issues and engaged in business
activities for more than twenty years, Wu Dingchang was determined to
avoid the pitfalls of state-owned enterprises by requiring the GZEC to create
a corporate structure. In a May 1939 speech, Wu stated that the provincial
government could adopt one of two forms for organizing planned economic
activities: joint government–merchant management or joint management by
the central government and the provincial government. After considerable
deliberations, the provincial government adopted the form of joint govern-
ment–merchant management because Wu and others believed that business
growth required continuity that the formal administrative bureaucracy could
not provide.58 As Wu explained, it would take a long time before the planned
enterprises could generate returns. Given frequent personnel transfers within
government administration, however, it was difficult to ensure personnel
continuity and policy consistency. What they needed was a business organ-
ization that was impervious to frequent changes of government personnel to
assure continuity of business operations. The GZEC was meant to be such an
organization.
Following Wu Dingchang’s plan, the GZEC was entrusted with the
responsibility of exploiting Guizhou’s resources and developing Guizhou’s
industry, with state-run banks and Guizhou provincial government providing
most of the investment capital. On the other hand, following the provisions
of the 1929 Company Law, the GZEC created a corporate structure that
consisted of a shareholder meeting, a board of directors, a standing board of
directors, a team of managers, and divisions and departments under the
general manager. One could argue that the GZEC developed multiple
57
贵州省档案馆档案, MG41/36, November 1949.
58
Wu Dingchang 吴鼎昌,花溪闲笔 (Reminiscences of My Time in Charge of Guizhou
Provincial Administration, 1938–1940) (Guiyang, Guizhou ribaoshe, 1940), pp. 2, 24–5.
200

identities: it was a regional state enterprise in terms of ownership and source

of investment. It possessed the core characteristics of a modern Western
shareholding company because of its corporate structure. And it was
a holding company or investment management company seen from the
perspective of capital organization and mode of operation.
Following such a development strategy, the GZEC divided enterprises
associated with it into “wholly owned,” “partially owned,” and “sponsored”
enterprises. Wholly owned enterprises were enterprises in which the corpor-
ation held all equity as well as appointing key management personnel.
Partially owned enterprises were enterprises in which the corporation held
more than 50 percent of all equity. By contrast, sponsored enterprises referred
to enterprises in which the corporation held less than 50 percent of all equity.
As a result, the corporation did not exercise control over sponsored enter-
prises; instead, it enjoyed rights and privileges similar to those of other
shareholders of the sponsored enterprises.59
The GZEC had thirteen units at the time of its establishment in 1939. The
number of units increased to twenty-two in 1940, twenty-four in 1941, and
twenty-eight in 1942. The number of units began to decrease after 1942, to
twenty-six in 1943, twenty in 1944, eighteen in 1945, and sixteen in 1946. After
1946, however, the downward trend was reversed: the number of units
increased to seventeen in 1948 and twenty in 1949.60
The GZEC sustained a loss of ¥25,252 in 1939 (¥8,417 in 1936 constant
price).61 However, the corporation earned a profit every year for the next
six years, reaping the largest amount of annual profit in 1941: ¥692,393 in 1940
(¥86,549 in 1936 constant price); ¥2,525,854 in 1941 (¥120,278 in 1936 constant
price); ¥3,239,772 in 1942 (¥50,621 in 1936 constant price); ¥4,639,973 in 1943
(¥18,634 in 1936 constant price); ¥7,844,044 in 1944 (¥10,147 in 1936 constant
price); ¥9,439,609 in 1945 (¥3,680 in 1936 constant price).62 As late as 1947, in
part because of the effect of inflation, the corporation made a profit of
¥2,310,542,515 (¥23,376 in 1936 constant price).63
59
He Jiwu 何辑五, 十年来贵州经济建设 (Guizhou Economic Reconstruction for the
Last Decade) (Nanjing, Nanjing yinshuguan, 1947), pp. 60–1.
60
贵州省档案馆档案, MG41/36, November 1949; MG 41/16, 1942; 贵州企业季刊
(Guizhou Enterprise Quarterly) 1.2 (February 1943), 90–106; 重庆市档案馆档案,
0101/4112, June 1943; 0101/1308, 1944; 0101/1452, June 1948.
61
贵州省档案馆档案, MG 41/16, 1942.
62
贵州省档案馆档案, MG 41/16, 1942; MG 41/23, 1942; 重庆市档案馆档案, 0101/4112,
1943.
63
Guizhousheng danganguan 贵州省档案馆 (ed.), 贵州企业股份有限公司 (Archival
Materials of the Guizhou Enterprise Limited Liability Corporation) (Guiyang, Guizhou
renmin chubanshe, 2003), pp. 503–29.
201

morris l. bian
After the Communist takeover of Guiyang, the Chinese Communist Party

(CCP) dismantled the GZEC’s corporate structure and replaced it with
a bureaucratic structure. It also removed retained staff from positions of
authority and imposed control over them. Still, Guizhou’s regional state
enterprises continued to expand as regional state enterprises under CCP
authority and control. While the CCP dismantled the enterprises’ corporate
structure and created a bureaucratic structure, it was less successful in
destroying the connective tissues that sustained, bound, and gave shape
and support to Guizhou’s regional state enterprises. Contrary to conven-
tional wisdom, the sovietization of Chinese enterprise management was not
a sudden process; rather, the national campaign to transplant the Soviet
model during the 1950s represented the culmination of efforts that had
begun two decades earlier. And, despite its efforts to implement Soviet-
style economic accounting by discrediting Western-inspired practices, the
CCP continued to rely on an indigenized Western accounting system as late
as the mid-1950s. Finally, the CCP inherited and expanded the institutions of
social services and welfare that had developed during the Sino-Japanese War.
In short, the changes in Guizhou’s regional economic institutions after 1949
were simultaneously revolutionary and evolutionary.64 In a paper presented
at the Joint Conference of the Association for Asian Studies and the
International Convention of Asian Scholars held in Honolulu, Hawaii in
2011, I described the changes in Guizhou’s regional economic institutions
after 1949 as “transforlutionary” in light of the fact that these changes were
characterized by both transformation and evolution.65
The Significance of Public Enterprise in Overall

Industrial Development
How significant was the expansion of central and regional state enterprises in
China’s overall industrial development? To address this question, I turn now
to a comparison between public and private enterprise in heavy industries.66
64
M.L. Bian, “Redefining the Chinese Revolution: The Transformation and Evolution of
Guizhou’s Regional State Enterprises, 1937–1957,” Modern China 41.3 (May 2015), 313–50.
65
As I explained elsewhere, no word in the English language is capable of capturing the
sense of transformative and evolutionary change that occurs simultaneously. We can
denote this type of change by coining a new word – transforlution – that combines the
meaning of transformation and evolution. This type of change is transforlutionary by
definition. See M.L. Bian, “Interpreting Enterprise, State, and Society: A Critical Review
of the Literature in Modern Chinese Business History, 1978–2008,” Frontiers of History in
China 6.3 (September 2011), 423–62.
66
Public enterprise included both central state enterprises and regional state enterprises.
202

I will first discuss the structure of Chinese industry and the position of heavy
industry within that structure before the Sino-Japanese War. I will then
compare public enterprise with private enterprise in heavy industry during
the war.
In the spring of 1933, the National Defense Planning Commission launched
a survey of Chinese industry as part of its mission of resource investigation.
Between April 1933 and October 1934, a group of researchers from the China
Economic Statistics Institute inspected firms in fourteen provinces and more
than 120 municipalities and counties. The statistics they gathered were then
compiled and published by the National Resources Commission in early 1937
as a Report on the Conditions of Chinese Industry. The report did not cover
Gansu, Xinjiang, Yunnan, Guizhou, Ningxia, and Qinghai provinces; nor did
it include Manchuria, which was under Japanese occupation. Moreover, the
report excluded government arsenals and firms that did not use mechanized
power and employed fewer than thirty workers.67 Still, the Report on the
Conditions of Chinese Industry was the most comprehensive industrial census
for prewar China.
The census takers found that China had 2,435 factories under Chinese
ownership, which had a capital of ¥406,872,634 and employed 435,257
workers.68 At the time of the census, the Chinese economy was predomin-
antly private, and the role of the state sector, though increasing over time,
remained small. Within what was essentially a market-oriented private
economy, light industry dominated. Food/beverage/tobacco and textile
products made up more than 70 percent of the total output of China’s
manufacturing industry in 1933, whereas basic metals (iron and steel) and
other metal products such as machinery accounted for less than 9 percent.69
Contemporary statistics from the Ministry of Industries also show the
dominant position that light industry occupied in the overall structure of
Chinese industry. Available sources reveal that, between 1932 and 1937, the
number of registered factories reached 3,885, with a capital of ¥377,848,000
and employing 457,143 workers. As far as the proportions between heavy and
light industry are concerned, light industry (such as the food processing,
textiles, and chemical industries) accounted for 3,305 factories or 85.1 percent
of the total, ¥308,643,000 in capital or 81.68 percent, and 432,049 workers or
67
Liu Dajun 刘大钧 (ed.), 中国工业调查报告 (Report on the Conditions of Chinese
Industry) (Shanghai, Jingji tongji yanjiu suo, 1937), vol. 1, pp. 1–5.
68
Liu, 中国工业调查报告, vol. 2, pp. 33–64.
69
T.G. Rawski, Economic Growth in Prewar China (Berkeley, University of California Press,
1989), pp. 3, 85, 360–1.
203

morris l. bian
94.5 percent. In sharp contrast, heavy industry (such as the hydroelectric

generating, metallurgical, machine, electrical, and ordnance industries)
accounted for only 580 factories or 14.83 percent of the total, ¥69,205,000 in
capital or 18.32 percent, and 25,094 workers or 5.49 percent.70
Within the industrial structure dominated by light industry, the proportion
of public enterprise was even smaller than that of private enterprise. The
Report on the Conditions of Chinese Industry notes that out of the 2,435 factories
surveyed, only sixty-six were public enterprises, and most of them were in the
hands of provincial authorities.71 Data compiled in 1935 by the Ministry of
Industries reveal an even clearer picture of the share of public enterprise. The
data show a total of seventy-two public enterprises in China, with ¥30,297,726
in capital, 40,669 workers, 38,779 horsepower, and ¥74,828,733 in value of
output. Among them, fifty-one were in either heavy industry or the trans-
portation industry. These fifty-one enterprises had ¥18,075,979 in capital,
26,966 workers, 33,169 horsepower, and ¥61,870,831 in value of output.
What is more, of all the public enterprises, only seventeen were in heavy
industry proper (the metallurgical, metals, machine, and energy industries),
with ¥12,320,992 in capital, 8,258 workers, 23,376 horsepower, and ¥53,365,723
in value of output.72 The data show that during the mid-1930s, public enter-
prises in heavy industry made up only 2.9 percent of all factories, 17.8 percent
of all capital, and 32.9 percent of all workers.73
The decade between 1935 and 1945 witnessed a fundamental change in
the existing position of public and private enterprise. According to
a statistical survey conducted by the Bureau of Statistics of the Ministry
of Economic Affairs, the relative position of public and private enterprise
had changed radically by 1942. The result is more revealing when we
compare the average capital, number of workers, and power equipment in
public and private enterprise in heavy industry. By 1942 public enterprise
had clearly overtaken private enterprise in the amount of capital, number
70
Weng Wenhao 翁文灏, “中国工商经济的回顾与前瞻” (The Retrospect and
Prospect of Chinese Industrial and Commercial Economy), 资源委员会公报 5.2
(August 16, 1943), pp. 59–68; Huang Bingwei 黄秉维, “五十年来之中国工矿业”
(Chinese Industry and Mining in the Last Fifty Years), in Zhongguo tongshang yinhang
中国通商银行 (ed.), 五十年来之中国经济 (The Chinese Economy in the Last Fifty
Years, 1896–1947) (Shanghai, Zhongguo tongshang yinhang bian, 1947), reprint (Taipei,
Wenhai chubanshe, n.d.), p. 173.
71
Liu, 中国工业调查报告, vol. 2, pp. 33–64.
72
Guomin zhengfu zhujichu tongjiju 国民政府主计处统计局 (ed.), 中华民国统计
提要 (Statistical Abstract of Republican China) (Chongqing, Guomin zhengfu
zhujichu tongjiju, 1940), Table 50, p. 85.
73
Weng, 中国工商经济的回顾与前瞻, pp. 59–68; Huang, 五十年来之中国工矿业,
p. 173; 中华民国统计提要, p. 85.
204

of workers, and power equipment. On average the amount of capital of

a public enterprise was eighteen times that of a private enterprise, the
number of workers 3.7 times that of a private enterprise, and the power
equipment 2.2 times that of a private enterprise.74 It is estimated that by
1942, the National Resources Commission already had under its control
¥8 billion, which constituted roughly 40 percent of the modern industrial
capital in China’s interior.75
Statistics on the increase in output of major products in heavy industry
provide further evidence of the dominant position that public enterprise
achieved during the Sino-Japanese War. The output of public enterprise in
some industries was very small one year after the war began, yet by the time
the war ended their share of total output had increased dramatically. For
instance, in 1938 the share of output of the state-owned iron and steel
industries was only 5.8 percent and 20 percent respectively. By the end of
1945, however, their share of output had increased to 64.8 percent and
96.4 percent respectively. Moreover, the manufacturing of many heavy
industrial products was from the very beginning a monopoly of state-
owned industries. Finally, the average figures of public enterprise’s share of
product show a steady increase during the eight-year period: 55.6 percent in
1938, 61.9 percent in 1939, 62.7 percent in 1940, 65.5 percent in 1941, 70.5 percent
in 1942, 75.5 percent in 1943, 77.9 percent in 1944, and 79.9 percent in 1945.76
To summarize, all the evidence points to a significant expansion of state-
owned heavy industry during the war. One could even argue that public
enterprise’s share of product in heavy industry, as well as light industry,
expanded significantly during the war. After analyzing the output statistics of
seventeen major heavy and light industrial products, Xu Dixin and Wu
Chengming concluded that public enterprise’s share of product was 21.2 per-
cent in 1938. By 1944, however, public enterprise’s share of product had
reached 53.7 percent.77 In other words, public enterprise had clearly achieved
74
Guomin zhengfu jingjibu tongjiju 国民政府经济部统计局 (ed.), 后方工业概况统计
(A Statistical Survey of Chinese Industry) (Chongqing, Guomin zhengfu jingjibu
tongjiju, 1943), p. 11.
75
Wu Taichang 吴太昌 et al., 中国国家资本的历史分析 (A Historical Analysis of
China’s State Capital) (Beijing, Zhongguo shehui kexue chubanshe, 2012), p. 232.
76
Wu Taichang 吴太昌, “抗战时期国民党国家资本在工矿业的垄断地位及其与民
营资本比较” (The Monopolistic Position of the Nationalist State Capital in Industry
and Mining and Its Comparison with Private Capital during the Sino-Japanese War),
中国经济史研究 (Studies in Chinese Economic History) 3 (September 1987), 133–50.
77
The seventeen products used in their estimates are coal, pig iron, steel, nonferrous
metals, petroleum, electricity, alcohol, gas substitutes, acid, alkali, cement, machinery
and electrical products, cotton yarn, cotton cloth, flour, matches, and paper. See
Xu Dixin 许涤新 and Wu Chengming 吴承明 (eds.), 新民主主义革命时期的中国
205

morris l. bian
a dominant position by the end of the Sino-Japanese War.78 After 1949, public
enterprises – central state enterprises and regional state enterprises – would
serve as the foundation for the CCP effort to bring about China’s industrial-
ization and modernization for the rest of the twentieth century.79
Further Reading
Bian Linan 卞历南, 制度变迁的逻辑：现代中国国营企业制度之形成 (The
Dynamics of Institutional Change: The Making of the State Enterprise System in
Modern China) (Hangzhou, Zhejiang daxue chubanshe, 2011).
Bian Linan 卞历南, 西方学界最近 40 年对中国企业史研究的述评 (A Critical
Examination of Western Studies of Chinese Business and Economic History, 1978–
2018) 经济社会史评论 (Social & Economic History Review) 16.4 (November 2018),
104–24.
Bian, M.L., “Explaining the Dynamics of Change: Transformation and Evolution of
China’s Public Economy through War, Revolution, and Peace, 1928–2008,” in
B. Naughton and K.S. Tsai (eds.), State Capitalism, Institutional Adaptation and the
Chinese Miracle (Cambridge, Cambridge University Press, 2015), pp. 201–22.
Bian, M.L., The Making of the State Enterprise System in Modern China: The Dynamics of
Institutional Change (Cambridge, MA, Harvard University Press, 2005).
Bian, M.L., “Redefining the Chinese Revolution: The Transformation and Evolution of
Guizhou’s Regional State Enterprises, 1937–1957,” Modern China 41 (May 2015), 313–50.
Bian, M.L., “The Sino-Japanese War and the Formation of the State Enterprise System in
China: A Case Study of the Dadukou Iron and Steel Works, 1938–1945,” Enterprise &
Society 3.1 (March 2002), 80–123.
Kirby, W.C., “Engineering China: Birth of the Developmental State, 1928–1937,” in
W.-H. Yeh (ed.), Becoming Chinese: Passages to Modernity and Beyond (Berkeley,
University of California Press, 2000), pp. 137–60.
Wu Taichang 吴太昌, 抗战时期国民党国家资本在工矿业的垄断地位及其与民营
资本比较 (The Monopolistic Position of Nationalist State Capital in Industry and
Mining and Its Comparison with Private Capital during the Sino-Japanese War), 中国
经济史研究 (Studies in Chinese Economic History) 3 (September 1987), 133–50.
资本主义 (Chinese Capitalism during the Period of New Democratic Revolution, 1921–
1949) (Beijing, Renmin chubanshe, 1993), pp. 521, 541–5.
78
M.L. Bian, The Making of the State Enterprise System in Modern China: The Dynamics of
Institutional Change (Cambridge, MA, Harvard University Press, 2005), p. 75;
Zhang Zhongmin 张忠民 and Zhu Ting 朱婷, 南京国民政府时期的国有企业
(A Study of State-Owned Enterprises in Republican China, 1927–1949) (Shanghai,
Shanghai caijing daxue chubanshe, 2007), p. 167.
79
M.L. Bian, “Explaining the Dynamics of Change: Transformation and Evolution of
China’s Public Economy through War, Revolution, and Peace, 1928–2008,” in
B. Naughton and K.S. Tsai (eds.), State Capitalism, Institutional Adaptation and the
Chinese Miracle (Cambridge, Cambridge University Press, 2015), pp. 201–22; Bian,
“Transforlution and Hybridization: Reinterpreting Changes in China’s Public
Economy, 1918–2018,” unpublished book manuscript.
206

Wu Taichang et al. 吴太昌等人合著, 中国国家资本的历史分析 (A Historical Analysis

of China’s State Capital) (Beijing, Zhongguo shehui kexue chubanshe, 2012).
Xu, Dixin 许涤新 and Wu Chengming 吴承明 (eds.), 新民主主义革命时期的中国资
本主义 (Chinese Capitalism during the Period of New Democratic Revolution, 1921–
1949) (Beijing, Renmin chubanshe, 1993).
Xue Yi 薛毅, 国民政府资源委员会研究 (A Study of the National Resources
Commission) (Beijing, Shehui kexue wenxian chubanshe, 2005).
Zhang Shouguang 张守广, 大变局: 抗战时期的后方企业 (Great Change: A Study of
Business Enterprises in China’s Interior during the Sino-Japanese War) (Nanjing,
Jiangsu renmin chubanshe, 2008).
Zhang Shouguang 张守广, 抗战大后方工业研究 (A Study of Industry in China’s
Interior during the Sino-Japanese War) (Chongqing, Chongqing chubanshe, 2012).
Zhang Zhongmin 张忠民 and Zhu Ting, 南京国民政府时期的国有企业 (A Study of
State-Owned Enterprises in Republican China, 1927–1949) (Shanghai, Shanghai caijing
Zhao Xingsheng 赵兴盛, 传统经验与现代理想: 南京国民政府时期的国营工业研究
(A Study of State-Owned Industries in Republican China) (Jinan, Qilu shushe, 2004).
207

7
Money and the Macro-economy
Introduction
An understanding of how the money market developed is vital because
money serves as the blood of an economy. From 1800 to 1937, the Chinese
money market transitioned from a highly fragmented bimetallic system to
a gradually integrated silver yuan system in tandem with a silver-backed
fiduciary paper-money system until a fiat money system was established. As
a consequence, the economy became increasingly monetized as the growth
rate of the money supply gradually surpassed the overall economic growth
rate without evident inflation pressure on general price trends. This develop-
ment resulted both from the efforts of governments and private institutions
in response to various types of shock separately and from the outcomes of
competition and co-operation between the two stakeholders over time.
During the mid-Ming dynasty (1364–1644), a “parallel bimetallism” monet-
ary standard consisting of silver and copper was established in China. Unlike
the bimetallic system consisting of gold and silver in the history of many
European countries – in which the two precious metals were usually minted
with rates regulated by states or by the market – silver and copper in China
complemented (rather than substituted) each other. In China, (unminted)
silver bullion served as the major medium of exchange in wholesale com-
merce, long-distance trade, and tax payment, while copper cash (zhiqian
制钱)1 served as a currency in petty exchange. High transaction costs led to
this functional exclusiveness of silver and copper in the market: in retail, the
cost of assessing silver bullion probably easily surpassed the value of petty
exchange; likewise, low-valued copper cash was too expensive and weighty
We are grateful to the editor, Debin Ma, for his insightful comments and suggestions for
improving this chapter.
1
Copper cash were moulded as a circle with a square hole in the center, through which
a string could go to tie them together.
208

to be used for wholesale and/or long-distance trade. Hence, to serve different

market levels and purposes, people had to exchange copper cash for silver
bullion or vice versa at an exchange rate that fluctuated according to relative
changes in the supply and demand of the two commodity monies.
The Qing government (1644–1912) managed the supply of copper cash but
left the monetization of silver entirely in the hands of private institutions.
China did not have many silver lodes, and foreign silver had been flowing
into China since the sixteenth century through international trade in the
private sector. Soon the business sector welcomed silver bullion as a high-
denomination means of payment, and the government further accommo-
dated and promoted the silverization of the economy by accepting silver for
taxes and distributing it for expenditures. Hence silver weight (tael or yinliang
银两)2 evolved into one of the most important units of account until 1933,
when the Republican government abolished it. However, there were no
unified silver bullion standards or tael units because they varied across
different locations.3
Overall, the Chinese bimetallic monetary system was split into manifold
regional currencies, which hindered interregional trade and hence slowed
economic development. The journey towards an integrated money market
began with the import of Spanish pesos (along with silver bullion), namely
Spanish silver coins minted in Mexico. By the early 1800s, Spanish pesos had
become so widely circulated that they became a new de facto monetary
standard, expressed in Chinese as yuan,4 and gradually complemented or
replaced silver bullion as a means of payment in commerce. Inspired by the
success of foreign silver coins, the Qing government started to mint silver
coins in its late years to compete with foreign coins. The official silver coins
minted by the Beiyang government finally gained a foothold in the money
market during the warlord era (1912–1927) with help from the private financial
sector, which tirelessly promoted the circulation of official silver coins.5
2
In Chinese, it is called the liang 两, which was about thirty-eight grams of silver in the
Qing dynasty.
3
H.B. Morse, The Trade and Administration of the Chinese Empire (London, Longmans,
Green, and Co., 1908), pp. 145–7. As observed by Morse, in many cases, the weight “tael”
even had several standards in one locality.
4
R. von Glahn, “Foreign Silver Coins in the Market Culture of Nineteenth-Century
China,” International Journal of Asian Studies 4.1 (2007), 51–78.
5
“Beiyang government” refers to the Republican government with a capital in Beijing
from 1912 to 1927. The Warlord Era in China featured a nominal central government in
Beijing. There were approximately 1,300 warlords, who controlled provinces, counties,
or only several districts, depending on their power. Scholars have not reached
a consensus on the exact dates of the Warlord Era.
209

Despite the turmoil in our study period, including numerous riots, civil
wars, foreign invasions, regime shifts, natural disasters, and other chaos, the
Chinese pace towards money market integration due to the spread of official
silver coin was unstoppable. A key contributor to (and benefactor of) this
process was the banking sector. The banking sector adopted the silver yuan
(or dollar, yinyuan 银元) as the most important unit of account for business.
Both foreign and domestic banks competed to issue banknotes convertible
with silver yuan under a free banking system; hence the money supply –
mainly the banknote supply – increased. Moreover, the domestic banks
aggressively branched across the nation, which made interregional money
transfer, exchange, and remittance – mainly denominated in silver yuan –
more affordable for businesses. Therefore a nationwide and integrated
money market came into being. In 1935, China went off silver and adopted
a fiat money standard.
As money market integration continued, there was an obvious power shift
in money supply from the private financial sector to the government.
However, this progress was far from linear, with both progress and twists
resulting from co-operation and competition between the government and
the private sector in the face of constraints and shocks. Examples of the
interplay between the two vividly show how the government gradually took
control of the monetary sector at the expense of private institutions, which
eventually lost ground.
An increasingly unified and expanding money market benefited economic
growth in prewar China from at least four perspectives. First, it largely
reduced transaction costs, which facilitated the commercialization of the
Chinese economy, and trade led to specialization; second, it mobilized the
very scarce savings of an underdeveloped country to support the develop-
ment of a modern industrial sector, which was small but sowed the seeds for
Chinese industrialization until the Second Sino-Japanese War in 1937; third,
the expansion and integration of the money market gradually lowered the
interest rate, which facilitated more investment and capital formation for
economic growth; finally, the increasing supply of paper money alleviated
monetary shocks caused by the changing market conditions of the monetary
metal silver, which was a commodity to the rest of the world, and stabilized
economic development.
The Chinese experience in the money market from 1800 to 1937 is of
particular interest because it provides a typical case of how a traditional
fragmented bimetallic money market transitioned to a modern centralized
fiat money system. This process includes how the money market was
210

affected by China’s participation in the world trade network, how modern

monetary and financial technologies spread in a large but underdeveloped
economy, how local private institutions took opportunities to improve
business practice and develop a modern banking industry from scratch, and
how various governments managed the monetary system in the face of
a series of internal and external challenges. This chapter will outline not
only the development process but also how a better monetary system works
well for economic development.
Three Media of Exchange

This section presents the main nature and characteristics of the major
Chinese currencies of copper, silver, and paper money and summarizes the
evolution and changing importance of the three media of exchange from 1800
to 1937.
Copper
The Qing court actively controlled the supply of copper cash. Its system for
minting copper cash consisted of two central minting bureaus – the Bao Quan
Mint (宝泉局) and the Bao Yuan Mint (宝元局) affiliated with the Ministry
of Revenue (hubu 户部) and the Board of Work (gongbu 工部) respectively –
and approximately eighteen local mints in the early nineteenth century.6 The
smallest cash was called one wen (文) and weighed approximately five
grams.7 Usually, the larger the denomination was, the heavier the cash.
A string could go through the square hole in the cash and so they could be
tied together. One thousand wen could be tied together in what was called
one guan (贯)8 and exchanged for approximately one silver tael around
the year 1800.9
Copper cash might have varied in pattern but they were always inscribed
with the denomination, the imperial reign name at the time of production,
and the mint of origin. Coins minted by a local government were confined to
6
The number of local mints could vary over time. Usually there was one mint in each
province.
7
The weight of one wen varied slightly over time too. For details, please see Peng Xinwei
彭信威, 中国货币史 (The Monetary History of China) (Shanghai, Shanghai People’s
Publishing House, 1987; first published 1954), pp. 557–74.
8
It could also be called one diao 吊 or chuan 串.
9
See Table 4 in Hongzhong Yan, “Economic Growth and Fluctuation in the Early Qing
Dynasty: From the Perspective of Monetary Circulation,” Frontiers of History in China 4.2
(2009), 221–64. The conversion rate fluctuated over time.
211

circulation in the province of their origin and were rarely taken outside
provincial borders. The cash produced by the two central minting bureaus
achieved relatively wide circulation since the government distributed them,
together with silver ingots, to cover fiscal expenditures on troops, public
projects, construction, and other items in minute detail. The total output of
cash in 1865 was approximately 2.46 million guan, 55 percent of which were
produced by the two central mints.10 Counterfeiting was illegal and subject to
severe punishment (e.g. beheading). Nonetheless, when severe debasement
occurred, the counterfeited copper cash flooded in, which drove the seignior-
age revenue for the government down to a negligible level. It was private
competition that effectively constrained the extent of the government’s
power to debase copper cash for a prolonged period.11
In 1900, Guangdong province first imported machines to make copper
coins (tongyuan 铜元): unlike copper cash, copper coins had no square hole in
the center. Machine-made copper coins looked more delicate than cash, were
difficult to counterfeit, and gained popularity in the market. Many other
provinces joined Guangdong in producing machine-made coins.
Approximately 97 to 98 percent of the machine-made coins in circulation
were inscribed with a domination value – “1 cash is equivalent to 10 wen.”
Although it was possible for machine-made coins to be more standardized,
there were numerous types of coin circulating in the market, varying across
regional and even local markets.12 Overall, the market in copper currency
was fragmented despite being under government control and regardless of
what production technology was adopted.
The importance of copper currency (cash and coins together) in the total
money supply declined over time: during the 1800s–1850s, the share of copper
currency in the total money supply was approximately 25 percent,13 declining
to merely 3–4 percent in the 1930s.14 On the one hand, this was due to the
increase in the supply of paper money over time (to be described later), which
significantly enlarged the monetary base; on the other hand, since 1914 the
Beiyang government started to mint silver coins, and, in addition to the one-
yuan coin,15 it minted silver coins with denominations of fifty cents, twenty
10
E. Kann, The Currencies of China (Beijing, Commerce Publishing House, 1926), p. 411.
11
S. Qian and L. Wu, “Who Defended Monetary Stability in a Specie Regime? Evidence
from the Chinese History,” Frontiers of Economics in China 13.3 (2018), 397–435.
12
Peng, Monetary History, pp. 569–73. 13 Yan, “Economic Growth,” 221–64.
14
1989), Table c16, p. 394.
15
The one-yuan official silver coin was nicknamed “Yuan’s Big Head” because the image
of the head of President Yuan Shikai was inscribed on it.
212

cents and ten cents. As the official silver coins gained popularity across the
nation (see below), silver dimes and cents even eroded the realm of copper
currency as small change.
Silver
Although the Qing court left the supply of silver to the market, it promoted
the use of silver as a medium of exchange through taxation and fiscal
expenditure. Except for some tribute grain being paid in kind, all other
taxes, including land tax, salt tax, tariffs, and customs duties, were paid in
silver. For government expenditures, silver, together with rice and copper
cash, was used to pay the salaries of government officials, soldiers, workers
hired for public projects, stipends for royal family members, government and
military expenditures, royal expenses, and so on. Despite the ratio of the
components of silver, rice, and copper cash varying for different purposes and
over time, silver was no doubt the most important means of payment.16 On
the eve of the first Opium War (1840–1842), 74 percent of total fiscal revenues
worth 65.8 million silver taels was paid in silver, and 84 percent of the total
fiscal expenditure of 46 million taels was paid out in silver.17 Fiscal silveriza-
tion helped make silver, despite its not being under government control, the
most influential means of payment and unit of account. Fiscal silverization
accompanied the silverization of economic activities overall. Up until the
1840s–1850s, 75 percent of the total money supply of 436 million silver taels
was silver, and the rest was copper cash.18
However, instead of forming a unified monetary standard, the silver
system became divided into manifold regional currencies: the types and
qualities of silver varied across the nation. Hence the value of a “tael” in
one place differed from that in other places, and numerous exchange rates
among “taels” in different places were needed for interregional trade. Native
financial institutions emerged, first mainly to assess the quality of silver
bullion by putting a stamp on it and to exchange “taels” across localities.
Later, they extended their business scope to taking deposits, making loans,
remitting and transferring money across the nation, and extending credit by
issuing paper monetary instruments mainly backed by silver. The most
16
Yan, “Economic Growth,” 221–64.
17
Revenue and expenditure in rice were 80 million and 34.6 million dan (a unit of rice of
approximately fifty kilograms) respectively. The rice price was estimated at 2.16 taels per
dan. Peng, Monetary History, p. 851. See also Tang Xianglong 汤象龙, 中国近代财政经济
史论文选 (Selected Papers on Modern Chinese Financial and Economic History) (Sichuan,
Southwestern University of Finance and Economics Press, 1987), pp. 204, 221.
18
Yan, “Economic Growth,” 221–64.
213

important native financial institutions in the early nineteenth century were

called qianzhuang 钱庄, known as “native banks,” which differed from the
Western-style modern banks that emerged in China later.
In addition to silver bullion, Spanish pesos, namely Spanish silver coins
(yinyuan) minted in Mexico, flooded into China and achieved popularity in
southern coastal provinces. These silver coins formed a satisfactory medium
of exchange and enjoyed a premium generally more than 30 to 40 percent
over their intrinsic value.19 By the 1820s, Chinese merchants were occasion-
ally found to export silver bullion in exchange for foreign silver coins to make
a profit out of this premium.20 In the 1850s, the lijin 釐金 duty, an internal
transit duty on commodities, emerged and was collected in silver coins
instead of bullion, which was different from other government taxes.21
Consequently, the transit duties officially promoted the usage and spread of
coins in commercial activities at the time.
Inspired by the success of foreign silver coins, the Guangdong government
started to mint silver coins in 1889. However, the silver coins minted by the
provincial government did not have the prestige of Spanish pesos: not
because they contained less silver but because they were easier to counterfeit
than Spanish pesos. Chinese craftsmen could easily produce coins identical to
government-minted coins with less silver content, but they failed to mimic
the fine details of the pattern and the Spanish letters inscribed on the foreign
coins. The debased counterfeited coins drove the full-bodied official coins out
of circulation. In the end, the status of foreign silver coins in the Chinese
market remained unshaken until the early twentieth century.
In 1914, the Beiyang government started to mint silver coins, which
circulated together with foreign silver coins in the market. Surprisingly,
official coins gradually gained a foothold in the market, despite it being
a highly fragmented era politically – the warlord era. Official coins became
the most popular currency in the market: in 1924, approximately 960 million
silver yuan circulated in China, and less than 4 percent of them were foreign
coins.22 This success was partially due to good timing: the foreign silver coins
minted after Spanish American independence lost their universal standard
and thus became less appealing to Chinese merchants, and the import of
19
For details, see Morse, Trade and Administration, pp. 163–5.
20
Von Glahn, “Foreign Silver Coins,” 51–78.
21
See Kaske and Lin’s chapter in this volume for details.
22
See the survey conducted by the Domestic Exchange Office of the Shanghai
Commercial and Savings Bank 上海商业储蓄银行国内汇兑处, 国内商业汇兑要览
(An Overview of Domestic Commercial Exchange) (Shanghai, Shanghai Commercial
and Savings Bank, 1925).
214

foreign coins dropped.23 Moreover, after Mexico left the silver standard in
1903, it almost stopped minting silver coins. The shortfalls in foreign coins
were readily filled by high-quality and standardized official coins, which were
produced by machines at the time and were more resistant to counterfeiting.
Paper Money
Although China pioneered the large-scale circulation of paper money as early
as the Northern Song dynasty (960–1127), its modern vintage started in 1845
following the entry of foreign banks,24 which began to issue banknotes
circulating in treaty ports and some foreign spheres of influence. These
notes were backed either by silver – namely banknote holders could
exchange notes for silver ingots in taels or coins in yuan in the issuing
bank – or by foreign paper money, such as Hong Kong dollars,25 Russian
roubles, or Japanese yen. The Commercial Bank of China, the first modern
Chinese bank, established in 1896, also rushed to issue banknotes two years
later. More newly established domestic modern banks entered the arena of
paper money issuance.
It was not until the waning years of the Qing dynasty that the court
realized that it was necessary to consolidate the right to issue paper money
in the hands of the government and established a government bank – the
Bank of the Great Qing26 – in 1905, on which was bestowed the right to
represent the government to issue paper money and manage the national
treasury. Another government bank, the Bank of Communications, was set
up in 1907. However, there was no law stipulating that only government
banks had the right to issue paper money until two laws were issued in 1909
and 1910.27 These laws clearly stated that only the Bank of the Great Qing, as
the central bank, could issue paper money, and other banks or financial
institutions should stop banknote issuance immediately. However, these
laws were no longer enforced following the collapse of the Qing dynasty in
23
A. Irigoin, “The End of a Silver Era: The Consequences of the Breakdown of the
Spanish Peso Standard in China and the United States, 1780s–1850s,” Journal of World
History 20.2 (2009), 207–44.
24
The introduction of modern banking to China was led by a British bank – the Oriental
Bank, which was a British–Indian joint venture established in Bombay in 1842. It set up
a branch in Hong Kong and an agency in Guangdong in 1845 and expanded to Shanghai
in 1848.
25
Hong Kong was a British colony at the time.
26
This was the predecessor of the Bank of China.
27
One was the Provisional Regulations on Paper Money in 1909, and the other was the
Rules of the General Bank in 1910.
215

1911. Despite this, as we will see in the following section, paper money was to
take off in the next two decades under a largely free banking system.
Money, Financial Institutions, and the State

This section describes the money market transition from highly fragmented
to unified by focusing on the roles played by two institutions: financial
institutions, especially the domestic modern banking industry, and various
governments during the period of study.
The Role of Financial Institutions in the Money Market

The evolution of the Chinese monetary system can only be understood as part of
a remarkable transformation of China’s financial system during this period.28
Figure 7.1 provides us with a glimpse of this extraordinary transformation. It
shows the capital power of three categories of financial intermediaries: native
financial institutions, domestic modern banks, and foreign banks in 1894, 1925,
and 1936. The total capital power of financial intermediaries quadrupled from
merely 863 million yuan in 1894 to 3,557 million in 1925 and increased by another
2.5 times to 8,925 million in 1936.29 Among these three types, modern domestic
banks performed strikingly well. In 1894, there was no modern Chinese bank. In
1925, modern domestic banks had already taken the largest share, at 40.5 percent,
of the total capital power in the banking industry, and they grew to take the lion’s
share at 77.7 percent in 1936.
In addition to the dramatic growth in the capital power of domestic
modern banks, the banking network expanded rapidly: after the establish-
ment of the first domestic modern bank in 1896, the number of modern banks
grew to 164, with 1,627 branches in over 500 localities throughout China in
28
See Sheehan and Zhu’s chapter in this volume for a comprehensive survey of China’s
banking sector.
29
We use the nominal value rather than the real value because a consistent price index
series for the whole period from 1894 to 1936 is not available. However, we believe that
price changes did not change the magnitude of the banking expansion observed here to
a significant extent. For instance, according to the wholesale price indices constructed
by Yuru Wang 王玉茹, “城市批发物价变动与近代中国经济增长” (Urban
Wholesale Price Change and Economic Growth in Modern China), Journal of Shanxi
University, Philosophy and Social Science Edition 5 (2006), 29–36, given the base year of
1930 = 100, the corresponding indices for 1925 and 1936 are 92.88 and 94.78 respectively.
The price index for 1894 is not available. According to the same estimation, the
wholesale price index increased by 40 percent from 1912 to 1925. If we assume
a similar magnitude of price change for the 1894–1912 period, the total price change
from 1894 to 1925 is slightly more than 100 percent. Thus, even after accounting for this
magnitude of price change, the real increase in the banking capital from 1894 to 1925 was
still fast.
216

1936
1925
1894
Total capital power Total capital power Total capital power

in 1894 = 863 in 1925 = 3,557 in 1936 = 8,925
Domestic modern banks Native financial institutions Foreign banks
Figure 7.1 The capital power of China’s banking industry in 1894, 1925, and 1936 (million
silver yuan)
Note. Native financial institutions in 1894 consisted of piaohao and qianzhuang. In 1894, piaohao
took a 32.5 percent share of the total capital power, only slightly less than qianzhuang’s
35 percent share. However, piaohao totally disappeared after the Qing dynasty collapsed, since
it relied on official patronage for transferring government funds across the nation and failed to
adapt to the new political and economic environment. Therefore the category native financial
institutions in 1925 and 1936 only refers to qianzhuang, also called native banks.
Source: data are from Table 8.1 in L. Cheng, Banking in Modern China: Entrepreneurs, Professional
Managers, and the Development of Chinese Banks, 1897–1937 (Cambridge, Cambridge University
Press, 2003), p. 241. The capital power defined there consisted of capital and other funding in
1894 (p. 19) and of capital, notes, and deposits in 1925 and 1936 (pp. 34, 241)
1936.30 As early as the 1920s, three cities – Shanghai in the south, Tianjin in the
north, and Hankou in the center – emerged as financial centers connecting
regional markets. Map 7.1 shows the locations of 104 cities in nineteen
provinces surveyed by the Shanghai Commercial Savings Bank in 1924, and
lines indicate that there existed channels for one city to transfer funds directly
to another city. It was evident that Shanghai was already the national
financial center, to which the majority of the cities had a direct financial
link, while Tianjin and Hankou served as regional financial centers. A city
30
Research Division of the Bank of China 中国银行经济研究室, 全国银行年鉴 (Bank
Statistical Yearbook) (Shanghai, 1937), pp. 26–32.
217

Beijing
Hankou Shanghai

Map 7.1 Cities with a direct money-transferring link with the major financial centers of Shanghai, Hankou, and Tianjin in 1924
Source: the Domestic Exchange Office in the Shanghai Commercial Savings Bank conducted a survey of financial markets in 104 cities in 1924 and
published the results in 国内商业汇兑要览 (An Overview of Domestic Commercial Exchange) (Shanghai, Shanghai Commercial and Savings
Bank, 1925). This map was constructed by the authors according to information on whether a city could transfer funds to the three major financial
centers directly or not
without direct access to one of the major cities could always transfer funds to
the regional center and from there to another major city or its intended
destination. This resembles a hub-and-spoke system in air transportation, and
enabled money to flow freely across the whole nation.
The rapid growth of China’s financial institutions and intermediation led to
China’s monetary transformation in the rise of first the official silver coins, then
paper money. Many native banks in Shanghai united to promote the circulation
of official coins by gradually phasing out the tael–yuan exchange rate quotations
for foreign coins and published quotations for official coins only from June 11,
1919.31 This gesture established the predominant status of official coins in the
national market, since Shanghai was the national financial center.32 The domestic
modern banks also tried hard to promote the usage of official coins. For them,
one standardized yuan system was far better than the complicated and frag-
mented tael system, which was simply out of their control due to lack of human
capital in assessing silver bullion and exchanging different silver bullion across
localities or with coins. The complication of the tael system made the rapid
branching of a modern bank almost impossible and simply did not fit the
structure and procedure of a modern bank.33 Hence the promotion of official
silver coins well suited the developmental goal of domestic modern banks.
The silver yuan, as a unit of account, further spread across the nation along
with the expanding business scope of banking. In addition to taking deposits and
making loans, the business scope of banking extended to issuing commercial
papers and banknotes, transferring bills, discounting and rediscounting bills, and
so on, which were all mainly denominated in yuan. As the domestic banking
sector grew stronger and more mature, the yuan replaced the tael as the major
medium of exchange, and silver coins spread extensively in local markets, even in
remote rural areas. According to the survey mentioned above, in the early 1920s
31
The daily tael–yuan exchange rate quotations in Shanghai can be found in People’s
Bank of China, Shanghai Branch 中国人民银行上海分行, 上海钱庄史料汇编 (The
Archival Materials of Shanghai Native Banks) (Shanghai, Xinhua Bookstore Press, 1961),
pp. 610–27. The quotation of the exchange rate started from January 1865 for the
exchange rate of Shanghai silver tael for Mexican silver coins until June 11, 1919.
Afterwards, the quotations were for taels–official silver coins.
32
As shown in D. Ma and L. Zhao, “A Silver Transformation: Chinese Monetary
Integration in Times of Political Disintegration, 1898–1933,” Economic History Review,
2020, 513–39, there was a clear trend of increasing monetary integration based on the
empirical analysis of monthly and daily prices of silver yuan in Shanghai, Tianjin, and
eighteen other cities in northern and central China from 1898 to 1933.
33
It is true that some native banks failed due to business lost in assessing bullion and
exchanging taels. However, many native banks successfully shifted their business to
lending, and some even remodeled to become modern banks. This type of structural
shift in the banking industry was productive.
219

official silver coins became the most widely circulated currency in most of the 104
places surveyed, including remote areas such as Guiyang in Guizhou province
and San Yuan in Shanxi province.34
The Role of the State in the Money Market

The ascendency of official silver coins prepared the Chinese market for the rise of
a single silver yuan standard, which was delinked from the silver tael.35 As early as
1917, the business world started to discuss the feasibility of abandoning the tael
and adopting the yuan as the only silver standard.36 However, due to the lack of
a powerful central government and the deficiency of official coins at the time, this
proposal from the business world was temporarily neglected until March 10, 1933,
when the Nationalist government (hereafter the Government37) was able to
respond to appeals from the business world and issued the Decree to Abolish
the Tael and Adopt the Yuan to stipulate that the yuan was the only unit of
account for transactions and contracts. Those who possessed silver bullion could
either turn silver bullion in to the central mint to mint coins or exchange it for
coins at the Central Bank, the Bank of China (BoC), and the Bank of
Communications (BoCom) at an exchange rate of 0.715 tael to the yuan.
A unified money market with a single yuan standard had finally arrived in China.
Despite the official silver coins achieving great success during the warlord era,
as described above, the Beiyang government could not control the paper money
supply, which became an increasingly important component of the total money
supply. The two government banks (BoC and BoCom), together with domestic
and foreign modern banks and native banks, competed to issue banknotes under
the free banking system. Moreover, the two government banks became increas-
ingly independent: the BoC became more of a private bank than a government
bank as the private share increased from 50 percent in 191538 to 80 percent in 1928,39
34
Both provinces were economically underdeveloped and located in inland China.
35
D. Ma, “Chinese Money and Monetary System, 1800–2000, Overview,” in G. Caprio
(ed.), Handbook of Key Global Financial Markets, Institutions, and Infrastructure, vol. 1
(Oxford, Elsevier, 2013), pp. 57–64.
36
See articles on the monetary situation in China published by the Bankers’ Weekly before
1921 and collected in Xu Cangshui 徐沧水 (ed.), 中国今日之货币问题 (Monetary
Issues in Today’s China) (Shanghai, Bankers’ Weekly Press, 1921).
37
After 1927, the capital moved to Nanjing when the Nationalist Party took power.
38
The Revised Rules of the Bank of China, passed by the Senate on September 30, 1915.
See Bank of China 中国银行总行 and Second Historical Archives of China 中国第二
历史档案馆, 中国银行史资料汇编上编, 1912–1949 (Compilation of the Archives of
the Bank of China, 1912–1949) (Beijing, The Archive Press, 1991), p. 115.
39
The Rules of the Bank of China, enacted by the Republican government on
December 26, 1928. Bank of China and Second Historical Archives of China, 中国银
行史资料汇编上编, p. 124.
220

2500 60
50
2000 40
30
20
1500
10
0
1000
1890
1892
1894
1896
1898
1900
1902
1904
1906
1908
1910
1912
500
0
1890
1894
1896
1898
1902
1904
1906
1910
1912
1914
1918
1920
1922
1926
1928
1930
1934
1936
1892
1900
1908
1916
1924
1932
Foreign banks Domestic banks
Figure 7.2 Banknote issuance in China from 1890 to 1936 (million silver yuan)
Sources: data on banknote issuance for both foreign and domestic banks from 1912 to 1936
are from T. Rawski, Economic Growth in Prewar China (Berkeley, University of California
Press, 1989), p. 380, Table c9. We extend Rawski’s note issuance for foreign banks to 1890
based on the same source – Xian Ke 献可, 近百年来帝国主义在华银行发行纸币概况
(Issue of Paper Currency by Imperialist Banks in China during the Past Century)
(Shanghai, Shanghai People’s Publishing House, 1958). Domestic modern banks started to
issue banknotes in 1898. Issuance by domestic banks from 1898 to 1911 is based on
Commercial Bank of China 中国通商银行, 五十年来之中国经济—中国通商银行创
立五十周年纪念册 (Fifty Years of the Chinese Economy: The Commemorative Book of
the 50th Anniversary of the Establishment of the Commercial Bank of China) (Shanghai,
Commercial Bank of China, 1947), pp. 4–5. This book records the volume of banknotes
issued by domestic modern banks in 1898, 1904, and 1912. We fill in missing values by linear
interpolation between the values for these three years
and the bank occasionally refused to obey government orders; the BoCom
followed suit and became more privatized as well.40 It was the Beiyang govern-
ment’s partial loss of control over these government banks that made them
behave like commercial banks and seek to build a good reputation in the
banknote market. Figure 7.2 shows the volumes of circulating banknotes issued
by domestic and foreign banks from 1890 to 1936. It was during the warlord era
that the banknotes issued by domestic banks began to outnumber those issued by
foreign banks, which indicates that people came to trust domestic banknotes.
Moreover, the gap was enlarging over time: the volume of foreign banknotes was
40
For the BoCom, the government share dropped from 40 percent in 1914 to 20 percent in
1928. See Bank of Communications 交通银行总行 and Second Historical Archives of
China 中国第二历史档案馆, 交通银行史料第一卷 1907–1949 (Archives of the Bank of
Communications, vol. 1, 1907–1949) (Shanghai, China Financial Publishing House, 1995),
pp. 19, 23.
221

212.7 million, only approximately 10 percent of the volume of domestic banknotes

in 1936. Among all the domestic banknotes, usually more than 65 percent were
issued by the two government banks in the 1910s and 1920s.41
The credibility of domestic banknotes was built on the fact that these
banknotes were convertible with silver coins since, by law, the notes were
backed by 60 percent silver and 40 percent securities, mainly consisting of
government bonds. After the Nationalist Party took power and established
the capital in Nanjing in 1927, in 1928 the two government banks moved their
headquarters from Beijing to Shanghai, which was nearer to the new capital
city, following the government order. A natural concern was whether the
Government would intervene in the operations of the government banks,
which would probably threaten the convertibility of their banknotes.
The BoC responded quickly. In March 1928, it founded an eleven-member
committee of public representatives to monitor the reserves for banknote issu-
ance to assure the public of the convertibility of its banknotes. The committee
consisted of two members from each of three organizations, namely the Shanghai
General Chamber of Commerce, the Shanghai Bankers’ Association, and the
Shanghai Native Bankers’ Association; two representatives from any other finan-
cial institution that used its banknotes; and three members from the Board of
Directors and Supervisors of the BoC.42 These people were all reputable busi-
nessmen and bankers whose business would be severely affected if the BoC’s
banknotes became inconvertible. They hired an auditing house to check the
bank’s silver reserves against its notes on a monthly basis and published the
auditing results in the Shanghai Newspaper (Shenbao, 申报) after each examination.
This practice assured the public about the convertibility of the BoC’s banknotes.
Soon, this practice was followed by other note-issuing banks, including the
BoCom and even the Central Bank, newly established in 1928.
The Government started to consolidate monetary power and aimed to
establish a unified money market. The Government achieved its goal by “walk-
ing on two legs”: on one leg, it gradually encroached on the private share of the
banking sector until almost the whole sector was nationalized; on the other leg, it
took control of the money supply step by step until a fiat money system was
established with the Central Bank as the sole issuer of paper money.
The Government gained control of the banking sector in three ways. First, it
issued significantly more government bonds than its predecessors had:
2,325 million silver yuan worth of bonds in less than ten years, compared to
41
Rawski, Economic Growth, Table c2, p. 371.
42
Bankers’ Weekly 银行周报, 1917–1950, Shanghai Bankers’ Association, April 3, 1928.
222

only 630 million issued by its predecessors in Beijing in more than fifteen years.43
The domestic modern banks in Shanghai became the largest bondholder:
approximately one-third to two-thirds of all extant bonds were in their hands.44
The banks held government bonds both for investment and as reserves for
banknote issuance. A by-product of the bond policy was that the banking industry
was closely tied to the government.45 Second, the Central Bank was established in
1928. This fully government-owned bank was initially small in capital, with only
20 million yuan paid in government bonds. Despite its small capital power
compared with that of the other two government banks, the Central Bank was
granted four important monopoly rights: banknote issuance, official coin mint-
age, state treasury management, and the underwriting of treasury bonds. It soon
gained importance in the banking sector. Finally, the government took the bold
step of nationalizing the banking sector in 1935. In March 1935, the government
suddenly nationalized the BoC and the BoCom – the two government banks all
but dominated by private shareholders – by injecting government bonds as
capital. Then, within less than half a year, other major private banks were
nationalized.46 The government share in the banking industry increased from
less than 12 percent in 1934 to 72.8 percent in 1936.47
Simultaneously, the Government adopted monetary policies towards estab-
lishing a fiat money system that was fully government-controlled. After abolish-
ing the tael and establishing the yuan as the only unit of account and medium of
exchange in 1933, the Government significantly empowered its position in the
money supply since it controlled the official coin mintage and partially controlled
the supply of paper money pegged to and convertible with silver coins through
government banks. Two years later, the Government implemented another
monetary reform: it left the silver standard and adopted a fiat money system.
Before this reform, there was a free banking system in which the Central Bank,
43
See C.-Y. Ho and D. Li, “Reputation Building of a Nascent Government in Debt
Finance: Lessons from the Bond Market in China, 1912–1937,” working paper (2018),
Fudan University.
44
See Wu Chengxi 吴承禧, 中国的银行 (The Chinese Banks) (Shanghai, The
Commercial Press, 1934), pp. 68–73; Zhang Naiqi 章乃器, 中国货币金融问题
(Monetary and Financial Issues in China) (Shanghai, Shenghuo shudian, 1936), pp. 68–
9; K.N. Chang, “Toward Modernization of China’s Currency and Banking, 1927–1937,”
in P.K.T. Sih (ed), The Strenuous Decade: China’s Nation-Building Efforts, 1927–1937
(New York, St. John’s University Press, 1970), p. 147; P.M. Coble, The Shanghai
Capitalists and the Nationalist Government, 1927–1937 (Cambridge, MA, Harvard
University Council on East Asian Studies, 1986), pp. 74–5.
45
Coble, Shanghai Capitalists, p. 77.
46
Including the Commercial Bank of China, the Ningbo Commercial and Savings Bank,
and the National Industrial Bank of China.
47
Research Division of the Bank of China, Statistical Yearbook, pp. A55, B11, B16.
223

three government banks,48 and private banks all could issue banknotes. After the
reform, all private banks were forbidden to issue paper money, and the govern-
ment banks increased the paper money supply. As shown in Figure 7.2, the
banknotes issued by domestic banks increased dramatically from 747.7 million
yuan in 1934 to 1,976.7 million yuan in 1936, in which the notes issued by the
Central Bank increased by four times, the notes issued by the BoC increased by 1.8
times, and the notes issued by the BoCom increased by 2.7 times. In 1942, paper
money issuance was further consolidated solely in the hands of the Central
Bank.49
Competition and Co-operation between the

Government and the Private Financial Sector in the
Money Market
Many phenomena in the money market during our period of study were
consequences of co-operation and/or competition between the government
and the private financial sector when facing constraints, internal turmoil,
external invasions, and many other shocks. Competition from the private
sector disciplined the government and prevented it from overusing its power
over the money supply, and co-operation helped the government expand its
monetary power and achieve its goal of building a unified and growing
money market. Although an increasingly powerful authoritarian regime
could achieve the success of unifying the money market, it could also weaken
the private sector and block or even stop further development in the money
and financial markets. This section illustrates how the interplays between the
two parties contributed to the evolution (and sometimes regression) of
the money market by presenting one case from each of three different eras:
the late Qing dynasty, the warlord era, and the Nationalist prewar era.
No Debasement Please!
To address the fiscal crisis caused by the Taiping Rebellion in 1850–1860, the
Qing ruler, the Xianfeng Emperor, decided to generate government revenue
by debasing copper cash and issuing paper notes. The government first
started to issue a “big coin” equivalent to ten wen in 1853. For instance,
a copper cash with a stamped value equivalent to ten wen might only contain
48
First, only the two government banks, the BoC and BoCom, could issue banknotes.
Later, the Bank of Farmers was added.
49
See Hong Jiaguan 洪葭管, 中央银行史料 1928.11–1949.05 (The Archival Materials of
the Central Bank of China: 1928.11–1949.05) (Beijing, China Finance Publishing House,
2005), pp. 1–14, for a brief history of the Central Bank of China.
224

copper content equal to two original one-wen coins. The government later
issued several more types of larger cash equivalent to 20, 50, 100, 500, and
even 1,000 wen, but with certainly far less copper content, as indicated by
their nominal value. Simultaneously, the government cast iron cash with the
same stamped value as that of copper cash (another form of debasement) and
issued paper notes. The government pushed these new currencies through
the market with the coercive power of the state, e.g. by distributing them as
part of the salary of government officials and soldiers.
The private sector responded to these monetary policies promptly. On the
one hand, “big cash” made counterfeiting so profitable that people were
willing to risk their lives to melt down the original one-wen cash to produce
big cash. Soon the number of counterfeited big cash equal to 500 or 1,000 wen
surpassed the number of official big cash. Similarly, counterfeiting paper
notes was probably easier, and some counterfeits were even made by greedy
government officials. The government itself became a victim of these coun-
terfeit currencies since governmental agencies occasionally received them as
payment. In the end, some governmental agencies refused “big cash” and
paper notes as payment, which openly denied the legitimacy of these
currencies.
On the other hand, businessmen priced their products differently accord-
ing to what currency a shopper used. If a shopper used the original one-wen
copper cash, the price of the goods remained more or less the same as before
the rebellion. If a shopper used big cash, the price would increase dramatic-
ally until big cash was traded at a value close to that of its intrinsic copper
content. If a shopper used paper notes or iron cash, shop owners might refuse
to sell products to him or her. In an extreme case, on January 11 (according to
the lunar calendar) of the seventh year of the Xianfeng reign,50 more than half
of the rice shops and grocery stores in Beijing shut down to protest that they
did not want the big iron cash equivalent to ten wen as payment. The high
prices due to the usage of big cash made soldiers and other governmental
staff, who received big cash as their salary, barely able to make ends meet,
and the government had to give them a salary raise. This further encroached
on the already limited seigniorage revenue of minting debased coins, which
were mainly circulated in Beijing and its surrounding areas.51
50
According to the Western calendar, it was February 4, 1857. It was during the Chinese
Spring Festival.
51
For a more detailed description of the case, please see Peng, Monetary History, pp.
616–20.
225

Counterfeiting and the resistance to big cash and paper notes from the
private sector forced the Qing court to stop the practice less than four years
after it had started. It was exactly the competition and resistance from the
private sector that prohibited the government from massive debasement.
Interestingly, for a one-wen copper cash, the size, design, and weight might
vary across dynasties, across different emperor reigns within one dynasty,
and sometimes across different periods within one emperor’s reign, but the
copper content of a one-wen cash never deviated far from one qian 钱.52
Therefore the circulating copper cash at a given time could consist of cash
from previous dynasties, cash minted during the current dynasty, foreign
copper cash, counterfeit cash, and private cash. As long as these cash con-
tained copper content close to one qian and were traditionally accepted by
the local market, they all functioned well as small change. Ironically, compe-
tition from the private sector helped the government maintain a fairly good
standard in the supply of official copper cash,53 which solved the “big problem
of small change” in China.54
Bank Reputation or Government Reputation

Although Yuan Shikai’s attempt to restore the monarchy failed in
March 1916,55 the damage to the government’s fiscal status was evident.
The headquarters of the two government banks – the BoC and the
BoCom – located in Beijing were pressured to lend enormously to the
government. Consequently, the Beijing branches increased the issuance of
banknotes without an appropriate amount of silver reserve. These banknotes
were called the Beijing currency (jingchao 京钞) because they were issued by
Beijing branches and circulated mainly in Beijing and its immediate environs.
Rumors about Beijing currency becoming inconvertible with silver coins
brewed soon after Yuan’s failure to restore the monarchy, and bank runs
were imminent. On May 12, Yuan Shikai ordered the two banks to stop
cashing their banknotes. The Beijing currency soon became irredeemable,
and its market price in terms of silver yuan dropped to approximately
60 percent of face value.56 Even governmental agencies, such as railway
52
One qian is 3.75 grams. 53 Qian and Wu, “Monetary Stability,” 397–435.
54
Ma, “Monetary System,” pp. 57–64.
55
Yuan Shikai, the president of the nascent Republican government from 1912, pro-
claimed himself the new emperor, the Hongxian Emperor, on January 1, 1916. Due to
widespread opposition and revolts by both his former supporters and regional gover-
nors, he stepped down from the throne on March 22 and acted as the president again
until his death on June 6 in the same year.
56
Bankers’ Weekly, May 4, 1918.
226

companies, the Maritime Customs, post offices, and tax agencies, refused the
banknotes issued by these two government banks, regardless of which
branch issued the banknotes. The panic soon spread to banknotes issued by
other banks. The specter of a serious financial crisis was haunting China.
At this critical moment, the managers of the BoC’s Shanghai branch
heroically stood up to resist the government order and insisted on the
convertibility of banknotes issued by the Shanghai branch.57 This branch
issued more than half of the total banknotes issued by the whole bank. The
remaining convertibility of the Shanghai banknotes of the BoC not only
helped prevent the spread of financial crisis across the nation but also
established a sterling reputation for the branch’s banknotes.
The BoC did not leave the inconvertible Beijing currency unattended: the
BoC and the BoCom urged the government to repay their loans so that they
could use the advance to buy back the inconvertible Beijing currency. In 1918, the
government agreed to issue two bonds to pay back bank loans: one was a seven-
year, 6 percent short-term bond with a volume of 48 million backed by Maritime
Customs revenue; the other was a seven-year, 6 percent long-term bond with
a volume of 45 million backed by unspecified government revenue and lijin 厘
金. All Beijing currency holders could exchange their paper money at face value
with a composition of half of each bond. Thus a 100-yuan banknote could be
exchanged for fifty yuan of the short-term bond and fifty yuan of the long-term
bond.58 Another bond, the Financial Reorganization Short-Term, was issued in
1920 with a total amount of 60 million backed by Maritime Customs revenue to
swap for the remaining Beijing currency.59 All the inconvertible Beijing currency
was finally redeemed by government bonds.
As both the BoC and the BoCom acted more like private banks due to the
declining governmental share, they started to resist loaning to the govern-
ment and its officials. For instance, after two crises on the convertibility of
BoCom banknotes due to government default on loans to the bank in 1916
and 1921, the bank set up a new board of directors in June 1921 to reform the
bank. One important direction for the reform was to promulgate the rules for
57
See D. Ma, “Financial Revolution in Republican China during 1900–1937: A Survey and
New Interpretation,” Australian Economic History Review, 2019, 242–62, on how the
Shanghai branch of the BoC successfully resisted the order from Beijing and kept the
convertibility of its banknotes.
58
Shanghai Commercial and Savings Bank 上海商业储蓄银行, 内国公债要览 (An
Overview of Domestic Bonds) (Shanghai, Shanghai Commercial and Savings Bank,
1931), pp. 1–6.
59
Qian Jiaju 千家驹, 旧中国公债史资料, 1894–1949 (Archival Sources on Domestic
Government Bonds in China, 1894–1949) (Beijing, Zhonghua Book Company, 1984),
pp. 61–3.
227

lending, including no more loans to any government agencies from then on

and no loans to government officials unless proper collateral was provided.60
Behind these two government banks stood their private shareholders –
private banks. Almost all major private banks held shares in either the BoC or
the BoCom or both.61 These private banks pushed the government banks to
act like real commercial banks while enjoying privileges granted by the
government. These private shareholders helped the government banks
maintain a good reputation and win trust among the people in the warlord
era when the central government was weak. The sterling reputation enabled
these banks to attract ever-increasing deposits and simultaneously to lend out
more, issue more banknotes, and underwrite dramatically swelling govern-
ment bonds after the Nationalist Party seized power in 1927. However, the
fruit of this good reputation was harvested by the Government, which
gradually increased its influence over the two banks and finally nationalized
them in 1935.
Fragile Co-operation between Private Bankers and the

Nationalist Government
One striking feature of the financial market in prewar China was that the
secondary market for government bonds thrived in terms of its increasing size
and liquidity from 1927 to 1931 – the nascent years of the Nationalist
government.62 This is a puzzle given that the Nationalist government was
new and authoritarian, without any investor-friendly institutions in place.
Moreover, it faced numerous internal and external military threats, which
added significant risks to holding its bonds. Ho and Li point out that the
government tried hard to build its reputation by settling its predecessors’ debts
and consistently servicing the debts, which successfully distinguished it from its
predecessors and helped win trust among individual investors.63 However,
60
Bank of Communications and Secondary Historical Archives of China, The Bank of
Communications, p. 9.
61
For details on the shareholders of the two government banks, please see Bank of China
and Secondary Historical Archives of China, The Bank of China, pp. 88–94, for the BoC;
and Bank of Communications and Secondary Historical Archives of China, The Bank of
Communications, vol. 1, pp. 25–32, for the BoCom.
62
Over the fifteen years of the Beiyang government in Beijing, it issued thirty-five bonds
worth a total of 643 million silver yuan, among which six bonds could be floated on the
secondary market. In contrast, throughout the four years from 1927 to 1931, the
nationalist government issued twenty-six bonds, with the amount of the bonds reach-
ing 1,069 million, which exceeded the total amount of bonds issued in the Beiyang era
by 66 percent. Moreover, given the significant amount of bonds issued, more than
70 percent could be floated on the secondary market. Ho and Li, “Reputation Building.”
63
Ho and Li, “Reputation Building.”
228

building the government’s reputation would not have been successful without
help from private bankers. Nonetheless, co-operation became fragile when the
government deviated from its goal of maintaining its reputation due to other
urgent needs. The rise and fall of the Sinking Fund Commission for the Shanghai
Maritime Customs 2.5 Percent Surtax Treasury Note (hereafter the Commission)
is a typical example to illustrate this relationship between the two.
After the Nationalist government issued its first bond in Shanghai – the
Shanghai Maritime Customs 2.5 Percent Surtax Treasury Note of 30 million
silver yuan – the Commission was set up to handle the sinking fund and bond
repayments. The Commission consisted of fourteen members, among whom
three were officials from the central government and the rest mainly local
bankers and/or businessmen.64 However, the government seemed rarely
involved: first, no government officials were included in the five members of
the standing committee,65 which represented the Commission in handling all
related businesses; second, the three officials were always absent from
Commission meetings;66 finally, the government refused to cover operating
costs for the Commission despite its being nominally under the Bureau of
Domestic Bonds – a government organ – hence the Commission had to raise
funds to cover operating costs by itself.67
Domestic modern banks in Shanghai held the largest number of govern-
ment bonds, ensuring that bond credibility was aligned with the
Commission’s interests. The specified portion of customs revenue earmarked
for paying the note was handed to the Commission directly from the
Maritime Customs.68 The Commission publicized the information on the
64
Among the remaining eleven members, two were from the Shanghai local government
and were important local businessmen, and the rest were all from the local financial
industry and various chambers of commerce.
65
The five members were Li Fusun, Xie Taofu, Lin Kanghou, Xu Jingren, and Wu Linshu.
The first four were bankers, and the last was in the textile industry. Li Fusun was the
chairman of the Shanghai Bankers’ Association.
66
See the meeting minutes of the Commission of the Second Historical Archives of China
中国第二历史档案馆, 中华民国史档案资料汇编第五辑第一编 (Collection of
Archives and Documents of the Republic of China, vol 5.1, Public Finance and
Economics) (Nanjing, Jiangsu Ancient Books Publishing House, 1991), pp. 471–519.
67
The committee raised funds for covering operating costs by depositing some sinking
funds in non-designated private banks and native banks to earn higher interest, and
sometimes even making loans. See Shijuan Song 宋时娟, “江海关二五附税国库券基
金保管委员会研究” (Research on the Sinking Fund Commission for Shanghai
Maritime Customs 2.5% Surtax Treasury Note), unpublished master’s thesis, Fudan
University, 2000.
68
Unlike Francis Aglen, the inspector-general of the Customs from 1911 to 1927 who
controlled the sinking funds and influenced or even decided on the debt reimburse-
ment, Frederick Maze, the inspector of the Shanghai Maritime Customs and later the
229

revenue of the sinking fund and bond repayments monthly in newspapers. It

hired a local accounting firm to audit the fund quarterly, and similarly the
auditing results also appeared in newspapers.69 The transparency of the
sinking fund increased the Commission’s accountability. The Commission
achieved unprecedented success, as the 2.5 Percent Surtax Treasury Note was
duly serviced and fully reimbursed right on schedule in 1929.
Nonetheless, clashes between the government and the Commission emerged.
One dispute rested on their different standpoints: the government desperately
needed to raise funds by floating debts with less concern about debt service, while
the Commission stood up for the bondholders’ sake to ensure that bonds be duly
serviced. According to its name, the Commission was supposed to be responsible
only for ensuring the repayment of the note, and it was supposed to be disbanded
after the note was fully reimbursed. However, the Commission took over
management of the sinking funds of another twenty bonds (and notes), some-
times voluntarily, sometimes reluctantly or under government pressure. For
instance, because a newly issued bond without secure collateral would undercut
the credibility not only of the new bond itself but also of extant bonds as a whole,
the Commission requested that the government consult it before issuing any new
bonds. However, the government did not always do so, and intended to manage
the sinking funds of the new bonds by itself, but failed to find a market. Therefore
the Commission was requested by the government (sometimes together with
institutional bondholders) to take over the management of the new bonds, and it
had no choice but to try hard to secure the collateral.70 Even though the
Commission took on more responsibility, transparency remained, and the repay-
ments of the bonds in its custody were usually well carried out.71 Li Fusun, the
head of the Commission, stated proudly, “A bond without its sinking fund
managed by our commission cannot win the trust of society.”72
Another dispute regarded the Commission’s right to designate banks for
depositing the sinking funds. To attract more investment in government
new general inspector, was not included in the commission at this time due to the rise
of nationalism.
69
See Shanghai Newspaper 申报, various issues, 1921–1942, for the information.
70
For instance, the commission secured the amount of revenue earmarked for bond
repayment from the Excise Duty Bureau and the Internal Transaction Tax Bureau. In
addition, tobacco firms directly bought certificates of tobacco tax payment from the
commission, and the proceeds were used as collateral for bonds.
71
For some bonds, including the Tianjin Maritime Customs 2.5% Surtax Treasury Note in
1928, the 20th-Year Jiang-Zhe Silk Industry Bond in 1931 and the Hebei Hai River Project
Short-Term Bond in 1929, the respective sinking-fund commissions were established
following an institutional arrangement similar to the commission.
72
See the meeting minutes of the twentieth meeting of the Commission on April 17, 1929.
See Second Historical Archives of China, Documents & Archives, pp. 489–91.
230

bonds from various banks, the Commission deposited funds not only in large
banks but also in various small or native banks according to whether the bank
purchased bonds up to a certain threshold. Small banks could use bonds and
silver at a ratio of four to six to exchange for banknotes from large banks,73
and they could receive coupon payments of the bonds. This arrangement did
not take up their scarce funds. The additional benefit of receiving deposits
from the Commission provided them more incentive to buy bonds.
However, when the Central Bank was established in November 1928, the
Department of Finance required the Maritime Customs to send the sinking
fund for domestic bonds to the Central Bank (Shanghai Newspaper,
November 15, 1928) without consulting the Commission. Obviously, this
encroached on the Commission’s rights. The Commission negotiated with
the government for several rounds, and a compromise was finally reached:
the Central Bank took 40 percent of the sinking funds as deposits, and the
Commission took the remaining 60 percent.74
One last noticeable clash occurred concerning the bond repayments. The
government wanted to stop debt services temporarily after the Japanese
occupied Manchuria in September 1931 and bombed Shanghai in
January 1932, which reduced the financial resources pledged for debt
repayments.75 The Commission protested against this proposal. The two
parties negotiated for many rounds. Given that there was obviously not
enough funding for debt repayments, both sides yielded: the Commission
accepted the bond reorganization scheme of reducing the yearly amount of
principal service by half and hence extending the remaining maturity two
times longer; in return, the government promised that it would not issue
bonds for the next four years.76 This deed saved the government from fiscal
crisis in 1932, and it revealed that the Commission was not as hardline as it
claimed to be.
73
The amount of banknotes that could be exchanged for bonds was based on the bonds’
market value, not face value.
74
Song, “Sinking Fund Commission,” p. 23.
75
The cession of Manchuria to Japan led to an estimated annual loss of 70 million silver
yuan in customs revenue and transaction tax revenue – approximately one-tenth of the
government revenue from these two tax items, which served as the major funding
source for bond repayment. The amount of loss arising from the Shanghai incident was
estimated at 1.56 billion yuan. More than 16,000 civilians were killed or missing, one-
fourth of Shanghai’s factories were destroyed and many were damaged, and 80 percent
of industrial workers became unemployed. Shanghai Newspaper, March 20, 1932.
76
For the detailed negotiations between the bankers and the government during this
bond reorganization, see Coble, Shanghai Capitalists, pp. 102–9.
231

Soon after the 1932 bond reorganization, the Commission was restructured
to be the Sinking Fund Commission for National Bonds – a name that had
been rejected by Li Fusun in the early years but was accepted at this time. In
addition to the name, more government officials were added to the new
Commission. The government’s promise not to issue new bonds in the
following four years turned out to be empty: it resumed issuing bonds within
the year to keep itself financially afloat. In total, another thirteen bonds were
issued in the 1932–1935 period, only two of which could be floated on the
secondary market.77 The new Commission had little voice to protest the
government’s blatant violation of the 1932 deed. In 1936, the government
conducted another bond reorganization without any legitimate pretext and
totally bypassed the new Commission.78 Together with the bank coups in
1935, the private financial sector was emasculated by the ever more powerful
Nationalist government.
Money Market Integration and Its Impact on the

Economy
How monetary and financial development may benefit the economy has
been discussed both theoretically and empirically in an excellent review
article by Levine.79 Although it is impossible for us to pin down to what
extent money and financial market development contributed to GDP growth
in China during the period of our study, their positive effects were certain:
money market integration and financial developments facilitated trade, chan-
neled financial resources for modern industry, reduced interest rates, and
stabilized the economy.
Since the silver tael had never reached the same standard across localities,
silver in one locality was treated as merchandise in distant places and was
“bartered” for other merchandise. This type of costly barter exchange, in
addition to high transportation costs and other market frictions, prevented
a trading network from coming into being in premodern China – a large,
populous, and diverse country where trade should have benefited the econ-
omy enormously if it could have developed. Perkins notes that for almost
77
Ho and Li, “Reputation Building.”
78
For detailed information on the 1932 and 1936 bond consolidations, see Li Dan 李丹, 历
史 “大数据”：民国证券市场之量化研究 (“Big Data” in History: The Development
of Security and Bond Markets in Republican China) (Beijing, Peking University Press,
2016), pp. 199–207.
79
R. Levine, “Finance and Growth: Theory and Evidence,” Handbook of Economic Growth 1
(2005), 865–934.
232

1,000 years before 1900, the share of total agricultural output dedicated to
long-distance trade remained steady at a level of no more than 7 to 8 percent
in China.80
As the silver yuan gradually replaced the silver tael to become the most
widely accepted unit of account, these easily recognizable silver coins signifi-
cantly reduced transaction costs and hence promoted trade. In the 1920s, the
total volume of interprovincial long-distance trade had increased to more than
three times what it had been in the late nineteenth century.81 Rawski also finds
that the domestic inter-port trade (inbound and outbound) increased by 6.3
times from 1891 to 1931 based on the data compiled by the Maritime Customs.82
Certainly this dramatic expansion in domestic long-distance trade was caused
not solely by money market integration but also by other factors, such as
transportation development, the rise of modern industry, and integration into
global markets.83 However, the effect of money market integration on trade is
not negligible: transaction costs probably not only fall once when the economy
moves to money, but may continue to fall through financial innovation, which
continuously stimulates trade. More trade leads to greater specialization,
which drives productivity growth. These productivity gains feed back into
financial market development. As the virtuous circle continues, the economy
grows, and GDP per capita increases.84
One example of how money market integration affected the economic
landscape at the time through its impact on trade is the formation of treaty
port–hinterland economic zones. Wu identifies six treaty port–hinterland
economic zones in the 1920s.85 These economic zones had treaty ports as
hubs with vast hinterlands attached through trade. Increasingly more land
gradually transformed to producing agricultural goods for trade rather than
80
D.H. Perkins, Agricultural Development in China, 1368–1968 (Edinburgh, Edinburgh
81
Perkins, Agricultural Development in China, pp. 145–73.
82
Rawski, Economic Growth, p. 193,
83
International trade grew by almost ten times in value from 1891 to 1931. Rawski,
Economic Growth, Table 4.2, p. 193.
84
Levine, “Finance and Growth,” pp. 865–934.
85
The Shenyang–Dalian-centered zone in northeast China, the Beijing–Tianjin-centered
zone in north China, the Qingdao–Jinan-centered zone in Shandong and its
surrounding areas, the Shanghai-centered zone in the Yangzi delta, the Xiamen–
Fuzhou-centered zone in Fujian and its surrounding areas, and the Hong Kong–
Guangzhou-centered zone in south China. See Wu Songdi 吴松弟, “中国近代经济
地理格局形成的机制与表现” (The Formation of Economic Geography in Modern
China: Mechanism and Pattern), Journal of Historical Science 8 (2009), 65–72. Wu Songdi
吴松弟, 中国近代经济地理：绪论和全国概况第一卷 (Economic Geography in
Modern China: Introduction and Overview, vol. 1) (Shanghai, East China Normal
233

for self-sufficiency, and production in the hinterlands became pegged to

market conditions in their hub cities. For instance, in the 1920s, the
Shenyang–Dalian-centered economic zone in northeast China produced 60
to 70 percent of the world output of soybean, 90 percent of which was
exported abroad.86 Shenyang and Dalian became the world trading center
for soybean, and vast areas of the hinterland were devoted to producing this
economic agricultural product.
These economic zones could not have come into being without money
market integration. In the absence of the gradual circulation of standardized
official silver coins in villages and the expansion of the financial network as
documented above in this chapter, it would have been highly costly for
millions of peddlers, vendors, and traders to come to villages to buy agricul-
tural goods to be sold in cities and the international market. Only when
a standardized medium of exchange – silver coins – became widely acceptable
did transaction costs decline, which attracted more people into the trading
business. Financial development further facilitated trading, for instance by
providing credit, transferring bills, and exchanging international currencies.
Trade linked villages to cities and induced specialization; hence an economic
circle emerged. When the money market developed more as regional finan-
cial centers emerged, trade expanded, and an economic zone appeared.
In addition to facilitating trade, monetary and financial developments may
change people’s incentives to save and invest. In the case of China’s under-
developed economy during the period of our study,87 it was crucial for
economic development to pool the very limited savings together and lend
them out to ease the external financing constraints that impede modern
industrial development. In traditional China, hoarding was a common prac-
tice that allowed rich families to save because depositing silver tael in native
financial institutions would have incurred a high assessment cost, which
86
Hu Xuemei 胡雪梅, “东北大豆出口贸易与近代中国东北开发, 1860–1931”
(Soybean Exports in Northeast China and the Development of Northeast China in
Modern Times, 1860–1931), Northern Cultural Relics 3 (2002), 93–9.
87
It is true that China was a poor country at the time. However, this does not imply that
the people had no extra wealth to save. Wang cites Zhang Zongli’s estimation that the
gentry class (approximately 2 percent of the overall population) at the end of the
nineteenth century earned an income sixteen times that of an average family. The
compradors in the five treaty ports, although numbering no more than several hundred
in each port, earned approximately 530 million silver taels (approximately 741 million
silver yuan) from 1842 to 1894 – twice the national tax revenue. Additionally, the total
savings deposited in foreign banks in Shanghai in prewar China amounted to between
400 million and 500 million yuan. See Wang Yejian 王业键, 中国近代货币与银行的
演进, 1644–1937 (The Evolution of Money and Banking in Modern China, 1644–1937)
(Beijing, Institute of Economics, Academia Sinica, 1981).
234

impeded these institutions from taking small deposits from the general
public. For instance, native banks were only willing to take deposits larger
than a certain threshold, which varied from bank to bank, from cofounders
and a few wealthy friends and relatives.88 Occasionally, a large amount of
silver bullion was deposited by a wealthy family in a financial institution – not
for interest revenue but for safety purposes. Sometimes, depositors even had
to pay a fee for depositing silver bullion in a native bank. Moreover, it was not
relatively much safer to deposit money in native banks than it was to hoard
since native banks were vulnerable to shocks due to their small scale and
limited capital. Hence the total savings in financial institutions in historical
China could be small.
Depositing money in financial institutions gradually became an option for
rich families, and even the general public, to restore the value of their wealth
when the money market evolved and the banking sector expanded. On the
fund supply side, when the money market was transitioning from a highly
fragmented silver tael system to an integrated silver yuan system in tandem
with a silver-backed fiduciary paper money system, financial institutions
became more willing to take deposits from a broader social base, since taking
standardized coins or paper money as deposits was much less costly than
taking silver bullion. Moreover, households, especially urban residents, could
easily access banking services due to the rapid branching of domestic modern
banks across the nation, and depositing money in large nationwide banks,
such as the BoC and the BoCom, was much safer than putting money in small
native banks. Consequently, bank savings increased dramatically: national
deposits increased by almost eight times in less than three decades from 1910
to 1936, and reached a staggering amount of 5.9575 billion silver yuan in 1936,
approximately 54 percent of the total money supply (M2) in the year.89
On the demand side, capital could flow more freely and at less cost across
county, city, and provincial borders since the standardized money was univer-
sally accepted and the banking network was extensive. The transaction costs and
information asymmetry involved in the capital flow across localities caused by
the fragmented silver tael system were eliminated. Capital would flow to places
where it was in high demand and most efficiently used, such as cities with firms
clustered in modern industries. The modern industry was no doubt the largest
beneficiary of the capital flow. As the most efficient and fastest-growing sector,
the modern sector absorbed capital and became the main driver of productivity
88
The People’s Bank of China, Shanghai Branch, 上海钱庄史料汇编, p. 10.
89
Rawski, Economic Growth, Table c16, p. 394.
235

1600
1400
1200
1000
800
600
400
200
0
1903
1905
1907
1909
1911
1913
1915
1917
1921
1925
1927
1929
1931
1933
1935
1919
1923
Figure 7.3 The modern-oriented gross domestic fixed-capital formation in China from
1903 to 1936 (measured in 1933 million silver yuan)
Note. The modern-oriented gross domestic fixed-capital formation was constructed as
a measure for the apparent consumption of cement, iron and steel products, and
machinery. For details, please see Rawski, Economic Growth, pp. 242–8.
Source: Rawski, Economic Growth, p. 245, Table 5.2
growth in China. Figure 7.3 shows the fixed-capital formation in the modern-
oriented sector growing from 223 million in 1910 to 1,398 million in 1936, measured
by 1933 silver yuan, an increase of more than six times (less than the growth rate of
approximately eight times in overall deposits during the same period).
As more investment went to the modern sector, its output value grew
handsomely: in the thirty-three years from 1887 to 1920, the proportion of output
value of modern industries in the total output value increased from zero to
6.4 percent. In the following sixteen years, the proportion skyrocketed to
13.6 percent in 1936, with total industrial output reaching 3.9 trillion silver
yuan, and the annual growth rate was as high as 7.2 percent.90 Yan further
shows that the correlation coefficients among M2 money supply, modern fixed-
capital formation, and gross modern industrial output all exceeded 0.96 from
1912 to 1936.91
Money market integration and financial developments boosted both fund
supply and demand, which left the direction of change in interest rate
ambiguous. Despite there being no reliable nationwide interest rate for
90
Yan Hongzhong 燕红忠, “货币供给量、货币结构与中国经济趋势, 1650–1936”
(Money Supply, Monetary Structure, and Economic Development Trends in China,
1650–1936), Journal of Financial Research 7 (2011), 57–69.
91
Yan, “货币供给量.”
236

China in the period under study, the interest rate in Shanghai was readily
available and of good quality. This interest rate could be regarded as the basic
rate of the Chinese capital market at the time.92 Figure 7.4 shows the yearly
interbank loan interest rate – an interest rate on loans stipulated by the
Association of Native Banks in Shanghai, by which modern banks also abided
– from 1872 to 1936. Despite the fluctuation of the interest rate over time, the
declining trend was obvious. Moreover, this declining trend occurred not in
a peaceful period but in an era full of political chaos, numerous external
invasions, countless civil wars, and many other natural as well as man-made
disasters, which all would have probably put upward pressure on interest
rates due to the high uncertainty. Simultaneously, the various governments
also borrowed from the financial market significantly, either as bank loans or
as government bonds, which increased fund demand and therefore put
upward pressure on interest rates. Despite all these negative odds, the
interest rate continued to drop, which implies that the increase in fund supply
continuously exceeded the increase in fund demand over the period under
study. As Wang points out,93 the key for industrialization in prewar China did
not depend on whether or not China had enough savings, but on how
efficiently the surplus could be channeled to investment in the modern
sector.
14.00
12.00
10.00
8.00
6.00
4.00
2.00
0.00
1872
1875
1878
1881
1884
1887
1890
1893
1896
1899
1902
1905
1908
1911
1914
1917
1920
1923
1926
1929
1932
1935
Figure 7.4 The interbank annual interest rate in Shanghai from 1872 to 1936
Source: the annual interest rate is calculated as the average monthly interest rates
provided in People’s Bank of China, Shanghai Branch 中国人民银行上海分行, 上海钱
庄史料汇编 (Archival Materials of Shanghai Native Banks) (Shanghai, Xinhua Bookstore
Press, 1961), pp. 630–42
92
Yan, “货币供给量.” 93
Wang, Money and Banking, p. 86.
237

Given that the majority of modern banks and firms were located in
treaty ports and/or in concession areas in a treaty port, these modern
institutions were better shielded from the aforementioned chaos.94 Banks
in relatively safe places, especially Shanghai, attracted funds from all over
China. Modern firms clustered in these places due to the availability of
capital supply and security, which were crucial for modern firms to
emerge and thrive. The declining interest rate further eased the external
financing constraints that impede firm expansion in modern industry.
Moreover, the trade boom and industrialization process facilitated the
urbanization process: compared with 1900–1910, in 1938, the number of
cities with a population of more than 100,000 had increased by more than
70 percent.95
Last, the transition from a mainly silver money system to a fiat money
system helped shield the Chinese economy from the monetary shocks
caused by the changing market conditions of silver, which was
a commodity to the rest of the world, and stabilized the economy. This
impact can be well illustrated by the experience of the Chinese economy
around the Great Depression.96 When the world silver price increased
from 1931, caused by the great powers consecutively abandoning the gold
standard, and skyrocketed after the American Silver Purchase Act in 1934,
China suffered a massive silver outflow, and deflation kicked in while the
rest of world was recovering from the Depression. Only after the
Nationalist government implemented currency reform in
November 1935 and moved from silver to a fiat money system did the
three government banks increase note issues,97 and the deflationary spiral
then stopped and was reversed. By mid-1937, the general price level had
returned to the 1931 average.98
The Chinese money market obviously became deeper due to money
market integration and financial market developments. As shown in
Table 7.1, from 1887 to 1920, the M2 supply in China grew by 82.52 percent,
which was exceeded by the GDP growth rate of 88.82 percent. This indicates
94
See Ma, “Financial Revolution,” for detailed description on the role of concession areas
in fostering the development of the financial industry at the time.
95
See Perkins, Agricultural Development, pp. 171, 388–95, Appendix, Table 5.1.
96
For a detailed description and analyses of various industries and economy as a whole in
China during the Great Depression, see T. Shiroyama, China during the Great Depression:
Market, State, and the World Economy, 1929–1937 (Cambridge, MA, Harvard University
Council on East Asian Studies, 2008).
97
For a detailed description of the currency reform, see A.N. Young, China’s Nation-
Building Effort, 1927–1937 (New York, Hoover Press, 1971), pp. 216–39.
98
Young, China’s Nation-Building Effort, p. 109
238

Table 7.1 M2/GDP in China, the UK, and the US in the late
nineteenth and early twentieth centuries (million US$)
China 1887 1920 1936

M2 635 1,159 2,908
GDP 4,346 8,206 8,865
M2/GDP (%) 14.62 14.13 32.8
UK 1880 1913 1929
M2 3,486 6,960 15,696
GDP 11,310 22,860 24,762
M2/GDP (%) 30.82 30.45 63.39
US 1880 1913 1929
M2 2,030 15,730 46,600
GDP 18,765 53,942 90,308
M2/GDP (%) 10.82 29.16 51.6
Notes: M2 and GDP in China are measured at the current value in 1936 with
the exchange rate of yuan to US$ at 0.303. M2 and GDP in the UK and the US
are measured at the current value of 1929 with the exchange rate of sterling to
US$ at 6.
Sources: this table is adjusted by the authors based on Yan Hongzhong 燕红忠,
中国的货币金融体系, 1600–1949 (The Monetary and Financial System in China,
1600–1949) (Beijing, China Renmin University Press, 2012), pp. 305–6, Table 13–1.
Yan cites the money supply in China from Peng Xinwei 彭信威, 中國貨幣史
(Chinese Monetary History) (Shanghai, Shanghai renmin chubanshe, 1958),
pp. 888–9 for 1887; and from Rawski, Economic Growth, p. 394, Table c16, for 1920
and 1936 (he uses the series with the larger estimators, namely series B in the
table); GDP in China for 1887 from Liu Foding 刘佛丁, Wang Yuru 王玉茹, and
Yu Jianwei 于建玮, 近代中国的经济发展 (Economic Development in
Modern China) (Jinan, Shandong People’s Press, 1997), p. 95; and from Wu
Chengming 吴承明, 中国的现代化：市场与社会 (Chinese Modernization:
Market and Society) (Beijing, Joint Publishing, 2001), pp. 109–10 for 1920 and 1936;
and the money supply and GDP in the UK and the US from Milton Friedman and
Anna J. Schwartz, Monetary Trends in the United States and the United Kingdom: Their
Relations to Income, Prices, and Interest Rates (Chicago, University of Chicago Press,
1983), pp. 144–61
that the money supply in China only matched economic expansion, which
left the M2/GDP ratio stable at approximately 14 percent. The financial
deepening sped up afterwards as the money supply grew more than 150 per-
cent, while GDP grew by merely 8 percent from 1920 to 1936. Hence the
depth of the money market measured by M2/GDP more than doubled, from
239

14.13 percent in 1920 to 32.8 percent in 1936. This dramatic monetization was
unprecedented in Chinese monetary history. However, we should not over-
estimate this achievement, as both the UK and the US experienced dramatic
monetization in the earlier 1913–1929 period similar to that in China during
1920–1936. Moreover, compared with the money market in the UK and the
US, the Chinese money market was very shallow, as the M2/GDP ratio in
China in 1936 was approximately half of that in the UK and two-thirds of that
in the US in 1929. Hence there would still probably have been plenty of room
for money and financial markets to continue to play a more active role in
facilitating economic development in China if progress had not been halted
by the Second Sino-Japanese War.
Conclusions
The period from 1800 to 1937 witnessed China’s journey towards a unified
money market. It started from the market response to and the govern-
ment’s accommodation of the influx of foreign silver. The two types of
foreign silver – bullion and coins – coexisted and competed in the money
market, and silver coins ultimately became the dominating currency due
to their convenience. Consequently, the yuan, as the unit of account for
foreign silver coins, became the most important unit of account, based on
which an integrated money market began to come into being. The
Beiyang government started to mint official silver coins in the early
1910s, and with help from the private financial sector, especially the
expanding domestic modern banking sector, official coins gradually out-
paced foreign silver coins and became the dominant medium of exchange
in the very complex and fragmented money market in China. In 1933, the
silver yuan officially became the only unit of account, and hence a unified
silver yuan standard was established. The supply of both silver coins and
banknotes exchangeable with coins was largely under government con-
trol. As China went off silver two years later, a fiat money system fully
under government control was founded. As the money market evolved, it
facilitated banking expansion, trade, industrialization, and economic sta-
bilization, which promoted economic growth.
The money market integration under study here was an outcome of
both a bottom-up evolution from the private sector and a top-down
reform from the government. The demand for a standard medium of
exchange from the goods market together with the tireless promotion of
silver yuan from an ever-expanding financial industry formed the major
240

market force pushing China towards a unified money market. However,

the journey did not progress linearly over time but sped up when
a powerful central government with modern monetary knowledge took
power.99 The decisive actions taken by the Nationalist government were
crucial steps towards monetary unification. The bottom-up force and the
top-down reform could be complementary, as the government responded
to the market’s demand, and the private sector helped the government
expand its monetary power. However, the two forces could also be
confrontational, as the government gained more power in a unified
money market and the private sector lost its ground in banknote issuance.
Even worse, the private financial sector was emasculated by the ever more
powerful Nationalist government.
This chapter has provided only a very brief summary of the evolu-
tion of the money market in historical China. Studies of historical
money market development in China are much less common than
those of the US, Britain, France, and other European countries.100
However, there are still many important questions and puzzles that
remain to be answered: for instance, why did China, as a centralized
country unlike disintegrated Europe, fail to create a unified money
market much earlier in history? Why did China, as a country where
99
The success achieved by the Nationalist government in such a short period can be
attributed to two factors. One is that the money and financial markets had become
integrated through the expanding modern banking network, where large government
banks, such as the BoC and BoCom, dominated the network before 1927. Hence the
government could easily affect the national money market insofar as it could influence
and/or control these major banks. The other is that there was a group of financial
experts in the Nationalist government, for instance Song Ziwen and Kong Xiangxi, the
ministers of finance consecutively from 1928 to 1937, who knew how to build a modern
unified money market. They knew that the key to maintaining an active bond market
was to keep a good reputation (see Ho and Li, “Reputation Building”); they knew the
importance of having a central bank; they sought advice, consultancy, and assistance
from foreign experts, including Arthur Young, Oliver C. Lockhart, F.B. Lynch,
Frederick W. Leith-Ross and many others (see Young, Nation Building) when they
encountered difficulties in economy and finance.
100
A few exceptions include Kann, The Currencies of China, R. von Glahn, Fountain of
Fortune: Money and Monetary Policy in China, 1000–1700 (Berkeley, University of
California Press, 1996); and von Glahn, “Foreign Silver Coins,” 51–78; Ma, “Monetary
System,” pp. 57–64; Ma, “Financial Revolution”; Ma and Zhao, “Silver
Transformation,” in English; in Chinese see Peng, Monetary History; Wang, Money
and Banking; Wu Jingping 吴景平, 上海金融业与国民政府关系研究, 1927–1937
(Research on the Relationship between the Shanghai Financial Industry and the
Nationalist Government, 1927–1937) (Shanghai, Shanghai University of Finance and
Economics Press, 2002); Yan Hongzhong 燕红忠, 中国的货币金融体系, 1600–1949
(The Monetary and Financial System in China, 1600–1949) (Beijing, China Renmin
University Press, 2012); Zhang, 中国货币金融问题 and others on the prewar China.
241

the earliest fiat money appeared in human history, as early as the

Northern Song dynasty (960–1279), only establish a fiat money system
by 1935? Why did China lag far behind Europe and the US in financial
development in modern history? More research on this topic is desired
to shed light on these questions and to offer a better understanding of
contemporary Chinese money and financial markets due to path
dependence.
Further Reading
Chang, K.-N., “Toward Modernization of China’s Currency and Banking, 1927–1937,” in P.
K.T. Sih (ed.), The Strenuous Decade: China’s Nation-Building Efforts, 1927–1937 (New York,
St. John’s University Press 1970), pp. 129–65.
Cheng, L., Banking in Modern China: Entrepreneurs, Professional Managers, and the
Development of Chinese Banks, 1897–1937 (Cambridge, Cambridge University Press,
2003).
Coble, P.M., The Shanghai Capitalists and the Nationalist Government, 1927–1937 (Cambridge,
Harvard University Council on East Asian Studies, 1986).
Irigoin, A., “The End of a Silver Era: The Consequences of the Breakdown of the Spanish
Peso Standard in China and the United States, 1780s–1850s,” Journal of World History 20.2
(2009), 207–44.
Ma, D., “Chinese Money and Monetary System, 1800–2000, Overview,” in G. Caprio et al.
(eds.), Handbook of Key Global Financial Markets, Institutions, and Infrastructure, vol. 1
(Oxford, Elsevier, 2012), pp. 57–64.
Ma, D., “Financial Revolution in Republican China during 1900–1937: A Survey and New
Interpretation,” Australian Economic History Review, 2019, 242–62.
Peng Xinwei 彭信威, 中国货币史 (Chinese Monetary History) (Shanghai, Shanghai
People’s Publishing House, 1987; first published 1954).
Von Glahn, R., “Foreign Silver Coins in the Market Culture of Nineteenth-Century
China,” International Journal of Asian Studies 4.1 (2007), 51–78.
Von Glahn, R., Fountain of Fortune: Money and Monetary Policy in China, 1000–1700 (Berkeley,
University of California Press, 1996).
Wang Yejian 王业键, 中国近代货币与银行的演进, 1644–1937 (The Evolution of Money
and Banking in Modern China, 1644–1937) (Beijing, Institute of Economics, Academia
Sinica, 1981).
Wu Chengxi 吴承禧, 中国的银行 (The Chinese Banks) (Shanghai, The Commercial
Press, 1934).
Wu Jingping 吴景平, 上海金融业与国民政府关系研究, 1927–1937 (Research on the
Relationship between the Shanghai Financial Industry and the Nationalist
Government, 1927–1937) (Shanghai, Shanghai University of Finance and
Economics Press, 2002).
Xu Cangshui 徐沧水 (ed.), 中国今日之货币问题 (Monetary Issues in Today’s China)
(Shanghai, Bankers’ Weekly Press, 1921).
242

Yan, Hongzhong, “Economic Growth and Fluctuation in the Early Qing Dynasty: From
the Perspective of Monetary Circulation,” Frontiers of History in China 4.2 (2009), 221–64.
Yan Hongzhong 燕红忠, “货币供给量、货币结构与中国经济趋势, 1650–1936”
(Money Supply, Monetary Structure and Economic Development Trend in China,
1650–1936), Journal of Financial Research 7 (2011), 57–69.
Yan Hongzhong 燕红忠, 中国的货币金融体系, 1600–1949 (The Monetary and Financial
System in China, 1600–1949) (Beijing, China Renmin University Press, 2012).
Zhang Naiqi 章乃器, 中国货币金融问题 (The Monetary and Financial Issue in China)
(Shanghai, Shenghuo shudian, 1936).
243

8
Public Finance
At the turn of the nineteenth century, the Qing dynasty entered a phase of
social and economic decline. By 1850, mounting crises had exploded in
a devastating series of rebellions (best known for the Taiping Rebellion,
1850–1864). By 1880, up to a quarter of the population had perished, although
the numbers are debated. The civil wars revealed the bankruptcy of the
dogma of fixed tax quotas that had governed China’s fiscal thought since the
Ming dynasty (see the chapter by von Glahn and Lamouroux in Volume 1).
New commercial taxes, most prominently foreign customs and lijin 釐金
(literally “one-thousandth”) trade tariffs, soon exceeded agricultural taxes and
increased state revenue.1 Fiscal recovery was short-lived, however, as the
double defeat in the First Sino-Japanese War (1894–1895) and the Boxer
Rebellion (1900–1901) once again threw Qing finances into turmoil.
Servicing the war loans and indemnities while simultaneously promoting
costly “New Policy” (xinzheng 新政) reforms (1901–1911), the imperial gov-
ernment gradually lost control of the provinces and was unable to check the
nationalist awakening of its citizenry. This led to the 1911 Revolution and,
eventually, national disintegration during the warlord era. In 1928, the
Nationalist Party partly reunified the country and dedicated itself to fiscal
consolidation and state developmentalism, despite persistent warlordism and
a fledgling communist insurgency. However, the short era of growth known
as the “Nanjing decade” was again interrupted after 1937 by the Japanese
invasion, which placed the wealthiest revenue-generating regions outside
the control of the Nationalist government, and the ensuing civil war with the
E. Kaske whishes to thank Ziyan Zhu, Gus Chan, and Rewert Hoffer for pre-screening
some of the literature and help with final editing.
1
Zhou Yumin 周育民, 晚清財政與社會變遷 (Late Qing Fiscal Policies and Social
Change) (Shanghai, Shanghai renmin chubanshe, 2000), pp. 238–9; Yeh-chien Wang,
Land Taxation in Imperial China, 1750–1911 (Cambridge, MA, Harvard University Press,
1973), p. 80.
244

Public Finance
rising Communist Party. The history of institutional change in modern China

has to be understood as a response to these historical crises and turning
points.
The central conflict of the era was the ability of the center to obtain
revenue from the provinces. Under the Qing dynastic system, the center
controlled the revenues extracted at the local level through distributional
authority rather than direct institutional intervention, but this authority
gradually eroded after the mid-nineteenth-century rebellions. Beijing’s
control of provincial revenues was all but lost after 1916, and only incom-
pletely restored after 1928 by the Nanjing government through military
force, nimble bank finance, and revenue-sharing negotiations. The history
of a decentralized fiscal system cast a long shadow, leaving the center
increasingly dependent on revenues from foreign trade, salt, and the fledg-
ling modern industries of a few coastal provinces. The Japanese occupation
cut off the Nationalist government from these sources of revenue and
forced it to build up its own revenue-extracting capacities on the ground,
but the fruits of these efforts would be reaped by the Communist regime
after 1950.
Given the limitations of space and available literature, this chapter focuses
on revenue and cannot exhaustively explore expenditures or the relationship
between public finance and the wider economy. Assessments of the share of
government expenditure in the economy depend on the underlying estimates
of national income. Extrapolating from Liu and Yeh’s computation of China’s
GDP for 1933, Wang Yeh-chien has estimated that the share of government
revenue and expenditure in GDP remained roughly the same until 1908,
about 2.4 percent, even as nominal revenue grew eightfold and real tax
revenue tripled between 1861 and 1911, with most of the growth happening
after 1895.2 Between 1908 and 1933, total central and local government
expenditure grew slightly faster than GDP, from 2.4 percent to about
4 percent.3 By 1933, the central government alone consumed about 2.3 percent
of GDP, or 2 percent if only non-borrowed revenue is considered. The deficit
2
T. Liu, K.C. Yeh, and T. Chong, The Economy of the Chinese Mainland: National Income and
Economic Development, 1933–1959 (Princeton, NJ, Princeton University Press, 1963), pp. 94,
375–9; Wang, Land Taxation in Imperial China, p. 133. For a higher estimate based on
a lower assessment of “national income,” see A. Feuerwerker, “The State and the
Economy in Late Imperial China,” Theory and Society 13.3 (May 1, 1984), 300.
3
Y. Xu, Z. Shi, B. van Leeuwen, Y. Ni, Z. Zhang, and Y. Ma, “Chinese National Income, ca.
1661–1933,” Australian Economic History Review 57.3 (November 2017), 368–93. For a higher
estimate see Kubo Tōru 久保亨, “財政史” (Fiscal History), in Kubo Tōru 久保亨 (ed.)
中国経済史入門 (An Introduction to China’s Economic History) (Tokyo: Tōkyō
daigaku shuppankai, 2012), p. 129.
245

was 12.3 percent.4 While this is a far cry from that of Britain, the most
developed fiscal state at the time, it may be roughly compared to that of
the US before the New Deal, which served as a policy model for Nationalist
fiscal planners. In the same year, the non-borrowed revenues of the US
federal government were 3.4 percent of GDP, but since the US had a larger
deficit during the Great Depression, spending was at 7.9 percent.5 Despite
fiscal modernization, the extractive capacities of the Chinese government
remained limited throughout our period.
The Crisis and Restoration of Qing Fiscal

Governance, 1800–1894
The White Lotus Rebellion (1796–1804) was a turning point for the Qing. Its
suppression lasted for eight years and consumed the total government
revenue of almost two years (up to 120 million taels). While statutory
revenue had been frozen at the quotas of 1766 due to a hardening dogma of
“benevolent” governance, the war demonstrated that it was easier to fix
revenues than to control expenditures. Since the Qing paid for wars from
accumulated treasure, the silver reserves of both Beijing and the provinces
declined drastically, never to fully recover.6
The court responded to the crisis with an austerity program starting from
the imperial house itself. The grand imperial tours and building projects of
the previous century stopped. This restored legitimacy, but could neither
improve the deteriorating economic and social conditions of a growing
population nor stop the increasing opium imports from (and silver drain
to) an expanding British Empire. Recurring floods and the collapse of the
Yellow River–Grand Canal system in 1823 aggravated the fiscal crisis. The
restoration of the tribute grain shipments to Beijing along the Grand Canal
came at huge expense (the less costly option of shipping by sea was tried
briefly but soon given up).7 With an increase of the land tax (nominally
4
Jiang Liangqin 姜良芹, 南京國民政府內債問題研究:以內債政策及運作績效為中心
(A Study of Domestic Debt under the Nanjing Nationalist Government: Domestic Loan
Policy and Performance) (Nanjing, Nanjing daxue chubanshe, 2003), p. 314.
5
Table 1.2 – Summary of Receipts, Outlays, and Surpluses or Deficits (–) as Percentages of
GDP: 1930–2025, Historical Tables, Office of Management and Budget, The White
House, www.whitehouse.gov/omb/historical-tables.
6
Y. Dai, The White Lotus War: Rebellion and Suppression in Late Imperial China (Seattle, WA,
University of Washington Press, 2019); Z. Shi, Central Government Silver Treasury:
Revenue, Expenditure and Inventory Statistics, ca. 1667–1899 (Leiden, Brill, 2016).
7
J.K. Leonard, Controlling from Afar: The Daoguang Emperor’s Management of the Grand Canal
Crisis, 1824–1826 (Michigan, University of Michigan Press, 1996).
246

Public Finance
73 percent of total revenue) ruled out by the emperors’ claim to benevolence,

public debt was also not an option, as the fixed-quota tax system and the high
interest rates would have made future repayments impracticable. The only
legitimate measure available to finance the growing deficit was to cut
spending by withholding the “silver to nourish integrity” emoluments
received by officials and to expand “contributions,” i.e. exactions from
holders of imperial privileges like officials or salt merchants as well as the
sale of rank and office.8 Government-licensed salt syndicates became
a particular target of exactions. Unlike the land tax, salt tax arrears were
almost never forgiven, and the statutory salt tax quota (nominally 8.6 percent
of revenue in 1766) was only one-quarter to half of what was actually
collected.9 As contributions piled up, while the high salt price, a looming
currency crisis, and rampant smuggling depressed sales, tax debts ballooned
in the 1820s, and the salt tax all but collapsed.10
By the time of the First Opium War (1839–1842), which firmly established
British military dominance in Chinese waters, the imperial treasury was
bankrupt. The burden of the war of about 52 million taels (including indem-
nities), almost a year of statutory revenue, was borne by the provinces.11 The
war and another breach of the Yellow River dikes in 1842 deepened
the recession which already plagued the economy due to the silver crisis.
The calamity soon erupted into a series of rebellions, of which the Taiping
Rebellion was only the most infamous. The total direct costs of these wars
from 1851 to 1874 have been estimated at around 800 million to 850 million
taels, about twelve times the annual statutory revenue of the empire.12 These
estimates do not account for the loss of human life or the destruction of
productive capacity and ecology. As the dynasty could no longer invest in
8
E. Kaske, “Austerity in Times of War: Government Finance in Early
Nineteenth-Century China,” Financial History Review 25.1 (April 2018), 71–96.
9
Saeki Tomi 佐伯富, 清代鹽政の研究 (A Study of the Qing Salt Administration)
(Tokyo: Tōyōshi kenkyūkai, 1956), p. 223; Ni Yuping 倪玉平, 博弈與均衡：清代兩
淮鹽政改革 (Games and Equilibrium: The Reforms of the Liang Huai Salt
Administration during the Qing Dynasty) (Fujian, Fujian renmin chubanshe, 2006),
pp. 62–70.
10
Chen Feng 陳鋒, 清代鹽政與鹽稅 (Salt Administration and Salt Tax during the Qing
Dynasty) (Wuhan, Wuhan daxue chubanshe, 2013), pp. 258, 264; T. Metzger, “The
Organizational Capabilities of the Ch’ing State in the Field of Commerce: The Liang-
Huai Salt Monopoly, 1740–1840,” in W.E. Willmott (ed.), Economic Organization in
Chinese Society (Stanford, CA: Stanford University Press, 1972), p. 18.
11
Mao Haijian 茅海建, “鴉片戰爭清朝軍費考” (The Military Expenses of the Qing
Empire during the Opium War), Jindaishi yanjiu 6 (1996), 34–80; Zhou, 晚清財政與社
會變遷, pp. 79–80.
12
Zhou, 晚清財政與社會變遷, pp. 153; Zhou Zhichu 周志初, 晚清财政经济研究
(Late Qing Government Finance and Economy) (Jinan, Qi Lu shushe, 2002), p. 205.
247

maintaining the dikes, the Yellow River changed course in 1855, thus render-
ing the Grand Canal largely defunct.13
The wars produced a new generation of provincial leaders, who had raised
militias to fight the rebellions. The system of governance they aimed to
restore, hand in hand with the imperial government, professed to be a slightly
modified version of the prewar economic and fiscal regime.14 However, the
balance between the center and the provinces changed forever, materialized
in new institutional settings. For one thing, in the chaos of the rebellions, new
military institutions – army secretariats, paymaster and reconstruction
bureaus – took over civil roles in resettling deserted land, rebuilding des-
troyed cities, and repairing irrigation infrastructure.15 They became the core
of a growing permanent provincial administration staffed with “deputies”
(weiyuan 委員) who stood outside the statutory bureaucracy but, unlike the
latter, had more clearly defined jurisdictions and fixed monthly salaries
(xinshui 薪水).16 Second, the war also gave rise to new internal transit duties,
known as lijin, which paid for the institutional expansion, as well as a foreign-
managed customs administration in charge of the treaty port trade that had
been opened up by the two Opium Wars.
The growth of commercial taxation and the rise of foreign debt is the
important and often-told story of late Qing fiscal reforms. Before the 1850s,
only one customs system existed, with thirty-four tollhouses along the sea-
board, major trade routes, and the northern frontiers.17 By the 1860s, three
types of revenue were collected from moving trade, each with its own tariff
schedule and collection agencies: Native Customs (changguan 常關), the
Maritime Customs Service (haiguan 海關, hereafter MCS), and the lijin
duties. By 1893, according to an estimate by British consul Jamieson, total
revenue had increased to 89 million taels and the share of the land tax fallen to
13
I. Amelung, Der Gelbe Fluß in Shandong (1851–1911): Überschwemmungskatastrophen und ihre
Bewältigung im China der späten Qing-Zeit (Wiesbaden, Harrassowitz, 2000).
14
M.C. Wright, The Last Stand of Chinese Conservatism: The Tʻung-Chih Restoration, 1862–
1874 (New York, Atheneum, 1969).
15
C. Wooldridge, City of Virtues: Nanjing in an Age of Utopian Visions (Seattle, University of
Washington Press, 2015), pp. 117–49; E. Schluessel, “Water, Justice, and Local
Government in Turn-of-the-Century Xinjiang,” Journal of the Economic and Social
History of the Orient 62.4 (May 16, 2019), 599–625.
16
Xiao Zongzhi 肖宗志, 候補文官群體與晚清政治 (Expectant Officials and Politics in
Late Qing China) (Chengdu, Bashu shushe, 2007); Guan Xiaohong 关晓红, 从幕府到
职官: 清季外官制的转型与困扰 (From Private Secretary to Official: The
Transformation and Predicament of the Provincial Bureaucracy in Late Qing China)
(Beijing, Sanlian shudian, 2014).
17
H.J. van de Ven, Breaking with the Past: The Maritime Customs Service and the Global
Origins of Modernity in China (New York, Columbia University Press, 2014), pp. 54–60.
248

Public Finance
36 percent. Over 40 percent came from customs and lijin duties; 15 percent from
salt.18 The relative shift from agricultural to commercial taxes has been lauded
as a sign of modernization and state strengthening, but it also led to a shift in
the balance of power between the central government and the provinces.
From this perspective, not the more famous MCS but the lijin collectorates
were the more revolutionary institutions, because they carved out an inde-
pendent fiscal space for the provinces.19
Lijin duties emerged in the early 1850s out of a flurry of imposts on trade,
commodities, and property levied by the army headquarters during the chaos
of the civil wars. Most were abolished after the wars, especially levies on land,
but provincial leaders defended and eventually salvaged the trade duties – in
exchange for a sharing deal with the central government.20 The term lijin was
used to refer to at least three types of duty that were accounted separately,
namely stationary excises and transit duties on general merchandise (“com-
modities lijin”), inland duties on foreign and native opium after its legaliza-
tion in 1858 (“opium lijin”), and, finally, extra levies on salt in addition to the
regular salt tax (“salt lijin”). The transit duties on commodities usually stand
as synecdoche for all lijin duties until their abolition in 1931. In general, duties
for a large variety of commodities had to be paid at least twice – at the place of
origin and the place of sale – within each province. In reality, toll stations
along the road could be more frequent, to say nothing of crossing provincial
boundaries. Tax rates, names, and modes of collection differed widely.
Further duties were levied from the producers (mostly tea and silk) and as
a sales tax (luodishui 落地稅).21
The lijin duties were reviled by Chinese and Western merchants alike,
because they imposed a regressive tax burden on the people and made
exports more expensive.22 However, as an institution they also introduced
important innovations. First, although some of the stationary excises were
18
G. Jamieson, The Revenue and Expenditure of the Chinese Empire (Shanghai, Shanghai
Mercury, 1897), pp. 21–2.
19
Iwai Shigeki 岩井茂樹, 中国近世財政史の研究 (A Study of the Fiscal History of
Modern China) (Kyoto: Kyōto daigaku gakujutsu shuppankai, 2004), Chapter 4;
Shi Zhihong 史志宏 and Xu Yi 徐毅, 晚清財政: 1851–1894 (Late Qing Fiscal Policy:
1851–1894) (Shanghai, Shanghai caijing daxue chubanshe, 2008), pp. 239–40, 249–50.
20
E.G. Beal, The Origin of Likin, 1835–1864 (Cambridge, MA, Harvard University Press,
1958), pp. 121–4; Xu Yi 徐毅, 江蘇厘金制度研究: 1853–1911 年 (A Study of the Lijin in
Jiangsu, 1853–1911) (Shanghai, Shanghai caijing daxue chubanshe, 2009). On land see Shi
and Xu, 晚清財政: 1851–1894, p. 95.
21
Luo Yudong 羅玉東, 中國釐金史 (The History of Lijin in China) (Shanghai, Shangwu
Yinshuguan, 1936), pp. 55–61; Jamieson, The Revenue and Expenditure of the Chinese Empire,
pp. 16–18, 33–8.
22
Luo, 中國釐金史, p. 135.
249

self-declared (renjuan 認捐) by guilds or handled by tax farmers, most

transit duties were levied by salaried staff of new provincial bureaus.
Unlike regular bureaucrats, tax collectors served under the direct control
of the governor and fulfilled specialized tasks, unencumbered by multiple
administrative duties. This improved efficiency and monitoring. Second,
unlike other taxes the lijin duties were assessed and collected in coin instead
of silver. This was not only an advantage at the time of their invention
when silver was expensive; it also reduced opportunities for collectors to
profit from arbitrage to the disadvantage of taxpayers. The lijin bureaus
were among the first government institutions to regularly report currency
exchange rates.23
Most importantly, the lijin duties became the main source of provincial
revenue. The central government only partially succeeded in bringing lijin
revenue under control. Even though the provinces reported lijin income from
1869, they soon settled on a fixed quota of less than 15 million taels for the
commodities lijin alone (Figure 8.2), whereas the real revenue was closer to
21 million taels (the balance became known as waixiao 外銷, meaning outside
the auditing of the Board of Revenue).24 Lijin on native opium was never
reported, and on foreign opium only reluctantly (see below). As the rebellions
blurred the boundaries of government-controlled salt markets, salt lijin duties
(both stationary and transit) became another source of waixiao income for
some provinces. Most famously, Sichuan’s cheaper salt, manufactured from
brine wells, broke the monopoly of expensive Southern Huai sea salt and
flooded Hubei and Hunan, whose provincial coffers also benefited from lijin
on the salt imports. The once powerful Southern Huai salt division appealed
to the court in Beijing, but failed to restore complete market dominance.
Sichuan in turn reformed its salt administration and greatly increased its salt
tax revenue.25 Growing revenues allowed the provinces to expand their
bureaucracies, enlarge their footprint in society, and finance modernization
projects. This marked an end to the state of minimal government that had
23
W. He, Paths toward the Modern Fiscal State: England, Japan, and China (Cambridge, MA,
Harvard University Press, 2013), pp. 164–9. For less optimistic assessments, see Luo,
中國釐金史, pp. 125–7; Zhou, 晚清财政经济研究, pp. 78–9, 85; Zhou, 晚清財政與社
會變遷, pp. 260–1.
24
Luo, 中國釐金史, pp. 119–20; He, Paths towards the Modern Fiscal State, pp. 169–72;
Zhou, 晚清財政與社會變遷, pp. 344, 352–8.
25
Ni, 博弈與均衡, pp. 134–42; Chen, 清代鹽政與鹽稅, pp. 143–54; Zhou, 晚清財政與
社會變遷, pp. 297–301; M. Zelin, The Merchants of Zigong: Industrial Entrepreneurship in
Early Modern China (New York, Columbia University Press, 2005), pp. 81–4, 98–9, 152–8;
S.A.M. Adshead, Province and Politics in Late Imperial China: Viceregal Government in
Szechwan, 1898–1911 (London, Curzon Press, 1984), pp. 37–42.
250

Public Finance
35
Million Taels
30
25
20
15
10
0
58
60
62
64
66
68
70
72
74
76
78
80
82
84
86
88
90
92
94
96
98
00
02
04
06
08
18
18
18
18
18
18
18
18
18
18
18
18
18
18
18
18
18
18
18
18
18
19
19
19
19
19
Commodity Opium (without tongjuan) Tea Salt Total
Figure 8.1 Lijin revenue, 1858–1908

Source: Zhou Yumin 周育民, “晚清厘金歷年全國總收入的再估計” (A Recalculation
of the Total Annual Lijin Revenue during the Late Qing), Qingshi Yanjiu 3 (2011), 1–24
characterized the Qing ideal.26 The potential of the lijin duties to finance
provincial fiscal autonomy would also make them a favorite target of aboli-
tion for all centralizing movements in the twentieth century.
The rise of autonomous provincial finances has been interpreted as a sign
of a devolution of power. However, as Hon-wai Ho has argued, the growth
of regionalism was not a zero-sum game, as long as total revenues were
growing and the center continued to negotiate sharing arrangements.27
Marianne Bastid speaks of a “deconcentration” of an already decentralized
fiscal system.28 It is true that the growing influence of provincial governors
forced Beijing to negotiate, sometimes tenaciously, to restore and maintain
the traditional remittances to the capital (jingxiang 京餉) or poorer provinces
26
Iwai, 中国近世財政史の研究, pp. 128–37; Yang Mei 楊梅, 晚清中央與地方財政關
係研究：以厘金為中心 (A Study of the Central and Local Fiscal Relationship in Late
Qing China: With a Focus on Lijin) (Beijing, Zhishi chanquan chubanshe) pp. 84–94; Shi
and Xu, 晚清財政 : 1851–1894, pp. 260–1.
27
S. Spector and F.H. Michael, “Regionalism in Nineteenth-Century China,” in
S. Spector, Li Hung-Chang and the Huai Army: A Study in Nineteenth-Century Chinese
Regionalism (Seattle, University of Washington Press, 1964), pp. 23–43; Ho Hon-wai
何漢威, “從清末剛毅、鐵良南巡看中央和地方的財政關係” (Late Ch’ing Center–
Province Fiscal Relations as Seen in the Imperial Missions of Kang-i and T’ieh-liang in
1899 and 1904), Bulletin of the Institute of History and Philology, Academia Sinica 68 (March
1997), 55–115.
28
M. Bastid, “The Structure of the Financial Institutions of the State in the Late Qing,” in
S.R. Schram (ed.), The Scope of State Power in China (Hong Kong, Chinese University of
Hong Kong, 1985), pp. 51–79.
251

(xiexiang 協餉), and that the provincial auditing reports (zouxiao 奏銷) were
more fiction than real. The Board of Revenue never gained complete know-
ledge of provincial finances despite a cautious attempt at accounting reform
in 1884. Nonetheless, Beijing’s revenues continued to improve, at least until
the Sino-Japanese War, as the center managed to mobilize an increasing share
of both old and new provincial revenues. In addition to a fixed quota of
8 million (later 10 million) taels of traditional jingxiang remittances for the
maintenance of the capital, the provinces were also allotted payments to so-
called “dedicated funds” (zhuanxiang jingfei 專項經費). These included
defense-related funds (for the capital, the northeastern border and the mod-
ern navy), augmentations of the metropolitan salaries, subsidies for the
imperial household, and allocations for new government expenditures (the
Foreign Legations Fund established in 1876 and the Railway Fund established
in 1889). According to Zhou Yumin, total assignment orders to the central
government in 1898 amounted to 22 million taels, about a quarter of rev-
enues, even as not all of it actually arrived. Much of these funds came from
“provincial” lijin revenues.29
Customs revenue, on the other hand, was considered central government
revenue, but frequently also financed projects in the provinces, especially those
related to defense and military industrialization. The newer of the two customs
services, the MCS, goes back to 1854 when foreign merchants in Shanghai,
opened as a treaty port by the Nanjing Treaty of 1842, came to self-organize
their duty payments to the customs tollhouse during a local rebellion. The model
was regularized and extended to the other treaty ports (increased from five to
sixteen) after 1861. From 1864, Inspector General Robert Hart resided in Beijing.
Each treaty port was furnished with a commissioner and a staff of mostly foreign
career bureaucrats. Notwithstanding its foreignness and “semi-colonial” charac-
teristics, the MCS was not yet the imperium in imperio that would mark the
Beiyang and warlord periods after the 1911 Revolution.30 The foreign customs
commissioners in the treaty ports served as middlemen between the foreign
merchants, their consuls, and the Chinese customs superintendents, a “frontier
regime,” in the words of Van de Ven.31 However, under the Qing the funds were
still stored in traditional Chinese customs banks, albeit under separate accounts,
and controlled by superintendents embedded in the usual bureaucratic
hierarchies.
29
Zhou, 晚清財政與社會變遷, pp. 242–4, 370–1. Cf. Iwai, 中国近世財政史の研究, p. 90.
30
Cf. Dai Yifeng 戴一峰, 近代中國海關與中國財政 (The Maritime Customs and Fiscal
Policies in Modern China) (Xiamen, Xiamen daxue chubanshe, 1993), pp. 28–36.
31
Van de Ven, Breaking with the Past, p. 65.
252

Public Finance
EMPEROR
Board of Revenue Foreign Office/

Board of Customs
Control (after 1906)
Governor, governor-
general I.G. of customs
Customs Customs
superintendent commissioner
Chinese-built ship Foreign-built ship
Figure 8.2 Foreign and Native Customs

Source: Ren Zhiyong 任智勇, 晚清海關再研究：以二元體制為中心 (A New Study of
the Late Qing Maritime Customs: With a Focus on the Dual System) (Beijing, Zhongguo
renmin daxue chubanshe, 2012), p. 88. Dotted line: tax assessment (down) and tax payment
(up); solid line: hierarchical relationship, report (up), order (down); dashed line: equal
relationship
Because Maritime Customs funds were treated as an extension of the

Native Customs system, they were at the disposal of the imperial govern-
ment and served as a counterbalance to the greater independence of the
provinces. The insertion of a foreign institution changed information flows
more than revenue flows: the Native Customs continued to operate accord-
ing to the quota system, with extra-quota income spent locally or ending in
the pockets of the collectors. In contrast, fixed quotas were abandoned for the
“foreign tax” (yangshui 洋稅), and the expansion of revenue, tightly con-
trolled by modern statistical methods, was allowed to follow the growth (and
occasionally contraction) of trade. Foreign supervision of the Maritime
Customs revenue thus performed an important function of checking the
253

acquisitive habits of the Chinese tax collectors and solving the information
asymmetry between the imperial center and regional fiscal extraction.32
The fact that the revenue under MCS supervision was subject to statistics
rather than fixed quotas, and thus protected from revenue leakage, made it an
ideal security for China’s indemnities and foreign loans. One of the service’s
first tasks after its regularization in 1861 had been to oversee the collection of
revenue to pay the indemnities for the Second Opium War. In 1867, when
Zuo Zongtang borrowed the first out of six war loans totaling 15.95 million
taels to recover the northwestern provinces and Xinjiang from Muslim rebels,
the Zongli Yamen (Foreign Office) and Inspector General of Customs Robert
Hart devised a formal process that required imperial endorsement for all
foreign loans secured with customs revenue, which thus became a sovereign
debt. This strengthened the position of the central government, as governors
could no longer borrow without central approval.33 However, the close
association of the central government with customs revenue and foreign
loans created a trajectory of deepening dependence well into the 1930s.
By 1894, the Qing had largely recovered from the civil wars. The land tax
had been restored to its prewar quotas, and the central government had
a reasonably firm grip (by Qing standards) on provincial resources, was
almost debt-free, and, officially at least, had no deficits. Steven Halsey has
argued that Qing China had emerged as a military–fiscal state able to mobil-
ize indirect taxes for national defense. Compared to Brewer’s ideal-type
British fiscal–military state a century earlier, the transformation of the
Chinese state was far from complete. It lacked not only tools of sustainable
public deficit finance but also a professional tax administration in the
Weberian sense, with the exception of the foreign-staffed MCS, which in
1893 contributed about 25 percent to the total budget. Strictly speaking, the
central government did not even have the power of the purse as the
decentralized fiscal system delegated revenue extraction to the provinces.34
32
Jamieson, The Revenue and Expenditure of the Chinese Empire, pp. 3, 6, 18–20; Dai, 近代中
國海關與中國財政, pp. 94–7; Ren, 晚清海關再研究, pp. 133–5.
33
Ma Jinhua 馬金華, 外債與晚清政局 (Foreign Loans and Late Qing Politics) (Beijing,
Shehui kexue wenxian chubanshe, 2011), p. 73; Hamashita Takeshi 濱下武志, 中国近
代経済研究：清末海関財政と開港場市場圈 (Economic History of Modern
China: Maritime Customs Finance and Open Port Market Zones in Late Qing
China) (Tokyo: Institute of Oriental Culture, Tokyo University, 1989), pp. 71–2;
Z. Ren, 晚清海關再研究, pp. 117–35.
34
S.R. Halsey, Quest for Power: European Imperialism and the Making of Chinese Statecraft
(Cambridge, MA, Harvard University Press, 2015), pp. 81–112; J. Brewer, The Sinews of
Power: War, Money, and the English State, 1688–1783 (Cambridge, MA, Harvard University
Press, 1990), pp. xiii, 75; He, Paths toward the Modern Fiscal State.
254

Public Finance
It is, therefore, useful to follow Ja Ian Chong’s suggestion to distinguish

authority from capacity.35 The imperial government in Beijing had authority,
through the power to appoint officials and as the symbolic center of the
empire. But it had the capacity neither to actually know the full extent of
provincial finances nor to force the governors to follow orders. The erosion
of the dynasty’s authority began when the indemnities of two military defeats
forced the center to encroach upon provincial fiscal interests and unhinged
the balance that had emerged from the dual system of trade taxation.
The Success and Failure of Qing Centralization,

1895–1911
The defeats in 1895 and 1900 pushed the Qing back into a fiscal predicament.
The war loans and indemnities of the Sino-Japanese War left the empire with
annual debt payments of roughly 24 million taels out of a total revenue of
89 million. The Boxer indemnities six years later brought annual payments to
42 million taels.36 The crisis generated some appreciation for Western ideas of
governance. In 1894 and 1898, the Board of Revenue experimented with
domestic bonds, but the bonds were soon converted into old-style contribu-
tions to be repaid with rank and office instead of principal and interest. The
name “trust bonds” (zhaoxin gupiao 昭信股票) did as much to destroy trust in
the government as their failure itself.37 On the eve of the Boxer Rebellion, the
board for the first time produced a rudimentary budget which anticipated
future revenue and differentiated between central and provincial finances.
The budget revealed substantial growth during the postwar years but also
a large deficit (see Table 8.1), even though these numbers do not include the
shares allocated to the provinces to repay the various war loans.38 In response
to the deficit, the throne sent an imperial commissary to the provinces;
35
J.I. Chong, “Breaking Up Is Hard to Do: Foreign Intervention and the Limiting of
Fragmentation in the Late Qing and Early Republic, 1893–1922,” Twentieth-Century China
35.1 (November 2009), 75–98.
36
Liang Qichao 梁啟超, 中國國債史 (A History of Chinese National Debts), in Liang, 飲
冰室合集 (Collected Works from the Ice Drinker Studio), vol. 6 (ed. Lin Zhijun 林志
鈞) (Beijing, Zhonghua shuju, 1989), pp. 5967–8.
37
Li Wenjie 李文杰, “息借商款與晚清財政” (The “Xijie Shangkuan” Bond and Late
Qing Fiscal Policies), Lishi Yangjiu 1 (2018), 68–86; Iwo Amelung 阿梅龍, “國債概念的
接受和中國早期發行之公債” (The Concept of Government Debt and the Earliest
Public Debt in China), in I. Amelung 阿梅龍, 真實與建構: 中國近代史及科技史新
探 (Truth and Construct: New Explorations on the History of Science in Modern
China) (Beijing, Shehui kexue wenxian chubanshe, 2019), pp. 274–90.
38
Peng Yuxin 彭雨新, “清末中央與各省財政關系” (Central and Provincial Fiscal
Relationships in the Late Qing), Shehui kexue zazhi (Peking) 1 (1937), 83–110.
255

Table 8.1 Budget for the 26th Year of Guangxu (1900) in kuping taels
Revenue Expenditure Balance
Central government 18,764,600 23,036,200 −4,271,600

Provinces 79,500,000 92,000,000 −12,500,000
Total 98,264,600 115,036,200 −16,771,600
Source: Luo Yudong 羅玉東, “光緒朝補救財政之方策” (Strategies to Salvage

Government Finance in the Guangxu Period), Zhongguo Jindai Jingjishi Yanjiu Jikan
中國近代經濟史研究集刊 1.2 (May 1933), 216
however, the goal was not to reform provincial institutions but to negotiate
higher remittance quotas to Beijing.39
More ambitious New Policy reforms began in the provinces after the Boxer
Rebellion, where they drove up the costs of government. Left alone to carry the
double burden of the indemnities and the costly modernization of schooling
and policing, the provinces broke all traditional limitations on raising taxes and
invented countless “miscellaneous imposts,” including surcharges on old taxes
and new excises. Provincial governments imported machines to mint silver and
copper coins, established banks, and issued paper money or government bonds
(see the chapter by Li and Yan in this volume).40 They also began a gradual
process of intra-provincial centralization by reorganizing their diverse and
poorly co-ordinated fiscal bureaus into a hierarchical structure.41
Fiscal reorganization at the center did not start before a government reform
commission traveled to Japan, Europe, and the USA in 1905–1906. On
September 1, 1906, an imperial edict announced the plan to institute
a constitutional monarchy. The Board of Revenue changed its name to
Ministry of Finance (duzhibu 度支部) – literally “ministry in control of expend-
iture,” symbolizing the departure from the traditional focus on fixed-quota
39
Ho, “從清末剛毅”; Liu Zenghe 劉增合, “財” 與 “政” ：清季財政改制研究 (“Finance”
and “Politics”: A Study of Late Qing Fiscal Reforms) (Beijing, Sanlian shudian, 2014), pp.
83–6.
40
Wang Shu-hwai 王樹槐, 庚子賠款 (The Boxer Indemnities) (Taipei, Institute of
Modern History, Academia Sinica, 1974), Chapter 2; Hamashita, Economic History
of Modern China, pp. 58–64, 166–70; Ho Hon-wai 何漢威, “從銀賤錢荒到銅元泛
濫：清末新貨幣的發行及其影響” (From Silver Glut and Coin Shortage to Surplus
of Copper Coins: The Issuance of New Currencies during the Late Qing Period and Its
Effects), Bulletin of the Institute of History and Philology, Academia Sinica 62.3 (1993),
389–494; Pan Guoqi 潘國旗, 近代中國國內公債研究 (1840–1926) (A Study of Public
Bonds in Modern China (1840–1926)) (Beijing, Jingji kexue chubanshe, 2007).
41
Liu, “財”與“政”, pp. 352–66.
256

Public Finance
Table 8.2 National revenue for 1911 (Ministry of Finance)

Source In kuping taels Percentage of total
Land tax 48,101,346 16.2

Public enterprises 46,600,899 15.7
Salt tax 46,312,355 15.6
Lijin duties 43,187,907 14.5
Miscellaneous incomes 35,244,750 11.9
Maritime Customs 35,139,918 11.8
Miscellaneous taxes 26,163,842 8.8
Native Customs 6,999,370 2.4
Contributions 5,652,333 1.9
Domestic debt 3,560,000 1.2
Total national revenue 296,962,720
Source: H. Ho, “A Final Attempt at Financial Centralisation in the Late

Qing Period, 1909–11,” Papers on Far Eastern History 32 (September 1985), 53
revenues – and converted its subdivisions from mixed territorial-cum-functional

into modern functional departments. In 1907, a statistical department was added.
The nine-year schedule for constitutional preparation promulgated in 1908 stipu-
lated the clearing (qingli 清理) of provincial finances in 1909–1910.42 The ministry
dispatched fiscal supervisors (jianliguan 監理官) to establish a hierarchical chain
of command between the ministry and the provincial treasurers, with the goal of
limiting the autonomy of the governors. While this was not entirely successful,
the investigation managed to uncover local fees and extra-account (waixiao)
revenues and forced the provinces to accelerate the centralization of their fiscal
bureaucracies. In 1910, the provinces submitted their reports to the Ministry of
Finance, which published the first national budget for 1911 (Table 8.2).43
The budget revealed a national revenue of almost 297 million taels, four
times the statutory revenue of 71 million taels at the beginning of the
nineteenth century. The importance of foreign customs revenue had been
eclipsed by salt taxes and income from public enterprises (industries, tele-
graphs, mines, mints, etc.). There still was a deficit of 42 million taels which
42
R.S. Horowitz, “Breaking the Bonds of Precedent: The 1905–6 Government Reform
Commission and the Remaking of the Qing Central State,” Modern Asian Studies 37.4
(October 2003), 775–97; N. Meienberger, The Emergence of Constitutional Government in
China (1905–1908) (Bern: Peter Lang, 1980), pp. 42–5; M.E. Cameron, The Reform
Movement in China, 1898–1912 (New York, AMS Press, 1974), pp. 103–4; Chen Feng
陳鋒 and Cai Guobin 蔡國斌, 清代財政史 (A Fiscal History of the Qing Dynasty)
(Changsha, Hunan renmin chubanshe, 2013), pp. 171–6.
43
Liu, “財” 與 “政”, pp. 142–98, 373–8; see also Zhou, 晚清財政與社會變遷, pp. 428–9.
257

the National Assembly – newly convened in 1910 as a consultative organ to the

government – converted (on paper at least) into a surplus mainly by cutting
the budgets of the Ministries of War and of Post and Communications. But the
major unresolved question in the budget was the division between the central,
provincial, and local revenues. The debate over how to share tax revenues had
been going on in the provinces since 1906. Many favored the Japanese model,
which distinguished between “national” (guojia 國家) and “regional/local”
(difang 地方) taxes, but no agreement was reached as to how to fit the province
into this dichotomy. The budget subsumed all revenues under central control,
making them “national taxes,” and kept 62 percent of expenditures at the disposal
of the Beijing ministries, the army, and the court; only 21 percent went to the
provinces for their expenses (not counting the 37 million taels that were added to
the total budget as expenditures of local governments). This in fact eliminated
the provinces as autonomous fiscal agents. In reality, the Beijing government did
not have the effective power to control the provincial budgets.44 Before a similar
budget for 1912 could pass the Assembly in the fall of 1911, the revolution upended
the dynasty.45
By the summer of 1911, the Qing still appeared relatively successful. The
provinces had substantially increased their extractive capacities, while the center
seemed to have more knowledge and receive a larger share of the fiscal bounty.
What went wrong? Arguably, the seeds for the secession of the provinces and the
collapse of the Qing dynasty were sown early on in the division between lijin and
customs duties and the dependence of the central government on the foreign-
managed MCS. The conflict of interests between the center and the provinces
was exacerbated by the war indemnities, which eroded the authority of the
imperial government faster than it was able to increase its capacity to rule.
The relationship between the MCS and the Qing government grew closer
after 1867, when all customs-secured loans required central government
sanction. In the same year, the responsibilities of the MCS were expanded
to include the collection of duties from all steamships, Chinese or foreign-
owned, while non-treaty ports were closed to steamship traffic. By legalizing
the practice of lijin evasion by Chinese merchants using foreign-flagged
steamships, this provision increased the share of revenue that the central
government could control through the MCS. The Chefoo Convention of 1876
44
Ho, “A Final Attempt at Financial Centralisation in the Late Qing Period, 1909–11,”
p. 38; Liu, “財” 與 “政”, pp. 226–30, 296–347. The remaining 17 percent went on
servicing debt.
45
“前清宣統四年全國歲入歲出總預算” (Annual Revenues and Expenditures for 1912),
Yinhang Zhoubao 銀行週報 (Bank Weekly) 26.27–8 (1942), 24–8.
258

Public Finance
(ratified only in 1887) not only opened additional treaty ports to expand steamer
shipping, but also brought the lijin on foreign opium under the MCS,46 thus
depriving the provinces of more extra-account revenue (see Figure 8.2).
The indemnities for the Sino-Japanese War and the Boxer Rebellion forced
the Qing government to bring more of the provincial revenues under central
control, without benefiting from these revenues to enhance its own capacity
to govern. Only the former involved a formal loan issued as bearer bonds in
the markets. The Treaty of Shimonoseki required China to pay 200 million
taels in indemnities and 30 million for Japan’s withdrawal from Weihaiwei
within seven years at 5 percent interest. In order to facilitate interest-free
early repayment within three years, the Qing government borrowed three
gold loans each of ₤16 million (in 1895, 1896, and 1898) from Russo-French and
Anglo-German banking consortia at interest rates between 4 and 5 percent
payable within thirty-six to forty-five years (the last instalment would be paid
in 1943). Borrowing bought time but increased effective costs due to loan
discounts and exchange rate losses (known as bangkui 鎊虧, “sterling loss”).
Worse were the political costs, since the lending powers seized concessions
and leases for themselves.47
The indemnities expanded the authority of the MCS, which evolved into “a
debt-collection agency for foreign financial interests,” and thus “a typical ‘infor-
mal empire’ institution,” in Van de Ven’s words.48 The contract for the second
Anglo-German loan stipulated that the MCS should take over the revenue of
seven lijin collectorates as security.49 While Inspector General Robert Hart
refused to take out new loans to pay the Boxer indemnities, his MCS took
charge of the revenues of the Native Customs stations located within fifty li
(roughly eighteen miles) of the treaty ports.50 Provincial resistance to the loss of
vital revenues destabilized the already vulnerable revenue-sharing system.
46
A. Reinhardt, “Treaty Ports as Shipping Infrastructure,” in R.A. Bickers and I. Jackson
(eds.), Treaty Ports in Modern China: Law, Land and Power (London, Routledge, 2016), pp.
104–5.
47
Liang, 中國國債史, pp. 5943–54. See also A.G. Coons, The Foreign Public Debt of China
(Philadelphia, University of Pennsylvania Press, 1930), p. 12; Pan, 近代中國國內公債
研究, pp. 85–91.
48
Van de Ven, Breaking with the Past, pp. 9, 134.
49
F.H.H. King, C.E. King, and D.J.S. King, The Hongkong Bank between the Wars and the
Bank Interned, 1919–1945, vol. 2 (Cambridge, Cambridge University Press, 1988), pp. 287–
8; Van de Ven, Breaking with the Past, p. 142; Hamashita, Economic History of Modern
China, p. 53.
50
R. Hart, “Memorandum Concerning the Indemnity to Be Paid by China, 25th
March 1901,” in Maritime Customs (ed.), Documents Illustrative of the Origin,
Development, and Activities of the Chinese Customs Service, vol. I: Despatches, Letters,
Memoranda, etc. 1842–1901 (Shanghai, Statistical Department of the I.G. of Customs,
1938), pp. 604–10; Liang, 中國國債史, pp. 5967–8; W. Tsai, “The Inspector General’s
259

The Boxer indemnities were crucial in changing the balance between the
center and the provinces. During the Rebellion, the southeastern governors
had maintained neutrality while foreign powers invaded the capital. They
saw the indemnities as a responsibility of the central government and mostly
refused to mobilize more lijin duties for their repayment. Instead, almost
a third came from salt price surcharges (yanjin jiajia 鹽斤加價), i.e. a surtax
on the fixed sales price of salt in the monopoly trade, and one-fifth from
surcharges on the land tax, both revenues of the central government.
A further fifth came from new excises and imposts, which became a pretext
also to greatly expand provincial and local levies. The Boxer indemnities thus
opened the floodgates for the chaotic proliferations of taxes and surcharges
that would form the basis of warlord finance in Republican China and make
fiscal standardization difficult (see below).51
Loath to lose the mainstay of their fiscal independence, the provinces also
resisted the Mackay Treaty of 1902, which replaced multiple lijin toll stations
with a single excise of 7.5 percent ad valorem in exchange for the powers’
consent to raise customs tariffs from 7.5 to 12.5 percent. Only seven provinces,
mostly poorer ones with little lijin revenue to lose, implemented the provin-
cial “consolidated tax” (tongjuan 統捐 or tongshui 統稅).52 As a result, the
treaty was never ratified. The “consolidated tax” was more successful as
a replacement for the opium lijin, but it was soon obliterated by the political
will to abolish the vice of opium. The imposts offered by Beijing to make up
the loss, another salt price surcharge and a newly created stamp duty
(yinhuashui 印花稅), covered less than 20 percent of the estimated
28 million taels in lost revenue, while further alienating the provinces.53
The straw that broke the camel’s back was the imperial government’s
belated attempt to nationalize railway development (see the chapter by Köll
in this volume). Absent guidance from Beijing, provincial leaders had been at
Last Prize: The Chinese Native Customs Service, 1901–31,” Journal of Imperial and
Commonwealth History 36.2 (June 2008), 243–58.
51
Chong, “Breaking Up Is Hard to Do,” 75–98; Hamashita, Economic History of Modern
China, pp. 52–3, 85–7; Wang, 庚子賠款, Table 26, p. 151; Zhou, 晚清財政與社會變遷,
pp. 368–70.
52
Luo, 中國釐金史, pp. 137, 53–4; Yang, 晚清中央與地方財政關係研究, pp. 67, 72,
241–4; Zhou, 晚清財政與社會變遷, pp. 365–6; D. Wang, China’s Unequal Treaties:
Narrating National History (Lanham, Lexington Books, 2005), pp. 19–22.
53
Liu Zenghe 劉增合, 鴉片稅收與清末新政 (Opium Tax Revenue and Late Qing New
Policy Reforms) (Beijing, Sanlian shudian, 2005), pp. 142–7, 181–264; Ho Hon-wai 何漢威,
“清季國產鴉片的統捐與統稅” (The Consolidated Tax on Native Opium during the
Qing Period), in 薪火集：傳統與近代變遷中的中國經濟 (Passing on the Torch:
China’s Economy from Tradition to Modern Transformation) (Taipei, Daoxiang chu-
banshe, 2001), p. 586 and passim.
260

Public Finance
the helm of railway nationalism in China since the 1890s. The redemption of
the American concession for the Canton–Hankou railway and the favorable
loan terms for the Tianjin–Pukou railway (the “Pukou terms”) became
landmark cases towards railway sovereignty. Many provinces made plans
for railway construction and mobilized patriotic sentiment to raise native
funds. By 1911, however, the Ministry of Communication decided to nation-
alize these regional lines with the help of a loan from a four-power banking
consortium (Germany, Britain, France, the USA) secured by lijin and salt
revenues of Hubei and Hunan (hence “Huguang loan”). The move met fierce
resistance, which became a major trigger of the 1911 Revolution.54
Beijing Finance under Warlord Regimes, 1912–1926

That the newly founded Republic of China eventually ended up in the hands
of provincial warlords shows that fiscal–military integration had been more
successful on the provincial than on the national level. This made it difficult
for any new national regime to establish a secure fiscal basis. During the four
decades following the revolution, no regime was able to hold all of China’s
territory. Any contender for central government was forced to renegotiate
the allocation of revenues from more or less independent regional powers.
During the Beiyang and warlord eras (1912–1927), budgetary imagination and
fiscal reality diverged widely. The land tax, though still in the budget, hardly
ever arrived in Beijing and was formally yielded to the provinces by the
Nationalists in 1927 (see below). Foreign-managed customs and salt tax
administrations secured the most reliable revenue sources for the center.
Thanks to securities provided by these institutions, domestic debt flourished
despite political instability, and borrowing (including foreign loans) maintained
a level of over 20 percent of revenue until the Sino-Japanese War (Table 8.3).
Since, during the revolution, most provinces had cut their remittances, the
ability to borrow became vital for the young republic. After Yuan Shikai
replaced revolutionary leader Sun Yat-sen as provisional president in
March 1912 and moved the capital from Nanjing to Beijing, his government
instantly began negotiations with the Huguang loan banks and in April 1913
signed a £25 million reorganization loan agreement with a regrouped five-
power consortium. More than two-thirds of the loan sum was retained for
debt service, reparations to foreigners harmed during the unrest, and the
54
E. Köll, Railroads and the Transformation of China (Cambridge, MA, Harvard University
Press, 2019); E. Kaske, “Taxation, Trust, and Government Debt: State–Elite Relations in
Sichuan, 1850–1911,” Modern China 45.3 (May 1, 2019), 239–94.
261

Table 8.3 Revenue structure of the central government, 1913–1945 (%)

Salt Agricultural Taxes on industry Government
Customs taxes taxes and commerce enterprises Debt
1913–1919 15.1 17.2 17.7 7.6 – 20.6

1928–1932 44.8 18.5 – 9.0 0.6 20.2
1933–1937 31.2 16.6 – 11.6 3.7 23.6
1938–1945 3.1 3.0 3.6 2.6 3.7 78.5
Source: Kubo Tōru 久保亨, Kajima Jun 加島潤, and Kigoshi Yoshinori 木越義則
(eds.), 統計でみる中国近現代経済史 (Chinese Modern and Contemporary
Economic History as Seen in Statistics) (Tokyo, Tōkyō daigaku shuppankai, 2016),
p. 162
reform of the salt tax administration which served as security.55 The loan was
contentious from the beginning. Opposition by the Nationalist Party contrib-
uted to a breakup of the volatile coalition and the “second revolution.” It has
also been seen as symbolic of China’s loss of sovereignty. In fact, Yuan’s
negotiators were quite successful in minimizing foreign political influence
while turning foreign support and diplomatic recognition to the advantage of
his government.56
As part of the loan terms, Yuan established a Sino-Foreign Salt
Inspectorate with foreign deputies at the chief and district inspectorates
as well as audit offices. Reforms under the British chief inspector, Sir
Richard Morris Dane, moved tax collection to the site of production,
created a professional corps of administrators, improved tax assessment,
lowered collection costs, and centralized tax receipts and statistics. This
brought a greater share of salt revenues under the control of the central
government, making the salt tax into the largest revenue stream of the
government in Beijing ahead of customs duties. Similarly to the Maritime
Customs Service, the Sino-Foreign Salt Inspectorate managed to insulate
itself from the political chaos of the early Republican era and survived as
55
Jia Shiyi 賈士毅, “五十年來中國之財政” (The Fiscal Policy of China during the Last
Fifty Years), in Liang Qichao 梁啟超 (ed.), 晚清五十年來之中國 (China during the
Fifty Years since the Late Qing) (Hong Kong, Longmen shudian, 1922), pp. 107–27.
56
Lin May-li 林美莉, “善後大借款析論：民國財政的奠基與民族主義的激蕩” (An
Analysis of the Reorganization Loan: Fiscal Foundation of the Republic of China and
Catalyst of Nationalism), in Wang Jianlang 王建朗 and Huang Ko-wu 黃克武 (eds.),
兩岸新編中國近代史 (New Studies on the History of Modern China from Both Sides
of the Taiwan Strait) (Beijing, Shehui kexue wenxian chubanshe, 2016), pp. 597–626.
262

Public Finance
an institution, even as revenues were frequently seized by warlord

regimes after 1917.57
The importance of the MCS as revenue agency for the central government
increased after the revolution, not least because it shook off the control of the
Chinese superintendent.58 In 1913, Yuan established a National Loans Bureau
with Inspector General Francis Aglen as vice chair, expanding the activities of
the MCS from securing foreign debt to supporting a market for domestic
bonds. The successful floating of domestic debt was an important novelty of
the Republican era, especially for the central government. Shortly before the
revolution, the Qing had issued “Patriotic Bonds” (aiguo gongzhai 愛國公債)
and included some 3.5 million taels of public debt as a legitimate revenue
in the budget of 1911 (see Table 8.3). With the debacle of the 1898 “trust
bonds” still in everyone’s memory, subscriptions to the bonds were
almost exclusively involuntary payments by government officials and
Manchu nobles. Nonetheless, Yuan Shikai’s government assumed respon-
sibility for their repayment, and issued a number of new public bonds. The
fact that the MCS stepped in to provide security greatly enhanced their
popularity.59
Yuan Shikai’s regime was reasonably effective in restoring fiscal authority.
The government produced a modern budget. Customs and salt surplus
(guanyu 關餘 and yanyu 鹽餘), i.e. the income from customs and salt taxes
after servicing China’s foreign debt, provided a moderately stable income.
Government bonds were oversubscribed. Silver coins bearing President
Yuan’s image were hugely popular and facilitated monetary integration.60
Beijing also managed to persuade the provinces to provide fiscal information
and remit revenue quotas. While the postrevolutionary coalition govern-
ment had adopted Sun Yat-sen’s idea of “distributed powers” (junquan 均權)
and divided all tax revenues into “national” and “regional/local” taxes,
57
S.A.M. Adshead, Modernization of the Chinese Salt Administration, 1900–1920 (Cambridge,
MA, Harvard University Press, 1970); J.C. Strauss, Strong Institutions in Weak Polities:
State Building in Republican China, 1927–1940 (Oxford and New York, Clarendon Press,
1998), Chapter 3.
58
D. Ma, “The Rise of a Financial Revolution in Republican China in 1900–1937: An
Institutional Narrative,” Australian Economic History Review 59.3 (November 2019),
242–62.
59
Pan, 近代中國國內公債研究, p. 115; Zhang Guisu 張桂素 (ed.), “宣統年間發行愛
國公債史料” (Historial Material on Issuing the “Patriotic Government Bonds” during
the Xuantong Reign), Lishi dang’an 4 (1997), pp. 67–76; Van de Ven, Breaking with the
Past, Chapter 5.
60
A. Kuroda, “The Collapse of the Chinese Imperial Monetary System,” in K. Sugihara
(ed.), Japan, China, and the Growth of the Asian International Economy, 1850–1949 (Oxford,
Oxford University Press, 2005), pp. 103–26.
263

Yuan’s increasingly authoritarian regime turned to centralization. Beijing

directly appointed the heads of the provincial finance departments, restored
the Qing remittance system from general tax revenue, and additionally
designated five exclusive imposts for the central government (zhongyang
zhuankuan 中央專款), namely the tobacco and liquor tax, the stamp duty,
the contract enforcement fee, the tobacco and liquor license fee, and the
brokerage fee.61 In 1914, Yuan’s Ministry of Finance made another attempt to
abolish the lijin transit duties, which had been designated “national” tax in
1912. In order to signal to the international community China’s willingness to
ratify the Mackay Treaty, Jiangsu became a test case to replace its multiple
lijin stations with a single excise of 7.5 percent levied at the place of produc-
tion (chuchang shui 出廠稅) and sale (xiaochang shui 銷場稅). Other
provinces followed reluctantly, but readjusted their tax assessment and
used tax farming in order to offset the steep fall in revenue seen in Jiangsu.
In February 1916, the Ministry of Finance ordered all provinces to replace lijin
with the consolidated tax (tongjuan).62
These advances were upended by Yuan Shikai’s ill-fated self-coronation
and soon by his death in 1916. The provinces seceded, and regional warlords
became largely independent rulers of their territories. They not only stopped
remittances of tax revenues, but also seized large parts of the salt tax,
especially after 1920. Several salt-secured loans, including the Huguang loan
of 1911, went into default. The reorganization loan remained secure, because
it had been placed under Customs security in 1917, when the German and
Austro-Hungarian shares of the Boxer indemnities were canceled. The
drying up of sources for foreign debt during World War I strengthened
demand for domestic debt. As the chapter by Sheehan and Zhu in this
volume shows, the relationship between the government and the banking
sector grew much closer after the war, and the stock exchanges emerging
after 1918 increasingly relied on government bonds. In 1921, the MCS helped
Finance Minister Zhou Ziqi to design the Consolidated Debt Service, which
combined the customs-secured bonds with unsecured bonds, including the
Late Qing Patriotic Bond, and brought them under the combined security of
the Customs, the Salt Tax Administration, and the alcohol and tobacco
monopolies. The insulation of the service from onslaughts by the changing
Beijing governments made the consolidated bonds very secure and greatly
61
Jiao Jianhua 焦建華, 中華民國財政史 (A Fiscal History of the Republic of China)
(Changsha, Hubei renmin chubanshe, 2015), pp. 141–2.
62
Lin May-li 林美莉, 西洋稅制在近代中國的發展 (The Development of Western
Taxation Methods in China) (Taipei, Academia Sinica, 2005), pp. 98–103.
264

Public Finance
enlarged the capital reserves of China’s banks.63 This only deepened the
dependence of whoever controlled the central government on the MCS,
even though very little of the customs revenue was directly available.
Between 1897 and 1910, the Qing government had used 30 to 40 percent of
the annual customs revenue for military expenditures, and 20 to 30 percent
for indemnities and foreign loans.64 The Beiyang government already used
three-quarters of customs revenue for debt service. Only 2 percent was
remitted to the government as Customs surplus and became a target of
competition by various regimes.65 From 1912 to 1928, the leadership of the
Ministry of Finance changed thirty-two times (although several ministers
served more than once), with tenures from less than a month to one year.
Often, the main qualification for the job was the ability to borrow. The only
budget drafted after 1915 was that of 1919 when an attempt was made to
appease the warring factions, but it remained little more than fiction. By 1922,
state bankruptcy was imminent.66
Following two civil wars between the Zhili and Fengtian cliques,
a semblance of order returned to Beijing in 1924 with the installation of
Duan Qirui as provisional executive of government. Duan established
a Financial Rehabilitation Committee which soon undertook the third
attempt to abolish the lijin transit duties. The committee created formal
bodies on the national and provincial level to investigate the myriad of transit
duties and stationary excises summarized under the term lijin. The plan was
to abolish the former within two years and expand business and consumption
taxes in compensation. The move was predicated on a renewed push for
higher customs tariffs. As one of the victorious countries of World War I,
Chinese representatives had demanded full tariff autonomy and an immedi-
ate rise of import tariffs to 12.5 percent during the Paris Peace Conference of
1919 and the Washington Naval Conference of 1921. In October 1925, the
Beiyang government assembled representatives of twelve nations in Beijing
at a Special Tariff Conference which estimated the lost lijin revenue at
63
Van de Ven, Breaking with the Past, pp. 183–5; A. Feuerwerker, “The Foreign Presence in
China,” in J.K. Fairbank (ed.), The Cambridge History of China, vol. 12, Republican China
1912–1949, part 1 (Cambridge, Cambridge University Press, 1983), p. 192; A.N. Young,
China’s Nation-Building Effort, 1927–1937: The Financial and Economic Record (Stanford, CA,
Hoover Institution Press, 1971), pp. 21, 115–18.
64
Tang Xianglong 湯象龍, 中國近代海關稅收和分配統計 (1861–1910) (Statistics of
Customs Revenues and Their Distribution in Modern China (1861–1910)) (Beijing,
Zhonghua shuju, 1992), pp. 27, 34.
65
Zhang Xianwen 張憲文 (ed.), 中華民國史 (A History of the Republic of China)
(Nanjing, Nanjing daxue chubanshe, 2005), vol. 2, p. 128.
66
Jia, “五十年來中國之財政”, 127; Jiao, 中華民國財政史, pp. 143–50.
265

C$70 million and proposed to replace this income with a customs surtax of
2.5 percent. Even though no formal agreement was reached until the collapse
of Duan’s regime in April 1926, the new Beijing government in February 1927
one-sidedly implemented the 2.5 percent customs surtax on imports, estimat-
ing potential revenue of C$40 million. By this time the Nationalist Party had
just began its Northern Expedition and denied all legitimacy to the
negotiations.67
Ja Ian Chong has argued that foreign intervention was largely responsible
for the persistence of the Beijing government during the period of fragmen-
tation and warlordism. Instead of settling in their respective territorial
footholds, various warlord factions continued to compete for control over
Beijing, because this could gain them diplomatic recognition and political
legitimacy, as well as access to customs and salt tax surpluses and foreign
loans, especially those secured by assets like railroads under the Ministry of
Communications.68 According to Ma Jinhua’s calculation, various Beiyang
regimes borrowed 633 loans from abroad with a total loan sum of
C$1.556 billion. The reorganization loan of 1913 comprised only 16 percent
of the total.69 Dependence on foreign loans was a double-edged sword. Weak
fiscal governance at the center and lack of authority to persuade regional
regimes to remit tax revenues forced each new regime to make an effort to
secure the loans of its predecessor and gain new loans from abroad. Given the
rise of public opinion and nationalism as a social force since the early 1900s,
regional powers could use this dependence on foreign loans to delegitimize
the central government. This happened, for example, with the reorganization
loan of 1913, which was used by Sun Yat-sen to mobilize against Yuan Shikai’s
government, and with the Nishihara loans to Duan Qirui’s regime in
1916–1917, which became one of the triggers of the May Fourth Movement
in 1919 after the case was leaked to the public.
67
Lin, 西洋稅制, pp. 114–19; Kubo Tōru 久保亨, 戦間期中国自立への模索: 関稅通
貨政策と経済発展 (China’s Quest for Sovereignty in the Inter-war Period: Tariff
Policy and Economic Development) (Tokyo: Tōkyō daigaku shuppankai, 1999), p. 23;
Van de Ven, Breaking with the Past, pp. 202–4, 208–9. The Nationalists introduced
a similar customs surtax, called “inland tax” (neidishui 內地稅), but such surtaxes
soon became obsolete when full tariff autonomy was achieved. See below.
68
Chong, “Breaking Up Is Hard to Do,” 89–90.
69
Ma Jinhua 馬金華, 中國外債史 (A History of Foreign Loans in China) (Beijing,
Zhongguo caizheng jingji chubanshe, 2005), pp. 109, 194, 235–7.
266

Public Finance
Nation-Building during the Nanjing Decade,

1927–1938
Nationalism and careful financial planning were crucial for the success of
Chiang Kai-shek’s Northern Expedition. The Central Finance Conference in
Nanjing in June 1927 solicited the support of the leaders of the modern
financial and industrial sectors based in Shanghai (known as the Jiangsu–
Zhejiang group), as well as the provincial leaders of Jiangsu, Zhejiang, Anhui,
Fujian, Guangdong, and Guangxi. In December 1928, a new government was
formed in Nanjing. The new Finance Ministry proved more stable than its
predecessor. Two of the ministers – the American-trained economists Song
Ziwen (T.V. Soong) and Kong Xiangxi (H.H. Kung) – had long tenures and
considerable political influence. Their American training and collaboration
with foreign experts like the Kemmerer commission and the League of
Nations shaped their fundamentally modern outlook. This was reflected in
more ambitious fiscal policies that not only aimed at fiscal survival but also
conceived the developmental state.70 Even though the Nationalists never
regained full control over the provinces and some regions remained out of
their reach altogether, the Nanjing government ameliorated center–province
fiscal relationships by convening several national finance conferences (1927,
1928, 1934, 1941) to renegotiate revenue-sharing arrangements.
Overall, the Nationalist Party’s nation-building process was successful.
Continued reliance on Sino-foreign co-operation in the customs and salt
administrations aside, the Nationalists were committed to modernizing
their tax structure and cutting their dependence on foreign loans. They finally
managed to abolish the lijin duties and replace them with a nationwide
consolidated tax levied on modern machine-produced products, which
became the third pillar of non-borrowed revenue of the central government
(see Table 8.4). They introduced modern direct taxes like income, estate, and
business taxes. They successfully utilized national banks to consolidate
China’s foreign and domestic debt, stabilize government revenue by bond
purchases, and establish a fiat currency to insulate China’s economy from the
effects of international silver price fluctuations during the Great Depression.
The most momentous change, however, was the central government’s
decision to relinquish control of the land tax, which for millennia had been
70
W.C. Kirby, “Engineering China: Birth of the Developmental State, 1928–1937,” in
K. Pomeranz (ed.), The Pacific in the Age of Early Industrialization (Burlington, VT,
Ashgate, 2009), pp. 215–38; P.B. Trescott, “Western Economic Advisers in China,
1900–1949,” Research in the History of Economic Thought and Methodology 28.1 (2010), 1–37.
267

the mainstay of China’s government revenue. This focused Nationalist tax

policies on a more modern, urban-focused tax structure, but would become
a strategic vulnerability once Japan occupied the eastern seaboard.
The reassignment of the land tax was part of a new division between
national and provincial taxes decided at the 1927 Central Finance Conference.
The list of national taxes, i.e. the revenue of the national government, now
included the salt tax, foreign and native customs duties, and tobacco and
alcohol excises, as well as special taxes on cigarettes and kerosene, the mining
tax, stamp duty, and the lijin transit duties until their abolition. Among taxes
designated provincial revenue, commercial excises and levies had tradition-
ally belonged to the provinces. However, it was an unusual step for the
central government to officially surrender the land tax to the provinces, and
thereby give up fiscal control over agricultural production that made up
65 percent of GNP. Scholars have argued that the move only recognized the
political realities established under the various Beiyang regimes and aimed at
winning the support of the regional powers.71 However, the actual fiscal
policies of the Nanjing government were staunchly centralist. The economist
Li Quanshi has explained the contradiction with the legacy of Sun Yat-sen’s
National Development Plan and the constitution of the Nationalist Party,
which both conceived the relationship between center and provinces as one
of “distributed powers.” The central government should only be charged
with affairs of national relevance, while the provinces were ultimately
responsible for local self-government and the provision of public goods.
The handling of the land tax by the Nationalist government enjoyed wide
support among contemporaries.72
However, ceding the land tax to the provinces also made it difficult to
alleviate the heavy burden of the farmers caused by tax practices inherited
from the warlord regimes. In the dual land tax system established during the
late Qing New Policy reforms, both local and national surtaxes (fujiashui
附加稅) were added to the traditional base tax quota (zhengshui 正稅). Even
though the Beiyang government had tried to limit the total sum of surtaxes to
71
J.K. Fairbank 費正清, 劍橋中華民國史 (Cambridge History of the Republic of China),
trans. Zhang Jiangang 章建剛, vol. 1 (Shanghai, Renmin chubanshe, 1992), pp. 115–17;
Iwai, 中国近世財政史の研究, pp. 381–2.
72
Li Quanshi 李權時, “國地財政劃分近況” (The Current Situation of the Fiscal
Division between Central and Local Governments), Xintuo Jikan 1.4 (1936), 1–20;
Zhang Lianhong 張連紅, 整合與互動：民國時期中央與地方財政關係研究
(Integration and Interactions: A Study of the Fiscal Relationship between Central and
Local Government in Republican China) (Nanjing, Nanjing shifan daxue chubanshe,
1999), pp. 169–70.
268

Table 8.4 Income and expenditure of the Nationalist government, 1928–1937. Unit: C$ million (1929–1935 silver yuan, 1936–
1937 fabi 法币)
1928.7– 1929.7– 1930.7– 1931.7– 1932.7– 1933.7– 1934.7– 1935.7– 1936.7–
Item 1929.6 1930.6 1931.6 1932.6 1933.6 1934.6 1935.6 1936.6 1937.6
Balance
Revenue (excl. debt) 333 438 498 553 559 622 745 817 870
Expenditure 413 539 714 683 645 769 941 1073 1167
Deficit 80 101 216 130 86 147 196 256 297
Deficit finance
Bonds, treasury notes 69 91 193 125 26 80 164 148 223
Bank loans, advances 32 10 125 5 86 91 36 128 113
Foreign debt (cotton– 8 25
wheat loan)

Tax revenue
Customs 179 276 313 370 326 352 353 272 379
Salt tax 30 122 150 144 158 177 167 184 197
Consolidated tax 30 41 53 89 80 106 105 135 158
Total 323 462 535 616 587 660 649 624 769
Source: Ma Jinhua 馬金華, 中國外債史 (A History of Foreign Loans in China) (Beijing, Zhongguo caizheng jingji chubanshe, 2005), p. 305
20 percent of the base tax, these regulations were largely ignored. Provinces,
counties, and even villages took an “entrepreneurial stance” towards taxation
and added dozens of surtaxes. Even though each was levied at an incremen-
tally small amount earmarked for specific expenditures, such as education,
river conservancy, military, police, or famine relief, they could add up to as
high as twenty to thirty times the base tax quota.73 One way to reduce the tax
burden was to improve the financial situation of the counties. At the Second
National Finance Conference in 1934, the Ministry of Finance asked the
provinces to drastically reduce the surtaxes in exchange for returning the
licensing fees for tobacco and alcohol producers to provincial control and
sharing 40 percent of the stamp duty (10 percent for the provincial govern-
ment, 30 percent for the county).74 The ministry also urged local govern-
ments to expedite a land survey in order to collect a land value tax on urban
property and make rural land tax rates more equitable. Since a land survey
would have required considerable mobilization of resources, few communi-
ties complied, and the outbreak of the Second Sino-Japanese War further
stalled progress.
The successful abolition of the lijin duties after various failed attempts
since 1902 has been celebrated as a major achievement of Nationalist fiscal
policies. The Nationalists’ approach – first proposed in 1926 – radically
differed from their predecessors’ in that they decoupled the lijin abolition
from the international customs tariff negotiations. Two taxes were desig-
nated to replace the lijin, the consolidated tax for the central government and
the business tax (yingyeshui 營業稅) for the provinces. The First National
Finance Conference convened in Shanghai in July 1928 ratified the plan. With
the new year of 1931, the lijin finally ceased to exist.75 The consolidated tax
(tongshui) expanded the intra-provincial tongjuan of the late Qing and Beiyang
regimes to a national scale. Chinese products were taxed at the place of
73
Zhu Xie 朱偰, “中國田賦問題參考資料：中國附加稅之沿革” (References for the
Land Tax Problem in China: The Evolution of Surtaxes in China), Guoli Zhongyang
Daxue Shehui Kexue Congkan 1.1 (1934), 1–19; Patricia Thornton, “Beneath the Banyan
Tree: Popular Views of Taxation and the State during the Republican and Reform
Eras,” Twentieth-Century China 25.1 (November 1999), 6–7; Zhang, 整合與互動, pp.
117–18.
74
Second Historical Archives (ed.), 中華民國工商稅收史料選編 (Selected Archives on
Industrial and Commercial Taxes in Republican China) (Nanjing, Nanjing daxue
chubanshe, 1996), vol. 1, p. 1168; “財政部河北印花菸酒局特刊” (Special Issue of the
Stamp, Tobacco and Alcohol Bureau of the Hebei Fiscal Bureau) (1935), in Chen Zhanqi
陳湛綺 (ed.), 國家圖書館藏民國稅收稅務檔案史料匯編 (Archival Materials con-
cerning Tax Revenue and Taxation Stored in the Collection of the National Library)
(Beijing, Quanguo tushuguan wenxian suowei fuzhi zhongin, 2008), p. 523.
75
Lin, 西洋稅制, pp. 128–31.
270

Public Finance
production (producers for export could file a tax return); imports at the place
of sale. The first such new-style consolidated tax had been the cigarette tax
started in 1918 in five southeastern provinces. After 1931, this tax was
expanded to cotton, matches, cement, and machine-milled wheat flour and
became known as the “five-product consolidated tax.” By 1936, the Ministry
of Finance had expanded the tax to almost all Chinese provinces; enlarged the
range of products to cured tobacco leaves, foreign liquors, beer, and pure
alcohol; and unified administration under a central Tax Office (shuiwushu
稅務署). In this year, the consolidated tax yielded C$161.580 billion in
revenue, about 60 percent of the lijin transit taxes prior to their abolition,
and contributed 22.11 percent of the national taxes which constituted the
backbone of Nationalist revenue.76
Customs and salt taxes continued to be the most important revenue
sources for the central government, contributing 44.43 percent and 29.81
percent respectively in 1936. As the powers gradually agreed to full tariff
autonomy, the import tariffs were raised in four steps – in 1928, 1930, 1933, and
1934 – to 20 percent. In 1930, the MCS began to assess tariff rates in the virtual
customs gold unit instead of silver (Haiguan tael) to prevent a steep fall in
customs revenue due to the falling silver price during the Great Depression.
Beginning in January 1931, customs revenues were stored in the Bank of
China, Bank of Communications, and Central Bank, before being remitted to
the Hong Kong and Shanghai Bank for the repayment of the loans or
indemnities. In this way, the Chinese government gradually regained full
control of its customs revenues.77 The salt tax was restored to the central
government at the level of 60 percent by 1929 and almost completely by 1936.
During the Northern Expedition, the Nationalists had briefly abolished the
Sino-Foreign Salt Inspectorate as they rallied against the unequal treaties.
However, Finance Minister Song Ziwen quickly restored the service after
1928 under the control of the ministry and absolved of the responsibility to
guarantee and repay foreign debt. Salt tax revenue was directly remitted to
76
Lin, 西洋稅制, pp. 142–4; Guojia Shuiwu Zongju 國家稅務總局 (ed.), 中華民國工商
稅收史：貨物稅卷 (A History of Commercial and Industrial Taxation in Republican
China: Commodity Taxes) (Beijing, Zhongguo caizheng jingji chubanshe, 2001), pp.
166–7.
77
Young, China’s Nation-Building Effort, p. 48; Hamashita, Economic History of Modern
China, pp. 83–4; Dan Guanchu 單冠初, “南京國民政府收復關稅自主權的歷程”
(The History of Reclaiming Tariff Autonomy under the Nanjing Nationalist
Government), Ph.D. thesis, Fudan University, 2003; Caizhengbu Guanshui Zongju 財
政部關稅總局 (ed.), 中華民國海關簡史 (A Short History of the Maritime Customs
Service during the Republican Period) (Taipei, Caizhengbu guanshui zongju, 1995),
pp. 22–6.
271

the Bank of China or the Central Bank and managed by the Ministry of
Finance. Even as the new Salt Law promulgated in May 1931 failed to abolish
the monopoly of the salt trade syndicates inherited from the Qing, successive
reforms managed to rationalize and increase tax rates and improve revenue.78
Enhanced control over customs and salt revenues helped to restore
China’s credit by settling the sovereign loan defaults (among them the
Tianjin–Pukou loan in 1936 and the Huguang loan in 1937). However, the
Nationalist regime did not rely on foreign debt as much as its predecessors
had. By 1928, China’s credit in the international bond markets was shattered
and the costs of borrowing from abroad were rising under the condition of
falling silver prices. This may have been an advantage, given the thorny
political questions raised by foreign debt. As a result, domestic debt became
much more important. Finance Minister Song Ziwen inherited
C$1.347 billion in government bonds. During his tenure from May 1927 to
November 1933, his ministry issued twenty-seven bonds with a total value of
C$1.03 million. Between 1927 and 1937 there was an estimated total of
C$4.503 billion in domestic debt, as against C$876 million in foreign debt.
These bonds were managed by Chinese banks mostly located in Shanghai
and seen as evidence that the Jiangsu- and Zhejiang-dominated banking
syndicates supported Chiang Kai-shek’s regime, even though they were
often coerced. In contrast to the Beiyang regime, the Nationalist government
spent most of its borrowed revenue (40 percent of domestic debt and
75.65 percent of foreign debt) on productive purposes like railways, teleg-
raphy, electricity, aviation, and river conservancy.79
Wartime Finance, 1937–1949

The Japanese invasion of 1937 was a severe blow to the Nationalist
government, whose main sources of income – customs, salt, and
78
Yeh Mei-chu 葉美珠, “國民政府鹽務稽核所的興革試析, 1928–1936” (The 1928–1936
Reforms of the Inspectorate of Salt Revenue under the National Government of China:
A Preliminary Study), Bulletin of Academia Historica 49 (September 2016), 1–32; Strauss,
Strong Institutions in Weak Polities, Chapter 3; Guojia Shuiwu Zongju 國家稅務總局
(ed.), 中華民國工商稅收史：鹽稅卷 (A History of Commercial and Industrial
Taxation in Republican China: Salt Tax) (Beijing, Zhongguo caizheng jingji chubanshe,
1999), pp. 82–6; Young, China’s Nation-Building Effort, pp. 54–63.
79
Sun Di 孫迪, 民國時期經濟建設公債研究 (1927–1937) (A Study of Public Bonds for
Economic Development during the Republic of China (1927–1937)) (Shanghai, Shanghai
shehui kexueyuan chubanshe, 2015), pp. 211, 222; Jiang, 南京國民政府內債問題研究,
pp. 85–7; Young, China’s Nation-Building Effort, pp. 97–142; P.M. Coble, The Shanghai
Capitalists and the Nationalist Government, 1927–1937 (Cambridge, MA, Harvard
University, 1986), pp. 29–40.
272

Public Finance
consolidated tax – drew heavily from the provinces now under Japanese
occupation. In 1928, there were forty regular Maritime Customs stations
and sixteen Native Customs stations under the fifty-li rule (see above). By
1937, there were still thirty-six stations, but by 1942 fully half of them had
fallen under Japanese control. Left without its traditional revenue, the MCS
collected an interport duty, which was extended from steamships to all
modes of transport in 1937, and from 1942 a highly unpopular wartime
consumption tax (abolished in 1945). Under these pressures, the service
unraveled.80 The Sino-Foreign Salt Inspectorate also disintegrated after the
most important salt-producing regions were lost to Japanese occupation. In
January 1939, China stopped servicing all foreign debt secured by customs
and salt revenues. In the rural and impoverished hinterland the costs of
providing salt for the alimentation of the populace often exceeded income.
From May 1942 to January 1945, the government implemented a sales
monopoly (zhuanmai 專賣) and increased the tax rates several times.
However, under the impact of rampant inflation the share of the salt tax
in total revenue declined.81
The war provided the Nationalist government in its Chongqing exile with
both the need and the legitimacy to regain control of provincial revenues and
private incomes to extract revenue from the economically backward hinter-
land. The Third National Finance Conference in June 1941 effectively elimin-
ated the fiscal autonomy of the provinces. The land tax was converted from
a monetary tax to collection in kind under the management of the central
government. Two-thirds of the collected grain was designated for the army;
the rest was reserved for the salaries of civil bureaucrats or to be sold cheaply
to stabilize market prices.82 Critics have argued that this new tax regime
increased the burden of the peasants and contributed to the defeat of the
80
Jin Baoguang 金葆光 (ed.), 海關權與民國前途 (Customs Sovereignty and the Future
of the Republic of China) (Shanghai, Commercial Press, 1928), pp. 84–7; Sun Baogen
孫寶根, 抗戰時期國民政府關稅政策研究 (1937–1945) (A Study of the Customs
Policy of the Nationalist Government during the Sino-Japanese War (1937–1945))
(Beijing, Zhongguo shehui kexue chubanshe, 2014), pp. 6, 66; F. Boecking,
“Unmaking the Chinese Nationalist State: Administrative Reform among Fiscal
Collapse, 1937–1945,” Modern Asian Studies 45 (March 2011), 277–301.
81
Guojia Shuiwu Zongju, 中華民國工商稅收史, pp. 301–3; Dong Zhenping 董振平, 抗
戰時期國民政府鹽務政策研究 (A Study of the Salt Policy of the Nationalist
Government during the Anti-Japanese War) (Jinan, Qi Lu shushe, 2004), pp. 62–70.
82
Chen Yousan 陳友三 and Chen Side 陳思德, 田賦徵實制度 (The System of Levying
the Land Tax in Kind) (Chongqing, Zhengzhong shuji, 1945), p. 3; Jiang Yongjing 蔣永敬,
“孔祥熙與戰時財政—法幣政策與田賦徵實” (H.H. Kung and Wartime Finance:
Currency Policy and Levying the Land Tax in Kind), in Chin Hsiao-yi 秦孝儀 (ed.), 抗
戰建國史料: 田賦徵實 (Materials of the Anti-Japanese War: Levying the Land Tax in
Grain) (Taipei, Zhongguo Guomindang dangshi weiyuanhui, 1989), vol. 4, pp. 365–7.
273

Nationalists by the Communists in 1949. However, the Communist Party in

1950 not only copied the land tax collection in kind from the Nationalists but
also relentlessly increased the degree of extraction, leading to violent
uprisings.83 For Chiang Kai-shek and other leaders of the Nationalist Party
the recentralization of the land tax was not simply a wartime expedient but
constituted a “correction” of what they considered an aberration from the
traditional fiscal ideal of the land tax as the “orthodox tribute” (zhenggong
正貢) that tied the people to the state.84 Chiang’s state authoritarianism was
diametrically opposed to the political philosophy that had supported the
separation of central and local finances, but it was smoothly inherited by
the Communist Party.
For the consolidated tax, the Nationalist government increased the tax
rates and broadened the range of taxable commodities to include, newly,
beverages and sugar products. However, some areas haphazardly imple-
mented government monopolies on sugar, matches, or tobacco products,
leading to confusion about whether the consolidated tax or the monopoly
applied. Cash-strapped local governments frequently ignored the one-tax rule
and erected new tax stations resembling the old lijin system. The share of the
consolidated tax in the total tax revenue of the government increased during
the war, but this was mainly due to the sharp drop in customs and salt
revenues during wartime, and in general non-borrowed income was dwarfed
by inflation and debt. After the end of the war, the consolidated tax was again
expanded and, together with taxes on tobacco and alcohol, as well as mining,
given a systematic legal basis under the new name of “commodities taxes”
(huowushui 貨物稅).85
The Nationalists’ commitment to modernizing the tax structure is best
represented by the expansion of direct taxes. The stamp duty and business tax
had been shared with provincial and local governments as a replacement for
lost lijin revenues in 1931. The Sino-Japanese War afforded the Nationalists an
opportunity to bring these taxes back under central control (in 1940 and 1941)
and add modern income and estate taxes. Income and estate taxes had been
83
Gao Wangling 高王凌, 中國農民反行為研究 (1950–1980) (A Study of the Counter-
actions of Chinese Peasants (1950–1980)) (Hong Kong, Chinese University of
Hong Kong, 2013), pp. 297–304.
84
Lu Fang-sang 呂芳上 (ed.), 蔣中正先生年譜長編 (Chronological Biography of
Chiang Kai-shek) (Taipei, Guoshi Guan, 2014), vol. 6, p. 572.
85
Guojia Shuiwu Zongju, 中華民國工商稅收史, pp. 181–98, 204–5; Hou Kunhong 侯坤
宏, 抗戰時期的中央財政與地方財政 (Central and Local Fiscal Policies during the
Anti-Japanese War) (Taipei, Guoshi Guan, 2000), pp. 104–7; Zhang Sheng 張生, 南京國
民政府的稅收 (1927–1937) (The Tax Revenue of Nanjing Nationalist Government
(1927–1937)) (Nanjing, Nanjing chubanshe, 2001), p. 4.
274

Public Finance
discussed since Yuan Shikai’s regime, but their implementation was ham-
pered by an insufficient legal infrastructure and resistance from the elites.
Under the Nanjing government, collection of the income tax started only in
October 1936, when patriotic sentiment could be mobilized for its support in
the face of an impending Japanese invasion. The war period also brought
a temporary “wartime profit tax,” which extracted heavy fines from profits in
industry and commerce that were deemed excessive. This can be seen as
a special kind of income tax. The levying of the estate tax only began in
July 1940. After dismal results in the beginning, the Ministry of Finance
achieved more success by incentivizing people to report tax evasion.
Between 1940 and 1944, direct taxes made up 20 to 30 percent of non-
borrowed revenue, but their share of total revenue was low, peaking in
1943 at 6.54 percent. While the actual contribution of these new tax categories
to government revenue was limited, the Nationalist government repeatedly
stressed their importance for the process of fiscal modernization. Under the
call for national resistance, the government was able to use tax reform to
integrate the scattered individuals of the country under its political rule and
thus create a modern fiscal state. Most importantly, direct taxes equally
applied to both Chinese nationals and foreign nationals living in China, and
compliance by foreigners increased as China joined the allied forces and the
limitations of China’s fiscal sovereignty were gradually removed.86
Although the Nationalist government managed to centralize and even
modernize its tax system and grow its revenue, the gap between noninfla-
tionary cash receipts and huge expenditures kept growing under the condi-
tions of the war. Collection of taxes in kind, including the land tax and also
the business tax, somewhat reduced inflationary pressures, but debt still
made up between 70 and 90 percent of government revenue during the
war (Table 8.5).
As traditional means of deriving both non-borrowed and borrowed
income failed, most of this debt was in fact an expansion of the money
supply. It is true that China during the war borrowed huge sums from the
Soviet Union, the US, and Britain. Wu Chengming has estimated that China’s
total foreign debt between 1865 and 1948 was equivalent to US$2.5 billion,
two-thirds of which occurred after the outbreak of the Second Sino-Japanese
War.87 However, in the absence of collateral from tax revenues, these loans
largely shifted to barter arrangements. To take only the US as an example, the
86
Lin, 西洋稅制, pp. 55–7, 209–13, 221–4, 260–4, 296–302, 339.
87
Wu Chengming 吳承明, 中國資本主義與國內市場 (Chinese Capitalism and the
National Market) (Beijing, Zhongguo shehui kexue chubanshe, 1985), p. 47.
275

Table 8.5 Income and expenditure of the Nationalist government, 1937–1945. Unit: C$ million (fabi)
1937.7–1938.6 1938.7–1938.12 1939 1940 1941 1942 1943 1944 1945
Balance
Revenue (excl. debt) 815 315 740 1,325 1,310 5,630 20,403 38,503 1,241,389
Expenditure 2,091 1,169 2,797 5,288 10,003 24,511 58,816 171,689 2,348,085
Deficit 1,276 854 2,057 3,963 8,693 18,881 38,413 133,186 1,106,696
Deficit finance
Public debt 256 18 25 8 127 363 3,880 1,989 62,820
Bank advances 1,195 854 2,310 3,834 9,443 20,081 40,857 140,090 1,042,257

Tax revenue
Indirect taxes 427 200 451 190 500 1,165 4,299 20,932 79,234
Direct taxes 24 11 32 76 166 1,641 7,870 10,215 20,740
Monopolies －－－－－ 1,357 3,157 3,504 2,270
Total tax revenues 451 211 483 266 666 4,163 15,326 34,651 102,253
Source: Ma, 中國外債史, p. 330

Public Finance
four loans borrowed before Pearl Harbor, totaling US$117 million, were
earmarked to purchase military equipment in the US and repaid by deliveries
of wood oil, tin, tungsten, and other metals. During the Pacific War, the USA
provided US$1.5 billion in aid, the majority through the lend–lease program.
But while these loans were vital during the war, they were no solution to
fiscal crisis.88
In the face of escalating military expenses, the government had tried to fall
back on domestic debt and issued seven bonds with a total nominal value of
C$2.757 billion between August 1937 and June 1939, but subscription was
muted, and most of the bonds were absorbed by four government banks,
the Central Bank, the Bank of China, the Bank of Communications, and the
Farmers’ Bank of China (the “big four”).89 From 1937 to 1945, the huge
payment deficit of the government was almost completely made up by
these so-called “bank advances” (yinhang diankuan 銀行墊款). In reality,
the advances meant a huge increase of money supply in the fiat currency
fabi 法币 (“legal tender”) which, once it entered the market, caused runaway
inflation.90 During the civil war from 1945 to 1949, the central government
took control of commodity taxes and direct taxes and restored customs and
salt revenues, but income still proved insufficient, even more so as the
government had declared a tax holiday on the land tax to celebrate victory.
In July 1946, the tax holiday was hastily rescinded and collection in kind
restarted. But deficits were still over 60 percent and could only be covered by
inflation. Loss of control over finances was the main reason for the failure of
the Nationalist government.91
88
Ma, 外債與晚清政局, pp. 13–17; Lin May-li 林美莉, “戰時旳財政經濟” (Wartime
Finance and Economy), in Lu Fang-sang 呂芳上 (ed.), 中國抗日戰爭史新編 (戰時社
會) (A New History of the Anti-Japanese War (Wartime Society)) vol. 4 (Taipei, Guoshi
Guan, 2015), pp. 151–6.
89
Zhongguo lianhe zhunbei yinhang 中國聯合準備銀行 (ed.), 中國內外債詳編
(Internal and External Debts of China) (Beijing, June 1940), in Chen Zhanqi 陳湛綺
(ed.), 民國時期中國內外債史料詳編 (Comprehensive Anthology Concerning
Internal and External Debts in the Republic of China) (Beijing, Quanguo tushuguan
wenxian suowei fuzhi zhongxin, 2010), p. 258.
90
K. Chang, The Inflationary Spiral: The Experience of China, 1939–1950 (Cambridge, MA,
MIT Technology Press, 1958); A.N. Young, China’s Wartime Finance and Inflation, 1937–
1945 (Cambridge, MA, Harvard University Press, 1965); Dai Jianbing 戴建兵, 金錢與戰
爭：抗戰時期的貨幣 (Money and War: Currency during the Anti-Japanese War)
(Guilin, Guangxi shifan daxue chubanshe, 1995); Lin May-li 林美莉, 抗戰時期的貨幣
戰爭 (The Currency War in Wartime China, 1937–1945) (Taipei, Institute of History,
Taiwan Normal University, 1996).
91
Lin, “戰時旳財政經濟”; Jiao, 中華民國財政史, pp. 1044–63.
277

Conclusion
Overall, we can observe three major tendencies in fiscal development during
the period from 1800 to 1950. First, the capacity of the state to tax its subjects
increased after 1850, and the footprint of government in society grew in
proportion. Following the civil wars of the mid-nineteenth century, the
Qing gradually abandoned the traditional political ideal of measuring expend-
iture on the basis of fixed revenue (liang ru wei chu 量入為出) and was forced
to seek new revenues to cover expenditures (liang chu wei ru 量出為入). It is
true that the increase of foreign trade and the establishment of the Maritime
Customs Service was an important factor that moved China into fiscal
modernization. However, we should not overlook responses by Chinese
governments to political crises as modernizing forces in the fiscal structure,
from the lijin duties, which emerged as a response to the Taiping Rebellion,
to income taxes rolled out during the Second Sino-Japanese War. Lijin duties
opened the way for taxes on the consumption of commodities. Income taxes
expanded taxation to the incomes of industrialists, merchants, and individ-
uals. Their importance consisted in shifting the tax system away from its
traditional focus on agriculture.
Second, the capacity to tax grew unevenly between the central govern-
ment and the provinces. Provincial capacity building was way ahead of the
central government well into the 1930s. While the fiscal autonomy of
the provinces was growing gradually due to lijin duties since the 1850s and
the uncontrolled growth of miscellaneous excises and surtaxes after 1895, the
central government continued to rely on revenue-sharing negotiations with
the provinces to obtain income. To a large degree, especially during the late
Qing and the Beiyang era, the central government also utilized foreign loans
and “informal empire institutions” like the Maritime Customs or the Sino-
Foreign Salt Inspectorate as leverage against provincial interests. At worst,
the government had to fall back entirely on customs and salt funds left after
servicing those loans.
Third, what triggered fundamental reform was not fiscal crisis alone but
the arrival of Western ideas and know-how. Conventions and vested interests
posed formidable obstacles to the acceptance of Western fiscal technologies
throughout the nineteenth century, but they were firmly established by 1908.
By the 1930s, Chinese fiscal policies were based on cutting-edge economic
theories. The Nationalist government had all the trappings of a modern fiscal
state and skillfully used domestic borrowing and monetary policy for eco-
nomic stabilization and development. However, its capacity to tax did not
278

Public Finance
grow in sync with its fiscal know-how. The Nationalist model of revenue
sharing limited the central government to “modernized” taxes like customs,
salt tax, and the so-called “consolidated taxes” on certain machine-produced
goods, while unreformed rural taxes like the land tax, which represented
65 percent of the economy, were left to the provinces. Despite ten years of
modernization promoted by the Nanjing regime, China remained starkly
divided into the modern cities that served as the connection point between
China and the global economy and the autonomously self-reproducing
village society inland. This created a strategic vulnerability when the
Japanese occupation after 1937 cut the government off from the main sources
of its modernized tax revenue. The war forced the regime to focus its
attention back on the important land tax while continuing its modernizing
zeal by experimenting with new income and estate taxes. As Iwai Shigeki has
argued, the war taught the regime that warlord-like adaptability to living off
the land was necessary to survive. This village hinterland provided the
foundation for an eight-year struggle against Japanese occupation, but also
for the insurmountable Communist insurrection.92
Further Reading
Boecking, F., No Great Wall: Trade, Tariffs and Nationalism in Republican China, 1927–1945
(Cambridge, MA, Harvard University Asia Center, 2017).
Chong, J.I., “Breaking Up Is Hard to Do: Foreign Intervention and the Limiting of
Fragmentation in the Late Qing and Early Republic, 1893–1922,” Twentieth-Century
China 35.1 (November 2009), 75–98.
He, W. Paths toward the Modern Fiscal State: England, Japan, and China (Cambridge, MA,
Iwai Shigeki 岩井茂樹, 中国近世財政史の研究 (A Study of the Fiscal System in Late
Imperial China) (Kyoto, Kyōto daigaku gakujutsu shuppankai, 2004).
Kaske, E., “Austerity in Times of War: Government Finance in Early Nineteenth-Century
China,” Financial History Review 25.1 (April 2018), 71–96.
Lin May-li 林美莉, 西洋稅制在近代中國的發展 (The Development of Western
Taxation Methods in China) (Taipei, Academia Sinica, 2005).
Van de Ven, H.J., Breaking with the Past: The Maritime Customs Service and the Global
Origins of Modernity in China (New York, Columbia University Press, 2014).
Zanasi, M. Saving the Nation: Economic Modernity in Republican China (Chicago, The
University of Chicago Press, 2010).
Zhou Yumin 周育民, 晚清財政與社會變遷 (Fiscal Policies and Social Change during
the Late Qing) (Shanghai, Shanghai renmin chubanshe, 2000).
92
Iwai, 中国近世財政史の研究, pp. 515–16.
279

9
Financial Institutions and Financial
Markets
Many studies of the history of Chinese finance lack systematic empirical

investigation, are limited to the period before the outbreak of war in 1937, or
focus on Shanghai, skewing our understanding of the full scope of Chinese
finance in this period. This chapter draws heavily on new work, as well as on
the empirical work of the two authors, though many areas for further
research remain.
Finance and Institutional Development

By the mid-nineteenth century, a variety of financial institutions served
China, including rotating credit societies, pawnshops, cash shops, and remit-
tance houses. Available evidence indicates great variation in practice both
across and within regions. In rotating credit societies (qianhui 錢會/lunhui
輪會) members made contributions on a fixed schedule, and collected funds
would be distributed by rotation or sometimes by drawing lots or by
auction.1 Pawnshops (diandang 典當/dangpu 當鋪/xiangya 餉押/xiang’an
餉按/ya 押) loaned money based on possessions which borrowers left at the
shop as collateral.2 Those institutions specializing in exchanging money,
commercial transactions, commercial deposits, or commercial lending were
1
K.S. Tsai, “Banquet Banking: Gender and Rotating Credit Associations in South China,”
China Quarterly 161 (March 2000), 143–6; Yu Jiang 俞江, “清中期至民國的徽州錢會”
(Rotating Credit Societies in Huizhou from the Mid-Qing to the Republican Periods)
安徽大學學報 (Journal of Anhui University) 4 (2017), 1–22.
2
Chang Mengju 常夢渠 and Qian Chuntao 錢椿濤 (eds.), 近代中國典當業 (Pawn
Shops in Modern China) (Beijing, Zhongguo Wenshi chubanshe, 1995); Liu Qiugen 劉
秋根, 中國典當制度史 (An Institutional History of Chinese Pawnshops) (Shanghai,
Guji chubanshe, 1995); T.S. Whelan, The Pawnshop in China, based on Yang Chao-yü,
中國典當業 (The Chinese Pawnbroking Industry), with a historical introduction and
critical annotations (Ann Arbor, Center for Chinese Studies, The University of
Michigan, 1979). On types of articles pawned in the late Qing period and terms of
loans in Shanghai in the 1930s, see Liu, 中國典當制度史, pp. 150, 184–6.
280

Financial Institutions and Financial Markets
called qianpu 錢鋪, qiandian 錢店, qianzhuang 錢莊, or yinhao 銀號, with many
other regional variations.3 Here they will be called “cash shops,” a literal transla-
tion close to most of the Chinese versions which is preferable to the colonial and
unhelpful “native bank” used by most English-language authors. Remittance
houses, a term with resonance to their origins as trading houses, specialized in
long-distance remittance and were often called piaohao 票號.4
The second half of the nineteenth century witnessed the appearance of a new
kind of financial institution when foreign banks (yinhang 銀行) began offering
financial services in China. Chinese subsequently began forming institutions
based on these foreign models, adopting the same term. Yinhang are often called
“modern banks” in the literature, but “modern” also has colonial overtones of
superiority, so yinhang will be referred to simply as “banks.”5
China’s rotating credit societies, pawnshops, cash shops, remittance
houses, and banks were not as functionally distinct as the different terms
suggest. Nonetheless, Table 9.1 presents a rough schematic of the types and
functions of financial institutions in China by the late Qing period.
Organizational form and managerial governance varied considerably as well,
but a few generalizations can be made (Table 9.2). Rotating credit societies grew
out of local social and family networks and had membership rather than
ownership. Pawnshops, money exchange shops, and cash shops tended to
have only one office and were formed as sole proprietorships, family-owned
firms, or partnerships.6 It was not unusual for investors to own shares in multiple
shops, and these often shared similar names, making them a “family” (lianhao
3
Hong Xiaguan 洪葭管 (ed.), 中國金融史 (A History of Chinese Finance) (Chengdu,
Xinan Caijing daxue chubanshe, 1993), p. 81. On the importance of monetary exchange,
see Zhongguo renmin yinhang zonghang jinrong yanjiusuo jinrong yangjiushi 中國人
民銀行總行金融研究所金融研究室 (Finance Research Department of Finance
Research Institute of the People’s Bank of China), 近代中國的金融市場 (Modern
China’s Financial Markets) (Beijing, Zhongguo Jinrong chubanshe, 1989), pp. 10–11;
and B. Sheehan, “Unorganized Crime: Forgers, Soldiers, and Shopkeepers in Beijing,
1927, 1928,” in B.K.L. So and M. Zelin (eds.), New Narratives of Urban Space in Republican
Chinese Cities: Emerging Social, Legal, and Governance Orders (Leiden and Boston, Brill,
2013), p. 103.
4
Zhang Guohui 張國輝, 中國金融通史 (A General History of Chinese Finance), vol. 2,
(1840–1911) 清鴉片戰爭時期至清末時期 (From the Opium War to the End of the
Qing Dynasty (1840–1911)) (Beijing, Zhongguo Jinrong chubanshe, 2003), p. 180.
5
Zhang, 中國金融通史, Chapters 1, 4, 6, 7, esp. p. 15 on the terminology of cash shops;
Hong, 中國金融史, Chapter 3; L. Cheng, Banking in Modern China (Cambridge,
Cambridge University Press, 2003), pp. 10–11, omits foreign banks as a category, but
otherwise follows convention; Z. Ji, A History of Modern Shanghai Banking: The Rise and
Decline of China’s Finance Capitalism (Armonk, NY and London, M.E. Sharpe, 2003), p. xxi,
combines piaohao and “native banks,” but otherwise follows convention.
6
Ma Min 馬敏 and Zhu Ying 朱英, 中國經濟通史 (A General Economic History of
China), vol. 8, part 2 (Changsha, Hunan renmin chubanshe, 2002), p. 550.
281

Table 9.1 Basic functions of financial institutions in the late Qing period
Services primarily to the
Retail services primarily Services to ordinary individuals and/or the wealthy, wealthy, businesses, or
to ordinary individuals businesses, and government government
Rotating Cash shops Businesses providing Banks (for-
credit specializing in monetary exchange as Cash Remittance eign and
societies Pawnshops Pawnshops exchange a sideline shops houses Chinese)
Take deposits X X X X X X X X
Make loans X X X X X X X
Exchange money X X X X X X X
Issue drafts on goods X X X X

transactions
Issue paper money X X X X X X
exchangeable to
silver or copper
Purchase gold, silver, X X X X X X
or copper as
investments
Remit money across X X X
long distances
Table 9.2 Forms and ownership governance of financial institutions in the

late Qing period
Ownership Branching
Individual Joint-
or stock Single Groups linked by
partners Family corp. office common owners Branches
Rotating N/A N/A N/A N/A N/A N/A

credit
societies
Pawnshops X X X Sometimes Occasionally
Cash shops X X X Sometimes Occasionally
Remittance X Sometimes X
houses
Banks X X
聯號) of related financial institutions.7 Remittance houses were usually con-

trolled by one family, though they took investments from minority share-
holders, and, as with cash shops, some families invested in more than one
remittance house, making a “family” of related institutions. With expansive
branch networks, remittance houses tended to be larger than cash shops.
Foreign and Chinese banks were usually limited-liability corporations with
numerous shareholders, large size, and substantial branch networks. Most
foreign banks were branches of banks with head offices in their home countries,
though a few “foreign” banks existed only in China and were referred to as
“foreign” only because of the status of their owners and clientele.
The literature on Chinese financial history remains split about indigenous
Chinese financial institutions. Many scholars have indicted them as feudal,
poorly managed, part of a perceived weakness of the late Qing political
economy, or tools in the control of foreign imperialists.8 Some scholars,
however, look to these institutions, especially remittance houses, as evidence
that Chinese capitalism was not inferior to that of the West.9
7
A. McElderry, Shanghai Old-Style Banks (Ch’ien-chuang) 1800–1935 (Ann Arbor, Center for
Chinese Studies, the University of Michigan, 1976), pp. 51–2.
8
For examples, see Tong Yuansong 童元松, “晚清錢莊動蕩衰弱的原因初探” (The
Decline of Cash Shops in the Turmoil of the Late Qing), 黑龍江史志 (Heilongjiang
History) 11 (2008), 16–17; Zhang, 中國金融通史, p. 7; Ma and Zhu, 中國經濟通史,
p. 567; and McElderry, Shanghai Old-Style Banks, pp. 21–2.
9
L. Wang, “Introduction,” in Wang, Chinese Hinterland Capitalism and Shanxi Piaohao:
Banking, State, and Family, 1720–1910 (New York, Routledge, 2021)., forthcoming, provides
an excellent review and critique of these arguments in relation to remittance houses.
283

Important new studies paint a more complex picture. Relations between

foreign banks and cash shops were interdependent, and Chinese business-
people were adept at using foreign-bank funding for their own purposes and
profits.10 Individual cash shops remained relatively small with little capital.
Rather than a weakness, as pointed out by some critics, this practice could
function as a method of risk management. For example, one Shanghai
financier invested in partnership in seven cash shops, and after they all failed
in a financial crisis, continued to invest in several more cash shops with other
partners.11 By keeping capital in his own hands, instead of the firms’, he
managed his risk, and could even continue to invest after a crisis.
Estimates for numbers of financial institutions are shown in Table 9.3. The
precipitous decline in numbers of pawnshops over the nineteenth century has
been attributed to the chaos of the civil wars in the mid-nineteenth century and
to taxation, as well as to competition from cash shops, remittance houses, and
banks.12 Both cash shops and remittance houses grew in relation to the circula-
tion of goods during a long period of increasing commercialization over the
course of the Qing dynasty.13 The first business devoted to remittances probably
appeared in the 1820s.14 Merchants in Shanxi province were well positioned to
capitalize on the expanding trade with Russia, and they came to dominate the
remittance-house sector.15 Both remittance houses and cash shops grew as trade
with the West expanded again after about 1870. In addition, remittance houses
became involved in providing services to the Qing state, which contributed to
their growth. They peaked in number in about 1883. Six remittance houses with
the most extensive branch networks, and best-surviving historical records, had
a total of ninety-two branches in the 1870s and the 1880s.16 The remaining
twenty-two remittance houses were smaller, so perhaps at their peak the
remittance houses as a group had 200 or 300 branches. The first foreign banks
10
S. Nishimura, “The Foreign and Native Banks in China: Chop Loans in Shanghai and
Hankow before 1914,” Modern Asian Studies 39.1 (2005), 109–32; G. Moazzin, Networks of
Capital: German Bankers and the Financial Internationalisation of China (1885–1919), Ph.D.
thesis, University of Cambridge, 2017, Chapter 2.
11
Moazzin, Networks of Capital, Chapter 2.
12
Liu, 中國典當制度史, p. 259; Du Xuncheng 杜恂誠, 中國金融通史 (A General
History of Chinese Finance), vol. 3, 北洋政府時期 (Beiyang Government Period)
(Beijing, Zhongguo Jinrong chubanshe, 1996), p. 295; Chang and Qian, 近代中國典
當業, p. 28; Whelan, The Pawnshop in China, pp. 6–12.
13
Zhang, 中國金融通史, pp. 3, 19, 29; Nishimura, “The Foreign and Native Banks in
China,” 110; Du, 中國金融通史, p. 233.
14
Zhang, 中國金融通史, pp. 37–9. 15 Wang, “Introduction.”
16
Huang Jianhui 黃鑒暉 et al. (eds.), 山西票號史料, 增訂本 (Historical Materials on
Shanxi Remittance Houses), expanded ed. (Taiyuan, Shanxi Jingji chubanshe, 2002), pp.
213, 215–16, 1279.
284

Table 9.3 Estimated numbers of financial institutions in China
Beginning of
nineteenth End of nineteenth/beginning of 1930s on the eve
Type century twentieth century of World War I I 1947
Rotating credit Unknown Unknown Unknown Unknown

societies
Pawnshops 21,000 4,000? 3,386 Unknown
Cash shops Unknown 4,000? 4,000? 966 firms with 1,076 offices
Remittance houses None 28 firms with perhaps 200 or 300 Few or none None in their earlier form
offices
Banks None 34 firms (12 foreign) with 111 offices 193 firms (39 foreign) 622 firms (16 foreign) with 4,217
(43 foreign) with 1,738 offices offices (40 foreign)
(146 foreign)
Rural co-operative None None Unknown 335 offices (1947 publication); 479
treasuries (1944 per Table 9.5)
Trust companies Unknown 21 firms with 27 offices
Insurance Unknown 129 firms with 378 additional offices
companies

Sources: Liu Qiugen 劉秋根, 中國典當制度史 (An Institutional History of Chinese Pawnshops) (Shanghai, Guji chubanshe, 1995), p. 259;
Chang Mengju 常夢渠 and Qian Chuntao 錢椿濤 (eds.), 近代中國典當業 (Pawnshops in Modern China) (Beijing, Zhongguo Wenshi
chubanshe, 1995), p. 28; Du Xuncheng 杜恂誠, 中國金融通史 (A General History of Chinese Finance). vol 3, 北洋政府時期 (Beiyang
Government Period) (Beijing, Zhongguo Jinrong chubanshe, 1996), pp. 233, 295; Huang Jianhui 黃鑒暉, et al. (eds.), 山西票號史料, 增訂
本 (Historical Materials on Shanxi Remittance Houses, expanded edn) (Taiyuan, Shanxi Jingji chubanshe, 2002), pp. 213, 215–16, 1279;
Zhongguo Yinhang (Taiwan) Jingji Yanjiu Shi 中國銀行 (台灣) 經濟研究室, 全國銀行年鑑, 民國 26 年 (All China Bank Annual, 1937)
(Taipei, Wenhua chubanshe, 1987, reprint of 1937 ed.), vol. 1, 811–23 (original pagination S53–S65); Zhongyang Yinhang Jihe Chu 中央銀行
稽核處 (Central Bank of China Auditing Department), 全國金融機構一覽 (Chinese Financial Institutions at a Glance) (n.l., Liulian
yinshua gongsi, 1947); Jiang Jianqing 姜建清 and Jiang Lichang 蔣立場, 近代中國外商銀行史 (A History of Foreign Banks in Modern
China) (Beijing, Zhongxin chubanshe, 2016), pp. 98–102, 164–5
arrived in Hong Kong and the treaty ports in the 1840s, but foreign banking did
not really take off until the 1860s with the need for the Qing state to turn to
foreign financing, and then especially after the 1870s with the rise in foreign trade
after the conclusion of China’s mid-nineteenth-century civil wars.17 Chinese
began founding banks based on these foreign models in the very late nineteenth
century, but real growth in the Chinese sector did not come until the interwar
period in the late 1910s and the 1920s.18 Subsequently, remittance houses and cash
shops both declined in total number while banks saw dramatic expansion.19
The numbers of banks and bank offices do not tell the whole story of the rise of
Chinese, banking, however, because the size of many of these banks dwarfed that
of pawnshops, cash shops, and even remittance houses. Cheng Linsun estimates
that by 1936 Chinese banks held about 81 percent of the capital power in the
Chinese financial market, with foreign banks holding 11 percent and cash shops
9 percent.20 These numbers do not consider pawnshops or rotating credit
societies, and likely understate cash shop strength because cash shops kept their
affairs as secret as possible. Nonetheless, the dominance of Chinese banks by the
1930s was undeniable. Interestingly, China had a small number of very large banks
and a large number of very small banks, a fact that still begs full explanation.21
In the early twentieth century, many new kinds of financial institution
appeared in China. “Lottery”-type savings societies, savings associations,
savings departments of banks, a postal savings bank, and savings departments
at retail establishments like department stores all aimed at attracting small
deposits from ordinary individuals who had previously been excluded by
many of the institutions of the financial industry.22 In the 1940s, rural
17
On the rise of foreign banking in this period, see Jiang Jianqing 姜建清 and
Jiang Lichang 蔣立場, 近代中國外商銀行史 (A History of Foreign Banks in
Modern China) (Beijing, Zhongxin chubanshe, 2016), Chapter 3.
18
Zhang Jia’ao 张嘉璈, 各省金融概略 (Summary of Financial Matters in China’s
Provinces) (n.p., 1915).
19
Zhongguo Yinhang (Taiwan) Jingji Yanjiu Shi 中國銀行 (台灣) 經濟研究室, 全國銀
行年鑑, 民國 26 年 (All China Bank Annual, 1937) (Taipei, Wenhua chubanshe, 1987,
reprint of 1937 ed.), vol. 1, 811–23 (original pagination S53–S65). My figures, which are
based on a page-by-page compilation of the volume, differ slightly from the summation
in the book’s Appendix 2, which only specified 3,077 offices. Jiang and Jiang, 近代中國
外商銀行史, pp. 98–102, 164–5; Zhongyang Yinhang Jihe Chu 中央銀行稽核處
(Central Bank of China Auditing Department), 全國金融機構一覽 (Chinese
Financial Institutions at a Glance) (n.l., Liulian yinshua gongsi, 1947), Appendix 2.
20
Cheng, Banking in Modern China, p. 78.
21
B. Sheehan, “Myth and Reality in Chinese Financial Cliques in 1936,” Enterprise and
Society 6.3 (September 2005), 458.
22
On the savings industry, see B. Sheehan, “The Modernity of Savings,” in M.Y. Dong and
J. Goldstein (eds.), Everyday Modernity in China (Seattle, University of Washington Press,
2006), pp. 121–55.
286

co-operative treasuries were formed to provide credit to farmers. In addition,

a number of other kinds of financial institution came into existence, such as
insurance companies, trust banks, investment companies, and stock
exchanges.23
Finance and Geography

Map 9.1 shows the branch networks of six important remittance houses in the
1870s and 1880s. In general, branches were located along major transportation
routes such as the upper, middle, and lower Yangzi river; along the Grand
Canal through north China; and into northern border regions toward
Mongolia (Zhang Jiakou) and up the Gansu corridor toward Russia
(Qin’an, Lanzhou, Liangzhou). In addition, there were clusters of branches
in tea-producing regions in Jiangxi and Hunan, and in the ports, which acted
as centers of trade with the West, such as Guangzhou, Fuzhou, and
Shanghai.24 We have no comparable record of pawnshop or cash shop
locations in the nineteenth century, but there is reason to believe they
were widely dispersed.
A 1915 survey showed a different pattern for banks which were located in
east coast cities, including newly prominent treaty ports, such as Yantai,
Qingdao, and Yingkou (Map 9.2).
The rise of banking and the introduction of new kinds of financial institu-
tion in the decades after 1915 changed China’s financial geography again.
Map 9.3 shows how extensive bank networks had become by 1947 on the eve
of the Communist revolution.25 Bank offices were densely packed through-
out “China Proper” (south of the Great Wall, east of the Tibetan plateau, east
of the deserts). Bank offices now followed many established transportation
routes where remittance houses previously held sway, including the Gansu
corridor toward Central Asia, but extended much further, and also tracked
important new transportation routes, including the path of the South
Manchurian Railway in the northeast. In 1947, bank geography also showed
the impact of political events and structures. The large number of banks in
23
Zhongyang Yinhang Jihe Chu, 全國金融機構一覽, Appendix 2.
24
Remittance house location data for Map 9.1 from Huang, 山西票號史料, pp. 212–16.
All of the maps in this chapter were created using the China Historical GIS
(CHGIS) of the Harvard-Yenching Institute. More information is available at
www.fas.harvard.edu/~chgis.
25
Bank office location data for Map 9.4 from Zhongyang Yinhang Jihe Chu, 全國金融機
構一覽. There were sixty-nine bank offices for which we could not match place names
in the CHGIS, but these were less than 2 percent of the total offices.
287

Shenyang
Zhangjiakou
Beijing Tianjin
Liangzhou
Pingyao
Jinan
Lanzhou
Quwo
Qin'an
Sanyuan Kaifeng
Xi'an
Huai'an
Yangzhou
Suzhou Shanghai
Chengdu
Jingzhou Hankou
Hangzhou
Chongqing
Changde Nanchang
Changsha Hekou
Xiangtan
Fuzhou
1 Guilin
2
Guangzhou
3
5–6
Map 9.1 Locations of the branches of six large remittance houses, 1870s–1880s
the southwest, especially in Sichuan province, shows the impact of World

War I I when the province hosted the Nationalist regime’s wartime capital.
Shanxi province, under the quasi-independent rule of Yan Xishan, had very
few banks, almost all of which were concentrated in the capital, Taiyuan. The
288

Changchun
Yingkou
Beijing
Tianjin
Baoding
Yantai
Qingdao
Kaifeng
Xi'an Huai'an
Gaoqiao Zhoujiakou
Yangzhou
Nanjing Zhenjiang
Wuxi Shanghai
Suzhou
Hankou
Hangzhou
Wenzhou
2–3
4–6
7–12
13–16
17–21
Shantou
Map 9.2 Bank offices (foreign and Chinese), 1915
surprisingly large number of banks on China’s western periphery in Gansu

and Xinjiang were almost all branches of the state-run provincial banks of
those provinces.
289

Jinghe Yongji
Wulumuqi Zhenxi Changchun
Tulufan
Kuche Guisui Zhangjiakou Shenyang

Wushi Wuyuan Chengde Yingkou
Jining
Suzhou Beijing
Ganzhou Tianjin
Ningxia Ding
Liangzhou Taiyuan
Hetian Qingyang
Xining Ji'nan Qingdao
LanzhouGuyuan Zhengzhou Ji'ning
Bin Yuncheng
Qin Zhou Kaifeng Lianyungang
Xi'anShang Xuchang Xuzhou
Kang Huai'an
XichuanHengchuan Haimen
SongpanLifan
TongnanYichang Liu'an Chu Wuxing
MaoMianyang
Shizhu Huaining Nanjing
Chengdu Zhong Wuchang
Kangding Jiujiang She
HangzhouDinghai
1 Luzhou YouyangChangde Nanchang Qu

Sinan Changsha

2–10 Ningdu
Dading
Guiyang Jing HengyangJi'an
11–50 Chen
Nanxiong Fuzhou
Luxi Qujing
Guilin Longyan
51–100 Kunming Liuzhou Lian
Baise Fogang Mei
101–379 Wuzhou
Nanning Yulin Guangzhou
Qin Taishan
Qiongzhou
Map 9.3 Banks in 1947

Shenyang
BaotouGuisui
Tong
Shimen
Beijing Tanggu
Tianjin
Xin Ding JinBaoding
Zhao De
Taiyuan Shulu Ye
Pingyao Xingtai ZhoucunWei
Ji'nan
Pingliang
Lanzhou Ji'ning
Zhengzhou
BaojiYuncheng
Kaifeng
Xi'an Xuzhou
Bengbu
YangzhouZhenjiang
Nantong Changshu
Wuhu
Wan SuzhouSongjiang
SuiningWusheng Yichang Hankou Nanjing
Ya'anChengdu Dinghai
BishanFuling Jiujiang Hangzhou
Zizhong
Xi'an Ningbo
Luzhou Chongqing
Yongjia
Xuyong Nanchang
QianyangChangsha
Guiyang
Fuzhou
Kunming
1 Shantou
2–5 Guangzhou
6–10
11–50
51–125
Map 9.4 Cash shops in 1947
Map 9.4 shows both the contraction and the stubborn persistence of cash
shops in the face of the dominance of banks. In 1947, cash shops not only
existed in hinterland areas such as Shanxi where they outnumbered bank
offices, but also continued to thrive on the north China plain, in Sichuan
291

province, and in the lower Yangzi region, all areas of banking strength. Even
banking centers like Shanghai, Tianjin, and Guangzhou hosted large num-
bers of cash shops. The substantial number of cash shops, like the large
number of small banks, which coexisted with China’s small number of
behemoth state and private banks remains an area requiring further
research.26
Map 9.5 shows the financial geography of two new kinds of financial
institution that appeared in the twentieth century: rural co-operative treasur-
ies and insurance companies.27 The former were almost exclusively rural (a
few were located in provincial capitals and other administrative centers)
while the latter clustered in China’s largest cities. This split shows
a growing rural–urban divide in specific kinds of new financial institution at
the same time as bank networks worked toward greater national integration.
As bank offices grew in cities and on the coast, state-sponsored financial
institutions like rural co-operative treasuries and provincial banks extended
tendrils into the countryside, keeping the urban–rural divide from being
complete.
Maps 9.1 and 9.2 show that China had no single dominant financial center
in the nineteenth century, but instead had a dispersed, multi-nodal pattern of
financial centers ranging from the capital, Beijing; to Pingyao in Shanxi
province, home to many remittance houses; provincial capitals; important
transportation hubs on rivers, the Grand Canal, or overland trade routes; and
treaty ports such as Shanghai, Fuzhou, and Tianjin.
The pattern of dispersed, multi-nodal financial centers persisted well into
the twentieth century. Du Xuncheng argues for two major centers, Beijing
and Shanghai, by the 1910s and 1920s, based in part on the growth of the state
banks, but the evidence for this is not so clear.28 In terms of bank headquar-
ters, Tianjin rivaled Beijing, for example. More importantly, the locations of
headquarters did not correspond exactly with financial strength, as shown by
statistics from the Bank of China, the largest bank in China, often accounting
for as much as a quarter of total bank assets. The Bank of China published
statistics based on the twenty regions into which the bank divided its branch
26
Cash shop location data for Map 9.5 from Zhongyang Yinhang Jihe Chu, 全國金融機
構一覽. There were four cash shops for which we could not match place names in the
CHGIS, but these were less than 1 percent of the total offices.
27
Insurance company offices and rural co-operative treasury location data for Map 9.6
from Zhongyang Yinhang Jihe Chu, 全國金融機構一覽. There were twelve rural co-
operative treasury locations for which we could not match place names in the CHGIS,
but these were less than 3 percent of the total offices.
28
Du, 中國金融通史, pp. 5–6.
292

Changchun
Shenyang
Beijing
Tangshan
Ji'nan
Lanzhou
Kaifeng
Xi'an
WuhuWuxiNanjing
Chengdu Wan
Hankou Hangzhou
Leshan Fuling PujiangLanxi
Jianwei
Yibin Nanchang
Fushun Hejiang Changsha Xi'an Yongjia
Qianyang
Guiyang
Fuzhou
Kunming
Guangzhou
Insurance company
offices
1
2–5
6–10
11–50
51–134
Co-operative treasuries
1
2
Map 9.5 Insurance company and co-operative treasury offices, 1947
network for the three years from 1919 to 1921. The relative financial strength
of those regions can be seen by looking at total currency issue plus deposits,
total remittances in and out, and annual profit. Although Shanghai and
Beijing did indeed rank first in size (currency plus deposits), Shanghai, in
293

particular, lagged in remittances, ranked only eleventh among twenty. Less

important regions like Ningbo, Zhejiang province, and Changchun each had
surprisingly high amounts of inflow and outflow, showing the importance of
these places as financial centers as well. In addition, even by size, no one
region had more than 20 percent of currency issue and deposits.29
Subsequently, Beijing’s importance faded after the Nationalists moved the
capital to Nanjing in 1928, and Tianjin, second only to Shanghai in the
numbers of Chinese bank headquarters, diminished after Japanese aggression
in north China catalyzed a number of banks to move to Shanghai for greater
safety. Thus Shanghai became China’s domestic pre-eminent financial center
only under the Nationalist regime in the 1930s, and even then, the country’s
vast area supported a large number of secondary and tertiary centers.30
As for international urban systems, at least through the 1920s, Chinese
banks generally lacked well-developed international branches, knowledge,
and operations.31 Nonetheless, because of the presence of foreign banks, by
one scholar’s estimate, the colony of Hong Kong and the treaty ports of
Shanghai and Tianjin were three of the ten top international banking centers
in the world in 1905.32 This was part of a trend in the early twentieth century
which saw the emergence of a number of new regional and local centers.33
Financial Institutions and the State

Over the century from 1850 to 1950 the Chinese state became more involved
in the finance sector when state finances became intertwined with services
29
Zhongguo yinhang zonghang 中國銀行總行 and Zhongguo di’er lishi dang’an guan 中
國第二歷史檔案館 (eds.), 中國銀行行史資料彙編上編 (1912–1949) (Historical
Materials on the History of the Bank of China, Set 1 (1912–1949)) (Beijing, Dang’an
chubanshe, 1991), vol. 3, pp. 1851–1902.
30
Wu Jingping 吳景平, “近代上海金融中心地位與南京國民政府之關係” (The
Status of Modern Shanghai as a Financial Center and Its Relations to the Nationalist
Government in Nanjing) 史林 (Historical Review) 2 (2002), 92; B. Sheehan, “Urban
Identity and Urban Networks in Cosmopolitan Cities: Banks and Bankers in Tianjin,
1900–1937,” in J. Esherick (ed.), Remaking the Chinese City: Modernity and National Identity,
1900–1950 (Honolulu, University of Hawaii Press, 2000), pp. 51–3; Jiaotong yinhang
zonghang 交通銀行總行 and Zhongguo di’er lishi dang’an guan 中國第二歷史檔
案館 (eds.), 交通銀行史料 (Historical Materials on the Bank of Communications)
(Beijing, Zhongguo Jinrong chubanshe, 1995), vol. 1, pp. 95–7.
31
Jiang and Jiang, 近代中國外商銀行史, p. 307.
32
H.C. Reed, The Preeminence of International Financial Centers (New York, Praeger, 1981),
p. 131.
33
S. Nishimura, T. Suzuki, and R. Michie, “Introduction,” in Nishimura, Suzuki, and
Michie (eds.), The Origins of International Banking in Asia: The Nineteenth and Twentieth
Centuries (Oxford, Oxford University Press, 2012), p. 7.
294

from financial institutions in the last half of the nineteenth century, when
various state organs began founding banks at the end of the nineteenth
century, and when the state moved to high levels of government sponsorship
and control of some financial institutions in the 1930s and 1940s.34
Over the course of the seventeenth and eighteenth centuries, many
statutes and edicts recognized and regulated the functions of pawnshops
and cash shops on matters such as the amount of interest that could be
charged, the issuance of paper money backed by copper or silver, and the
need for guarantors to support liabilities.35 In the eighteenth century it
became common for government officials to invest public funds with pawn-
shops in order to earn interest.36 In the early nineteenth century, individuals
who purchased honorary titles from the Qing state often submitted those
funds to Beijing through remittance houses.37
Nonetheless, as the rise of remittance banks in the nineteenth century
made the transfer of funds cheaper and safer than ever before, officials at the
Qing court still insisted on the transfer of tax revenues by government
shipment of silver under state supervision and control. As late as 1848, an
official in Zhejiang province and his superior were both cashiered because the
former sent funds to Beijing by remittance rather than the expensive and
potentially dangerous shipment of silver ingots overland.38
The mid-century rebellions made physical transfer even more difficult, and
in 1862 the Qing court finally gave tacit approval to sending money through
remittance houses.39 This ad hoc approval ushered in a new era in relations
between the state and the finance sector, and simultaneously led to some of
the dramatic growth for the remittance houses recounted above. From 1863
to 1893, there were at least 3,000 instances of tax remittance via private
remittance houses for amounts totaling as much as 5.2 million taels
per year. Remittance houses also loaned money to local and provincial
officials to help them meet their tax quotas. For three decades, Qing fiscal
34
These milestones differ from standard timelines based on the beginning and ending of
political regimes. See Sun Jianhua 孫建華, “晚清金融創新與發展的階段性特徵及
其變遷” (Characteristics and Evolution of the Periods of Development of Chinese
Finance in the Late Qing Period) 學理論 33 (2010), 141–2.
35
Yan Hongzhong 燕红忠, 晋商与现代经济 (Shanxi Merchants and the Modern
Economy) (Beijing, Jingji kexue chubanshe, 2012), pp. 192–4. Yan sees this as
a relaxation of government regulation because of the legal recognition given to financial
institutions, but the nature of these laws made many functions subject to regulation.
36
Whelan, The Pawnshop in China, p. 10.
37
Zhang Guohui 張國輝, 晚清錢莊和票號研究 (Qianzhuang and Piaohao in the Late
Qing Period) (Beijing, Zhonghua shuju, 1989), p. 37.
38
Wang, Hinterland Capitalism, Chapter 3. 39 Wang, Hinterland Capitalism, Chapter 3.
295

operations and remittance houses were intertwined, although, as Luman

Wang argues, this interdependence was decentralized, ad hoc, wracked by
scandal, rife with bribery, limited by the local supplies of silver at the disposal
of the remittance houses, and constantly under pressure from officials at
court who had deep suspicions of private financial institutions.40 Suspicions,
however, ran deep on both sides, as the Qing state frequently asked the
remittance houses to make large “donations” to the state to help with fiscal
difficulties throughout the 1880s.41
Equally complex were relations between the Qing state and foreign finan-
cial institutions, as the former borrowed large amounts of money from the
latter. From 1877 to 1895, the Hong Kong and Shanghai Banking Corporation
(HSBC) alone extended loans of over £12 million to the Qing court and
various provincial governments.42 In addition, a variety of foreign bank
consortia negotiated more or less successfully to lend money to the Qing
government and its successors.43 Foreign banks also financed the huge
indemnities paid after defeat in war to Japan in 1895 and after the Boxer
Uprising in 1900.44
By then, however, the ongoing political and fiscal crisis sent the Qing
government into high gear of institutional reform, which included founding
a number of government-sponsored banks. In 1897 the official Sheng
Xuanhuai established a bank in Shanghai modeled on the foreign HSBC.45
Many provinces founded their own official banks, as did the Qing Board of
Revenue and the Communications Ministry.46 Often authorized to remit
government revenues, these new government banks posed an existential
threat to the remittance houses.47 In the end, the remittance houses did not
survive as an active group of financial institutions much beyond the fall of the
Qing dynasty in 1911.
In January 1908, the newly formed Ministry of Finance promulgated
regulations which categorized all domestic financial institutions, including
40
Wang, Hinterland Capitalism, Chapter 3; see also Zhou Yumin 周育民, 晚清財政與社
會變遷 (Government Finance and Social Change in the Late Qing Period) (Shanghai,
Shanghai renmin chubanshe, 2000), pp. 277–82.
41
Wang, Hinterland Capitalism, Chapter 4.
42
F.H.H. King, The History of the Hongkong and Shanghai Banking Corporation, vol. 1, The
Hongkong Bank in Late Imperial China 1864–1902: On an Even Keel (Cambridge, Cambridge
43
Du, 中國金融通史, p. 40.
44
W. He, Paths toward the Modern Fiscal State: England, Japan, and China (Cambridge, MA,
Harvard University Press, 2013), p. 177.
45
He, Paths toward the Modern Fiscal State, p. 175; Wang, Hinterland Capitalism, Chapter 4.
46
Wang, Hinterland Capitalism, Chapter 4. 47 Wang, Hinterland Capitalism, Chapter 4.
296

remittance firms, money shops, and banks, as “ordinary banks” (putong

yinhang 普通銀行).48 In spite of this law, finance professionals and the
general public alike continued to distinguish among remittance houses,
cash shops, and banks.
The close and complex involvement of the state with financial institutions
continued after the fall of the Qing dynasty in 1911. Most importantly, the Bank of
China and the Bank of Communications functioned for much of the next four
decades as appendages of the central state, often lending money directly to the
government or underwriting the issuance of public debt.49 Control of the two
banks became a common issue of dispute between various political factions and
private investors in the banks.50 At the local level, provincial and other local
government banks continued into the republic, but often with disastrous results
as they were used to finance the activities of the various warlords and political
factions vying for power in Republican China.51 Not all of the state banks were
completely subservient to government will. The Bank of China, in particular, had
moments of quasi-independence from the state.52 Nonetheless the general trend
was for government control and regulation of the financial system, and this
became particularly apparent with the rise of the Nationalist regime after 1927.
As one of its first acts, the new Nationalist regime created its own central
bank while also increasing the percentage of government ownership in the
Bank of China and the Bank of Communications, China’s two largest banks.53
The state created a Financial Supervision Bureau in 1927 to oversee and
inspect financial institutions; then, in 1931, the regime issued a series of laws
to regulate China’s financial sector. Its new banking law repeated many of the
provisions of both the late Qing law referred to above and a 1924 law issued
during the warlord period. This new banking law, like those earlier laws,
specified that all institutions which took deposits, made loans, or remitted
48
Wang, Hinterland Capitalism, Chapter 4.
49
G. Mickey, “Safeguarding National Credibility: Founding the Bank of China in 1911,”
Twentieth Century China 37.2 (2012), 139–60; Du 中國金融通史, pp. 97–100.
50
Du, 中國金融通史, pp. 107–35.
51
Du, 中國金融通史, pp. 222–31; Brett Sheehan, Trust in Troubled Times: Money, Banking,
and State–Society Relations in Republican China (Cambridge, MA, Harvard University
Press, 2001), Chapter 4.
52
Du, 中國金融通史, pp. 96, 104–6, 128–9. On the limits of independence during this
period, see Sheehan, Trust in Troubled Times, p. 145 and Chapters 3, 4.
53
P.M. Coble, The Shanghai Capitalists and the Nationalist Government, 1927–1937 (Cambridge,
MA, Harvard Council on East Asian Studies, 1986); Sheehan, Trust in Troubled Times,
Chapters 5, 6; Hong Xiaguan 洪葭管, 中國金融通史 (A General History of Chinese
Finance), vol. 4, 國民政府時期 1927–1949 (The Nationalist Period (1927–1949)) (Beijing,
Zhongguo jinrong chubanshe, 2008), pp. 48–9; Sheehan, Trust in Troubled Times, pp. 129–
31, 145.
297

money were banks (yinhang). Thus, in law, there were no differences

between cash shops and banks, though in practice most cash shops failed to
meet the capital requirements of the new law, refused to publish their
financial statements, and continued using one or another of the local vari-
ations for “cash shop” in their names through the 1940s.54
In 1935, against the backdrop of the American Silver Purchase Act, which
created a major crisis for China’s continued use of the silver standard, the
Nationalist regime took undisputed control of the Bank of China and the
Bank of Communications.55 For the rest of the Republican period, a large
sector of state-controlled banks coexisted with a smaller sector of private
banks. In 1936, for example, the assets of the four banks controlled directly by
the central state – the Central Bank, the Bank of China, the Bank of
Communications, and the Farmers’ Bank of China – stood at 59 percent of
the total assets of Chinese banks. The Bank of China alone held a whopping
25 percent of total bank assets. Eleven provincial and municipal government
banks held another 11 percent of bank assets, and the remaining 30 percent of
bank assets was divided among 139 private banks.56
Bank offices in areas controlled by the Communist regime often served as
government organs, but tended to be very small and will not be discussed at
length here.57 The Japanese invasion in 1937 reshaped relations between the
financial sector and the state. The Japanese occupation regime set up its own
reserve banks and took over the “enemy” foreign-bank offices in the occupied
areas.58 In turn, the Nationalist regime reorganized its central state financial
institutions on a wartime footing.59 After the Japanese surrender, one or another
of the Nationalists’ centrally controlled financial institutions took over nine
banks and a dozen other financial institutions run by the Japanese or their allies.60
Finance and Agriculture

China’s smallholder agriculture required small amounts of highly dispersed
financing with variation based on locality and season. Much of our
54
Sheehan, Trust in Troubled Times, pp. 147–8.
55
Coble, The Shanghai Capitalists, pp. 172–92; and Sheehan, Trust in Troubled Times, 166.
56
Zhongguo Yinhang (Taiwan), 全國銀行年鑑, vol. 1, pp. 811–23 (original pagination
S53–S65).
57
For a fuller account of finance in the Communist areas, see Jiang Hongye 姜宏業, 中國
金融通史 (A General History of Chinese Finance), vol. 5, 新民主主義革命根據地時期
(The Period of New Democracy and the Revolutionary Base Areas) (Beijing, Zhongguo
Jinrong chubanshe, 2008).
58
Sheehan, Trust in Troubled Times, p. 175; Du, 中國金融通史, pp. 345–6, 361–3, 367–70.
59
Hong, 中國金融通史, p. 385. 60 Hong, 中國金融通史, p. 477.
298

knowledge of rural finance comes from surveys made in the 1930s and 1940s,
and it is unknown how much these surveys reflect long-term practice.
Prior to the 1930s, rural finance relied on personal loans between people
and on pawnshops.61 Even in the 1930s, when banks had exploded on the
scene, total bank loans to rural China were less than 2 million yuan in 1933,
about 10 million-plus yuan in 1934, and about 14 million yuan in 1935. As for
cash shop lending in rural China, “although exact figures are not known, it is
reasonable to think that it was not significant.” In contrast, in 1937, 3,386
pawnshops had made loans to peasants, totaling 14,931,500 yuan.62 Interest
rates charged by pawnshops varied greatly by locality. One survey in 1931
noted that “more than half the pawnshops in urban and rural areas charged
two to three percent [monthly] interest.”63 Most loans were for less than
one year.
Rural finance changed dramatically in the 1930s and 1940s when the
Nationalist government became a significant provider of rural capital through
specialized banks, such as the Farmers’ Bank of China, and through the
promotion of rural co-operatives. In September 1941 the Nationalist govern-
ment expanded the Farmers’ Bank of China’s scope of operations, and
increased its capital to 20 million yuan. In September of 1942 “capital was
raised again from twenty million to sixty million yuan.”64 At the same time, the
government began heavily promoting the formation of co-operatives and co-
operative treasuries. Rural co-operatives in China date back to the reform
movements of the 1910s.65 As a result of the war, they became the central
61
Mi Gonggan 宓公幹, 典當論 (On Pawnshops) (Shanghai, Shangwu Yinshu Guan,
1936); Qu Yanbin 曲彥斌, 中國典當史 (A History of Chinese Pawnshops) (Shenyang,
Shenyang chubanshe, 2007); Huang Jianhui 黃鑒暉, 中國典當業史 (A History of
China’s Pawnshop Industry) (Taiyuan, Shanxi Jingji chubanshe, 2006); Liu, 中國典當
制度史; Yang Zhaoyu 楊肇遇, 中國典當業 (China’s Pawnshop Industry) (Shanghai,
Shangwu Yinshu Guan, 1933).
62
Qian Chengxu 錢承緒 (ed.), 中國金融之組織 – 戰前與戰後 (Chinese Finance before
and after the War) (Shanghai, Zhongguo Jingji Yanjiu Hui, 1941), pp. 95–6.
63
Mi, 典當論, pp. 192–3.
64
Yao Gongzhen 姚公振, 中國農業金融史 (A History of Chinese Rural Finance)
(Shanghai, Zhongguo wenhua fuwushe, 1947), pp. 320–1.
65
For documents about rural loans and the activities of rural co-operatives, see
Lu Guoxiang 陸國香, 湖南農村借貸之研究 (Research on Rural Loans in Hunan)
(n.l., Guomin zhengfu shiye bu maoyi ju, 1935); Wang Zhixin 王志莘 and Wu Jingfu
吳敬敷, 農業金融經營論 (On Rural Credit Management) (Shanghai, Shangwu
Yinshu Guan, 1936); Yao, 中國農業金融史; Li Jinzheng 李金錚, 民國鄉村借貸關
係研究 (Research on the Relationship between Lending and Borrowing in the
Countryside in the Republican Period) (Beijing, Renmin chubanshe, 2003);
Gong Guan 龔關, 國民政府與中國農村金融制度的演變 (Development of the
System of Chinese Rural Finance during the Nationalist Period) (Tianjin, Nankai
299

link in the Nationalist government’s attempts to solve the problem of military

and civilian food supply. The number of co-operatives increased fourfold and of
co-operative members tenfold between 1936 and 1945, to 172,053 and 17.2 million
respectively.66 At the same time, rural co-operative treasuries – the organiza-
tions that took on the major role in providing rural credit as part of the
co-operative system – grew quickly as well. Table 9.4 shows a high tide of co-
operative treasuries founded between 1938 and 1940. Sichuan province, the
location of the Nationalists’ wartime capital, started early and had the most,
a little more than a quarter of the total. During the wartime period, most were
county- or city-level co-operative treasuries with minimum capital of 100,000
yuan. Below the level of cities and counties, there were county treasury
representative offices, co-operative trust associations, and united co-operatives
(dailichu 代理處, xinyong hezuoshe 信用合作社, lianheshe 聯合社).67
Capital invested in co-operative treasuries increased as well, as shown in
Table 9.5. From 1937 to 1941, total capital invested in co-operatives grew from
5.3 million to about 48.3 million yuan, a nearly tenfold increase in four years.
Although co-operatives raised capital by themselves, much more capital came
from external sources. On the one hand, government organs subscribed to
purchase shares.68 For example, from January to November 1941, the four central
government banks purchased shares in 317 rural co-operative treasuries, totaling
59,304,493 yuan.69 On the other hand, capital remained only a small percentage
of loans made, amounting to 19 percent or less from 1937 to 1941, and a peak of
59 percent at the end of 1945 – indicating other external sources of funding.
Table 9.6 shows the impact of rural co-operative treasuries on changing
the supply of rural capital in 684 counties in fifteen provinces. The percentage
of rural households borrowing money increased steadily. By 1943, 61 percent,
or well over half, of all rural households needed credit. Individuals provided
43 percent of total rural credit in 1938. Subsequently, credit from individuals
began to decline as credit from co-operatives increased. By 1942 the co-
operative organizations together, directly and indirectly, accounted for
40 percent of rural credit. Banks also grew as a source of rural credit, reaching
22 percent in 1943. By then, that previous mainstay of rural credit, the
pawnshop, only provided 7 percent, and cash shops only 2 percent.
In spite of the growth of financial institutions, much rural credit still
remained personal. According to one survey in 1941, rich peasants were the
66
Zhongguo Hezuo Shiye Xiehui 中國合作事業協會, 抗戰以來之合作運動 (The Co-
operative Movement since the Beginning of the Anti-Japanese War) (1948), pp. 13–14.
67
Yao, 中國農業金融史, pp. 273. 68 Yao, 中國農業金融史, pp. 272–3.
69
For all the figures cited here, see Yao, 中國農業金融史, pp. 320–1.
300

Table 9.4 Establishment of county/city co-operative treasuries, 1937–1944
Province 1937 1938 1939 1940 1941 1942 1943 1944 Unknown Total Note
Sichuan 5 57 17 33 5 1 1 2 121
Xikang 7 1 2 10
Guizhou 15 26 12 1 54
Yunnan 1 8 1 25 2 37
Guangxi 16 13 13 16 2 3 2 2 65 Also 1 county-level preparation office
Hubei 1 9 2 12 12 treasuries had ceased operation
Hunan 6 6 5 9 1 1 1 29
Jiangxi 2 1 1 5 9 2 treasuries had ceased operation
Zhejiang 17 2 10 7 1 2 39 Also 1 county-level preparation office
Fujian 3 2 5
Henan 9 18 21 6 54
Shaanxi 1 6 8 4 1 20
Gansu 19 19

(worker co-operative) 2
Total 13 112 95 136 57 41 11 1 10 474
Note in original table: besides those listed here, there were two treasuries in Shandong, one in Hebei, and two in Anhui for a total of 493
(including those that had ceased operations).
Source: Zhongguo Hezuo Shiye Xiehui 中國合作事業協會 (China Co-operative Enterprises Association), 抗戰以來之合作運動 (The
Co-operative Movement in the Period since the Start of the Anti-Japanese War) (Nanjing, Zhongguo hezuo shiye xiehui, 1948), pp. 45–6
Table 9.5 Growth of the capitalization of co-operatives, 1937–1945 (yuan)

Capital Average Average co-operative Balance of Capital as
Year investment share value member shareholding loans made % of loans
1937 5,309,079 115.3 2.5 27,055,948 19

1938 7,994,055 122.8 2.6 61,948,345 13
1939 12,611,944 137.9 2.9 112,611,898 11
1940 25,513,370 191.1 3.5 155,578,662 16
1941 48,301,078 310.3 5.2 249,878,770 19
1942 93,291,530 513.1 9.2 387,694,457 24
1943 326,485,306 1,957.0 23.7 802,376,044 41
1944 707,380,719 4,120.3 44.7 1,187,853,797 59
1945 1,461,082,953 8,492.2 84.8 2,482,932,926 59
Source: Zhongguo hesuo shiye xiehui, 抗戰以來之合作運動, p. 34
largest individual providers of rural credit, accounting for more than 45 per-
cent in both 1940 and 1941. Landlords were the next-largest source of loans,
with 30 percent, followed by merchants with 25 percent. As for lending terms,
the most common lending was based on collateral. Among the fifteen
provinces in the survey, more than 50 percent of loans in most places were
based on collateral, and in some places the proportion was as high as
70 percent. We can conclude that without friends and family, other social
connections, or assets to pledge as collateral, the chances of being able to
borrow were extremely small. Interest rates were also very high, rarely less
than 2.8 percent a month and sometimes as high as 3.4 percent.70
Table 9.7 shows the interest rates on and length of loans in rural finance
from 1940 and 1941. Interest rates on loans – based on credit, guarantee, or
collateral – are similar to the rates on individual loans noted above. Loans
from co-operatives had lower rates, with a weighted average of 1.2 percent in
both 1940 and 1941. Interest rates on loans from co-operatives were even
lower than from the rotating credit societies founded among the people
themselves.
Continued differences in interest rates probably arose either from the
limited coverage of co-operatives; from the fact that control of
70
“Nonglin bu Zhongyang Nongye Shiyansuo Nongye Jingjixi Diaocha” 農林部中央農
業試驗所農業經濟系調查 (Investigation of the Agricultural Economics Department
of the Agricultural Institute of the Ministry of Agriculture and Forestry), Archives of the
Ministry of Agriculture and Forestry, Institute of Modern History, Academia Sinica,
Taiwan, 20-07: 55-2. Note: The original chart was named “Survey of rural finance across
China (continued), Table 1 cash loans (continued).”
302

Table 9.6 Survey of rural finance in 1943 (according to lending institution, weighted average)
Lending institution (%)
State organs
No. of counties % of rural households Cash Pawn- Co- (co-operative
Province reporting borrowing money Banks shops shops Shops operatives treasuries) Individuals
1938 681 51 8 3 13 14 17 2 43
1939 673 55 8 2 11 13 23 2 41
1940 621 50 10 2 9 13 26 2 38
1941 693 51 17 2 9 11 30 4 27
1942 716 55 19 2 8 10 34 6 21
1943 684 61 22 2 7 8 32 5 24
1943 detail by province
Zhejiang 23 56 33 8 3 6 8 20 22
Jiangxi 43 56 16 6 8 24 2 44

Hubei 18 57 30 31 4 35
Hunan 49 64 24 1 6 35 2 32
Sichuan 121 54 22 3 9 4 31 8 23
Henan 50 76 28 3 6 10 35 3 15
Shaanxi 62 62 30 2 4 9 30 1 24
Gansu 43 64 30 1 7 8 41 1 12
Qinghai 5 56 6 12 23 12 18 29
Fujian 46 58 15 9 15 44 17
Table 9.6 (cont.)
Lending institution (%)
State organs
No. of counties % of rural households Cash Pawn- Co- (co-operative
Province reporting borrowing money Banks shops shops Shops operatives treasuries) Individuals
Guangdong 42 63 34 5 7 13 13 28
Guangxi 72 59 15 2 11 6 36 10 20
Yunnan 49 64 15 8 8 31 6 32
Guizhou 54 50 7 5 5 42 9 32
Ningxia 7 58 12 4 16 28 40

Note from original table: (1) co-operatives borrowed money indirectly and loaned money directly; (2) “state organs” are co-operative
treasuries
Source: 農林部中央農業實驗所農業經濟系調查 (Survey by the Agricultural Economics Section of the Agriculture Laboratory of the
Agriculture and Forestry Department), Archive held at Zhongyang Yanjiu Yuan Jindai Lishi Yanjiusuo 中央研究院近代歷史研究所
(Academia Sinica, Modern History Institute), 20-07, Zonghao 宗號 55-2
Table 9.7 Survey of rural finance across China (cash lending) in 1941
Lending interest rate (monthly %) Lending term (percentage distribution)
13
Lending Co- 1–3 4–6 7–9 10–12 months
Province Credit Guarantee Collateral societies operatives months months months months or more
Ningxia 3.0 3.0 3.2 1.0 33 33 34

Qinghai 2.5 2.7 2.7. 2.0 10 10 52 28
Gansu 1.6 2.2 2.4 2.3 1.2 13 20 2 59 6
Shaanxi 1.8 2.1 2.5 2.1 1.3 26 33 1 39 1
Henan 2.1 2.4 2.5 2.2 1.3 16 26 1 57
Hubei 1.7 2.0 2.5 2.0 0.9 7 19 69 5
Sichuan 1.9 2.2 2.2 1.9 1.3 12 20 1 63 4
Yunnan 2.1 2.5 2.5 1.8 1.2 10 20 2 57 11
Guizhou 1.5 1.9 2.3 2.2 1.2 8 10 71 11
Hunan 1.5 1.9 2.0 2.0 1.1 10 14 5 66 5
Jiangxi 1.5 1.5 1.7 1.6 1.1 1 21 1 74 3
Zhejiang 1.3 1.2 1.3 1.4 1.1 12 32 50 6

Fujian 1.4 1.5 1.7 1.5 1.0 2 18 2 72 6
Guangdong 1.9 1.7 1.9 1.9 1.0 4 28 51 17
Guangxi 1.7 1.8 2.1 2.1 1.2 3 28 1 60 8
Weighted 1.8 2.1 2.2 2.0 1.2 11 23 1 59 6
average
1940 1.9 2.1 2.1 1.9 1.2 5 16 8 65 6
Source: 農林部中央農業實驗所農業經濟系調查
co-operatives often lay with specific individuals and co-operatives could not
necessarily become conduits for credit to the general populace; or from
problems inherent in the co-operatives themselves, such as the complexity
of procedures, limits on available funds, and so on.
Rural credit went to a variety of purposes. According to a 1941 survey by
the Chinese Economic Yearbook, in the four counties of Haining, Jiaxing,
Pinghu, and Haiyan, of peasants who borrowed money from pawnshops,
“55 percent did so for direct support of production.” Although this statistic
cannot necessarily represent all of China, “it is believable that more than one-
third” of China’s rural pawnshop loans “was used for production.”71 The
figures in Table 9.8 showing statistics from Guangxi province from 1938 to
1941 allow us to see the uses of rural loans in one Nationalist-controlled area
during the war. The bulk of rural loans for all four years went either to
production (plowing cattle, fertilizer, and seeds) or consumption (grain), but
the latter steadily decreased.
Table 9.8 Uses and percentage distribution for co-operative loans in

Guangxi province, 1938–1941
1938 1939 1940 1941
Plowing cattle 17.63 19.00 23.27 29.34

Fertilizer 21.70 22.45 21.69 16.61
Seeds 9.78 10.49 6.60 5.89
Grain 33.03 24.73 18.83 15.78
Tools 3.41 2.40 2.72 2.14
Wages 4.10 4.70 4.16 5.67
Loan repayment 0.66 1.19 4.13 8.51
Water improvements 1.60 4.19 3.76 4.41
Sideline production 6.23 6.54
Living needs 0.35
Raising animals 5.66 5.97 3.51
Other 2.43 4.88 5.10 4.76
Source: “廣西經濟建設統計提要” (Outline of Guangxi Economic Development

Statistics), February 1943, archive held at Zhongyang Yanjiu Yuan Jindai Lishi
Yanjiusuo 中央研究院近代歷史研究所 (Academia Sinica, Modern History
Institute), 20-07 Zong Hao 宗號 55-3
71
Qian, 中國金融之組織 – 戰前與戰後, pp. 95–6.
306

From existing research there is no evidence of reports of severe problems

in agriculture or the supply of grain in Nationalist areas. There certainly were
multiple factors involved, but rural finance, especially the development of co-
operatives and their work making rural loans for production and cultivation,
undoubtedly played an active role which changed the long-existing condi-
tions of rural finance.
Finance and Business, Commerce, and Industry

There is evidence to indicate that a significant amount of early industrial
capital came from financial institutions such as pawnshops, remittance
houses, and especially cash shops, as seen in case studies of two of the most
important early industrial firms.72
The statistics in Table 9.9 show that the China Merchants Steam
Navigation Company (China’s first mechanized transport firm, hereafter
China Merchants) received a substantial portion of its financing from cash
shops: 470,000 taels (calculated together with loans from individuals), 610,000
taels (calculated alone), and 590,000 taels (calculated with individuals) in 1876
to 1877. As one company manager noted, “When operating cash was short,
[we] frequently got accommodation from Shanghai’s cash shops.”73 The Qing
official Li Hongzhang noted, “The firm has had difficulty raising capital, so it
must temporarily borrow from cash shops.”74 The China Merchants execu-
tive Xu Run 徐潤 stated, “For ten years running at year end the company
owed more than a million taels to cash shops and individuals drawn from the
merchant-gentry class.”75 From these sources, it is clear that financing from
cash shops for China Merchants was extremely important, if not decisive, in
the company’s early years.
Likewise, development of the Hanyeping (漢冶萍) Company (the largest
mining and steel company during this period of China’s history) coal mine at
Ping county almost entirely relied on cash shops. The Ping county coal mine
opened in 1898 and during its first years of operation, all funds to meet
72
See Wang Yejian 王業鍵, 中國近代貨幣與銀行的演進 1644–1937 (The Development
of Money and Banks in Modern China 1644–1937) (Taipei, Zhongyang Yanjiuyuan Jingji
Yanjiusuo, 1981).
73
“Zhaoshangju Disanjie Zhanglue” 招商局第三屆賬略 (Third Accounting Report of
the China Merchants Steam Navigation Company), in Li Yadong 李亞東 (ed.), 招商局
創辦之初 (1873–1880) (The Early Years of the Founding of the China Merchants Steam
Navigation Company) (Beijing, Zhongguo Shuhui kexue chubanshe, 2010), pp. 98.
74
Li, 招商局創辦之初, p. 54.
75
Xu Run 徐潤, 徐愚齋自敘年譜 (Self-Recorded Annals of Xu of the Modest Studio)
(Taipei, Taiwan Shangwu Yinshuguan, 1981), p. 177.
307

Table 9.9 The China Merchants Steam Navigation Company assets and liabilities prior to 1880 (taels)
Deposits from Renhe
Loans from Insurance (Renhe
Year Capital government Loans from cash shops Loans from individuals Baoxian 仁和保險)
1873–1874 476,000 123,023

1874–1875 602,400 136,957 475,354 (cash shops and individuals together)
1875–1876 685,100 353,499 613,228 238,328 200,000
1876–1877 730,200 1,866,979 593,449 87,884 350,000
1877–1878 751,000 1,928,868 1,472,404 (cash shops and individuals together) 418,430

1878–1879 800,600 1,928,868 624,088 (cash shops and individuals together) 582,632
1879–1880 830,300 1,903,868 533,029 (cash shops and individuals together) 619,848
Source: Tang Tingshu 唐廷樞 and Xu Run 徐潤, “招商局第一至第七屆賬略” (The China Merchants Steam Navigation Company
Accounting Periods One through Seven), in Li Yadong 李亞東 (ed.), 招商局創辦之初 (1873–1880) (The Early Years of the Founding of
the China Merchants Steam Navigation Company (1873–1880)) (Beijing, Zhongguo Shuhui kexue chubanshe, 2010), pp. 39–174
expenses “were borrowed from cash shops . . . Invested capital came only
after 1899, and then only in small amounts to pay back loans, but not enough
to satisfy needs.”76 In January 1905, the Ping county coal mine borrowed
416,000 taels from cash shops.77 Besides the coal mine, the Hanyeping
company itself had large loans from cash shops. In August 1917, Hanyeping
borrowing from more than thirty cash shops in Hankou reached more than
350,000 taels, in addition to loans from Shanghai cash shops.78 At this time it
was common for a given large firm to have loans from dozens of cash shops.
A given cash shop could also have loans to many firms. Table 9.10 shows
loans made to industry by the notorious Fukang, which became closely
intertwined with a variety of government projects and eventually crashed
spectacularly.
For some firms, reliance on cash shops came because they were excluded
from funding at the new state banks founded in the late Qing period. One
cotton mill executive noted, “we could not even get in the door of a state
bank. Our relations were primarily with cash shops.”79 Another businessper-
son stated, “banks did not have relations with business and industry, which
completely relied on cash shops.”80 Spoken by individuals in Hubei province
76
Zhang Zanchen 張贊宸, “奏報萍鄉煤礦歷年辦法及礦內已成工程” (Memorial and
Report on the Management of the Coal Mine at Pingxiang and Engineering in Progress
in the Mine), in Hubeisheng Dang’anguan 湖北省檔案館 (ed.), 漢冶萍公司檔案史
料選編 (Selected Archival Documents on the Hanyeping Company), vol. 1 (Beijing,
Zhongguo shehui kexue chubanshe, 1994), p. 205. As of January 1905, capital shares in
the Pingxiang Coal Mine were one million kuping taels (p. 204). On Hanyeping, see
Zhang Zhidong 張之洞, “鐵廠招商承辦議定章程折” (Memorial on the Regulations
Decided for Merchant Capital for an Iron Factory), in Yuan Shuyi 苑書義 (ed.), 張之洞
全集 (Collected Works of Zhang Zhidong) (Shijiazhuang, Hebei renmin chubanshe,
1988), pp. 1167.
77
Zhang Zancheng 張贊宸, “奏報萍鄉煤礦歷年辦法及礦內已成工程” (Memorial
and Report on the Pingxiang Coal Mine Methods over the Years and Construction
Already Completed within the Mine), in Hubei Sheng Dang’an Guan 湖北省檔案館
(Hubei Provincial Archive) (ed.), 漢冶萍公司檔案史料選邊 (Collection of Historical
Materials on the Hanyeping Company), vol. 1 (Beijing, Zhongguo shehui kexue
chubanshe, 1994), p. 205.
78
Hubeisheng Dang’anguan 湖北省檔案館 (ed.), 漢冶萍公司檔案史料選編 (Selected
Archival Documents on the Hanyeping Company), vol. 2 (Beijing, Zhongguo shehui
kexue chubanshe, 1994), p. 727.
79
Zhang Songqiao, “1959 年张松樵回憶錄未刊稿” (unpublished 1959 memoir), in
Yudahua fangzhi ziben jituan shiliao bianjizu bian 裕大華紡織資本集團史料編輯組
(ed.), 裕大華紡織資本集團史料 (Historical Materials on the Yudahua Textile Group)
(Wuhan, Hubei renmin chubanshe, 1984), p. 61.
80
“錢遠聲, 王仰蘇訪問記錄” (Record of the Interview of Qian Yuansheng and
Wang Yangsu), in Zhongguo renmin yinhang Shanghai fenhang 中國人民銀行上海
分行 (ed.), 上海錢莊史料 (Historical Materials on Shanghai Qianzhuang) (Shanghai,
Shanghai renmin chubanshe, 1960), p. 170.
309

Table 9.10 Loans from Shanghai’s Fukang (福康) cash shop to industrial enterprises, 1899–1907 (taels)
Year Enterprise Loan amount Year Enterprise Loan amount
1899 Textile Bureau (紡織局) 20,246 (credit) 1904 Lunhua Silk Factory (綸華絲廠) 5,720 (credit)
1899 Ruilun Silk Factory (瑞綸絲廠) 5,112 (credit) 1904 Hengchang Silk Factory (恆昌絲廠) 44,000 (collateral)
1900 Hengchang Silk Factory (恆昌絲廠) 80,000 (credit) 1905 Hengchang Silk Factory (恆昌絲廠) 33,233 (collateral)
1902 Xiechang Match Factory (燮昌火柴廠) 5,000 (credit) 1906 Youxin Cotton Mill (又新沙廠) 10,317 (credit)
1902 Ruishun Silk Factory (瑞順絲廠) 65,000 (collateral) 1906 Hanyeping (漢冶萍局) 10,200 (credit)
1902 Fengji Oil Factory (豐記油廠) 22,259 (collateral) 1906 Youxin Cotton Mill (又新沙廠) 20,000 (collateral)
1902 Textile Bureau (紡織局) 20,000 (collateral) 1907 Gong Yi Cotton Mill (公益沙廠) 20,279 (credit)
1903 Lunhua Silk Factory (綸華絲廠) 10,315 (credit) 1907 Hanyeping (漢冶萍局) 20,267 (credit)

1903 Ruilun Silk Factory (瑞綸絲廠) 100,000 (collateral) 1907 Youxin Cotton Mill (又新沙廠) 10,337 (credit)
1903 Baochang Silk Factory (寶昌絲廠) 40,000 (collateral) 1907 Qixin Cement Company (啟新洋灰公司) 10,244 (credit)
1904 Ruishun Silk Factory (瑞順絲廠) 45,000 (credit) 1907 Huaxing Flour Company (華興麵粉公司) 10,184 (credit)
The original source divided loans into mining, credit loans, and collateral loans, but they have all been combined here, but with notes to
distinguish loans made on credit versus collateral. Some enterprises received two loans in a year, hence credit loans and collateral loans.
Source: Zhongguo Renmin Yinhang Shanghai Fenhang 中國人民銀行上海分行 (ed.), 上海錢莊史料 (Historical Materials on Shanghai
Qianzhuang) (Shanghai, Shanghai renmin chubanshe, 1960), pp. 784–5
and Shanghai respectively, these words show the functions of cash shops in
supporting the development of business and industry.
The importance of cash shops for modern industry and business
continued well into the Republican period. For example, in 1931 the
Rong Family Company purchased the Shanghai Housheng (厚生)
Cotton Mill for 3.4 million taels. Housheng was operating at a loss
and could not repay money owed to the Dunyu (敦裕) cash shop.
Rong Zongjing purchased it for the amount owed to the cash shops
plus a little more for Housheng’s owners. Rong borrowed the money
against collateral from some cash shops. As a result, “Rong did not
spend any money, but simply transferred balances among accounts at
cash shops.”81
During and after World War I the size and breadth of bank lending to
industry increased noticeably.82 For example, in 1919 the Kincheng (Jincheng)
Banking Corporation’s loans to industry and railroads reached a total of
2.81 million yuan. By 1923, they had grown to 7.59 million yuan, and by
1927 again to 15.32 million yuan.83 The Shanghai Commercial and Savings
Bank’s loans to industry and business totaled more than 3.6 million yuan by
the end of 1926, and “stood at 19.9 percent of all loans made.”84
An even clearer change, however, came in the ten years from 1927 to 1937. As
of 1930, fifteen important banks in Shanghai had loaned a total of 91.49 million
yuan to industry and mining. By 1933, that figure had grown to 163.38 million, and
by 1936 it had increased again to 291.25 million, increasing 2.18 times in seven
years.85 A survey made at the time examined the capital structures of 100
enterprises between 1932 and 1939 which had total financing of 262.2 million
yuan. Of this amount, 114.9 million yuan, or 43.8 percent, was borrowed from
banks (mostly) and a few cash shops.86 The Kincheng Bank’s loans to industry
and mining usually made up more than 20 percent of its lending. The Shanghai
Commercial and Savings Bank’s share was more than 30 percent. The National
81
Shanghai shekeyuan jingjisuo 上海社科院經濟所 (ed.), 榮家企業史料 (Historical
Materials on the Rong Family Enterprises), vol. 1 (Shanghai, Shanghai renmin chu-
banshe, 1980), p. 253.
82
See Li Yixiang 李一翔, 近代中國銀行與企業的關係 1897–1945 (Relations between
Banks and Enterprises in Modern China, 1897–1945) (Taipei, Dongda Tushu Gongsi,
1997), Chapter 1.
83
金城銀行史料 (Historical Materials on the Jincheng [Kincheng] Bank) (Shanghai,
Shanghai renmin chubanshe, 1983), p. 14.
84
上海商業儲蓄銀行史料 (Historical Materials on the Shanghai Commercial and
Savings Bank) (Shanghai, Shanghai renmin chubanshe, 1990), p. 161.
85
Li, 近代中國銀行與企業的關係, p. 65, Table 13.
86
Chen Zhen 陳真 (ed.), 中國近代工業史資料 (Historical Materials on the History of
Chinese Industry), vol. 4 (Beijing, Sanlian shudian, 1961), p. 67, Table 9.
311

Commercial Bank had the highest ratio with about 50 percent of its loans
normally made to business and industry.87 State banks like the Bank of China
and the Bank of Communications also saw a dramatic increase in loans to
business and industry in the 1930s. The Bank of China loaned 30 million to
50 million yuan to business and industry every year, reaching a total of
80.22 million yuan in loans to industry and 400 million to commercial enterprises
at the end of 1936.88 The Bank of Communications, “at the end of 1936 had loaned
69.22 million yuan to business and industry combined. This was an increase of
35.55 million over the previous year and a tenfold increase over 1932.”89
The rise of banking also changed the basis of credit with a multiplication of
types of collateral lending in contrast to credit loans, which were simply based on
the creditworthiness of individuals. Although pawnshops had long made collat-
eral loans in China, cash shops relied more on credit loans.90 In contrast, banks
often relied on collateral, and even built large warehouses to keep track of assets
pledged to support transactions. Numerous banks also established investigation
offices in major cities such as Shanghai, Tianjin, and Hankou to investigate the
creditworthiness of borrowers so they could lend on personal credit or make
overdrafts when appropriate.91
Finance and Capital Markets

The roles in capital markets of cash shops and banks have been discussed above,
so here we will examine the role and status of stock markets in modern Chinese
finance.92 In the years before and after 1880, almost forty firms raised
87
Li, 近代中國銀行與企業的關係, p. 67, Table 14.
88
Zhongguo yinhang hangshi bianjiweiyuanhui 中國銀行行史編輯委員會 (ed.), 中國
銀行行史, 1912–1949 (History of the Bank of China, 1912–1949) (Beijing, Zhongguo
Jinrong chubanshe, 1995), p. 255.
89
Jiaotong yinhang zonghang 交通銀行總行 and Zhongguo di’er lishi dang’an guan 中國
第二歷史檔案館, 交通銀行史料, vol. 1, p. 289.
90
Zhongguo renmin yinhang Shanghai fenhang, 上海錢莊史料, p. 215.
91
Yang Yinpu 楊蔭溥, 中國金融研究 (Research on Chinese Finance) (Shanghai,
Shangwu yinshuguan, 1936), p. 159.
92
On stockmarkets, see Liu Zhiying 劉志英, 近代中國華商證券市場研究 (Research on
Chinese Stock Markets in Modern China) (Beijing, Zhongguo shehui kexue chubanshe,
2011); and Liu Zhiying 劉志英, 近代上海華商證券市場研究 (Research on Chinese Stock
Markets in Modern Shanghai) (Shanghai, Xuelin chubanshe, 2004); Zhu Yingui 朱蔭貴, 近
代中國: 金融與證券研究 (Modern China: Research on Finance and Stock Markets)
(Shanghai, Shanghai renmin chubanshe, 2012), Zhu Yingui 朱蔭貴, “近代上海證券市場
上股票買賣的三次高潮” (Three High Tides of Buying and Selling on Shanghai Stock
Markets), 中國經濟史研究 (Research on Chinese Economic History) 3 (1998), 58–70,
Zhu Yingui 朱蔭貴, “抗戰時期的上海華商證券市場” (Stock Markets in Shanghai
during the Anti-Japanese War), 社會科學 (Social Science) 2 (2005), 88–97; and
Zhu Yingui 朱蔭貴, “試論近代中國證券市場的特點” (Characteristics of Stock
312

approximately 10 million taels in capital through issuing shares on China’s capital

market. The time of this surge gave birth to a private exchange firm buying and
selling shares of different companies on behalf of clients.93 In 1891 foreign
residents of Shanghai organized a stock exchange for trading foreign stocks
and bonds. Only in 1914, however, did the new Republican government in
Beijing formally promulgate its “stock market law.” Then it took until 1918 for
the Beijing Stock Exchange, the Shanghai Stock and Commodities Exchange,
and the Shanghai Stock Exchange to receive permission to open.
The opening of these three exchanges one after the other marked a new era in
the market for stocks in China. All the way up to the outbreak of full-scale war
with Japan in 1937, however, these stock markets focused on the buying and
selling of government bonds. When the Beiyang government (1912–1937) issued
bonds domestically, they were usually underwritten by banks and then circulated
to society on the stock market. The Beijing Stock Exchange’s request for
permission to operate from the Department of Agriculture and Commerce
said that “the buying and selling of public bonds and securities is gradually
increasing, but there is no central place to evaluate [them], so prices go up and
down without standards.”94 Thus it is clear that the reason for the founding of the
Beijing Stock Exchange was to solve problems in the circulation of securities,
especially government bonds. “From 1912 to 1926, the Beiyang government
issued 27 different domestic bonds for a total of 876,792,228 yuan.”95 As a result,
“the time when the government’s issue of bonds and treasury notes was at its
most excessive and chaotic was also the time when the Beijing Stock Exchange
flourished and prospered most.”96 During the period of rule by the Nanjing
Nationalist government from 1927 to 1937, the Nationalist government continued
to issue large amounts of domestic bonds, totaling more than 3.6 billion yuan.97 In
the securities market, “98 percent of the transactions were government bonds, so
sometimes stock exchanges were called government bond exchanges.”98
Markets in Modern China: A Preliminary Assessment), 經濟研究 (Economic Research) 3

(2008), 150–60; and Zhang Zhongmin 張忠民, “抗戰時期上海的產業證券與新興企業
集團 – 以新亞集團為例” (Shanghai Industrial Securities and New-Style Enterprise
Groups during the Anti-Japanese War: The Case of the Xinya Group), 上海經濟研究
(Research on the Shanghai Economy) 3 (2002), 72–9.
93
Zhu, 近代上海.
94
“北京籌設證券交易所” (Beijing Plans Stock Exchange), 銀行週報 (Bankers’ Weekly)
2.11 (March 26, 1918).
95
See Qian Jiaju 千家駒 (ed.), 舊中國公債史資料 (Documents on the History of Public
Debt in Old China) (Beijing, Caizheng Jingji chubanshe, 1955), pp. 10–11.
96
Zhongguo renmin yinhang zonghang, 近代中國的金融市場, p. 166.
97
Qian, 舊中國公債史資料, pp. 19, 23.
98
Zhu Sihuang 朱斯煌 (ed.), 民國經濟史 (Economic History of the Republican Period),
photocopied ed. (Taipei, Xuehai chubanshe, 1970), pp. 143, 152.
313

Table 9.11 comes from a Japanese survey made at the time. Table 9.12
comes from statistics kept by Chinese sources. The figures in these tables
show that the government bond price index reached its lowest point in 1932 at
about the time of the Japanese invasion of Manchuria and the attack on
Shanghai by Japanese forces. The fundamental reason for the dip, however,
grew from the Nationalist government’s desire to extend the period over
which the principal would be repaid and to reduce interest, policies which
caused trust in the bonds to drop precipitously. After reorganization of
government debt, however, there was recovery and stability by 1934. In 1935
Table 9.11 Index of domestic bond transactions and volume of transactions,

1926–1937 (thousands of yuan)
Commodity securities
Year Bond index Transaction volume transaction volume
1926 450,738
1927 238,169
1928 370,487
1929 1,320,555 97,703
1930 2,341,820 90,615
1931 85.62 3,362,540 555,022
1932 60.68 901,710 303,939
1934 78.48 3,182,685 230,090
1935 97.94 4,773,410
1936 98.25 4,909,980
Jan. 1937 93.94 146,365
Feb. 1937 94.91 118,360
March 1937 97.28 197,600
April 1937 102.12 296,035
May 1937 103.02 231,325
June 1937 110.35 485,815
July 1937 106.91 604,260
Aug. 1937 101.73 328,201
Notes. (1) The volume of securities traded refers to the volume on the Shanghai
Chinese Stock Exchange. The volume of commodities securities traded refers to
transactions on the Shanghai Commodities Exchange. Figures for transactions in 1934
are for the period after the merger of the two exchanges, and are for the fiscal, not
calendar, year. They represent transactions for January–May 1933. (2) Bond index
uses July 1931 = 100. (3) The bond index for 1931 is the average for the second half of
the year. (4) The bond index for 1932 is the average of January and April–December.
Source: Zhong Zhina Zhenxing Zhushi Huishe Diaochake 中支那振興株式會社調查
課 (ed.), “振興調查資料第二十八號” (Zhenxing Survey Number Twenty-Eight), in
上海華商證券業概況 (Shanghai’s Chinese Stock Exchanges) (1941), pp. 12–13
314

Table 9.12 Stock index and volume of stock and corporate bond transactions,
1931–August 1937
Stock transaction Corporate bond transaction
Year Stock index volume (shares) volume (thousands of yuan)
1931 99.76 7,269

1932 80.28 4,338 20,299
1933 71.36 8,534 51,422
1934 65.29 18,453 44,059
1935 57.11 898 12,437
1936 57.66 9,685 16,413
Jan. 1937 48.30 3,135 1,068
Feb. 1937 46.72 3,684 1,956
March 1937 48.50 4,271 965
April 1937 48.60 3,692 1,045
May 1937 47.10 1,229 1,493
June 1937 46.67 1,389 2,235
July 1937 46.22 542 2,167
Aug. 1937 44.33 177 181
Notes: (1) Stock Index July 1931 = 100. (2) Figures for 1931 are for the second half of
the year. (3) The stock index for 1931 is the average of January and April–December.
Source: Zhongguo Jingji Tongji Yanjiusuo 中國經濟統計研究所 (ed.), 經濟統計月志
(Economic Statistics Monthly) 4.10 (October 1937), 20, Table 5, and 22, Table 8
the index reached a new high, and it grew steadily through March of 1937,
breaking previous records. The stock share price index did exactly the oppos-
ite. After 1931 it continuously decreased, reaching a low in 1936, then it
continued to drop in 1937. The difference in the directions of the two indices
reflects the dynamism of the government bond market and a slump in the
stock market. Looking at the volume of transactions, the volume of govern-
ment bond trading grew rapidly after weathering the drop in 1932. It exceeded
4.7 billion yuan in 1935, 4.9 billion yuan in 1936, and 2.4 billion yuan for the first
eight months of 1937. The volume of stock trades varied greatly over time. At
the peak in 1934, 18,453,000 shares were traded, but in 1935 the volume was
only 898,000 shares for the whole year. In addition, the volume of corporate
bonds was minimal. At its greatest, the volume was a little more than
50 million yuan a year; at its least, just over 10 million yuan. The volume of
corporate bonds trades also generally declined over the course of the period.
Beginning in 1935, the Shanghai Stock Exchange started a market in the
shares of the finance industry and businesses, but the securities market
“became monopolized by government debt. The cold shoulder given to
315

stock shares was completely the opposite to that of public bonds.”99 The
economist Zhang Naiqi said that Shanghai had a “Chinese Stock Exchange,
and it was for the purchase and sale of government bonds – it was a public
debt market and not a market in the securities of productive businesses.”100
Finance and Society

Financial institutions were recognizable icons of town and city life in China.
In the early to mid-nineteenth century, pawnshops often dominated local
skylines in villages and cities. One foreign observer noted during a tour of
rural southern China that in each village there “stands one conspicuous great
solid square structure of granite, lined with brick, about four stories high . . .
[it is] the village pawnshop.” The same observer continued, in the large city
of Guangzhou, “there are upwards of a hundred first class pawn-towers,
besides a multitude of the second and third class.”101 Cash shops and remit-
tance houses tended to keep lower profiles, though they were often grouped
together along particular streets or in particular neighborhoods.102 As banks
grew in the twentieth century, they built large and imposing buildings, often
in the foreign concessions, which provided physical manifestations of their
strength and influence. The buildings along the Shanghai Bund are probably
the best-known example, but other financial centers like Tianjin had similar
streets.103
Financial institutions and financiers loomed large in the cultural imagin-
ation as well, though their symbolism varied significantly depending on the
point of view. The Dream of the Red Chamber, the Qing period’s most famous
novel, mentions pawnshops more than thirty times in episodes ranging from
the pathos of a poor relative of the Jia family who suffers from the cold
because she pawned her winter clothes to one somewhat tainted source of
the Xue family’s fabulous wealth.104 As banks came to dominate the Chinese
99
Wu Yitang 吳毅堂 (ed.), 中國股票年鑑 (China Stock Share Annual) (n.l., Zhongguo
gupiao nianjian she, 1947), p. 3.
100
Zhang Naiqi 章乃器, “中國貨幣金融問題” (The Problem of China’s Money and
Finance), in Zhang Lifan 章立凡 (ed.), 章乃器文集 (Collected Works of Zhang Naiqi)
(Beijing, Huafu chubanshe, 1997), p. 425.
101
C.F. Gordon Cummings, quoted in Whelan, The Pawnshop in China, p. 19.
102
Shen Danian 沈大年 et al. (eds.), 天津金融簡史 (A Brief History of Tianjin Finance)
(Tianjin, Nankai daxue chubanshe, 1988), pp. 10–11; McElderry, Shanghai Old-Style
Banks, pp. 42–3.
103
Sheehan, Trust in Troubled Times, 104–5.
104
Cao Xueqin, The Story of the Stone, trans. D. Hawkes (New York, Penguin Books, 1982),
vol. 3, The Warning Voice, Chapter 57, pp. 88–115, and vol. 4, The Debt of Tears,
Chapter 85, p. 133.
316

financial scene, banks and bankers became recognizable icons of cities. In Cao
Yu’s 1936 play Sunrise, the banker Pan Yueting joined other unsavory urban
archetypes such as the gangster, the society matron, the gigolo, and the
obsequious student.105 Bankers were not always perceived negatively. Wen-
hsin Yeh has shown how bankers and bank compounds became a model of
professionalism and modernity.106
Turning to the relationship between finance and personal networks, the
role of native place stands out. For the nineteenth century, most remittance
bankers came from Shanxi province and cash shops were often divided into
groups or bang 幫 based on the native place of the individuals owning or
running those firms. In Shanghai many cash shop owners and managers came
from Ningbo and Shaoxing in Zhejiang province. In Tianjin there was
a dominant local group of cash shops owned and managed by Tianjin natives
along with others run by sojourners from a number of other parts of China.107
Native place played an important role in networks of bankers as well, and
much of the literature in both English and Chinese focuses on the importance
of elite bankers from Jiangsu and Zhejiang provinces, often referred to as the
Jiangsu–Zhejiang clique.108 Table 9.13 shows the native places of 3,278 bankers
who held positions as middle managers, upper managers, and directors of
Chinese banks according to a banking directory from 1936.109 Unfortunately,
the native place of almost a third of the bankers in the volume is not stated,
thus conclusions can only be preliminary. Zhejiang and Jiangsu natives did,
indeed, have a large presence in the ranks of these bankers, but so did natives
of a number of other provinces, including Hebei, Sichuan, Guangdong,
105
Sheehan, “Urban Identity,” pp. 47–8.
106
W.-H. Yeh, “Corporate Space, Communal Time: Everyday Life in Shanghai’s Bank of
China,” American Historical Review 100.1 (1995), 97–122.
107
McElderry, Shanghai Old-Style Banks, pp. 52–3; Shen, 天津金融簡史, pp. 10–11.
108
S.M. Jones, “The Ningpo Pang [Ningbo Bang] and Financial Power at Shanghai,” in
M. Elvin and G.W. Skinner (eds.), The Chinese City between Two Worlds (Stanford, CA,
Stanford University Press, 1974), p. 73; Yao Huiyuan 姚會元, 江浙金融財團研究
(Research on the Zhejiang–Jiangsu Financial Clique) (Beijing, Zhongguo Caizheng
Jingji chubanshe, 1998), pp. 7–8; M.-C. Bergère, “The Shanghai Bankers Association,
1915–1927: Modernization and the Institutionalization of Local Solidarities,” in
F. Wakeman Jr. and W.-H. Yeh (eds.), Shanghai Sojourners (Berkeley, Institute of
East Asian Studies, Center for Chinese Studies, 1992), pp. 15–34; A. McElderry,
“Confucian Capitalism? Corporate Values in Republican Banking,” Modern China 12
(July 1986), 401–16; McElderry, “Robber Barons or National Capitalists: Shanghai
Bankers in Republican China,” Republican China 11 (November 1985), 52–67; Cheng,
Banking in Modern China, pp. 46–52.
109
Zhongyang yinhang jihe chu 中央銀行稽核處 (Central Bank of China Auditing
Department), 全國金融機構一覽 (Chinese Financial Institutions at a Glance) (n.l.,
Liulian yinshua gongsi, 1947).
317

Table 9.13 Native provinces of

bankers in 1936
Unknown 1,115 34.0%

Zhejiang 729 22.2%
Jiangsu 616 18.8%
Hebei 173 5.3%
Sichuan 159 4.9%
Guangdong 109 3.3%
Anhui 86 2.6%
Shandong 78 2.4%
Hunan 42 1.3%
Fujian 32 1.0%
Hubei 32 1.0%
Shanxi 32 1.0%
Liaoning 24 0.7%
Jiangxi 21 0.6%
Yunnan 7 0.2%
Henan 6 0.2%
Shaanxi 6 0.2%
Guizhou 4 0.1%
Jilin 2 0.1%
Guangxi 1 0.0%
Total 3,274
Source: Zhongyang Yinhang Jingji Yanjiu

Chu 中央銀行經濟研究處 (Research
Office of the Central Bank of China), 全國
銀行人事一覽 (Chinese Banking
Personnel at a Glance) (Shanghai, 1936)
Anhui, and Shandong. It is also possible that inclusion of the armies of lower-
level bank employees such as tellers and clerks would provide a different
distribution.
Common province alone tells us little about the way networks actually
functioned. For example, the density of interlocking directorship ties
among members of boards of directors of Chinese banks in 1936 was no
greater for bankers from the same province than it was for the population
as a whole. Density increased “among bankers from adjacent counties or
among important, or elite, bankers who each held seats on many different
banks. The former suggests local kinds of connections such as dialect,
early schooling or family [while the latter suggests] shared professional
318

interest.”110 Politics also affected personal networks of bankers because of

the close connection between banking and the state in China.111 Thus
generalizations about native place and networks of bankers need to be
supplanted by solid empirical evidence and much work remains to be
done.
Financial trade associations were fixtures in national and local power
structures. Some had long histories that date back to the late seventeenth
or the eighteenth century.112 These organizations also reflected the frag-
mented nature of the finance industry. Shanghai had two cash shop associ-
ations, one for shops in the Chinese city and one for those in the foreign
concessions.113 In a similar vein, Tianjin’s cash shop guild excluded firms
owned by nonnatives of the city until after 1930.114 When banks appeared on
the scene in the late nineteenth and early twentieth centuries, bankers, too,
formed associations in cities and towns around China, though foreign and
Chinese bankers formed separate associations. Sometimes these associations
came together in larger umbrella organizations, but the norm was to associ-
ate by type, location, and, often, native place of ownership.115 These profes-
sional associations co-ordinated operational relations between financial
institutions, such as the clearing of drafts drawn on each other, and provided
a platform to deal with a variety of government relations ranging from
negotiating the amounts of government demands for loans to protesting
the wording of regulatory laws.116 Financial trade associations also played
a role in responding to the financial crises which periodically erupted, though
not always with success.117
110
Sheehan, “Myth and Reality,” 470–3.
111
Sheehan, “Urban Identity,” pp. 54, 56; B. Sheehan, “Warlords, Cadres and Bankers:
Private Commercial Banking in the Republican and Post-Mao Periods,” Journal of Asian
Business 14.1 (1998), 16.
112
Ou Jiluan 歐季鸞, 廣州之銀業 (Guangzhou’s Cash Shop Industry) (Guangzhou, Jingji
zhi diaocha chu, 1932), p. 1; Hong Xiaguan 洪葭管, “略論山西票號, 上海錢莊的性質
和歷史地位” (A Brief Discussion of the Nature and Historical Status of Shanxi
Remittance Banks and Shanghai qianzhuang), Jindaishi yanjiu 2 (1983), 259; Chang and
Qian, 近代中國典當業, p. 139.
113
Jiang and Jiang, 近代中國外商銀行史, p. 308; Bergère, “Shanghai Banker’s
Association”; Cheng, Banking in Modern China, pp. 192–5; Sheehan, Trust in Troubled
Times, p. 81; McElderry, Shanghai Old-Style Banks, pp. 42–3.
114
McElderry, Shanghai Old-Style Banks, pp. 52–3; Shen, 天津金融簡史, pp. 10–11.
115
Jiang and Jiang, 近代中國外商銀行史, p. 309.
116
See, for example, B. Sheehan, “Urban Identity,” p. 60; and Sheehan, Trust in Troubled
Times, Chapter 4; Ye Shichang 葉世昌, “銀行公會, 錢業公會抵制 1931年 ‘銀行法’”
(Resistance to the 1931 “Banking Law” by the Bankers’ Association and Cash Shop
Guilds), 中國金融史集刊 (Journal of Chinese Financial History) 2 (2007), 94–105.
117
Sheehan, Trust in Troubled Times.
319

Pawnshops were by far the most likely institutions to provide credit to

ordinary individuals. A 1932 survey in Guangxi province showed “35 percent
of transactions less than one yuan, 40 percent one to three yuan, 15 percent
three to five yuan, 7 plus percent five to ten yuan, and 2 plus percent ten to
thirty yuan. Transactions of more than 30 yuan were very rare.”118 City
residents often used pawnshops for daily needs, as with Nanjing rickshaw
drivers surveyed in the fall of 1932. Among the 1,350 rickshaw drivers sur-
veyed, “547, or about 40 percent, had debts.” Besides those who borrowed
from family and friends, 220 owed money to pawnshops.119 A 1933 survey of
311 worker households in Guangzhou showed that “151 had borrowed money,
and 176 had used pawnshops . . . Of worker households which use pawnshops
92.61 percent do so to satisfy daily needs and only 7.39 percent did so for other
purposes.”120 One scholar from the Republican period believed, “The pawn-
shop industry worked out of the spotlight, but effectively acted as the
financial fulcrum for ordinary people.”121
In the early twentieth century, ordinary people, especially urban residents,
began to have access to financial institutions as places to deposit money as
well. Banks began offering special departments and programs to attract small
depositors and lottery-type savings societies which drew account numbers at
random for special rewards sprang up to compete with them.122 The govern-
ment-run postal savings bank extended retail banking services to many small
cities across China. In addition, many kinds of business in China had long
taken deposits, and by the late 1920s and early 1930s some of these began
competing with banks for savings from small depositors, much to the dismay
of China’s bankers. In a letter to the Shanghai Bankers’ Association dated
20 March 1930, the Shanghai Commercial and Savings Bank complained that
firms “take advantage of newspaper advertising to unhesitatingly tempt
customers with high interest rates to draw [funds] from all directions.”123
Many of these firms were retail businesses with large existing customer bases
118
Guanxi tongji ju 廣西統計局 (ed.), 廣西年鑑 (Guangxi Annual) (1936), p. 635.
119
Qian, 中國金融之組織 – 戰前與戰後, p. 95.
120
Qian, 中國金融之組織 – 戰前與戰後, p. 95.
121
Ou Jilan 歐季鸞, 廣東典當業 (Guangdong’s Pawnshops) (Guangzhou, Zhongshan
daxue jingji diaocha chu, 1934), p. 125.
122
Sheehan, “The Modernity of Savings,” pp. 121–55.
123
“上海商业储蓄银行致上海银行公会函” (Shanghai Commercial and Savings Bank
to Shanghai Bankers’ Association), Archives of the Shanghai Bankers’ Association, S173-
1-203, 20–1; Liu Qiugen 劉秋根, 明清高利貸資本 (High Interest Loans Capital in the
Ming and Qing Dynasties) (Beijing, Shehui Kexue Wenxian chubanshe, 2000), pp. 138–
9; Wang Zhixin 王志莘, 中國之儲蓄銀行史 (A History of China’s Savings Banks)
(Shanghai, Xinhua Xintuo Chuxu Yinghang, 1934), p. 319.
320

such as the Sincere, Yong’an, Xinxin, and Zhongyuan department stores.124

Others developed savings departments separately from their main line of
business. For example, the Rong family, which by 1928 had twelve flour mills
and six cotton mills, also founded a specialized savings office estimated to
save “200,000–300,000 yuan every year” over the interest charged by financial
institutions.125
Conclusion
Over the century roughly from 1850 to 1950, Chinese financial institutions and
markets witnessed a number of developments. There was remarkable
growth in the number and size of financial institutions, especially banks,
and a multiplication of kinds of financial institution, as banks based on
Western models, retail savings societies, savings banks, trust banks, stock
markets, rural co-operatives, rural co-operative treasuries, and insurance
companies joined the cash shops, remittance houses, pawnshops, and infor-
mal rotating credit societies which dominated China’s financial world for
much of the nineteenth century, and in some cases even into the twentieth.
The financial sector, especially banks, was highly involved with the state and
ties to the state increased over time. By the 1930s and 1940s, the vast majority
of financial assets were in large, state-controlled banks. As a corollary of the
increasing state role in finance, credit from financial institutions in rural
China shifted from reliance on pawnshops to rural co-operatives, though
the latter reached only a small number of rural households. Financial institu-
tions such as cash shops, remittance houses, and banks worked actively in the
business-to-business sector and even became involved to some extent in
financing early attempts at industrialization. In contrast, stock markets
focused on trading government bonds and had less to do with commerce
or industry. China’s financial geography changed as financial institutions
became more numerous, larger, more broadly situated across the peripheral
regions of China, and more densely packed in urban and commercial centers.
Although China remained a place with multiple financial centers, by the 1930s
Shanghai had emerged as the pre-eminent financial city of China. Financial
institutions and the people who ran them played important roles in society as
institutions and as symbols. Native place played an important role in
124
Wang Zhixin, 中國之儲蓄銀行史, p. 319.
125
Shanghai shekeyuan jingjisuo, 榮家企業史料, p. 276; 茂新福新申新總公司三十周
年紀念冊 (Thirtieth Anniversary Memorial Volume of the Maoxin, Fuxin, Shenxin
Company) (January 1929), Appendix, “勸告同仁儲蓄宣言” (Savings Announcement).
321

networks of financial professionals, but that role needs to be understood in

relation to other important political and institutional elements. Finally, the
growth of new institutions such as rural co-operative treasuries and deposit
services for ordinary individuals made the financial sector more and more
enmeshed in the lives of ordinary individuals, though unevenly so.
In the end, although financial professionals and government regulators
made very conscious attempts to “modernize,” “standardize,” “centralize,”
and “professionalize” finance, Chinese financial institutions as a whole
remained highly decentralized, unevenly distributed across space, and hugely
varied in form and function. Even as the growth of a vibrant banking sector
linked many parts of China, a rift widened between kinds of financial institu-
tion in urban and rural China. Financial institutions and markets, and the
financiers who ran them, remained generally fragmented, with periodic
moments of greater or lesser connection subject to the vagaries of economic
trends and political events.
Further Reading
Chang Mengju 常夢渠 and Qian Chuntao 錢椿濤 (eds.), 近代中國典當業 (Pawnshops
in Modern China) (Beijing, Zhongguo Wenshi chubanshe, 1995).
Cheng, L. Banking in Modern China (Cambridge, Cambridge University Press, 2003).
Du Xuncheng 杜恂誠, 中國金融通史 (A General History of Chinese Finance), vol 3,
北洋政府時期 (Beiyang Government Period) (Beijing, Zhongguo Jinrong chubanshe,
1996).
Gong Guan 龔關, 國民政府與中國農村金融制度的演變 (Development of the System
of Chinese Rural Finance during the Nationalist Period) (Tianjin, Nankai daxue
chubanshe, 2016).
He, W., Paths toward the Modern Fiscal State: England, Japan, and China (Cambridge, MA,
Hong Xiaguan 洪葭管 (ed.), 中國金融史 (A History of Chinese Finance) (Chengdu,
Xinan Caijing daxue chubanshe, 1993).
Huang Jianhui 黃鑒暉, 中國典當業史 (A History of China’s Pawnshop Industry)
(Taiyuan, Shanxi Jingji chubanshe, 2006).
Jiang Hongye 姜宏業, 中國金融通史 (A General History of Chinese Finance), vol. 5,
新民主主義革命根據地時期 (The Period of New Democracy and the Revolutionary
Base Areas) (Beijing, Zhongguo Jinrong chubanshe, 2008).
Jiang Jianqing 姜建清 and Jiang Lichang 蔣立場, 近代中國外商銀行史 (A History of
Foreign Banks in Modern China) (Beijing, Zhongxin chubanshe, 2016).
Li Jinzheng 李金錚, 民國鄉村借貸關係研究 (Research on the Relationship between
Lending and Borrowing in the Countryside in the Republican Period) (Beijing, Renmin
chubanshe, 2003).
322

Li Yixiang 李一翔, 近代中國銀行與企業的關係 1897–1945 (Relations between Banks

and Enterprises in Modern China, 1897–1945) (Taipei, Dongda tushu gongsi, 1997).
Liu Qiugen 劉秋根, 中國典當制度史 (An Institutional History of Chinese Pawnshops)
(Shanghai, Guji chubanshe, 1995).
Liu Zhiying 劉志英, 近代上海華商證券市場研究 (Research on Chinese Stock Markets
in Modern Shanghai) (Shanghai, Xuelin chubanshe, 2004).
Liu Zhiying 劉志英, 近代中國華商證券市場研究 (Research on Chinese Stock Markets
in Modern China) (Beijing, Zhongguo shehui kexue chubanshe, 2011)
Qu Yanbin 曲彥斌, 中國典當史 (A History of Chinese Pawnshops) (Shenyang, Shenyang
chubanshe, 2007).
Sheehan, B., “The Modernity of Savings,” in M.Y. Dong and J. Goldstein (eds.), Everyday
Modernity in China (Seattle, University of Washington Press, 2006), pp. 121–55
Sheehan, B., “Myth and Reality in Chinese Financial Cliques in 1936,” Enterprise and Society
6.3 (September 2005), 452–91.
Sheehan, B., Trust in Troubled Times: Money, Banking, and State–Society Relations in
Republican China (Cambridge, MA, Harvard University Press, 2001).
Sheehan, B., “Urban Identity and Urban Networks in Cosmopolitan Cities: Banks and
Bankers in Tianjin, 1919–1937,” in J. Esherick (ed.), Remaking the Chinese City: Modernity
and National Identity, 1900–1950 (Honolulu, University of Hawaii Press, 2000), 47–64.
Wang Yejian 王業鍵, 中國近代貨幣與銀行的演進 1644–1937 (The Development of
Money and Banks in Modern China 1644–1937) (Taipei, Zhongyang Yanjiuyuan Jingji
Yanjiusuo, 1981).
Wu Jingping 吳景平, “近代上海金融中心地位與南京國民政府之關係” (The Status
of Modern Shanghai as a Financial Center and Its Relations to the Nationalist
Government in Nanjing), 史林 (Historical Review) 2 (2002), 90–121.
Yan Hongzhong 燕红忠, 晋商与现代经济 (Shanxi Merchants and the Modern
Economy) (Beijing, Jingji kexue chubanshe, 2012).
Yao Gongzhen 姚公振, 中國農業金融史 (Agricultural Finance History of China)
(Shanghai, Zhongguo wenhua fuwushe, 1947).
Yeh, W., “Corporate Space, Communal Time: Everyday Life in Shanghai’s Bank of
China,” American Historical Review 100.1 (1995), 97–122.
Zhang Guohui 張國輝, 中國金融通史 (A General History of Chinese Finance), vol. 2,
(1840–1911) 清鴉片戰爭時期至清末時期 (From the Opium War to the End of the
Qing Dynasty (1840–1911)) (Beijing, Zhongguo Jinrong chubanshe, 2003).
Zhang Guohui 張國輝, 晚清錢莊和票號研究 (Qianzhuang and Piaohao in the Late
Qing Period) (Beijing, Zhonghua shuju, 1989).
Zhu Yingui 朱蔭貴, “抗戰時期的上海華商證券市場” (Stock Markets in Shanghai
during the Anti-Japanese War), 社會科學 (Social Science) 2 (2005), 88–97.
Zhu Yingui 朱蔭貴, “試論近代中國證券市場的特點” (Characteristics of Stock Markets
in Modern China: A Preliminary Assessment), 經濟研究 (Economic Research) 3 (2008),
150–60.
Zhu Yingui 朱蔭貴, “近代上海證券市場上股票買賣的三次高潮” (Three High Tides
of Buying and Selling on Shanghai Stock Markets), 中國經濟史研究 (Research on
Chinese Economic History) 3 (1998), 58–70.
Zhu Yingui 朱蔭貴, 近代中國: 金融與證券研究 (Modern China: Research on Finance
and Stock Markets) (Shanghai, Shanghai renmin chubanshe, 2012).
323

10
Chinese Business Organization
madeleine zelin
In this chapter we will examine how key institutions were mobilized to

shape China’s early modern business practices under weak state engage-
ment with the economy, and a growing foreign presence. Business prac-
tices in the late imperial period rested on four pillars, each a fundamental
part of the institutional framework that structured social and economic
life. The first, family, provided templates for the utilization of capital and
labor and the mobilization of trust, tools that proved as useful for China’s
late imperial commercial economy as for the early modern economy of
industrial enterprise and global engagement. The second might be termed
the system of private ordering that served generations of Chinese mer-
chants and others in combining capital and establishing the terms of
economic interaction, often through written contracts whose provisions
established highly flexible forms of partnership that continued to form the
basis of most Chinese business until the early PRC. The third, native
place, in significant ways mirrored the intangible assets provided by ties of
kinship, offering a predetermined basis for co-operation, nurturing and
protecting group interests and skills and, like the fourth, grounding these
intangibles in very tangible organizations catering to inhabitants of
a particular city, region, or province. The fourth pillar, the guild, drew
on many of the practices associated with the other three. The guild,
however, played a special role in the ordering of Chinese manufacturing
and trade, translating China’s complex and unregulated currency system
for its merchant constituency, and mediating both the legal and the fiscal
relationships between merchants and an often weak and poorly informed
state.
These four pillars were by no means unique to China. Nor was the
preference shown by many Chinese multi-owner firms for partnerships and
324

family firms,1 although the utility of these firms in addressing specific chal-
lenges facing Chinese enterprise helps explain the particular organizational
forms that emerged during the late imperial period and persisted into the
republic. At the same time we will eschew explanations based on “culture.”
While individual actors may have understood their behavior to have
reflected cultural commitments, here we will focus on the enactment of
those commitments in the choices made and the strategies deployed to
manage the challenges of China’s particular business environments.
Inasmuch as these four pillars retained their influence well into the twentieth
century, the last section of this chapter will address some of the ways in
which the pressures of foreign contact and global economic interaction
affected long-established modes of business organization in China. New
technologies, foreign business models, borrowed legal frameworks, and the
end of imperial rule could not but have had an impact on business organiza-
tion. However, the lesson of recent scholarship on early twentieth-century
Chinese business has been that their impact was far more diverse and uneven
than we might have anticipated based on standard theories of the firm and
that many of the key elements of business practice that emerged over the
length of the last dynasty continued to serve businesspeople and the entities
they created long after emperors were but a distant memory.
Family
Family as the node of economic activity is a universal phenomenon and
family firms have continued to play a significant role, numerically, if not in
terms of total capitalization, in most parts of the world.2 While scholars no
longer consider the “family firm” to have uniquely Chinese salience, charac-
teristics of Chinese family organization played a significant role in the
structuring of Chinese business in the early modern period. As a biological
and social construct, the Chinese family provided intersecting bases for trust
and the pooling of resources. Patriarchal authority and generational hier-
archy, key Confucian expressions of a well-ordered society, were baked into
1
For example, T.W. Guinnane and J. Schneebacher, Capital Structure and the Choice of
Enterprise Form: Theory and History, Economic Growth Center Discussion Paper No. 1061
(New Haven, Yale University, 2018), p. 2, demonstrate that as late as 1890 three-quarters
of all firms were ordinary partnerships in France, Germany, and Spain, despite the
availability for some time of general incorporation.
2
A. Colli and M.B. Rose, “Family Business in Comparative Perspective,” in F. Amatori
and G. Jones (eds.), Business History around the World (Cambridge, Cambridge University
Press, 2003), pp. 339–52.
325

madeleine zelin
both the imperial code and social practice. Among their key manifestations
was the joint household, whose common property was managed by the
patriarch, and only passed on to the next generation in the act of household
division, a predominately male affair that could occur during the lifetime of
the patriarch but ideally took place after his death. Household division itself
was conceived of as an egalitarian process in which each son received an
equal share of the estate. Partible succession created a shared stake in the
common property, while deferred division could supply a common pool of
resources available for profitable deployment.3 If culture played a role in
encouraging co-operation toward growth of the estate, it did so by sustaining
the belief that each male household member was linked through time and
placed on earth to bring honor to the ancestors and to lay the groundwork for
the success of future generations.
How did these distinctive features of the joint household, its commonly
held property and the moral vision that encased it, shape Chinese business
organization? Much has been made of the family as a basis for trust, and
indeed in China, as elsewhere, the face-to-face bonds of familiarity and
mutual dependence of family contributed to the willingness of people to
lend money, broker deals, and join in business endeavors. However, it was in
its ability to exercise authority over its members and to lay claim to common
resources that the family and its joint household economy most influenced
the evolution of Chinese business organization. The case of the Guo Family
Shop, a late Qing business of no particular distinction, illustrates this point.
The shop was founded in the late Qing by a member of the Guo family of
Shangyu county, Zhejiang. The founder, who had three sons who most likely
worked alongside their father, operated the shop as a single proprietorship
until his death, when the joint household of which he was the head under-
went division. We do not know what other assets comprised the Guo estate,
but as a result of a dispute with creditors that emerged following the death of
the elder brother, we know that the three brothers emerged from the process
of division as equal partners in what was now a single-surname partnership.
While the brothers still referred to their business simply as the Guo Family
Shop, our use of the term “partnership” underscores a change in the relation-
ship among stakeholders. When the family branch headed by the youngest
brother wished to leave the business, the two remaining branches had to buy
3
For a more detailed examination of the process and implications of household division
see D. Wakefield, Fenjia: Household Division and Inheritance in Qing and Republican China
(Honolulu, University of Hawaii Press, 1998). The distribution to adopted sons and sons
of concubines, as well as the portion to which daughters were entitled, varied over time.
326

him out, leaving, as recorded in court documents, the representatives of the

senior branch (a widow and her adopted son) and the middle branch each
partners with a 50 percent stake in the shop. The litigation that has left
a record of this firm took place at a moment when the law affecting business
was being moved from the realm of custom to that of code, a process that, we
will see, took decades and never fully escaped the ambiguities created by the
legacy of the joint household. However, throughout the case, the identity of
the firm as a partnership was never in doubt.4
The process by which household division created single-surname partner-
ships is often missing from our discussion of Chinese business, but it is a key
to understanding the organization and longevity of so-called family busi-
nesses. While many firms did not survive household division, their assets
(and their debts) being distributed among the founder’s heirs, many of the
most successful firms in the late imperial period were founded by one
household head and thrived for long periods as single-surname partnerships.
Because they began as closely held businesses we often do not know much
about their organization until they were embroiled in lawsuits, or changes in
their ownership or management structure brought them to public attention
in the last years of the Qing dynasty. Ruifuxiang, wholesalers and retailers of
native cloth, was founded in the seventeenth century by the Meng family of
Shandong, and although it encountered difficult times in the eighteenth
century, it was still a family partnership when its business once again took
off in the 1870s.5 In 1669 Yue Xianyang founded Tongrentang as a purveyor of
Chinese medicine. His son opened a shop in the Qianmen commercial district
of Beijing, and by 1723 the shop, still under exclusive Yue ownership, became
purveyors of medicines to the imperial court. In 1753 a fire destroyed the shop
and the family was forced to sell shares to outside investors in order to
reopen. In the 1880s a concerted effort to buy out nonfamily shareholders
restored the business to the Yues, who re-created the ideal of the family firm
by employing only family members.6 A similar story could be told about
another famous Beijing medicine shop. Founded in the early Ming dynasty by
Yue Fengyi, Wangquantang remained solely in Yue hands until the 1740s,
when debt forced them to enter into a shareholding partnership with a family
4
M.B. Young, “Law and Modern State-Building in Early Republican China: The Supreme
Court of Peking (1911–1926),” unpublished Ph.D. thesis, Harvard University, 2004, p. 174.
5
W.K.K. Chan, “The Organizational Structure of the Traditional Chinese Firm and Its
Modern Reform,” Business History Review 56.2 (1982), 222–3.
6
S. Cochran, Chinese Medicine Men: Consumer Culture in China and Southeast Asia
327

madeleine zelin
named Suo.7 Similar stories can be told about salt merchants from Tianjin,
long-distance merchant families from Huizhou and Shanxi, and the large-
scale salt manufacturers and wholesalers of Zigong, Sichuan.
Whether they remained solely family-owned, were forced by circum-
stances to bring in outside partners, constituted themselves from the start
as shareholding partners among kin,8 or invested jointly with other families,9
family-owned businesses relied on a number of organizational devices that
emerged out of the joint household and continued to contribute significantly
to the more complex business practices of the late Qing and the twentieth
century.
Two of the most important of these devices were the lineage trust and the
lineage accounting office. To understand the lineage trust let us return to the
joint household and the practice of family division. The process by which the
assets of a household were passed on to its coparceners in late imperial China
required their fragmentation as each son received an equal share of the estate.
Even for those with middling resources this contravened the need to devote
a portion of the estate to ritual expenses incurred in honoring their ancestors.
By the Song dynasty (960–1279), households with sufficient means were
encouraged to create endowments to which were contributed land and
other income-producing assets that were protected from family division
and could grow in value, and whose management and the purposes to
which the income was put created a focus of solidarity for the descendants
of its founders. The status of these trusts, most commonly called tang 堂, was
reinforced by custom, contract, and individual lineage regulations until 1756,
7
Liu Yongcheng 刘永成 and He Zhiqing 赫治清, “万全堂的由来与发展” (The Origin
and Development of Wanquantang), 中国社会经济史研究 (Research on the
Economic and Social History of China) 1 (1983), 8.
8
For example, two branches of the Zhao family began a business which in 1838 was
expanded to included four branches, each contributing cash for a total of 12,000 taels.
Two shops sold oil, salt, and grain in the Qianmen commercial district of Beijing. In this
case the investment was passive, with management left to others while profits and
losses were shared in proportion to each branch’s investment. Yang Guozhen 杨国祯,
明清以来商人‘合本’经营的契约形式,” (The Contractual Form of Merchant “Joint
Shares” Business since Ming and Qing,” 中国社会经济史研究 (Research on
Economic and Social History of China) 3 (1987), 4.
9
For example, during the mid-nineteenth century, Du Baotian, Yuan Tinglu, and Yuan
Baozhai opened the Zhengxunxiang as a partnership that appears to have operated
a workshop producing writing paper. When Yuan Baozhai died his four shares were
sold to Zhang Xiangzhi’s Tongtaihang, a medicinal herb shop that may have simply
viewed the shares as an income-producing asset. Xie Jing 谢晶, “设有法律的秩序：
晚清巴县工商业合伙研究” (Order in the Absence of Law: Research on Partnership
in Handicraft and Commerce in late Qing Baxian), unpublished MA thesis, Central
Nationalities University, 2012, p. 21.
328

when they were also protected from intra-familial predation in the Qing
Code.10 For Chinese business they became a unique mechanism through
which to effect some of the benefits of incorporation.11
The salt manufacturers of Zigong, Sichuan provide some of the best data
on the way in which lineage trusts functioned as business entities. The men
who began investing in drilling deep brine wells and moved from evaporating
salt using coal-fired furnaces to developing China’s first industrial natural gas
wells did not start out as well-capitalized business entrepreneurs.12 Most were
farmers and small merchants, some of their families had produced degree
holders, and many dabbled in well drilling in a region of southern Sichuan
dotted with salt derricks and abandoned wells. The Wangs, who emerged as
one of the most successful of China’s salt producers and wholesalers, were
typical in their use of the institutions of the joint household economy. By the
nineteenth century they had already established a modest trust to support an
ancestral hall and ritual activities. However, it was the success in the devel-
opment of salt wells and evaporation furnaces that led the head of the senior
branch of the Wang lineage, Wang Langyun, to create a trust that would be
central to the family’s economic juggernaut. Twenty operating wells and 600
mu of agricultural land were set aside as an initial endowment, income-
producing resources that would not only support elderly lineage members,
maintain ancestral graves, memorialize chaste widows, pay for ancestral
rites, and contribute to the education of lineage males.13 Its real significance
10
Yunsheng Xue 薛允升 and Jingjia Huang 黄靜嘉, 讀例存疑重刋本 (Doubtful Points
on Reading the Substatutes) (Taipei, Chengwen chubanshe, 1970), pp. 626–7; M. Zelin,
“The Firm in Early Modern China,” Journal of Economic Behavior & Organization 71.3
(2009), 626–7.
11
A similar logic facilitated the creation of shareholding associations devoted to sustaining
temples and various ritual activities, allowing non-kin to create partnerships in land,
irrigation works and other revenue-generating assets. Shares in such trusts could form
a component of joint household property. For examples of such associations see Li Li
李力, “清代民间契约中关于‘伙’的观念和习惯” (The Concept and Common
Practice of “Huo” in Qing Civil Contracts), 法学家 (Legal Profession) 6 (2003), 41–2.
Cohen has argued that by the end of the last dynasty about a third of the “best wet rice
land” in Minong county, Taiwan was owned by associations “mainly dedicated to the
worship of gods or ancestors.” M. Cohen, “Writs of Passage in Late Imperial China,” in
M. Zelin, J.K. Ocko, and R. Gardella (eds.), Contract and Property in Early Modern China
(Stanford: Stanford University Press, 2004), p. 43.
12
For a detailed analysis of the Zigong salt manufacturers, their technological and
business contributions, and the impact of China’s changing political landscape on
their business fortunes, see M. Zelin, The Merchants of Zigong: Industrial
Entrepreneurship in Early Modern China (New York, Columbia University Press, 2005).
13
M. Zelin, “The Rise and Fall of the Furong Saltyard Elite: Merchant Dominance in Late
Qing China,” in J. Esherick and M.B. Rankin (eds.), Chinese Local Elites and Patterns of
Dominance (Berkeley, University of California Press, 1990), pp. 94–6.
329

madeleine zelin
was as a mechanism to keep intact the wealth produced by the lineage

businesses and to manage reinvestment in those businesses while satisfying
the desire for benefit by the large numbers of lineage members who would
ultimately be only passive beneficiaries. In its founding regulations, after
accounting for the above-mentioned non-business expenses, it was stipulated
that half the income of the trust would be reinvested in the trust itself and half
would be allocated to constituent lineage branches to develop their own
business and landed assets.
Over time, the management of individual business entities within the trust
would change. In particular, like many late nineteenth-century Chinese firms,
we see in the Wang lineage trust a shift to non-kin professionals in key
management positions. The lineage trust as a corporate entity not only
managed wholly owned lineage properties. In the late nineteenth century
and the twentieth it increasingly became an investor in shares in business
ventures with other households and lineage trusts. Indeed, in this respect,
long before the introduction of a formal company law in China, the lineage
trust could be seen as having a form of legal personhood. However, the
legacy of the joint household remained. Most importantly, overall manager-
ial authority continued to be vested in the senior male member of the senior
lineage branch.14 And the kinship requirement for participation in profit-
sharing remained inviolate.15
How did entrepreneurial lineage trusts manage increasingly complex
assets while continuing to maintain their commitments to coparceners
whose numbers increased with each new birth? Not surprisingly, lineage
trusts, like most Chinese businesses for which we have data, maintained
a central accounts office (zhangfang 账房). Accounts offices differed in the
functions they performed, the degree to which they exercised hierarchical co-
ordination over all business assets, and the extent to which they employed
14
While the patriarchal principle does not seem to have been questioned in earlier years,
during the 1920s it became a source of intense competition as the lineage trust suffered
from mounting debt and political pressures. For an account that clearly favors the
opposition, see Benqing Chen 陈本清, “渝沙债团与王三畏堂债务始末” (The Yusha
Debt Group and the Whole Story of the Debts of the Wang Sanwei Tang), 自贡文史资
料选集 (Zigong Selected Historical Archives) 22 (1992), 1–10.
15
Unlike other forms of partnership, which distributed profit and loss according to one’s
contribution to the firm, a subject addressed below, one’s share in the profits of this and
other lineage-trust-based businesses depended on one’s position within a lineage
branch, the number of claimants in that branch, and the generational depth of the
lineage itself. For an example of a New Territories lineage trust whose shareholders
now span the globe, see J.L. Watson, “Presidential Address: Virtual Kinship, Real
Estate, and Diaspora Formation: The Man Lineage Revisited,” Journal of Asian Studies
63.4 (2004), 893–910.
330

non-kin specialists to carry out their work. However, most served as both
a business office and the main site for the administration of lineage affairs,
under the unified authority of the lineage head. Few lineage trusts with
significant business activities have left detailed records of their operations
but their traces abound in the genealogies of notable local families. At one
extreme, for example, were the locally powerful Wen lineage of Guisheng
county in Jiangxi, who owned thirteen mountains whose tenants either
mined coal or harvested timber. Rents collected by the lineage maintained
an ancestral hall and supported lineage members with monthly allowances.16
No elaborate internal business organization was necessary to enable this
lineage to generate sufficient wealth to allow lineage members to obtain
official degrees and lineage leaders to participate in regional politics.
At the other extreme were the large lineage-based salt manufacturers of
Zigong, Sichuan. They too deployed the institution of the lineage trust as
a mechanism to manage diverse assets as well as the ritual and philanthropic
concerns of increasingly wealthy single-surname partnerships. At the height
of the salt boom in the mid-nineteenth century, the wealthiest of the salt
entrepreneurs refined the use of the trust. While maintaining their founding
trust, the Hu Yuanhe tang, in 1867 the Hus established a separate trust,
the Shenyi tang, devoted exclusively to the salt business. The main office
(zongguifang 总柜房) of the Shenyi tang was usually headed by a family
member whose title was general director (zongzhanggui 总账柜). Under him
were five departments: (1) an accounts office (guifang 柜房) run by a chief
accountant (zongzhang 总账) and two assistants (bangzhang 帮账) in charge
of the overall productivity of the lineage’s wells and furnaces; (2)
a procurement department (huowugu 货物股) in charge of purchasing all
supplies needed for the daily operation of the wells and furnaces; (3) an
external affairs department (jiaojigu 交际股) in charge of buying brine for
the lineage’s furnaces and selling salt at lineage-owned retail shops; (4) a
department of agricultural estates (nongzhuanggu 农庄股) in charge of col-
lecting rents and selling grain; and (5) a cash department (xianjingu 现金股) in
charge of daily cash expenditures and silver–copper exchange transactions.17
Despite its focus on the family’s salt business it remained an appendage of the
main lineage trust, along with separate offices devoted to lineage schools, its
16
J. Hornibrook, “Local Elites and Mechanized Mining in China: The Case of the Wen
Lineage in Pingxiang County, Jiangxi,” Modern China 27.2 (2001), 202–28.
17
Shaoquan Hu 胡少权, “贡井胡元和的兴起与衰落” (The Rise and Fall of Gongjing
Hu Yuanhe), 自贡文史资料选集 (Selected Historical Archives of Zigong) 12 (1981),
55–6.
331

madeleine zelin
agricultural estate, and a wide array of family employees. In the early

twentieth century, when the three branches that had comprised the original
Hu Yuanhe trust decided to divide their assets and establish separate ancestral
trusts, the Shenyi tang was maintained intact but its constituent businesses
were turned over to outsiders to run.18
Other salt producers developed somewhat different models based on the
lineage trust. For example, the Li Siyou tang maintained overall supervision
of its business and family interests through its main lineage trust. However,
separate managerial hierarchies were established for each business division –
furnaces, pipes, brine wells, and wholesale distribution.19 Each operated as
a separate management unit reporting to the main accounts office of the
lineage trust and its general manager, who by the end of the century was no
longer a family member. At the beginning of each year representatives of the
four branches of the lineage met at the general accounts office to hear reports
on the state of the constituent firms, review their accounts and discuss
strategy for the coming year. In some respects the annual meeting of the
four branches at the lineage headquarters resembled a typical board of
directors meeting in which shareholders had the opportunity to examine
the work of their hired managers and assess the health of the business.
However, we are reminded of the thin line between lineage and business
by the fact that each branch was allocated a fixed allowance for family and
ritual expenses each year, with a warning that at no other time were they to
withdraw money from the business treasury.20
Private Ordering
As we saw above, when kinship combined with the institutions of the joint
household, individual households did not partake of the benefits of income-
producing assets as shareholders. Their interest in the income produced
could not be augmented or sold. And the manner in which they enjoyed
the benefit of that income might shift depending on whether division was
based on total number of lineage males or on division among branches
regardless of each branch’s male population. However, both individual
joint households and lineage trusts also engaged in business as participants
18
Hu, “贡井胡元和,” 70.
19
Zilin Li 李子琳, “自流井李四友堂由发轫到衰亡” (The Rise and Fall of Ziliujing Li
Siyou Tang), 四川文史资料选集 (Selected Historical Archives of Sichuan) 4 (1962–
1963), 150–1.
20
Li, “自流井李四友堂由发轫到衰亡.”
332

in shareholding partnerships with kin and non-kin alike.21 Chinese sharehold-

ing practices drew on three of the pillars noted at the beginning of this
chapter. While kinship could create natural partnerships, it also facilitated
formal shareholding arrangements. Whereas earlier examples may exist, at
least by the sixteenth century we have contracts memorializing the creation
of long-term, jointly owned and jointly managed property in support of
purposes other than the ritual functions of the ancestral line, or that brought
together investors whose association did not arise primarily out of kinship.
Such partnerships took many forms. Drawn up in 1568 in Qimen county,
Anhui, one such document states that its signatories are engaged in joint
management of timber resources. While the surviving document is in the
form of an oath signed by three branches of the Li lineage pledging not to cut
down trees or steal timber for personal use and establishing fines to be
imposed on those who do, the forested mountain is clearly thought of as
more than a simple lineage endowment. The warning is addressed to
“households with shares and households that do not have shares” and the
mountain itself is referred to as a “jointly owned mountain.”22 A 1432 contract
from Xiuning county, Anhui is even more explicit in its description of
a partnership in mountain resources. In this case three men of different
surnames established joint ownership of a mountain, allocating six shares
to Xie Dexiang, and one share each to Wu Yanduan and to Li Zhongjie and
his unnamed brother.23 Cohen describes an innovative family division in
Minong county, Taiwan, whereby instead of distribution of patrilineal
resources and the establishment of new household properties by each of
three brothers of the Liu family, the property accumulated by their father was
kept intact and each brother was assigned a number of shares in a jointly
owned sugar plantation. Farmed by a tenant, the brothers then received
a portion of the rent according to their shares.24 Allee points to similar
partnerships among entrepreneurial tea producers in late Qing Taiwan. In
21
For a detailed discussion of the roots of Chinese shareholding practices, see M. Zelin,
“A Deep History of Chinese Shareholding,” Law and History Review 37.2 (May 2019),
325–51.
22
Zhang Chuanxi 张传玺 (ed.), 中國歷代契約會编考釋 (Compilation and
Interpretation of Chinese Historical Contracts), 2 vols. (Beijing, Beijing daxue chu-
banshe, 1995), pp. 1078–9. A similar document in the unpublished collection of Huizhou
contracts compiled by the late Tian Tao memorialized a dispute over the exploitation
of forest land for firewood and resulted in the division of a jointly owned mountain into
a mountain in which owners had specified shares. Tian Tao collection, Ming Wanli 14,
Wang Hongqing et al., Qingjie Hemo contract (not published).
23
Zhang, 中國歷代契約會编考釋, pp. 1089–90.
24
Cohen, “Writs of Passage in Late Imperial China,” 82.
333

madeleine zelin
one case land was first purchased from the original aboriginal owners, and
then a partnership consisting of six shares was created, in which he estimates
approximately half of the shares were sold to non-kin.25
While kinship and/or propinquity/native-place relationships facilitated
the creation of trust that enabled these partnerships to be formed, China’s
strong tradition of contractually based private ordering served as the scaf-
folding upon which a robust shareholding tradition was built. Indeed, it is
through the rich collections of business contracts collected by Chinese
scholars and preserved in Chinese legal archives that we know much of
what we know about the organization of Chinese business.
Contracts memorialized the establishment or dissolution of a partnership
and the sale or transfer of shares, and were intended to serve as actionable
proof of the intent of investors. Partnership contracts followed patterns that
can be traced to an even deeper history of land sale and lease contracts going
back to the early years of the Chinese state. Models were recorded in
encyclopedias and merchant guidebooks,26 and by the nineteenth century
merchants seem to have deliberately emulated elements of the organiza-
tional models of prominent merchant communities, in particular the far-
flung traders from the western province of Shanxi. While varying in their
details, most founding contracts named the investors and their respective
contributions in cash and kind and the number of shares to which each was
entitled. Common also was the stipulation of a timetable for the settling of
accounts, at which the firm’s ledgers would be inspected and dividends
determined and distributed. By the nineteenth century, it was common for
founding documents to clearly state that profits and losses were allocated
according to the number of one’s shares or the proportion of total capital one
had invested. Some founding documents indicate that a manager, who may
or may not have been an investor, was engaged and named at the time the
partnership was formed.27
Before we look at specific contracts and what they can tell us about
business organization in the late imperial period it is worth noting what the
contract culture of China bequeathed to all Chinese shareholding partner-
ships. First, while the imperial code and state policy were concerned with the
25
M.A. Allee, “The Status of Contracts in Nineteenth-Century Chinese Courts,” in Zelin,
Ocko, and Gardella, Contract and Property, pp. 162–3.
26
R.J. Lufrano, Honorable Merchants Commerce and Self-Cultivation in Late Imperial China
(Honolulu, University of Hawaii Press, 1997).
27
We see both of these stipulations in a model contract in the 1895 compilation Shuji
Bianmeng translated in R. Gardella, “Contracting Business Partnerships,” in Zelin,
Ocko, and Gardella, Contract and Property, p. 332.
334

maintenance of patriarchal authority, status, and generational hierarchy

within households, they made almost no effort to regulate economic rela-
tionships that fell outside the joint-household framework.28 This meant that
contracts in China were in a real sense the record of the will of the parties
who wrote them and were primary evidence in litigation over real and
business property and obligations.29 Unless otherwise stipulated in the agree-
ment creating a shareholding partnership, shares could be freely bought and
sold.30 At the same time, the absence in China of a law requiring the
dissolution of partnerships upon the death or withdrawal of a shareholder
meant that Chinese businesses could survive numerous shifts in the compos-
ition of ownership. Indeed, a subset of partnership contracts could be cat-
egorized as share reorganization contracts as new partners were brought in
and old ones left.31 These two characteristics of shareholding practice help
explain the longevity of some of the family firms we have already examined,
particularly as they took on outside partners in times of financial stress, and in
some cases, like the Yues of the Tongren tang medicine firm, consolidated
family ownership when business turned around in the nineteenth century.
The flexibility afforded by shareholding in the absence of legal regulation
may have combined with the underdevelopment of formal credit institutions
in encouraging firms like the Tongren tang to rely on equity expansion when
in need of infusions of capital.
What can surviving partnership contracts tell us about the organization of
business in the late imperial period? First, we can see that investment took
many forms and had an impact on the relative rights and obligations of
partners. Particularly in industries that relied on natural resources, including
28
The state did attempt to regulate the freely contracted relationships that fell within the
kinship rubric, with varying success, as Sommer and Ransmeier show in their respect-
ive work on polyandry and the sale of wives and on the transactional family and the
market in people in Qing and Republican China. M. Sommer, Sex, Law, and Society in
Late Imperial China (Stanford, CA, Stanford University Press, 2000); Sommer, Polyandry
and Wife-Selling in Qing Dynasty China: Survival Strategies and Judicial Interventions
(Oakland, University of California Press, 2015); J. Ransmeier, Sold People: Traffickers
and Family Life in North China (Cambridge, MA, Harvard University Press, 2017).
29
For a discussion of contract in the establishment of property rights during the last
imperial dynasty, see M. Zelin, “A Critique of Rights of Property in Pre-war China,” in
Zelin, Ocko and Gardella, Contract and Property, pp. 17–36.
30
Some contracts stipulated a right of first refusal for fellow shareholders. See, for
example, K.M. Bun, “Custom, Code and Legal Practice, the Contracts of Changlu
Salt Merchants in Late Imperial China,” in Zelin, Ocko, and Gardella, Contract and
Property, p. 279.
31
Share reorganization could take many forms, ranging from simple transfer to the
complex reallocation of profit and losses found in Zigong shangxiajie agreements.
Zelin, Merchants of Zigong, pp. 42–8.
335

madeleine zelin
land, it was common to find businesses established between owners of

underexploited real property and entrepreneurs eager to develop mining
and commercial agriculture. This contract from the Mentou gou coalmines
outside Beijing was concluded in 1655 but already has many of the character-
istics of such businesses:32
The writers of this contract to form a partnership to open a pit are Wang
Conglian and others [these would include the signatories and the landowners
listed below]. In previous years we opened the Daxing pit at Jingming si.
Now are going to reopen the pit, but we are short of capital. We have come
to an agreement with the middleman Zhang Yingji to contact the Sun and
Ma families to put up money and enter into partnership. The pit will be
divided into forty-five shares. The owner of the land [on which the pit is
located], Ming Xiangxing, will receive five shares. The Zhang and the Wang
families will receive twenty-four shares. The Sun and Ma families will receive
sixteen shares. When the pit begins to produces coal, we will first deduct and
pay the new costs of production and deduct the Wangs’ old costs of produc-
tion in the amount of 10,300 cash. Once [production costs are] deducted and
paid, if there are profits they will be divided equally [according to shares?].
This is everyone’s wish and there are no regrets. Fearing that in the future
there will be no evidence, we are writing this contract and two copies will be
kept as proof.
Dated Tongzhi 12, 8, 3 [followed by signatures
of the parties and a middleman]
In an earlier partnership the Wangs and the Zhangs had put up capital and
joined forces with landowners sitting on potentially rich coal reserves to
excavate a mine. While they were possibly neighbors, they give no indication
of kinship. It is likely that the two households financing the excavation ran
out of money before they were able to start production and, with the help of
someone who had connections to people with liquid assets, they sought to
dig further. We do not know what the contribution of each party was, but
shares were most likely allocated on the basis of relative cash contributions.
Although we lack some information, this and other early coal mining
partnerships were important precursors to more complex business arrange-
ments in extractive industry. At Baxian, Sichuan, where coal lay close to the
surface and the costs of production were low, coal mining partnerships often
consisted of no more than a few local men who leased a potential mine from
32
Zhiping Chen 陈支平 and Zengrong Lu 卢增荣, “从契约文书看清代工商业合股委
托经营方式的转变” (The Transformation in Qing Joint-Stock Trust Management as
Viewed from Contracts), 中国社会经济史研究 (Research on Chinese Social and
Economic History) 2 (2000), 29–30.
336

a neighborhood farmer.33 We find similar arrangements among parties in the

timber industry. Indeed, Meng Zhang has shown that indigenous Miao
people participated in the commercial timber industry in southeastern
Guizhou by emulating landowner–planter share contracts used by Chinese
to develop timber plantations.34 At its most sophisticated, the combination of
landowner shares and investor shares is found in the Zigong salt industry,
where the development of what became one of late imperial China’s most
capital-intensive industries began when entrepreneurial middlemen joined
forces with local landowners and both local and extraprovincial investors to
exploit southern Sichuan’s rich brine deposits. Here the crucial role of the
middleman in bringing together investors and landowners, and in supervising
the initial drilling of the brine well, led to the granting of partnership shares to
the middleman as well as to the contributors of factors of production. At
Zigong we can see the gradual shift in importance of land versus capital
reflected in the structure of shareholding. Initially many investor partnerships
in effect rented potential well sites from landowners who hedged their bets
by retaining the ability to return their land to farming. By the mid-nineteenth
century, landlords had in effect relinquished control and received
a decreasing share in the well being drilled on their land as capital expenditure
on drilling deep wells increased.35
The granting of shares in the partnership to non-investors is seen as one of
the most important contributions of the Shanxi merchants mentioned above,
whose businesses ranged from wholesale distribution of salt and other
commodities, to retail marketing, to the creation of some of China’s earliest
banks for long-distance remittance of funds. “Body shares” addressed both
the problem of short-term working capital discussed below and principal–
agent problems of particular interest to merchants with business branches in
distant markets. Silver shares (yingu 银股) went to the investors whose
capital created the firm. Body shares (shengu 身股) were granted to those
hired staff, branch managers, and even clerks who rose to the top of their
ranks. The shift from salaried employee to equity stakeholder was seen as
33
For example, Yang Yingquan and Yao Chengxiu rented a piece of land from the Jinyun
temple and sublet part of it to a man named Zhang, who mined it with one hired
laborer. Baxian Archives, 6.3.16552
34
M. Zhang, “Financing Plantation Forestry in Southwest China: Securitization of
Timberlands and Shareholding Practices, 1700–1900,” AAS New Frontiers in Asian
Economic History Workshop, East Lansing, Michigan, May 11–15, 2017, 10–14.
35
For a detailed discussion of the allocation of shares among landowners, middleman
developers, and owners of capital and other critical inputs in Zigong well partnerships,
see Zelin, The Merchants of Zigong, pp. 24–49.
337

madeleine zelin
a way to ensure honesty and loyalty and continued hard work. Every three
years the accounts of the firm were settled and all holders of body shares,
including staff, received a portion of the profits commensurate with their
shares.36 Chinese scholars have noted that over time the number of body
shares could exceed the number of investor or silver shares. However, there
are no studies of the impact this may have had on the viability of the firm.
The practice of issuing body shares continued into the mid-twentieth cen-
tury, as noted in accounts of Shanxi merchants in Manchuria written in the
1940s.37 In their study of the Rishengchang, one of the early Shanxi remittance
banks, Randal Morck and Fan Yang argue that managers who were granted
body shares received dividends, but unlike equity shareholders they had no
claim to the equity of the firm, nor could their shares be inherited. However,
their kin did receive a death benefit of sorts, in the form of term-limited
dividend-generating death shares (gugu).38
The practice of ensuring the loyalty of managers by giving them a share in
the firm can be seen elsewhere as well. For example, in 1878 Qu Fulu and Li
Yingzhou wrote an agreement stating that they had purchased a fully
equipped rice mill for 1,700 taels.39 Together the partners would hold ten
shares. Qu Fulu would get 6.5 shares for an investment of 1,105 taels. Li
Yingzhou would get 3.5 shares for an investment of 595 taels. Two additional
shares were established for a total of twelve. One was designated the
property in perpetuity of the mill’s manager, Jiang Rongzhuang. The other
was to be assigned to the God of Wealth,40 a rather recent object of merchant
devotion and mercantile virtues.
The assignment of a share to the God of Wealth reminds us of another
problem addressed in shareholding partnerships, that of short-term finance
and working capital. There is a mistaken notion that the absence of modern
36
Xu Ke 徐珂, Qingbai Leichao 清稗类钞, 17 ce, Nongshanglei, 70–1, cited in
Jianhui Huang 黃鉴暉 and Zhongguo Renmin Yinhang Shanxisheng Fenhang 中国
人民银行山西省分行 (eds.), 山西票号史料 (Historical Archives of Shanxi Piaohao)
(Taiyuan, Shanxi Jingji chubanshe, 2002), p. 582.
37
Wu Xiyong 吴希庸, “近代东北移民史略” (A Brief History of Migrants to the
Northeast in Modern China), 东北季刊 (Northeast Quarterly) 38 (1941), 219–342.
I want to thank Martin Fromm for bringing this to my attention.
38
R. Morck and F. Yang, “The Rise and Fall of the Rishengchang Bank Model: Limiting
Shareholder Influence to Attract Capital,” in J.G.S. Koppell (ed.), The Origins of
Shareholder Advocacy (New York, Palgrave Macmillan, 2011), pp. 195–8. Unfortunately,
the one example they provide is from the 1940s so we do not know whether the practice
of issuing death shares was the result of foreign influences.
39
Li, “清代民间契约中关于‘伙’的观念和习惯,” 43.
40
R. von Glahn, “The Enchantment of Wealth: The God Wutong in the Social History of
Jiangnan,” Harvard Journal of Asiatic Studies 51.2 (1991), 651–714.
338

banks slowed Chinese industrialization by depriving businessmen of long-

term finance. Banks were not an important source of long-term venture
capital in the early modern West any more than they were in China. While
China’s so-called native banks (qianzhuang 钱庄) played a vital role in
clearing commercial accounts and both native banks and remittance banks
issued commercial paper, Chinese business partnerships had to find other
ways to provide liquidity to pay wages, purchase supplies, and market goods.
In the founding document of the rice mill above, it is likely that the one share
of profits set aside for the God of Wealth was really meant to be a form of
retained earnings for just such a purpose. Replacing wages with dividends for
senior staff also served this purpose.
We have other evidence that manufacturing firms did try to anticipate
ongoing expenses and working capital. Some, such as those formed to mill
sugar in nineteenth-century Taiwan, appear to have anticipated annual
infusions of working capital from shareholders who expected their advances
to be more than offset by profits at the end of the milling season.41 Once
again, shareholding practices mitigated the need for short-term credit. At
Shanxi firms, partners were sometimes called upon to contribute fuben 副本,
a form of deposit that remained in the firm, accrued interest, and could be
used if needs arose, but was not treated as equity investment and did not earn
dividends.42 One of the jobs of the middleman/manager at the brine wells in
Zigong was to ensure that equity partners made periodic payments to cover
the cost of drilling and that no profits were divided from the sale of brine
before a well was fully operational.43 On the other hand, not all businesses
were able to solve the problem of working capital. Robert Eng describes late
nineteenth- and early twentieth-century silk manufacturers who formed
partnerships to build filatures. Upon completion and finding themselves
lacking working capital, they often leased the factories to other partnerships
who actually produced yarn.44 Eng attributes this practice to weaknesses in
capital markets but acknowledges that particular characteristics of the silk
41
C.M. Isett, “Sugar Manufacture and the Agrarian Economy of Nineteenth-Century
Taiwan,” Modern China 21.2 (1995), 244.
42
Zhongmin Zhang 张忠民, 艰难的变迁，近代中国公司制度研究 (A Difficult
Transition: A Study of the Chinese Company) (Shanghai, Shanghai Academy of Social
Sciences Press, 2001), pp. 28–30.
43
44
R.Y. Eng, “Chinese Entrepreneurs, the Government, and the Foreign Sector: The
Canton and Shanghai Silk-Reeling Enterprises, 1861–1932,” Modern Asian Studies 18.3
(1984), 360–1.
339

madeleine zelin
market also contributed to partnerships treating filatures more as real estate

than as integrated manufacturing firms.
While our discussion of partnerships gives the impression of a business
world of active investor managers, late imperial Chinese were also well
acquainted with the roles of passive investors. On the one hand, foreign
observers and Chinese surveys of local customs point to the propensity of
people with small amounts of capital to invest as dormant partners, with
limited or no liability for debt. In a debate over Chinese notions of liability,
George Jamieson, a late nineteenth-century observer of Chinese business
practices based in China, noted that China had the equivalent of the dormant
partnership in which some partners did not play an active role in the firm and
were not liable for its debts.45 On the other hand, we have seen that as long as
the Chinese household was by law and custom a joint economy, any
individual investor named on a partnership agreement really represented
a group of closely related kin. What we have not foregrounded is that a large
number of investors in Chinese business were and would continue to be
shareholders in other businesses as well as the larger corporate bodies
represented by lineage trusts.46 Cross investment was also characteristic of
so-called liaison stores (lianhao 联号) in which investors and managers of one
store invested with those of other stores to establish new stores, often
diversifying their core business and providing a hedge against business
failure.47 The world of Chinese business was thus one in which tangled
webs of lineage trusts and multi-member households, as well as individual
businesses, invested in other businesses, local and sometimes distant, in
related and at times in very different trades. Diversified portfolios were
important hedges against loss, and cross investment, even in the late imperial
period, provided an important mechanism for horizontal and vertical
integration.
Native-place affiliations and guilds facilitated the business development
described above. Regional merchant groupings could be based on common
origins in a county, town, province, or region. Among the most successful
45
“Chinese Partnership: Liability of the Individual Members,” Journal of the North China
Branch of the Royal Asiatic Society 21.2 (1887), 48.
46
Elisabeth Köll notes that most of the shareholders in Dasheng Cotton Mill were other
businesses or lineage trusts. E. Köll, From Cotton Mill to Business Empire: The Emergence of
Regional Enterprises in Modern China (Cambridge, MA, Harvard University Asia Center,
2003), p. 128–9. A variation on cross investment has been identified in the liaison firms
47
M.-H. Lin, “Interpretive Trends in Taiwan’s Scholarship on Chinese Business History,
1600 to the Present,” Chinese Studies in History 31.3–4 (1998), 70–1. Most research on
lianhao has focused on Manchuria.
340

merchants of the late imperial period were long-distance traders from the
northwest provinces of Shanxi and Shaanxi and from Huizhou in Anhui
province, both groups benefiting from the opportunities presented by the
government system of licensing the long-distance trade in salt. Other regional
merchant groups are better known for their role in the expansion of China’s
maritime trade, particularly those based in Fujian and Guangdong. According
to Cheong, most of the prominent Guangdong merchants who comprised
the so-called Cohong that enjoyed a government monopoly in the Canton
trade with Western merchants traced their origins to Fujian and Zhejiang and
“went to their previous contacts in the producing regions to obtain their
export goods, finished the silks in their own Hongs, and used the same
network for the redistribution of imports.”48 To these long-distance mer-
chants could be added smaller but no less significant groups, such as the
Ningbo merchants who, like many of their counterparts from other prov-
inces, engaged in long-distance trade in foodstuffs, grain, and other staples, as
well as playing a critical role in the development of commercial banking in
southern China.
Not surprisingly, native place affinities became an important source of
trust and facilitated the kind of investment and cross investment mentioned
above. During the last dynasty a critical mass of fellow countrymen in
a distant market eased relations with local officials. There is also evidence
that the ability to deposit cash in the shops of fellow countrymen in distant
markets anticipated the development of remittance banking and the use of
new instruments of exchange.49 Native-place ties also served as a conduit for
the exchange of market information and the spread of organization strategies.
At a more local level, well into the twentieth century, propinquity also served
as a lubricant for entrepreneurs seeking investors in new business ventures.
Men like Chen Qiyuan utilized local ties to obtain credit and labor to start one
of China’s first steam silk filatures in his native Nanhai county, Guangdong.50
Sun Yingde, a blacksmith from the Canton region working around
the Shanghai docks, joined forces with a relative to repair steamships,
eventually building a machine and machine tools factory that employed
48
W.E. Cheong, The Hong Merchants of Canton: Chinese Merchants in Sino-Western Trade
(London, Curzon Press, 1996), p. 376.
49
S. Guo, “The Shanxi Merchants in Beijing in the Qing Dynasty: An Analysis Based on
136 Samples of Merchants and Their Activities,” Frontiers in History in China 4.2 (2009),
167–9.
50
D. Ma, “Between Cottage and Factory: The Evolution of Chinese and Japanese
Silk-Reeling Industries in the Latter Half of the Nineteenth Century,” Journal of the
Asia Pacific Economy 10.2 (2005), 203–5.
341

madeleine zelin
some 200 men.51 Indeed, investors in some of China’s most famous industrial
conglomerates, such as Zhang Jian (Dasheng Cotton Mill) and Fan Xudong
(Jiuda Salt Refinery), discussed below, were friends, family, and local
acquaintances.
Guilds served a different purpose. Although they sometimes originated as
native-place associations established to serve the needs of officials, scholars,
and long-distance merchant sojourners in far-flung parts of the empire, by the
end of the Qing many guilds had become specialized agencies joining
members of particular trades. Guilds expedited the system of private ordering
discussed above by authorizing and promulgating “customs of the trade” and
enforcing business standards among their member firms. Their ability to play
this role depended not only on the growing wealth and influence of more
prosperous merchant communities. The state itself came to rely on guilds to
collect taxes, to regulate the value of the tael in market transactions, and to
mediate merchant disputes. These roles would continue to be of importance
into the twentieth century.52
A New Deal for Business?

Beginning in the sixteenth century, Chinese merchants had increasing oppor-
tunities to interact with Western counterparts, who brought with them
Western modes of business organization. These early contacts do not appear
to have upended prevailing practices on either side. While Western partici-
pants in the Canton trade benefited from the development of the commenda,
similar modes of investment in long-distance trade had existed in China as
early as the Song dynasty, and Chinese familiarity with them encouraged
some Chinese to invest in Western trading ventures.53
With the end of the Opium War in the early 1840s and the loosening of
restrictions on foreign trade at Chinese ports, opportunities to engage with
foreign businessmen expanded. Foreigners took up residence in increasing
51
W.K.K. Chan. “Sources of Capital for Modern Industrial Enterprises in Late Ch’ing
China,” in R.A. Brown (ed.), Chinese Business Enterprise: Critical Perspectives on Business
and Management (London and New York, Routledge 1996), p. 51.
52
K.-C. Liu, “Chinese Merchant Guilds: An Historical Inquiry,” Pacific Historical Review
57.1 (1988), 15–18.
53
R. Harris, “The Institutional Dynamics of Early Modern Eurasian Trade: The
Commenda and the Corporation,” Journal of Economic Behavior & Organization 71.3
(2009), 612. For a detailed discussion of the business interactions of Chinese and
Western Traders in the Canton region, see P.A. van Dyke, The Canton Trade: Life and
Enterprise on the China Coast, 1700–1845 (Hong Kong and London, Hong Kong University
Press, 2005).
342

numbers as ports opened to international trade by treaty and Chinese

continued to invest in foreign ventures. David Faure cites a Chinese study
that claims that as much as 70 percent of investment in Western shipping in
China may have been financed by Chinese, while Hao Yen-ping argues for
a 50 percent position for Chinese capital in Western insurance ventures in
China.54 At the same time, Chinese merchants began to participate in foreign
business as so-called compradors. These Chinese agents of foreign firms acted
both as liaisons with Chinese business and primary producers, and as entre-
preneurs in their own right, constituting a key component of the investors in
Chinese efforts to develop modern industry, commerce, and transportation.
However, until the 1860s these contacts did not suggest a template for
a dramatic reimagining of Chinese business organization.
Two simultaneous events changed the meaning that Western business organ-
ization would have for China. One was the end of the Taiping Rebellion in 1864.
While Chinese officials had the opportunity to observe British steamships and
armaments during the Opium War it was not until the Taiping Rebellion that
the confluence of enhanced authority on the part of powerful provincial officials
and the opportunity for Chinese forces to use foreign ordnance created the
impetus for large-scale, albeit regionally based, programs directed at Chinese
military modernization. These efforts, known collectively as the Self-
Strengthening movement, began with Chinese investment in China’s first
modern arsenals and shipyards and soon stimulated efforts to produce steam-
ships for civilian use as well as a variety of manufactured products.
The legacy of the Self-Strengthening movement is credited with some of
China’s earliest forays into military modernization, the development of mechan-
ized mines, steam transportation, and the industrial production of textiles and
other consumer goods. Its contributions to business organization were fleeting,
but hold some lessons for the relationship between business practices before and
after the intensification of China’s economic engagement with the West. Most
important were the aforementioned efforts by provincial officials, in particular
men like Li Hongzhang, who, as organizer of the regional Huai Army, was
instrumental in defeating the Taiping rebels. Following the war Li Hongzhang
was rewarded with the civilian post of governor of Jiangsu and later governor-
general of Hunan and Hubei. These posts, when combined with the political and
fiscal authority he had accrued during the war, allowed him and other governors
54
D. Faure, China and Capitalism: A History of Business Enterprise in Modern China
(Hong Kong, Hong Kong University Press, 2006), p. 51; Y.-P. Hao, The Comprador in
Nineteenth Century China: Bridge between East and West (Cambridge, MA, Harvard
University Asia Center, 1970), pp. 252–3.
343

madeleine zelin
with similar histories to take a leading role in introducing modern technology

and production methods in both the civilian and military spheres.
Most of the modern industrial projects founded during the 1860s, shipyards
and arsenals, were government-funded and government-managed.55 By the 1870s
the focus of provincial development efforts shifted away from defense. One
“innovation” emerging from the new focus on mining, shipping, textiles, and
consumer goods was the government-supervised–merchant-managed or guandu
shangban 官督商办 enterprise. The guandu shangban enterprises of the 1870s and
1880s were envisioned as joint-stock companies. Merchants (and others) were
offered the opportunity to buy shares in return for dividends, often set at a fixed
rate as an inducement to investors.56 Sale of shares, even to parties without
strong personal ties, as a mode of finance and distribution of profits and losses
was not new. However, guandu shangban enterprises pioneered the public
offering, including their publication in the growing number of newspapers that
emerged in China in the 1870s and 1880s. The model for investment prospectuses
(zhaogu zhangcheng 招股章程) may have been that of the China Merchants
Steam Navigation Company. However, hundreds of these solicitations could be
found in newspapers like Shenbao by the early 1780s.57 As Wellington Chan has
noted, the term guandu shangban had a precedent in the organization of the salt
gabelle, and the employment of hired managers was common in Chinese
partnerships and even family businesses.58 Businesses patronized by Li
Hongzhang, such as the China Merchants Steam Navigation Company, the
Kaiping Mining Company, and the Shanghai Cotton Cloth Mill, all had merchant
managers, often former compradors in Western firms. These managers operated
with relative independence from their official patrons and their shareholders
were largely passive recipients of dividends, playing little role in oversight over
their investments. Managerial practices that included the movement of funds
between enterprises sponsored by a particular official patron, as well as more
questionable expenditure of firm funds for personal use, had an impact on the
55
See Chapters 5 and 6 of this volume.
56
Guaranteed rates of dividend payments, often around 8 percent of investment, blurred
the line between loans and equity investment, as did the term applied to these
dividends, “official interest” (guanli 官利). Government involvement in these busi-
nesses could provide tangible benefits, as in the promise to ship government tribute tax
rice exclusively on ships of the China Merchants Steam Navigation Company. S.
R. Halsey, “Sovereignty, Self-Strengthening, and Steamships in Late Imperial China,”
Journal of Asian History 48.1 (2014), 81–111.
57
See, for example, the zhaogu zhangcheng calling for investors in the Pingquan
Mining Bureau published in Shenbao 申报, 6th month, 11th day, 1882.
58
W.K.K. Chan, Merchants, Mandarins, and Modern Enterprise in Late Ch’ing China
(Cambridge, MA, Harvard University East Asian Research Center, 1977), pp. 70–1.
344

performance of guandu shangban enterprises and the willingness of merchants to

invest in their shares.59 Some of these practices, particularly cross subsidization of
firms connected through a family interest or a particular entrepreneur, however,
were part of the managerial repertoire of traditional Chinese business overall
and, as we will see, would continue into the twentieth century.
Another impetus toward change was the British adoption of the Joint-
Stock Companies Act in 1856 and the extension of most of its provisions to
Hong Kong in the Company Ordinance of 1865. By requiring accounting
transparency, annual shareholders’ meetings, and other shareholder protec-
tions, the new British laws (and similar laws emerging out of other European
countries doing business in China) could have addressed some of the prob-
lems that emerged under guandu shangban and later under more clearly
private entities. For British subjects doing business with subjects of the
Qing empire, the most important issue raised by these new laws concerned
limited liability and the handling of company liquidation, inasmuch as China
did not have a similar law and foreign courts had no jurisdiction over Chinese
investors in the European and American companies.60 The absence of
a company law in China was a major factor leading to British pressure for
Chinese legal reform. However, even before the promulgation of a company
law, Chinese private businesses began to partake of what they perceived as
useful elements of the Western business form. Often these adaptations were
superficial. For example, during the Self-Strengthening period businesses
established under government auspices had adopted the term ju 局, or
bureau, symbolic of their ties to the state bureaucracy and older corporate
forms, while most private businesses continued to use terms like hang 行 or
hao 号 as suffixes to business names. By the 1870s, the term gongsi 公司 or
“company” had come to signify a new understanding of the private business
firm and in 1875 the emperor himself abandoned the term ju in favor of gongsi
in his call for officials to encourage the development of business.61
59
Chan, Merchants, Mandarins, and Modern Enterprise, pp. 72–5, attributes the corruption
associated with these firms largely to the influence of Sheng Xuanhuai, an official who
took over supervision of Li Hongzhang’s business initiatives in the 1880s. See also
A. Feuerwerker, China’s Early Industrialization: Sheng Hsuan-Huai (1844–1916) and
60
E. Motono, Conflict and Cooperation in Sino-British Business, 1860–1911: The Impact of the
Pro-British Commercial Network in Shanghai (New York and London, St. Martin’s Press,
2000), pp. 66–72.
61
Liufang Fang 方流芳, “公司词义考：解读语词的制度信息” (The Etymology of
Gongsi: Institutional Information from the Reading of Words), Sino-Foreign Legal
Studies 3 (2000), 277–99. Fang argues that the term gongsi originally was understood to
apply only to the British East India Company.
345

madeleine zelin
A formal market in Chinese shares did not emerge until after the fall of the
Qing dynasty.62 However, Chinese businessmen were quick to imitate the
foreign use of advertisements for public share offerings. Founded by foreign-
ers in 1861, Shanghai xinbao, the first Chinese-language commercial journal,
was devoted almost entirely to advertisements for shares. By the 1870s
Chinese entrepreneurs were soliciting investors by publishing prospectuses
in major newspapers like the Chinese-language newspaper Shenbao. These
investment prospectuses (zhaogu zhangcheng) give us a glimpse of some of the
organizational aspirations of firms as they sought to create an impersonal
market for shares.
A prospectus soliciting investment in the Hangzhou Electric Light
Company appeared in 1897 in the Jicheng bao.63 As a wholly private public
utility that requested and received a monopoly to establish electric lighting in
that city it was under somewhat greater official scrutiny than would be other
kinds of endeavor. But its organizational provisions reflect the prevailing
practices of businesses reaching out to stranger-investors in the late nine-
teenth century.
The Hangzhou Electric Light Company sought investors with the goal of
raising approximately 150,000 taels to buy equipment for the manufacture of
electric lamps and cable and to operate coal yards and warehouses. Among
the organizational commitments made in the prospectus was the selection of
a ten-man board of general directors from among those holding the largest
number of shares and election from their ranks of one general manager and
one assistant manager. The managers would meet and deliberate with the
directors on important matters, requiring six out of the ten directors’
approval for major decisions. Income and expenditure were to be handled
exclusively by the company accounts officer, who would compile a summary
accounting each month, quarterly closings, and an annual general settlement
of accounts. After the payment of the shareholders’ guaranteed dividend
(guanli), salaries and wages, and other expenses, any surplus would be set in
a reserve fund and would go to providing bonuses for the founders as well as
additional dividends for shareholders.
Unlike traditional shareholding firms in which proof of investment was
based on holding a copy of the founding contract, the Electric Light
Company would follow the Western practice of issuing share certificates.
62
W.A. Thomas, Western Capitalism in China: A History of the Shanghai Stock Exchange
(Aldershot and Burlington, VT, Ashgate, 2001), pp. 245–7.
63
“杭州電燈公司招股章程” (The Hangzhou Electric Light Company Share-Offering
Articles of Association), Jichengbao 8.12 (1897), 25–29.
346

But like traditional firms, if a shareholder wished to sell his shares, he would
have to offer them first to the existing shareholders. If there were no takers he
could offer his shares to outsiders, informing the company so that the names
of shareholders would remain up to date. Finally, as a confirmation of the
role that personal relationships of family or native place traditionally played
in hiring, the prospectus assured investors that no one would be hired as
a sinecure, that all persons recommended to work for the company would be
vetted by the company managers, and that all personnel with responsibility
for finances or materials would be required to have a guarantor. In the event
that the managers recommended expansion of the business they would be
required to call a meeting of all shareholders, providing notice a month in
advance by placing an announcement in Shenbao and other newspapers.
The Hangzhou Electric Light Company was typical of firms founded in the
last years of the nineteenth century to introduce new and often expensive
technologies in the area of mechanized production and transportation. Their
large capital requirements differentiated them from earlier shareholding
partnerships. But their business models would not have seemed alien to the
merchants they sought to attract to their shareholder registers. These firms
were founded before the Qing dynasty provided a legal basis for incorpor-
ation. Promulgated in 1904, China’s first company law (Gongsi lü) acknow-
ledged the juridical status and new names for existing business forms, such as
the single proprietorship, the simple partnership, and the unlimited com-
pany, and for the first time provided the option for Chinese firms to register
as limited companies.64 While fairly rudimentary, the new law introduced
ideas that in various combinations would be taken up by twentieth-century
business founders, including limited liability, the rights and obligations of
shareholders, the establishment of formal boards of directors, and rules to
encourage financial transparency, such as the requirement that all firms
engage auditors independent from management. The new law also memor-
ialized practices that existed under the regime of private ordering, such as the
freedom to buy and sell shares unless otherwise stipulated in the founding
document, issuing shares as bonuses, and the ability of a company to assess
shareholders for additional funds.
What impact did the company law have on the organization of business?
Some scholars have argued that the impact was negligible, as evidenced by
the small number of firms that registered as limited-liability joint-stock
64
Shanghai Municipal Archives (ed.), 旧中国的股份制 (1868–1949) (The Joint-Stock
System in Old China) (Shanghai, Zhongguo dangan chubanshe, 1996), pp. 11–24.
347

madeleine zelin
companies.65 On the other hand, many businesses borrowed from the reper-
toire of organizational provisions found in this and subsequent company laws
without registering with the government. The absence in China of legal
restrictions on the longevity of firms that were not incorporated made
limited liability one of the few benefits available only through registration
as a corporation under the new law. By the early twentieth century, regis-
tered and unregistered firms were including in their founding documents
stipulations on the conduct of shareholder meetings, the formation of boards
of directors, the duties of managers, and so on, and using mechanisms such as
partnership contracts and ledgers to regulate shareholder rights and keep
track of the transfer of shares.66 At the same time, Chinese businesses in the
early twentieth century remained hybrid institutions, each making use in its
own way of earlier and more recently introduced managerial and organiza-
tional techniques.
Business Organization in Flux

The early twentieth century was a period of experimentation, especially for
those firms engaging in the production and sale of new kinds of products,
making use of new technologies, and expanding their reach into new mar-
kets. In some instances, family investment and control were central to the
business model and remained at its core despite expansion in both the scope
of business and its geographical location. The Meng family continued to run
its Ruifuxiang as a closely held family firm, working through “a personalized
network of trusted subordinates and a formal structure of meetings and
reports.”67 By the twentieth century it had expanded from retail and whole-
sale trade in native cloth to opening stores in several cities in eastern China. It
expanded its product line into imported fabrics, cosmetics, and luxury goods,
and opened pawnshops, Chinese medicine shops, and shops specializing in
handicraft weaving. According to Wellington Chan, each kind of firm was
part of its own managerial hierarchy and accounting structure. However, all
65
An early article that set the terms for this discussion is W.C. Kirby, “China
Unincorporated: Company Law and Business Enterprise in Twentieth-Century
China,” Journal of Asian Studies 54.1 (1995), 43–63. Kirby says only 227 companies had
registered as limited-liability companies by 1908 and only twenty-two of these were the
kind of large company the law was meant to enable.
66
Li, “清代民间契约中关于‘伙’的观念和习惯,” 42. Suzhou Archives, 乙 14-002-0059-
020, provides an example of a small shareholding partnership engaged in a dispute over
shares in which the company shareholder ledger was brought as evidence in court.
67
Chan, “The Organizational Structure of the Traditional Chinese Firm,” 222–5.
348

business accounts were handled through the same accounts office that dealt with
the Meng lineage estate.68 Family control persisted even in firms that registered
as limited-liability companies. The Nanyang Brothers Tobacco Company was
founded as a family firm by Jian Zhaonan and his uncle, Jian Kongzhao, in the
last years of the Qing and took advantage of the new company law to incorpor-
ate in the 1910s. In the 1920s it expanded its holdings to include newspapers, an
insurance company, and a remittance firm while retaining a majority of shares in
the name of family members.69 By the 1920s the Jians had joined other family
firms that had been using professional managers since the late Qing, but tensions
between the board of directors and the family reflect a familiar reluctance to
relinquish family operational control.
Whether dominated by a single family or a single entrepreneur, invest-
ment and management of many firms during this early period of industrial
development reflected what we might call relationship capitalism.70 The
Rong family of Wuxi are known for their development of modern flour
and cotton mills. By the early 1920s these mills were clustered into three
managerially independent groups, the Maoxin and Fuxin flour mills and the
Shengxin Cotton mills. While the Rongs had varying numbers of shares in
each mill, other shareholders, both individuals and lineage trusts, held shares
in more than one mill, creating interlocking ownership structures that
encouraged some co-operation, for example in the sharing of brand names
and in negotiating loans. A central office in the combined names of the three
groups performed some co-ordinating functions, including purchasing. And
individual mills kept deposits in the central office in part for this purpose.
While the nature of the relationship among the mills is often difficult to
discern, the importance of that relationship, bound by the Rong connections,
facilitated cross-subsidization within the group in the development of new
mills and temporarily aided mills in need of cash.71
68
Chan, “The Organizational Structure of the Traditional Chinese Firm,” 222–5.
69
D. Faure, “The Control of Equity in Chinese Firms within the Modern Sector from the
Late Qing to the Early Republic,” in R.A. Brown (ed.), Chinese Business Enterprise in Asia
(London, Routledge, 1995), pp. 68–72. Faure relates a similar story of family ownership
and control at the Guo’s Wing On department stores, the Hengfeng Spinning Mill, and
the Yans’ Dalong Machine works, which, despite its growing size and diversification
into the production of diesel engines and spinning and weaving machines, as well as the
operation of textile mills, remained a closely held family firm.
70
I use the term “relationship capitalism” because “network capitalism” implies more
formal ties among participants than appears to have been the case in China during the
early twentieth century.
71
This section relies on a new interpretation of the Rong enterprises in K.Y. Chan,
“Making Sense of the ‘Business Group’ in Modern China: The Rong Brothers
Businesses, 1901–37,” Australian Economic History Review 51.3 (2011), 219–44.
349

madeleine zelin
Individual entrepreneurs with social and political capital appear as central

in the development of many of China’s pioneering modern businesses, dating
back to the projects sponsored by powerful official patrons during the Self-
Strengthening period. Fan Xudong, considered the founder of China’s mod-
ern chemicals industry, benefited from investment from one of China’s
leading statesmen, Liang Qichao, who brought along family and friends as
shareholders in Fan’s modern Jiuda Salt Refinery. When Fan sought to
expand into the production of soda ash in the late 1910s, his position as
founder facilitated his reliance on capital and other resources from Jiuda to
build Yongli Soda Ash. These included loans and direct investment from
Jiuda, as well as shared land, dockyards, and senior management.72
Zhang Jian, one of the new group of literati businessmen of the late
nineteenth century, began his entrepreneurial career by promoting mechan-
ized cotton yarn production. Unable to raise sufficient capital from private
investors, he accepted a donation of machinery from the governor-general of
Liangjiang and established his Dasheng Number One Cotton Mill as a state–
private joint enterprise (guanshang heban 官商合办). Unlike earlier state–
private ventures, Zhang was able to retain managerial control and by the turn
of the century had infused his business with additional private investment,
particularly from local cloth merchants in the vicinity of the mill in Nantong,
Jiangsu. Zhang Jian followed the cotton mill by founding flour and oil mills,
shipping lines, land reclamation companies, a publishing house, and
a distillery.73
Although Zhang Jian and his family never had a majority of shares in the
businesses he established, Zhang relied on family members for their day-to-
day management. Overall co-ordination and control were exercised through
Zhang’s personal Shanghai-based accounts office, which managed accounts
of affiliated companies as well, acted as their paymaster and broker, and
handled their diverse financial transactions. This office both functioned as the
business office for what Köll calls Zhang’s “conglomerate” and was the office
that handled all of Zhang and his family’s personal accounts. This allowed the
movement of money between accounts, shifting money from one business to
another and between Zhang’s own accounts and those of his companies. The
central accounts office, while it did not co-ordinate the business activities of
72
M.B. Kwan, Beyond Market and Hierarchy: Patriotic Capitalism and the Jiuda Salt Refinery,
1914–1953 (New York, Palgrave Macmillan, 2014), pp. 27–30, 73–4. Kwan notes that by
providing loan guarantees from Jiuda, Yongli was also able to get interest-free and
below-market-rate loans from the main bank with which Jiuda was associated.
73
Köll, From Cotton Mill to Business Empire, pp. 64–8.
350

each unit, also allowed a degree of vertical integration, something we see in

the Zigong saltyards as well, where brine from one family’s wells might
move through that same family’s pipes, to be evaporated at the family’s
furnaces and sold by its wholesalers, all of which were managed as separate
vertical business lines.74
While the backgrounds of the most notable entrepreneurs of this period
differed, the use of the central account office as a co-ordination mechan-
ism was a central feature of early twentieth-century Chinese business. Liu
Hongsheng began his career as comprador for the London-registered
Chinese Engineering and Mining Company. Liu formed partnerships on
his own with other coal distributors, investing in mining and wharfs and
eventually diversifying into matches, coal briquettes, textiles, banking,
and real estate. While he was never able to bring his business interests
together under a single management structure, they were nevertheless
linked horizontally through Liu and his personal account office.75
Following the privatization of the Hubei Textile Bureau founded by
Governor-General Zhang Zhidong, its branch factories established inde-
pendent management structures, each handling its own accounts.
However, with the demise of the dynasty a new management agreement
was made with Xu Rongting and his Chuxing Company, under whose
central account office certain co-ordinating functions, including share-
holder relations and dealings with foreign machine parts suppliers, were
managed. Xu’s success attracting investors for these factories as private
enterprises was assisted by his close relationship with the erstwhile presi-
dent of the Chinese republic, Li Yuanhong.
Conclusion
How should we understand the continuation of key organizational characteris-
tics of Chinese businesses as China entered a period of legal reform accompanied
by rapidly changing politics, technology, and international relations? Of course
things did not stay the same. The businesses we have foregrounded in the post-
imperial period were increasingly large in scale and scope, with a central focus on
manufacturing. Single proprietorship and partnerships continued to dominate
the business landscape in China as in the West, but it is in this modern sector that
we see Chinese entrepreneurs grappling with new business opportunities in
74
75
K.Y. Chan, Business Expansion and Structural Change in Pre-war China: Liu Hongsheng and
His Enterprises, 1920–1937 (Hong Kong, Hong Kong University Press, 2006), pp. 27–46.
351

madeleine zelin
a context that was nothing if not unstable. When China finally opened its first
stock exchange in 1920, it was dominated by government and military bonds and
closed with the Japanese occupation, having been plagued by speculation for the
span of its short history.76 Passage of a company law in 1904 and its revision in
1914 introduced new standards of business practice that were not always enforced
by changing governments and a weak modern judiciary. In many trades guilds
retained power by setting silver exchange rates in a state that did not establish
a fiat currency until the 1930s and in which courts of law relied on chambers of
commerce and their member guilds to assist in the adjudication of complex
commercial cases. These are just a few of the constraints that confronted modern
firms.
It is therefore not simply a matter of cultural preference that many of the
institutions that served late imperial businesspeople found new life in the
ways we have noted above. Family, friends, and influence networks con-
tinued to encourage co-operation and co-investment. Lineage trusts, with
their claims to collective assets and longevity, continued to be major invest-
ors in modern firms, and the lineage accounts office served as a model for the
central accounts office and its tools for flexible co-ordination, including the
movement of funds between related but not necessarily co-invested busi-
nesses. And the diversified portfolio, as a technique for spreading risk and
influence, continued to play a major role in the business strategies of what
has come to be seen as the model of the modern Chinese conglomerate.
Further Reading
“The Canton Trade: Life and Enterprise on the China Coast, 1700–1845,” at www.jstor.org/
stable/j.ctt1xwbc2, accessed September 14, 2020.
Chan, K.Y., Business Expansion and Structural Change in Pre-war China: Liu Hongsheng and
His Enterprises, 1920–1937 (Hong Kong, Hong Kong University Press, 2006).
Chan, K.Y., “Making Sense of the ‘Business Group’ in Modern China: The Rong Brothers
Businesses, 1901–37,” Australian Economic History Review 51.3 (2011), 219–44.
Chan, W.K.K., “The Organizational Structure of the Traditional Chinese Firm and Its
Modern Reform,” Business History Review 56.2 (1982), 218–35.
Cheong, W.E., The Hong Merchants of Canton: Chinese Merchants in Sino-Western Trade
(Richmond, Curzon, 1997).
Cochran, S., Big Business in China: Sino-Foreign Rivalry in the Cigarette Industry, 1890–1930
Cochran, S., Encountering Chinese Networks: Western, Japanese, and Chinese Corporations in
China, 1880–1937 (Berkeley, University of California Press, 2000).
76
E. Hertz, “The Shanghai Stock Market, an Institutional Overview,” in Brown, Chinese
Business Enterprise, p. 117.
352

Fang, L., “Chinese Partnership,” Law and Contemporary Problems 52.3 (1989), 43–68.
Faure, D., “The Control of Equity in Chinese Firms within the Modern Sector from the
Late Qing to the Early Republic,” in R.A. Brown (ed.), Chinese Business Enterprise in Asia
(London, Routledge, 1995), 60–79.
Feuerwerker, A., China’s Early Industrialization: Sheng Hsuan-Huai (1844–1916) and Mandarin
Gardella, R., “Contracting Business Partnerships in Late Qing and Republican China:
Paradigms and Patterns,” in M. Zelin, J.K. Ocko, and R. Gardella (eds.), Contract and
Property in Early Modern China (Stanford, Stanford University Press, 2004), pp. 327–47.
Kim, K., Borderland Capitalism: Turkestan Produce, Qing Silver, and the Birth of an Eastern
Market (Stanford, Stanford University Press, 2016).
Kirby, W.C., “China Unincorporated: Company Law and Business Enterprise in
Twentieth-Century China,” Journal of Asian Studies 54.1 (1995), 43–63.
Köll, E., From Cotton Mill to Business Empire: The Emergence of Regional Enterprises in Modern
China (Cambridge, MA, Harvard University Asia Center, 2003).
Kwan, M.B., “Managing Market, Hierarchy, and Network: The Jiuda Salt Industries, Ltd.,
1917–1937,” Enterprise & Society 6.3 (2005), 395–418.
Lai, C.-K., “The Qing State and Merchant Enterprise: The China Merchants’ Company,
1872–1902,” in J.K. Leonard, J.R. Watt, and Cornell University (eds.), To Achieve Security
and Wealth: The Qing Imperial State and the Economy, 1644–1911 (Ithaca, NY, East Asia
Program, Cornell University, 1992), p. 146.
Liu, K.-C., “Financing a Steam-Navigation Company in China, 1861–62,” Business History
Review 28.2 (1954), 154–81.
Lufrano, R.J., Honorable Merchants: Commerce and Self-Cultivation in Late Imperial China
(Honolulu, University of Hawaii Press, 1997).
Moll-Murata, C., “Chinese Guilds from the Seventeenth to the Twentieth Centuries: An
Overview,” International Review of Social History 53, Supplement S16 (2008), 213–47.
Motono, E., Conflict and Cooperation in Sino-British Business, 1860–1911: The Impact of the Pro-
British Commercial Network in Shanghai (New York and London, St. Martin’s Press, 2000).
Pomeranz, K., “‘Traditional’ Chinese Business Forms Revisited: Family, Firm, and
Financing in the History of the Yutang Company of Jining, 1779–1956,” Late Imperial
China 18.1 (December 22, 1997), 1–38.
Qiao, Zhijian, “The Rise of Shanxi Merchants: Empire, Institutions, and Social Change in
Qing China, 1688–1850,” Ph.D. thesis, Stanford University, 2017.
Qiu Pengsheng 邱澎生, “由公产到法人–清代苏州，上海商人团体的制度变化”
(From “Common Property” to Corporation: The Institutional Changes of the
Merchant Associations of Suzhou and Shanghai in Late Qing), 法制史研究 (Research
on Legal History) 10 (2006), 117–40.
Rowe, W.T., Hankow, Commerce and Society in a Chinese City, 1796–1889 (Stanford, CA,
Stanford University Press, 1984).
Ruskola, T., “Conceptualizing Corporations and Kinship: Comparative Law and
Development Theory in a Chinese Perspective,” Stanford Law Review 52.6 (2000), 1599–1729.
Zelin, M., The Merchants of Zigong: Industrial Entrepreneurship in Early Modern China
353

11
The Economic Impact of the West
A Reappraisal
james kai-sing kung
Introduction
For the Chinese, the nineteenth century was a period of waking up and
realizing why the Middle Kingdom had fallen behind the West in economic
growth.1 Not only had this large and once prosperous country fallen behind
economically (with its apparent failure to industrialize), it also fell prey
during the two Opium Wars to the same country that first embarked upon
the Industrial Revolution – Britain. Consequently, a long period of autarky
came to an end.2 While initially China was forced to open up only several
“treaty ports” for trade and commerce, eventually the entire country was
subjected to the influences of the West, and in spheres that went far beyond
trade and commerce to also include industry, education, and even politics.3
By assembling data from a variety of previously untapped historical sources,
I thank the editor, Debin Ma, Ghassan Moazzin, Dwight Perkins, and Thomas Rawski for
useful suggestions on an earlier draft. I am most grateful to Ting Chen for compiling the
various data sources, and to Bin Huang and Yan Zhou for excellent research assistance.
Generous financial assistance from Sein and Isaac Souede is gratefully acknowledged.
1
For two millennia, China regarded itself as the Middle Kingdom, the most civilized state
under heaven, around which it developed a hierarchical system, the tributary system, in
dealing with international relations. Specifically, China was at the center of the system,
and surrounded by tributary states in Asia (e.g. Korea, Vietnam, Burma). These periph-
eral states were required to honor China by sending tribute periodically, and their kings
were expected to adopt the Chinese culture. J.K. Fairbank, The Chinese World Order:
Traditional China’s Foreign Relations, vol. 32 (Cambridge, MA, Harvard University Press,
1968).
2
To restrict the spread of Christianity, Qing China had implemented a closed-door policy
since around the early eighteenth century (c. 1721), following an ongoing dispute that
started as early as in 1645 between Roman Catholic missionaries and the Qing emperors
concerning whether Chinese ritual practices of honoring family ancestors and other
deities were compatible with Catholic beliefs.
3
Many Chinese historians thus see this period as marking the beginning of a century of
humiliation – a “dark and shameful period” in China’s history.
354

this chapter attempts to analyze the Western influences that shaped the
economic trajectories of late imperial China.
Clearly, for a topic as broad as the economic impact of the West, it is
impossible to encompass every single aspect in which China had been
affected in the process. Our choice of which aspects to examine, therefore,
is unavoidably subjective and selective to some extent. Nonetheless, our
choice is not arbitrary, but premised importantly on the key changes that
occurred in the economy, society, and polity of late imperial China.
To a large extent, this chapter was inspired by Robert Dernberger’s
important question regarding the role of the foreigner in China’s economic
development4 – a question prompted by the forced opening up of China by
the Western powers (including Japan) on the one hand, and its failure to
industrialize even after foreign firms and banks had entered the local market
via the treaty ports on the other. In the absence of relevant systematic data,
the mere juxtaposition of these two historical events unwittingly led to
a largely negative assessment of the effects of treaty ports on China’s
economic development. For example, Dernberger laments an “overwhelm-
ing acceptance of a single major theme” among “leading Western-trained
anthropologists, sociologists, and economists,” that the treaty ports had
destroyed “the economic and social fabric of China’s countryside.”5 And
while some speak favorably of the effects of trade and resource transfers as
they occurred in the treaty ports, they see such benefits as necessarily
confined within the treaty ports; there were virtually no “trickle-down”
effects on the vast hinterland. China remained essentially a “dual” economy
consisting of a small “modern” sector and a sizeable “traditional” sector, the
totality of which resembles the stylized economy described by Lewis’s “two-
sector” model.6
4
R.F. Dernberger, “The Role of the Foreigner in China’s Economic Development, 1840–
1949,” in D.H. Perkins (ed.), China’s Modern Economy in Historical Perspective (Stanford,
Stanford University Press, 1975), pp. 19–47.
5
Dernberger, “The Role of the Foreigner in China’s Economic Development,” p. 22.
Dernberger was referring to Chen Han-seng, H.D. Fong, Fei Hsiao-tung, Franklin Ho,
and so forth. Philip Huang’s account of how China’s indigenous weaving cottage
industry was devastated by the Japanese iron gearwheel, which was a thousand times
more productive, represents a prominent case in point. P.C.C. Huang, The Peasant
Economy and Social Change in North China (Stanford, Stanford University Press, 1985).
6
W.A. Lewis, “Economic Development with Unlimited Supplies of Labor,” Manchester
School 22.2 (1954), 139–91. Also see, for example, C.-M. Hou, Foreign Investment and
Economic Development in China, 1840–1937 (Cambridge, MA, Harvard University Press,
1965); R. Murphey, The Treaty Ports and China’s Modernization: What Went Wrong? (Ann
Arbor, University of Michigan Center for Chinese Studies, 1970).
355

james kai-sing kung
Of course not everyone sees the foreigner as the culprit in China’s lack of
economic development in the late Qing period and beyond. An ailing
government, inadequate savings for capital formation, technical backward-
ness, and perhaps also the unique characteristics of the Chinese family and
the social institutions in which it was embedded were all considered “indi-
genous obstacles” hindering development.7 Indeed, it is from this point of
view that Dernberger credits the foreigner with catalyzing the modernization
of the Chinese economy. Specifically, he views trade, technology transfer,
and foreign direct investment (FDI) as the foreigner’s “largest contribution to
China’s economic development.”8
A thorough, unbiased analysis and thus conclusive assessment is hard to
reach, however, in the absence of systematic data. To address the kind of
questions raised by economic historians and other social scientists nearly half
a century ago, we make use of newly assembled data to examine the role
played by FDI and foreign debt (FD) in treaty port development (a classic
question posed earlier), and the extent to which markets had developed or
integrated in facilitating the distribution of goods for import and export (i.e.
whether or not treaty port development produced “spillover effects”). We
find that between the two, FDI outsized FD by a wide margin, and that of all
the countries involved, Japan invested the most, followed by the United
Kingdom.9 To identify more precisely the sectoral and geographic
7
Mandarin Enterprise (Cambridge, MA, Harvard University Press, 1958); M.J. Levy,
“Contrasting Factors in the Modernization of China and Japan,” Economic Development
and Cultural Change 2 (1953), 161–97; D.H. Perkins, “Government as an Obstacle to
Industrialization: The Case of Nineteenth-Century China,” Journal of Economic History
17.4 (1967), 478–92. Regardless of whether the foreigner was a culprit, Cohen criticizes
this paradigm pioneered by John K. Fairbank and others that views Qing China as
playing a merely passive role in response to its contact with the West, to the extent that
it was incapable of change without Western interference. P. Cohen, Discovering History
in China: American Historical Writing on the Recent Chinese Past (New York, Columbia
8
Dernberger, “The Role of the Foreigner in China’s Economic Development,” p. 46.
While certain sectors of the population or economy may have been “impoverished,” it
was not due to foreign penetration; rather, the peasant economy had been struggling
with growing land shortages amid continuing population growth, not to mention the
competition it faced from China’s own modernization efforts. Using the textile and
food processing sectors as examples, Dernberger argues (at pp. 36–7) that “the major
damage done to the handicraft industries was due largely to the Chinese response to
the challenge of modernization and not to the competition of the foreigner,” and that
“the foreign goods were of a different type and were sold in a different segment of the
market from those issuing from the traditional handicraft sector.” Evidence on these
claims, however, remains scant.
9
Wu Chengming 吴承明, 帝国主义在旧中国的投资 (Investment in Old China by
Imperialists) (Beijing, Renmin chubanshe, 1955). Xu Dixin 許滌新 and Wu Chengming
356

distributions of foreign investments, we exploit a unique data set compiled by

Chang,10 who documents the number of foreign firms and banks entering
China for the 1841–1916 period. We find that the food and beverage and
machinery sectors outsized each of the other sectors established by foreign-
ers, with Britain and Japan having established the largest cumulative number
of firms. Insofar as the development of markets was concerned, our analysis
finds that the twelve earliest treaty ports covered the majority of the trading
networks, with Shanghai and Nanjing having much bigger rippling effects
than those established later.
Our goal, however, is to provide a fresh perspective from which to
understand the role of treaty ports in spurring the development of modern
institutions – both business and financial – in the final chapter of the imperial
Chinese economy. An important economic influence of the West is that,
under the rubric of the “Self-Strengthening movement,” it effectively elicited
a concerted response from the political elites (with the likes of Zeng Guofan
曾國藩 and Li Hongzhang 李鴻章, and later Zhang Zhidong 張之洞) to
develop the country’s industrial and military capabilities. This is reflected in
the establishment of wide-ranging enterprises from steamships and engines
to arsenal, naval, and military academies, and a host of industrial concerns
such as coal mines, cotton mills, and iron and steel complexes to serve these
enterprises, as well as basic infrastructure such as railways and the telegraph
to facilitate development.11 While these concerted efforts represented mostly
state endeavors or specifically those of the regional political elites,12 the
industrialization effort that grew out of the “Self-Strengthening movement”
undoubtedly created a qualitatively new social and political atmosphere in
China’s continuing efforts toward modernization within the institutional
context of the treaty ports and the so-called “self-initiated ports” (zikai
shangbu 自開商埠, hereafter SIPs) – the latter established for the dual
purposes of avoiding granting further “concessions” or specifically a range
of extraterritorial rights to the Western powers, and competing with the
treaty ports that were dominated (at least initially) by foreigners.
吳承明, 中國資本主義發展史 (A History of Capitalist Development in China), vol 3

(Beijing, Renmin chubanshe, 1993).
10
Y. Chang 張玉法, “清末民初的外資工業” (Foreign Industries in Late Qing and Early
Republican China), Bulletin of the Institute of Modern History, Academia Sinica, 1987,
129–249.
11
For example, see T.Y. Kuo and K.C. Liu, “Self-Strengthening: The Pursuit of Western
Technology,” in J.K. Fairbank (ed.), The Cambridge History of China, vol. 10, Late Ch’ing,
1800–1911, part 1 (Cambridge, Cambridge University Press, 1978), pp. 491–542;
J.D. Spence, The Search for Modern China (New York, W.W. Norton & Company, 1990).
12
Feuerwerker, China’s Early Industrialization.
357

james kai-sing kung
To fully appreciate the influence of the foreign firms and banks introduced
into China (all of which had their operations set up in treaty ports), we
perform a multivariate analysis and confirm that a good majority of the
modern domestic firms and banks had chosen to set up operations in close
proximity to their foreign counterparts inside the treaty ports, suggesting the
possible existence of a “demonstration effect.” Moreover, this effect is further
confirmed by the finding that, of those which had chosen to locate outside
the treaty ports, the vast majority were located in the SIPs. The fact that only
a handful of modern Chinese banks were located in cities with no special
administrative status suggests that businesses paid a “premium” for setting up
shop in the treaty ports and to a lesser extent the SIPs.
But treaty ports may have produced positive, dynamic effects that went
beyond the private firms’ choice of locations. A recent study has documented
a strong positive correlation between greater military investment during the
Self-Strengthening movement and greater output in civilian industrial sectors
in the 1930s through input–output linkages.13 On this basis we conduct
a multivariate analysis, and find that the earlier treaty ports, particularly the
first five established in the 1840s, are strongly correlated with the locations of
munition factories, confirming the existence of a dynamic effect of the ports.
Another hallmark of modern economic development pertains to techno-
logical adoption. For instance, in order for modern industrial firms to use the
steam engine, electricity was essential. We find that private electricity plants,
while small in scale, had unwittingly sprung up all over China. While these
small private power plants appeared in every nook and cranny of the country,
they were especially concentrated in south and south-central China – regions
where public efforts were visibly lacking. On the whole, the wide geographic
availability of electrical power supply was an important precondition for the
increased adoption of the steam engine among firms in both types of port
cities from 1880 onwards. A similar growth trend was observed among firms
adopting machinery.
Treaty ports and SIPs served as the engines of growth historically, but how
lasting are their impacts? To find out, and following Jia,14 we compare the
treaty ports and SIPs with the non-port cities, and find that both types of
historical ports outperform their non-port counterparts in terms of GDP, per
capita GDP, fiscal income, and manufacturing output of 2010, suggesting that
13
S. Bo, C. Liu, and Z. Yan, “Military Investment, Industrial Linkage, and the Rise of
Cluters: Evidence from China,” Working Paper, Jinan University, 2020.
14
R. Jia, “The Legacies of Forced Freedom: China’s Treaty Ports,” Review of Economics and
Statistics 96.4 (2014), 596–608.
358

they might have had a long-term effect on economic development.

A potentially fruitful topic of future research concerns the possible channels
through which their effects persist over time.
Another angle from which to assess the contributions of treaty ports to
economic development is to examine the extent to which human capital or
education developed in that context vis-à-vis the non-port environment. We
begin with the role of the missionaries. By introducing an entirely new
curriculum that incorporated not only the modern sciences of mathematics
and physics but also such social science subjects as geography, the missionar-
ies gradually but surely overhauled the Chinese educational system under-
pinned by the Confucian classics. In an attempt to replace the civil service
examination system and modernize the education system, the Qing govern-
ment established a set of “new schools” ranging from primary to middle
schools and equipped with a largely modern curriculum close on the heels of
the missionary endeavors. Complementing this effort, the Qing government
further decreed that at least one modern university be established in every
single provincial capital alongside the private and missionary universities.
Our analysis finds that both treaty ports and SIPs are correlated with the
establishment of universities in general, and of universities with an engineer-
ing school in particular. We give special emphasis to engineering schools as
they provided the specific human capital required by many industrial firms
established back then. Not surprisingly, there were more engineers in the
treaty ports and SIPs, who presumably were employed by the modern
industrial firms located therein.
It is certainly no exaggeration to suggest that, by making a last-ditch effort,
Qing China had made major strides on the economic front in its last fifty
years or so of imperial rule, to the extent that the empire could have been
successfully transformed down the road had it not been brought down
abruptly by the 1911 Xinhai Revolution 辛亥革命. Perhaps this reflects
how history is made. While the political elites (the Constitutionalists) had
mustered the will and energy to strengthen a significantly weakened China,
for instance by spearheading the development of modern firms and banks
and establishing a national assembly, by adopting a modern education system
based on the corpus of knowledge allegedly “useful” for modernization, and
by sending students overseas to acquire new knowledge deemed useful for
reforming China, the revolutionaries saw the Manchu empire as incapable of
creating a new and stronger China. With Western ideas and political ideolo-
gies infiltrating China rapidly via the media (newspapers) and the universities,
among other channels, it was the overseas students – especially those who
359

james kai-sing kung
studied in Japan and were supposed to return to help build a new China –
who were most inclined to join the revolutionary organizations, whose
alliance eventually succeeded in overthrowing the Qing government. The
rest, as they say, is history.
The remainder of this chapter is organized as follows. In the next
section, we first provide a brief introduction to the treaty ports and
SIPs. This is followed immediately by an examination of their impact
in terms of FDI and FD, market integration, “demonstration effects”
for the domestic modern firms and banks, and technology adoption.
Without claiming causality and identifying the specific channels, we
also make an attempt to assess their effects over the long run. We then
examine the development of human capital or education in the treaty
ports vis-à-vis the non-port cities, by emphasizing the initial role of the
missionaries in introducing to the Chinese a brand new curriculum
based on Western education, and the birth of the modern university
and specifically the training of engineers amid China’s nascent indus-
trialization. Finally, we examine the growing popularity of studying in
Japan after China’s defeat in the Sino-Japanese War, and how that may
have made possible the 1911 Xinhai Revolution. The chapter ends with
a few brief concluding remarks.
Treaty Ports
Treaty ports were ports conceded by the Qing government after its defeat by
Great Britain during the First Opium War in 1840. A treaty (the Treaty of
Nanking of 1842) was signed, according to which five ports located on the
south and southeast coast of China – Shanghai, Guangzhou (previously
Canton), Fuzhou (Foochow), Xiamen (Amoy), and Ningbo (Ningpo) –
were forced to open up to the West for international trade (panel A of
Figure 11.1).
In the treaty ports, Great Britain successfully negotiated a set of privileges
beyond the restricted rights of the previous “Canton system,” whereby
foreigners were allowed to conduct business with the Chinese only in the
port city of Canton and only via the institutional arrangement known as the
“Thirteen Hongs.”15 The treaty now allowed the foreigners to reside and
15
For example, see J.K. Fairbank (ed.), The Cambridge History of China, vol. 11, Late Ch’ing
1800–1911, part 2 (Cambridge, Cambridge University Press, 1980). Essentially the
“Thirteen Hongs” monopolized trading with the foreigners; in other words, anyone
who wished to trade with China could only do so via the “hongs.”
360

A B
1845 1865
C D
1885 1910
Figure 11.1 Expansion of the treaty port system (indicated by the prefectures in which
they were located)
Source: Zhang Hongxiang 張洪祥, 近代中國通商口岸與租界 (Treaty Ports and
Concessions in Chinese History) (Tianjin, Tianjin renmin chubanshe, 1993)
conduct business in the port cities, to spread Christianity, and even to lease
land in perpetuity in the so-called “concessions” or zujie 租界. Moreover, the
extraterritorial privileges conferred upon the foreigners allowed them essen-
tially diplomatic immunity as they were exempted from Chinese jurispru-
dence but instead abided by the legal system of their mother country.16
But it was the Second Opium War of 1856 that culminated in the further
expansion of these ports from a handful to approximately forty between 1860
and 1894, after China was forced to sign two additional treaties (the Treaty of
Tientsin (now Tianjin)) and the Treaty of Peking (now Beijing)) upon its
defeat – treaties that essentially reaffirmed and expanded the initial privileges
conferred in 1842. As Figure 11.1 shows, China was forced to open up eleven
16
For example, see J.K. Fairbank, Trade and Diplomacy on the China Coast: The Opening of
the Treaty Ports, 1842–1854 (Cambridge, MA: Harvard University Press, 1953).
361

james kai-sing kung
80
First Opium War
Second Opium War

60
First Sino-Japanese War
40
20
0
1840 1850 1860 1870 1880 1890 1900 1910 1920
Treaty port Self-initiated port
Figure 11.2 Growth in the number of treaty ports and self-initiated ports
Source: Chang Yufa 張玉法, 近代中國工業發展史, 1860–1916 (A History of Modern
Chinese Industrial Development: 1860–1916) (Taipei, Gui guan tu shu you xian gong si,
1992)
more ports, this time along the coast and in particular along the Yangzi river –
the country’s main navigable river, and also up in the far north (panel B).
With the passage of time, more ports were added, especially in both the
northeast and the northwest (panel C).
The subsequent treaties did not merely expand the number of ports in
which foreigners could conduct trade and business in China. These new
treaties gave foreigners access also to the non-treaty ports, namely the inland
regions, as explorers, businessmen, and missionaries, among other roles.17
Finally, after half a century of development, treaty ports were further
expanded with the signing of the Treaty of Shimonoseki in 1895, after China
was defeated by the Japanese in the First Sino-Japanese War a year earlier. It
was during the 1895–1910 period that the number of treaty ports peaked,
reaching a total of seventy-seven, excluding Taiwan and Hong Kong, which
17
Spence, The Search for Modern China.
362

were colonies of Japan and the UK respectively. In particular, more ports had
now been added in both the northeast and the southwest alongside further
expansion elsewhere, enlarging the ports’ trading networks with those of the
hinterland (see panel D in Figure 11.1). Figure 11.2 shows the expansion of
treaty ports over time.
“Self-Initiated Ports” (SIPs)

Existing literature has focused almost exclusively on the role of the treaty
ports, while neglecting that played by the “self-initiated ports,” which essen-
tially represented a concerted response by the Chinese to the escalating
demand made by the foreign powers to open up more ports for trade and
business after China’s defeat in the Sino-Japanese War.18 What the Chinese
feared was not treaty ports per se, which they recognized had attracted trade
and other investment opportunities. Rather, they were concerned that some
of the treaty ports might end up becoming “concessions” to the foreign
powers, as in the case of Shanghai, which would necessitate their conceding
the associated fiscal (taxes), judicial, and policing rights – the so-called
“extraterritoriality,” all of which contravened the Qing legal code and
administrative practices.19 While the humiliation that resulted from granting
extraterritoriality to the foreign powers had led to a new sense of national
indignation and heightened concerns about sovereignty, the awareness that
treaty ports might bring trade and investment benefits led the Qing govern-
ment to willingly open up more ports, but on their own terms.
To convince the foreign powers that their business interests would not be
compromised in the SIPs, they were guaranteed the same freedom to trade
and invest in these ports, which were now overseen by the provincial
governor or xunfu 巡撫. More importantly, the Chinese promised to set up
a new legal system that would better protect foreign business interests, and to
invest in public infrastructure and employ measures to strengthen public
health. Unlike in the concessions, however, the Chinese government would
maintain the rights to collect taxes, adjudicate disputes, and provide public
18
For an in-depth discussion of the self-initiated ports, see Zhang Hongxiang 張洪祥,
近代中國通商口岸與租界 (Treaty Ports and Concessions in Chinese History)
(Tianjin, Tianjin renmin chubanshe, 1993); Zhang Jianqiu 張建俅, 清末自開商埠之
研究 (1898–1911) (Research on the Self-Initiated Ports of the Late Qing (1898–1911))
(Taipei, Huamulan wenhua chubanshe, 2009).
19
For example, Zou Rong 鄒容 – a revolutionary – was pursued by the Qing government
for his inflammatory writings against the Manchu regime. By living in the foreign
concession of Shanghai, he was spared arrest and possibly execution, as the foreign
authorities steadfastly refused to hand him over to the Qing administration. See Spence,
The Search for Modern China.
363

james kai-sing kung
security – the set of rights that collectively define sovereignty. And instead of
perpetuity, land rights would only be extended to foreigners for a maximum
of sixty years – a measure presumably designed to retain sovereign rights.
Altogether thirty-five SIPs were established, and about half (seventeen)
were explicitly established for the purpose of avoiding the need to give
concessions to the foreign powers. For instance, the “Ningbo model”20 –
a prefecture located in southeast China and which is in close proximity to
Shanghai – was proposed by Zhang Zhidong, the governor-general of the
provinces of Jiangsu and Jiangxi after easing Emperor Guangxu’s concern
about the appropriateness of setting up SIPs. Yuezhou prefecture in Hunan
province is another example of one of the earlier SIPs. Additionally, as soon as
the Qing government became convinced that the proactive establishment of
SIPs could in fact help preserve its power, another eighteen of these zones
were further established as “trade hubs” in direct competition with the treaty
ports. A related strategic concern was that, in anticipating the construction of
the railway, the Qing government feared that failure to establish a SIP in each
of these “economic zones” might run the risk of losing their control rights to
the foreigners. Figure 11.2 shows the development of SIPs in the context of
the overall development of treaty ports. Although both types of ports had
experienced sharp growth since 1895, it is clear that the SIPs were essentially
a product of the Treaty of Shimonoseki.
An interesting question arises in regard to the locational choice of the two
types of ports. Were the SIPs set up strategically to capture the benefits of
trade? If so, were they chosen in close geographic proximity to the treaty
ports? Figure 11.3 shows that by and large the SIPs were established near the
treaty ports. It would appear that their specific locations were so chosen as to
maximize the prospects of trade. To address this issue more systematically
we perform a simple multivariate analysis. As reported in Table 11.1, we find
that treaty ports and SIPs are indeed significantly correlated, regardless of
whether we are using locations of the treaty ports – defined by the prefecture
in which a port was located (column 1), or distance to the treaty ports as the
pertinent measure (columns 3–4). Additionally, Table 11.1 also reveals that the
SIPs were located close to the railway line, lending weight to the claim that
20
While Ningbo was one of the five earliest treaty ports, the Taiping Rebellion signifi-
cantly weakened the foreigners’ dominance there. This unwittingly enabled the
Chinese to regain much of the de facto control rights in Ningbo, despite its treaty
port status. Inspired by the Ningbo experience, Zhang Zhidong thus proposed to
Emperor Guangxu that China should establish more ports to capture the benefits of
trade, but retain more control over fiscal and other property rights in the newly
established ports.
364

Treaty port
Self-initiated port
Figure 11.3 Geographical distribution of treaty ports and self-initiated ports

Source: Chang, 近代中國工業發展史
they were initiated because the Chinese did not want to see cities with easy
access to the railway line falling into the hands of foreign powers (columns
2–4).21
The Impact of Treaty Ports

Foreign Direct Investment and Foreign Debt
Conventionally, treaty ports were viewed either as a carrier of technology
via trade and investments,22 or as a destructive force capable of impoverish-
ing certain sectors of the population and economy, particularly the
21
In Table 11.1 we also compare several key features of the two types of port cities,
ranging from geographic characteristics (e.g. distance to major navigable river and to
coast) to population density and growth rate in various periods where such data are
available (1776, 1820, and 1851), but we fail to find any significant statistical difference.
Thus we do not report their results separately.
22
See Hou, Foreign Investment and Economic Development in China; Murphey, The Treaty
Ports and China’s Modernization; Dernberger, “The Role of the Foreigner in China’s
Economic Development.”
365

james kai-sing kung
Table 11.1 Locations of self-initiated ports

(1) (2) (3) (4)
Self-initiated ports
Treaty ports 1.409*** 0.913**

(0.411) (0.455)
Railway 0.599*** 0.652*** 0.631***
(0.107) (0.105) (0.133)
Distance to treaty port
0 kilometer 1.897*** 1.691**
(0.643) (0.739)
1–50 kilometers 1.739** 1.442*
(0.718) (0.800)
50–100 kilometers 2.183*** 2.089***
(0.734) (0.765)
100–200 kilometers 1.417* 1.378*
(0.747) (0.822)
Distance to main river (log) −0.056
(0.040)
Distance to coast (log) −0.047
(0.049)
Distance to provincial capital (log) 0.001
(0.042)
Observations 377 377 377 367

Robust standard error Yes Yes Yes Yes
Note. These are logistic regressions. Robust standard errors in parentheses, * p < 0.1,
**
p < 0.05, *** p < 0.01.
countryside.23 Despite the fact that treaty ports had existed for a long time
(since the 1840s), systematic data on FDI and FD are available only for a much
later period (c. 1902–1936).24 Figure 11.4, which shows these trends, reveals
that FDI always exceeded FD, and with a widening trend since around the
late 1910s, whereas FD (which did not include government bonds) remained
basically stable during the entire period in which such information is avail-
able. Figure 11.5 further decomposes the sources of FDI and FD by host
23
See H.D. Fong, Rural Industries in China (Shanghai, China Institute of Pacific Relations,
1933); H. Chen, Landlord and Peasant in China: A Study of the Agrarian Crisis in South China
(New York, International Publishers, 1936); H.-T. Fei, Peasant Life in China: A Field Study
of Country Life in the Yangtze Valley (London, Routledge & Keegan Paul, 1939);
A. Feuerwerker, The Chinese Economy, 1912–1949, No. 1 (Ann Arbor, University of
Michigan, Center for Chinese Studies, 1968).
24
The data for 1902, 1930, and 1935 are extracted from Wu, 帝国主义, pp. 52–3; and the
rest re-estimated by Xu and Wu, 中國資本主義發展史.
366

3000
Amount (US$ million)
2000
1000
0
1902 1914 1920 1930 1936
FDI Foreign debt
Figure 11.4 Foreign direct investment and debt
Source: Xu Dixin 許滌新 and Wu Chengming 吳承明 (eds.), 中國資本主義發展史
(Developmental History of Chinese Capitalism), 3 vols. (Beijing, Renmin chubanshe, 1985)
country. Among the six countries with major investment/debt relations with
China, Japan ranked first, with disproportionate growth in FDI in the 1930s,
followed by the United Kingdom. FDI for the remaining four countries,
namely the United States, France, Germany, and Russia, was much smaller
in magnitude.25
While these data are informative, a major limitation is that they are
reticent on both geographic and sectoral distributions, although intuition
would suggest that the bulk of these investments and debts had likely flowed
into the Japanese and British spheres of the treaty ports and to a lesser extent
the SIPs. To find out more about their geographic and sectoral distributions
we turn to a data set compiled by Chang on the number of foreign industrial
firms entering China and their initial registered capital,26 and the nature of
their business. Figure 11.6 shows the geographic distribution of these foreign
private firms, whereas Figure 11.7 breaks down their investment by country
of origin. Consistent with the initial observation that Japan was the largest
source of FDI, these firms were indeed mostly located in the northeastern
part of China – a territory where Japan had a distinct “sphere of influence,”
followed by the southern coast, presumably representing Britain’s economic
interest. Figure 11.7 confirms that Japan and Britain were indeed the biggest
25
That Japan was the most enthusiastic investor in China might be seen as a harbinger of
Japanese military aggression later on.
26
Chang, “清末民初的外資工業.” There were many missing values, however.
367

Japan UK USA
1500 1500 1500
1000 1000 1000
US$ million
US$ million
US$ million
500 500 500
0 0 0
1902 1914 1920 1930 1936 1902 1914 1920 1930 1936 1902 1914 1920 1930 1936
FDI Foreign debt FDI Foreign debt FDI Foreign debt
France Germany Russia

1500 1500 1500

1000 1000 1000
US$ million
US$ million
US$ million
500 500 500
0 0 0
1902 1914 1920 1930 1936 1902 1914 1920 1930 1936 1902 1914 1920 1930 1936
FDI Foreign debt FDI Foreign debt FDI Foreign debt
Figure 11.5 Foreign investment and foreign debt by country

Source: Xu and Wu, 中國資本主義發展史
1 dot = 5
foreign firms
Treaty port
Self-initiated port
Figure 11.6 Geographic distribution of foreign private firms

Source: Chang, 近代中國工業發展史
foreign investors in China. What Figure 11.7 also tells us is that, while Japan’s
foreign investments in China remained low for the first fifty years or so
before surging after 1900 and surpassing those of Britain by around the mid-
1910s, Britain’s rise as a major foreign investor in China followed a smoother
gradual trend for more than half a century.27
Chang’s data also enable us to identify the sectors in which these foreigners
were investing.28 Figure 11.8 shows their distribution based on our own
classification. With the exception of ten firms which we find difficult to
classify,29 we sort the remaining 485 foreign firms that entered China between
1841 and 1916 into twelve sectors. As shown, the largest sector was food and
beverages, which included tobacco, alcohol, tea, sugar, salt, soda, flour, and
oil processing, followed (by a wide margin) by machinery, which included
shipping, motor vehicles, electrical appliances (both manufacturing and
27
Though referring to a much later period (c. 1936), Duus finds that the great majority of
cotton mills in China were owned by the Japanese, accounting for 40 percent and
57 percent of all of China’s machine-spun yarn and machine-woven cloth respectively.
P. Duus, “Zaikabo: Japanese Cotton Mills in China,” in P. Duus, R. Myers, and
M. Peattie (eds.), The Japanese Informal Empire, 1985–1937 (Princeton, Princeton
28
Chang, “清末民初的外資工業.”
29
For example, a couple of these firms were engaged in the production of musical
instruments, whereas several others were canneries.
369

james kai-sing kung
200
# Foreign firms (cumulative)
50 100 0 150
1850 1860 1870 1880 1890 1900 1910 1920

Year
Japan UK USA
France Germany Russia
Figure 11.7 Foreign firms, by country of origin
Source: Chang, 清末民初的外資工業
maintenance), and others. The next three largest sectors were utilities (pri-
marily electricity, water, and gas), textiles and silk, and chemicals (dyes,
matches and kerosene, paint, cosmetics, soap, rubber, tires, and so forth).
Virtually all of these sectors were related to consumption (with shipping
perhaps the only exception), suggesting that the foreign firms were likely
capitalizing on the changing consumption habits and possibly rising standard
of living of the Chinese people.
Treaty Ports and Market Integration

A particular perspective from which to assess the effects of treaty ports is to
examine the extent to which they helped develop or integrate markets, such
as by sourcing goods from all over China for export, and, conversely,
importing and disseminating goods from abroad. By making use of
a survey conducted in the mid-1930s by the postal office (Chunghwa Post),
which contains information on the geographic distribution of goods shipped
from virtually all the counties in China via both types of ports, we are able to
document not only trade flows but also their corresponding value. We begin
with panel A of Figure 11.9, which shows the already wide geographic
370

150
# Foreign firms
100
50
0
e
ry
es
lk
ce
r
he
al
rg
g
tin
in
oo
in
si
ne
iti
i
ra
ic
rv
ild
ic
lu
at
in
W
d
til
hi
ve
ed
Se
al
an
Le
Bu
Pr
U
ac
he
et
be
M
ile
M
M
C
d
xt
an
Te
od
Fo
Figure 11.8 Sectoral distribution of foreign firms

coverage of trade after the signing of the Treaty of Nanking of 1842. In

particular, the twelve earliest treaty ports were the ones where the majority
of trading networks were established.30 In terms of trade volume, Shanghai
and Nanjing had an approximately 27 percent larger value of shipments than
those ports founded later, and double the value of shipments of the SIPs. But
huge discrepancies also existed among these earlier treaty ports. Shanghai,
for example, was nearly three times the size of Guangzhou and Fuzhou and
nine times that of Xiamen. Their differences notwithstanding, by 1935 nearly
everywhere in China sold produce to these dozen treaty ports. In panel B,
more trading networks were established after the signing of the subsequent
two treaties (Treaty of Tientsin in 1858 and Treaty of Peking in 1861). In panels
C and D, we connect the flow of goods to those treaty ports opened later and
the SIPs, which when compared with the earlier ports had a very limited scope
30
These twelve ports were Shanghai, Xiamen, Ningbo, Guangzhou, and Fuzhou under
the Treaty of Nanking of 1842; Jiujiang, Nanjing, Shantou, Yantai, Yingkou, and
Zhenjiang under the Treaty of Tientsin of 1860; and Tientsin under the Treaty of
Peking of 1861.
371

1842 Treaty of Nanking 1858 Treaty of Tientsin
Trade network 1861 Treaty of Peking
Trade network
A B

Other forced-
open ports Self-initiated ports
Trade network Trade network
C D
Figure 11.9 Domestic trade network

Source: Chunghwa Post survey (1935)
of coverage, suggesting that the earlier treaty ports were indeed more effective
in facilitating the development or integration of markets in this period.
Treaty Ports and Modernization

The role of treaty ports in facilitating modern economic growth is a question
that had to wait until the early 1990s for a systematic answer, after historian
Du Xuncheng (1991) assembled a list of newly established industrial firms and
banks for the 1840–1927 period.31 As suggested by the title of his book, 民族資
本主義與舊中國政府 (National Capitalism and the Old Chinese
Government), the kinds of firms enumerated therein were all established
by the Chinese, and the majority of them were either state-owned (guanban
官辦) or privately owned, with the rest being privately funded but managed
by government officials (guandu shangban 官督商辦).32 In this context we use
the number of modern firms and the number of modern financial institutions
(or simply banks) enumerated in Du’s book as the two indicators of modern
economic growth. Our first step is to examine the role of treaty ports in
attracting these foreign institutions, before examining the “demonstration
effect,” if any, that they may have had on their domestic counterparts. Upon
establishing these lines of evidence, we then examine the locational choice of
the domestic firms and banks, i.e. whether they set up operations in the treaty
ports, the SIPs, or the non-port cities. Basically, we would expect all foreign
firms and banks to establish their operations in the treaty ports, and
a substantial proportion of the domestic firms and banks to do so in both
treaty ports and SIPs. Last but not least we examine the issue of technology
adoption; specifically, we are interested in finding out the extent to which
domestic firms adopted the Industrial Revolution technology of the steam
engine and/or electricity.
Figure 11.10 first shows the growth trend of foreign firms and banks.
Whereas foreign firms grew more linearly in the earlier period (between
1840 and 1890), foreign banks experienced two spikes of growth – one
31
To be sure, Feuerwerker has also estimated that at least 549 Chinese-owned modern
mining and manufacturing enterprises had been established by 1913, a mere two years
after the fall of the Qing dynasty. Du’s number, however, is in the thousands.
Feuerwerker, The Chinese Economy, 1912–1949; Du Xuncheng 杜恂誠, 民族資本主義
與舊中國政府, 1840–1937 (Chinese Capitalism and the Old Chinese Government,
1840–1937) (Shanghai, Shanghai shehui kexueyuan chubanshe, 1991).
32
Chang also enumerates firms for the 1841–1916 period, but he classifies them according
to whether they were foreign (1987) or domestic (1989). Chang, Foreign Industries in Late
Qing and Early Republican China; Chang Yufa 張玉法, “清末民初的民營工業”
(Domestic Private Industries in Late Qing and Early Republican China), Bulletin of the
Institute of Modern History, Academia Sinica, 1989.
373

james kai-sing kung
600
500
Foreign firms
400
300
200
100
0
1840 1850 1860 1870 1880 1890 1900 1910 1920
500
400
Foreign banks
300
200
100
0
1840 1850 1860 1870 1880 1890 1900 1910 1920
Figure 11.10 Numerical growth of foreign firms and banks
Source: Huang Guangyu 黃光域, 外國在華工商企業辭典 (Dictionary for Foreign
Industrial and Business Enterprises in China) (Chengdu, Sichuan renmin chubanshe, 1995),
for foreign firms; Yearbook of National Banks (1934) for foreign banks
between 1840 and 1890 and the other between 1900 and 1920. In Figure 11.11
we superimpose the locations of the foreign firms and banks on those of the
treaty ports. We find the two to be highly correlated spatially, which is as
expected given that it would be highly unlikely for foreign institutions to be
established in cities outside the treaty ports where the foreigners would be
distinctly less protected.
We then examine the development of domestic firms and banks based on
Du’s data.33 As Figure 11.12 shows, the latter experienced rapid growth only
33
Du, 民族資本主義與舊中國政府.
374

(a)
Treaty port
Foreign firms
(b)
Treaty port
Foreign banks
Figure 11.11 Geographical distribution of foreign firms and banks (1920s)

Data Source: Huang, 外國在華工商企業辭典, for foreign firms; Yearbook of National
Banks (1934) for foreign banks
375

james kai-sing kung
after the 1900s, presumably after the Treaty of Shimonoseki, as earlier

alluded. Figure 11.13, which geo-codes the locations of the domestic institu-
tions, shows that they are in strikingly close proximity to where the SIPs were
established, suggesting that the foreign firms and banks might have produced
a “demonstration effect” on the domestic institutions. To test this conjecture,
we perform a multivariate analysis. Regardless of whether we are using
treaty ports to proxy for “Western” economic impact (e.g. columns 1 and 3
of Table 11.2) or more specifically using the actual number of foreign firms
and banks (columns 2 and 4) as predictors, the relationship between these
measures and the number of domestic firms and banks is statistically signifi-
cant, suggesting that foreign firms and banks likely had a “demonstration
effect” on the domestic institutions.34 Using only the treaty ports sample for
the test similarly confirms this salient finding (columns 5 and 6).
Another angle from which to examine the role of the treaty ports and SIPs
is to ask where the modern Chinese firms and banks chose to set up their
operations.35 Did they, for instance, also choose to operate in the treaty ports
for possible agglomeration and other benefits conferred by their foreign
counterparts, or did they opt for the SIPs or even the non-port cities, to
avoid fierce competition? Figure 11.14 provides some clues to this important
strategic question. More firms had chosen to operate within the treaty ports
than in the SIPs or the non-ports, a trend that became more pronounced after
1890 (panel A).36 This difference is even sharper for the banks, especially after
1910 (panel B). Only a handful of banks had chosen to establish themselves in
cities with no special administrative status. The distinctly greater preference
on the part of the domestic banks to operate in the treaty ports may
underscore their reliance on the legal and accounting systems, which are
essential for the proper functioning of the modern financial system.
34
The idea of a “demonstration effect” hinges on whether there were factors other than
foreign firms and banks affecting the rise of the domestic institutions. If there were, and
we are unable to identify them, we would have overestimated the effect of the foreign
institutions.
35
Firms are considered modern provided that (a) their operations were powered by the
steam engine or electricity, (b) they were relatively large, (c) they had a registered
capital of at least 10,000 silver yuan or approximately 1,094 pounds sterling, (d) they
employed at least thirty workers, (e) they produced an annual output of at least 50,000
silver yuan in value, and (f) they adopted modern (hierarchical) management practices.
Similarly, the modern Chinese banks are not qianzhuang or native banks. Chang, “清末
民初的民營工業.”
36
For example, for the 1916–1920 period, a treaty port prefecture attracted an average of
fifteen modern domestic firms, whereas an SIP prefecture attracted six and a non-port
prefecture three.
376

1000 300000
800
200000
600
400
100000
Number of firms
200
Total firm capital (unit: 1,000 yuan)

1840 1860 1880 1900 1920 1840 1860 1880 1900 1920
I. Number of firms II. Firm capital
100 80000
80
60000
60
40000
40

20000
Number of banks
20
0 0
Total firm capital (unit: 1,000 yuan)
1840 1860 1880 1900 1920 1840 1860 1880 1900 1920
III. Number of banks IV. Bank capital
Figure 11.12 Rise of domestic modern firms and banks in China (total trend)
Source: Du, 民族資本主义与旧中国政府
james kai-sing kung
(a)
Treaty port
Self-initiated port
Modern firm
(b)
Treaty port
Self-initiated port
Modern bank
Figure 11.13 Geographical distribution of domestic modern firms and banks (1920s)
Source: Du, 民族資本主义与旧中国政府
Even though some modern Chinese firms and banks chose not to operate in
either the treaty ports or the SIPs, they could still have been connected to these
ports. We plot the locations of modern firms outside the treaty ports and SIPs
according to the distance to their nearest treaty port and SIP (Figure 11.15). As can
be seen, a good proportion of these institutions were actually located in close
proximity to either a treaty port or a SIP, despite not being right in these ports.
378

Table 11.2 Impact of foreign firms and banks on domestic modern firms and
banks
Full sample Treaty ports sample
Domestic Domestic Domestic Domestic
firms (log) banks (log) firms (log) banks (log)
Treaty ports 0.378*** 0.176* 0.105*** 0.083**

(0.090) (0.090) (0.032) (0.034)
Foreign 0.748*** 0.667***
firms (log) (0.068) (0.112)
Foreign 0.181*** 0.105***
banks (log) (0.057) (0.039)
Prefecture fixed Yes Yes Yes Yes Yes Yes
effects
Year fixed effects Yes Yes Yes Yes Yes Yes
Provincial spe- Yes Yes Yes Yes Yes Yes
cific time
trend
Observations 6786 6786 6786 6786 936 936
Adjusted R2 0.478 0.603 0.200 0.208 0.760 0.324
Note. Standard errors in parentheses, * p < 0.1, ** p < 0.05, *** p < 0.01
The extent to which treaty ports affected the locational choice of

private firms and banks can be further illustrated by the extent to
which they affected the locational choice of the munition factories estab-
lished during the Self-Strengthening movement (c. 1868–1890). Bo, Liu,
and Zhou find that some of these factories were located in or near the
ports,37 as Figure 11.16 shows, presumably to facilitate the import of raw
materials and machinery from abroad. To confirm this, we regress the
locations of munition factories first on the five earliest treaty ports
selected after the Treaty of Nanking (c. 1842), then on the next nine
agreed in the Treaties of Tientsin and Peking after China lost to Britain
again in the Second Opium War and added during 1860–1864. Reported in
columns 1 and 5 of Table 11.3, we find that the earliest treaty ports can
indeed significantly account for the locational choice of the munition
factories after controlling for a number of geographic variables.
37
Bo, Liu, and Yan, “Military Investment, Industrial Linkage, and the Rise of Clusters.”
379

james kai-sing kung
A
15
Number of native modern firms (average)
10
1840 1860 1880 1900 1920

Year
Treaty port Self-initiated port Non-port
B
1.5
Number of native modern banks (average)
.5
1840 1860 1880 1900 1920

Year
Figure 11.14 Development trend of domestic modern firms and banks

Source: Du, 民族資本主義與舊中國政府
380

A
1
.8
Number of native firms
.6
.4
.2
0
1840 1860 1880 1900 1920
Year
1 km–50 km 50 km–100 km 100 km–200 km
200 km–300 km >300 km
B
.8
Number of native firms
.6
.4
.2
0
1840 1860 1880 1900 1920
Year
1 km–50 km 50 km–100 km 100 km–200 km
200 km–300 km >300 km
Figure 11.15 Development of domestic modern firms in non-port regions
Note. In panel A, distance indicates a prefecture’s distance to the nearest treaty port.
Likewise, in panel B, distance indicates a prefecture’s distance to the nearest self-initiated
port.
Source: Du, 民族資本主義與舊中國政府
381

james kai-sing kung
1879
1878 1881
1883 1882
1867
1884 1883
1871 1875
1869
1865
1889 1865
1877 1884 1881
1875 1874
1880 1866
1872 1873 1883
Munition factory
Treaty port
(opened before 1865)
Figure 11.16 Spatial distribution of treaty ports opened before 1865 and munition factories
Source: Chang, 清末民初的外資工業; Fan Baichuan 樊百川, 清季的洋務新政 (The
Qing Westernization Movement) (Shanghai, Shanghai shudian chubanshe, 2003)
Last but not least, we examine the issue of technology adoption, for the simple
reason that the modernity of firms also depends crucially on whether they
adopted an Industrial Revolution technology – most notably the steam engine
and (subsequently) electricity. Figure 11.17 shows the overall growth trend of the
capacity of China’s power plants (1903–1931), with 1920 representing a year of
rapid growth. Figure 11.18 maps the geographic distribution of China’s earlier
electrical plants in 1929, including those constructed by foreigners. An interesting
feature of the development of electric power plants in that period is that the
government only played a small part; in fact most of these plants were built
through private endeavors. While most of the private electrical plants tended to
be smaller in scale (in terms of generation capacity), they were built all over
China. In particular, many were located in south and south-central China –
locations where government efforts were visibly lacking.38 Similar to the
government-funded plants, the bulk of the foreign-built power generators
38
These regions may have become more decentralized economically after the Taiping
Rebellion, when a province like Hunan (in south-central China) had built up its own
strong army with its tax income.
382

Table 11.3 Location of munition factories
Munition factories
Treaty ports (first 5) 4.066***

(1.145)
Distance to treaty port (first 5)
0 kilometer 3.920*** 2.568* 1.353
(1.152) (1.333) (1.562)
1–200 kilometers −0.557 −1.659 −3.743**
(1.059) (1.223) (1.619)
200–400 kilometers −0.384 −0.951 −0.726
(0.775) (0.741) (0.924)
400–600 kilometers −0.834 −1.441 −2.260*
(1.054) (1.288) (1.373)
Treaty ports (first 14) 3.785***
(0.662)
Distance to treaty port (first 14)
0 kilometer 3.556*** 3.825*** 2.871**

(0.718) (1.119) (1.168)
1–200 kilometers 0.200 0.368 -1.285
(0.650) (0.817) (1.268)
200–400 kilometers −0.736 −0.627 −1.281*
(0.818) (0.876) (0.768)
400–600 kilometers −1.146 −1.074 −1.579
(1.085) (1.097) (1.076)
Table 11.3 (cont.)
Munition factories
Distance to coast (log) −0.423* −0.782** 0.080 −0.285

(0.221) (0.370) (0.238) (0.368)
Distance to main river (log) −0.173 −0.033
(0.194) (0.194)
Distance to provincial capital (log) −0.868*** 0.837***
(0.184) (0.170)

Observations 301 301 301 301 301 301 301 301
Robust standard error Yes Yes Yes Yes Yes Yes Yes Yes
Note. These are logistic regressions. Robust standard errors in parentheses, * p < 0.1, ** p < 0.05, *** p < 0.01.
200000
180000
160000
Electricity capacity (kW)
140000
120000
100000
80000
60000
40000
20000
0
1903 1908 1913 1918 1923 1928
Year
Figure 11.17 Growth of electrical power plants in China, 1903–1931
Source: Shun pao yearbook (1936)
were concentrated in north and northeastern China, where private interests,

albeit small, were also well represented. Judging from the geographical coverage
of the private electrical plants, it is by no means an exaggeration to suggest that
private interests played a uniquely important role in spearheading the develop-
ment of modern industrial firms in China.
As elsewhere, electrification came much later than the steam engine, and
China was no exception in that regard.39 Before the advent of electricity,
firms in China were considered to be “modern” if they used machinery
without necessarily powering it by steam; prominent examples included
cotton gins in the textile industry, silk filatures, and bean cake fertilizer
plants, according to a survey of so-called modern firms established between
1841 and 1916 by Chang.40 This may explain why, when comparing the two
panels of Figure 11.19, which plots the average number of firms adopting the
39
Even in Japan, where industrialization occurred much earlier, many factories in the
nine dominant industrial sectors still used the internal combustion engine in non-
powered factories as late as the 1940s, according to R. Minami, “Mechanical Power in
the Industrialization of Japan,” Journal of Economic History 37.4 (1977), 935–58.
40
Chang Yufa 張玉法, 近代中國工業發展史, 1860–1916 (A History of Modern Chinese
Industrial Development, 1860–1916) (Taipei, Gui guan tu shu you xian gong si, 1992).
385

Electrification in 1929
Capacity of private power plants
3–412
413–1350
1351–3675
3676–20215
20216–44098
Capacity of foreign power plants

3–412 Capacity of official power plants
413–1350 3–412
1351–3675 413–1350
3676–20215 1351–3675
20216–44098 3676–20215
Prefecture boundary in 1920 20216–44098
Figure 11.18 Electricity generation in China, 1929

Source: Shun pao yearbook (1936)
A
0.5
Average number of firms in a prefecture
adopting the steam engine
0.4
0.3
0.2
0.1
0
1840 1850 1860 1870 1880 1890 1900 1910
B
Average number of firms in a prefecture
0.8
adopting machinery
0.6
0.4
0.2
0
1840 1850 1860 1870 1880 1890 1900 1910

Figure 11.19 Adoption of the steam engine and machinery in modern Chinese firms
steam engine (panel A) and machinery (panel B) over time, more firms
adopted machines than adopted the steam engine, and the trend toward
mechanization started much earlier (from around 1850) than that toward the
steam engine (panel A) – whose wider adoption had to wait until the 1880s
387

james kai-sing kung
Table 11.4 Adoption rate of the steam engine and machinery

Steam engine Machinery
Treaty Self-initiated Non- Treaty Self-initiated Non-
port port port port port port
Adoption 6.8% 7.6% 6.4% 17.9% 15.6% 16.1%

rate
(panel B).41 In terms of the geography of adoption, both machinery and steam
engines were more widely adopted by firms located in the treaty ports than in
the SIPs and non-port cities.42 However, if we consider the percentage rather
than the number of firms adopting the two new technologies, it was actually
similar across the three types of cities (Table 11.4).
The “Persistent” Effects of Treaty Ports

Using population and GDP growth as gauges in her study of the long-run
growth effect of China’s treaty ports, Jia finds that while treaty ports had no
discernible effect on development during the communist period, prefectures
in which treaty ports had been formerly established systematically experi-
enced faster economic development in the wake of China’s “open-door”
policy in the 1980s.43 To confirm this, we use four economic indicators in
contemporary times (c. 2010) – GDP, GDP per capita, a prefecture’s fiscal
income, and industrial output – to compare the performance of treaty ports
and SIPs with that of prefectures without any historical port status. We find
that both types of ports perform significantly better today regardless of
whether we control for a gamut of geographic factors that are likely correl-
ated with economic performance (Table 11.5). For example, in column 1, both
treaty ports and SIPs have higher GDP today than non-ports; controlling for
various geographic factors only reduces their magnitude but not level of
significance (column 2). The larger magnitude of treaty ports implies that
41
The later adoption of the steam engine can be accounted for by the fact that the first
steam engine in China was built by Xu Shou 徐壽 in 1862 based on technology brought
back from Europe, after which it took another decade or so before steam power began
to be adopted on a wider scale.
42
In the case of the steam engine, while firms in the SIPs had mostly caught up by 1890, this
lasted for a mere ten years before the gap widened again. The same trend applies to foreign
firms. By the same token, with the exception of a brief interlude (c. 1900), more firms in the
treaty ports embraced mechanization than did firms in SIPs or non-port cities.
43
Jia, “The Legacies of Forced Freedom.”
388

Table 11.5 Long-term effects of treaty ports and self-initiated ports
GDP Fiscal income Industrial output GDP per capita
(log) (log) (log) (log)
Treaty ports 0.780*** 0.552*** 1.008*** 0.730*** 0.796*** 0.609*** 0.486*** 0.417***
(0.158) (0.102) (0.185) (0.142) (0.206) (0.146) (0.094) (0.089)
Self-initiated ports 0.580*** 0.277** 0.712*** 0.362** 0.697*** 0.282* 0.292*** 0.211**
(0.143) (0.109) (0.174) (0.153) (0.196) (0.155) (0.095) (0.098)
Population (log) 0.761*** 0.693*** 0.718***
(0.059) (0.088) (0.090)
Geographic controls Yes Yes Yes Yes
Observations 287 250 287 250 287 250 286 250

Adjusted R2 0.148 0.598 0.173 0.491 0.093 0.454 0.116 0.147
Robust standard error Yes Yes Yes Yes Yes Yes Yes Yes
Notes. Standard errors in parentheses, * p < 0.1, ** p < 0.05, *** p < 0.01. Geographic controls include longitude, latitude, distance to main
river, and distance to coast.
james kai-sing kung
they have a larger effect than the SIPs. The long-run effects of these two types
of ports are consistently the same for fiscal income (columns 3 and 4),
industrial output (columns 5 and 6), and GDP per capita (columns 7 and 8).
Treaty Ports and Human Capital Development

We have shown that treaty ports are highly correlated with economic
development in both the short and the long run. In this section, we document
the relationship between treaty ports and human capital development via the
channel of education.
The Role of Missionaries

Joseph Needham once remarked that the traditional Chinese education
underpinned by Confucianism (the lynchpin of the civil service examination)
provided a set of skills and knowledge not conducive to economic
development,44 a remark strongly hinting at China’s failure to industrialize
at a time when Western Europe succeeded. Beginning with Simon Kuznets
and subsequently reiterated by Joel Mokyr,45 economists have adopted the
term “useful knowledge” to emphasize that some types of knowledge are
more useful than others in stimulating economic development. In the
Chinese context, the opening up of the country as a consequence of the
two Opium Wars brought back the Western missionaries (along with the
merchants and the diplomats), whose interests lay not merely in converting
China to Christianity but more so in introducing to the Chinese a brand new
system of values and knowledge fundamentally different from what the
Chinese had held dear for millennia.46 For instance, the missionaries had
translated and published many Western texts in a variety of disciplines
ranging from the pure sciences and medicine to the social sciences, erected
schools that adopted a curriculum fundamentally different from that of the
Confucian classics-based civil service examination, and put up hospitals
practicing Western medicine. As Jonathan Spence observes, “Through their
texts, their presses, their schools, and their hospitals, the effort of the
missionaries affected Chinese thought and practice. The strength of that
44
J. Needham, The Great Titration (Toronto, University of Toronto Press, 1969).
45
S. Kuznets, Economic Growth and Structure (New York, W.W. Norton, 1965). Joel Mokyr,
The Gifts of Athena: Historical Origins of the Knowledge Economy (Princeton, Princeton
46
Y. Bai and J.K.S. Kung, “Diffusing Useful Knowledge While Spreading God’s Message:
Protestantism and Economic Prosperity in China, 1840–1920,” Journal of the European
Economic Association 13.4 (2015), 669–98.
390

influence is impossible to calculate, but the missionaries did offer the Chinese
a new range of options, a new way of looking at the world.”47
Christianity returned to China after more than a century since it was banned
by the Kangxi Emperor in 1721. Initially, missionary activities were permitted
only in the treaty ports after China’s defeat in the First Opium War and with the
signing of the Treaty of Nanking in 1842. It was only after China’s second defeat
in 1856 and the signing of the Treaty of Tientsin that missionaries were allowed
to enter the inland regions. A nationwide survey of missionary activities con-
ducted by Milton Stauffer in 1922 confirmed that the Protestant missionaries
began to cover China far and wide only after the 1860s (Figure 11.20).48 By the
early 1920s, Stauffer’s survey revealed that more than 94 percent of Chinese
counties had records of a missionary presence. More specifically, 84 percent of
these counties kept records of the Protestant communicants, with 78 percent
having established Protestant congregations or evangelistic centers.
Where they had penetrated, the Protestant missionaries did far more than
disseminate the Christian texts. Believing that China was backward and needed
1000
Number of counties with initial
800
missionary activities
600
400
200
0
1800 1820 1840 1860 1880 1900 1920
Year
Figure 11.20 Growth of missionary activities in China
Source: M.T. Stauffer, The Christian Occupation of China: A General Survey of the Numerical
Strength and Geographical Distribution of the Christian Forces in China (Shanghai, China
Continuation Committee, 1922)
47
Spence, The Search for Modern China, p. 208.
48
Stauffer does not include data on the number of Catholic communicants, however.
Using the number of Catholic stations to proxy for the influence of Catholic missionar-
ies on economic prosperity (measured by the degree of urbanization), Bai and Kung
find no significant association between the two. M.T. Stauffer, The Christian Occupation
of China: A General Survey of the Numerical Strength and Geographical Distribution of the
Christian Forces in China (Shanghai, China Continuation Committee, 1922); Bai and
Kung, “Diffusing Useful Knowledge While Spreading God’s Message.”
391

james kai-sing kung
Western culture to help it develop, the missionaries built a large number

of schools – mostly primary but also some secondary ones. Figure 11.21 reveals
the close correlation between their presence and the density of primary schools
they erected – be that on the coastal seaboard in the east and southeast or up in
the north.49 By introducing for the first time a thoroughly Western-based
curriculum with subjects ranging from those in the pure sciences such as
mathematics to the social sciences and humanities, such as geography and
history, the missionary primary schools had an enormous influence on the
government schools at all levels in terms of curriculum. Table 11.6 reveals the
differences in the curriculum between the missionary schools (c. 1876) and the
“new schools” (c. 1904) set up everywhere in the country by the Chinese
Missionary primary school

(per 1000 people)
0–0.241
0.242–0.652
0.653–1.836
1.837–3.942
3.943–10
Missionary
(per 1000 people)
0–25
26–57
58–1000
Figure 11.21 Missionary activities and the primary schools they founded (1922)
Source: Stauffer, The Christian Occupation of China
49
The western region seems to represent an exception, where there were disproportionately
more primary schools than there was missionary presence.
392

Table 11.6 Comparison of curricula in the 1900s

Missionary New school curriculum
Curriculum (since 1876) (1904)
Bible Yes
Ideology and classics Yes
Chinese classics Yes Yes
Chinese characters Yes Yes
History Yes Yes
Geography Yes Yes
Mathematics Yes Yes
Physical education Yes
Physics Yes Yes
Music Yes
Painting and mapping Yes Yes
Source: Gu Changsheng 顧長聲, 傳教士與近代中國 (Missionary and Modern

China) (Shanghai, Shanghai renmin chubanshe, 2004); Qiu Xiuxiang 邱秀香, 清末新
式教育的理想與現實 (The Ideality and Reality of New Education in Late Qing: An
Exploration of the Rise of New Primary Schools) (Department of History, National
Chengchi University, 2000)
government.50 The two types of schools adopted similar curricula, except that
Bible study in the curriculum of the missionary schools was replaced by the
compulsory subject of ideology and classics aimed at indoctrinating the students
to obey authority, particularly the emperor (as such, the subject was still premised
on the Confucian classics). A great majority of the subjects in the curriculum of
the new schools were novel to the Chinese.
Bai and Kung make use of Stauffer’s data to examine whether Protestantism
helped promote economic prosperity by means of Max Weber’s “Protestant
ethic” (of working hard and saving for future investments).51 Using a cross-
sectional measure of urbanization rates in the early 1920s and a panel of data on
the growth of modern firms for the 1895–1930 period as proxies for economic
prosperity, they find that prefectures with more Protestant communicants were
more urbanized in the early 1920s and their modern firms exhibited more robust
growth during the period in question. However, they find that what was
50
To be sure there were fewer secondary schools than primary schools (Qing
Government First Education Census, 1907). Moreover, their distribution also appeared
to be more skewed. Geographically the secondary schools were concentrated dispro-
portionately on the southeast coastal seaboard, the south coast, along the Yangzi river,
and in the northeast (where junior primary schools abound).
51
Bai and Kung, “Diffusing Useful Knowledge While Spreading God’s Message.”
393

james kai-sing kung
effectively promoting economic prosperity back then was not a cultural ethic but
rather education and health care, as the Protestant missionaries had erected
schools (both primary and middle) and hospitals in prefectures where they
were stationed. In particular, by providing useful knowledge more effectively,
prefectures with more middle schools and hospitals were poised to grow dis-
tinctly faster.
Modern Universities
Not only had the opening up of China radically changed the fundamentals of
basic education (up to senior middle school), but it also paved the way for the
modern university. Immediately after 1860 (shortly after the Second Opium War
treaty was signed), a handful of military and language colleges were established
by the Qing government with the overriding aim of teaching new military
technologies and foreign languages.52 The number increased sharply after 1905
upon the abolition of the civil service examination (Figure 11.22). Still more
universities were established in the Republican period. By 1937, more than 200
modern colleges and universities had been established. As Figure 11.23 shows,
most universities in China were founded by the Chinese government. While
private universities also grew rapidly at the turn of the century and accounted for
more than a third (36 percent) of all universities present by the 1930s, and with
missionary universities also experiencing stable growth, in the Chinese context
public universities still accounted for approximately half of all universities in
China.53
We can see from Figure 11.24 that nearly all provincial capitals (the small
dots) had at least one university as per government mandate. It is also worth
noting that, with the exception of the southwest and to a lesser extent the
southeast, a great majority of the universities were also located in or around
the treaty ports and SIPs. This is not hard to understand. To the extent that
the treaty ports spearheaded the development of the local economy by virtue
of accommodating more foreign firms and banks, they increased the demand
for upper-tail human capital, which in the context of a modern economy
would be engineers, accountants, and lawyers.54 This surge in demand for
52
In our classification that follows we do not consider them as universities.
53
The earliest universities in China – those established before 1900 such as Saint Johns and
Dongwu – were invariably founded by the missionaries.
54
For the importance of upper-tail human capital in the French Industrial Revolution, for
example, see M. Squicciarini and N. Voigtlander, “Human Capital and Industrialization:
Evidence from the Age of Enlightenment,” Quarterly Journal of Economics 130.4 (2015),
1825–83.
394

200
Number of universities
150
100
50
0
1885 1890 1895 1900 1905 1910 1915 1920 1925 1930
Year
Figure 11.22 Growth of universities in China
Source: Second China Education Yearbook (1948)
120
100
80
60
40
20
0
1885 1890 1895 1900 1905 1910 1915 1920 1925 1930
Missionary universities Private universities
National universities
Figure 11.23 Growth of universities in China, by type/funding source
new forms of human capital would in turn stimulate the demand for univer-
sities, especially those associated with private initiatives.
To put the development of human capital in the treaty ports in context, we
conducted a multivariate analysis and report the results in Table 11.7. In
395

james kai-sing kung
Number of
universities
1
2–3
4–5
6–9
10–23
Provincial capital
Treaty port
Self-initiated port
Figure 11.24 Distribution of universities in 1937

column 1, we find that both treaty ports and SIPs are significantly correlated
with the development of universities at respectively a 1 percent and
a 5 percent level of statistical significance; the magnitude is larger for treaty
ports than for SIPs. In other words, higher education in twentieth-century
China was better developed in the treaty ports than in the SIPs, which is
hardly surprising. In column 2 we replace the two kinds of ports by both
foreign and domestic firms and banks, and find that only foreign firms and
domestic banks are significantly correlated with the development of univer-
sities (the former at the 10 percent level and the latter at the 5 percent level of
conventional statistical significance). In columns 3 and 4 we divide
the universities into those with an engineering school and those without,
assuming that engineers are more important for the growth of industrial
firms.55 Once again, both treaty ports and SIPs are significantly correlated
with universities with an engineering school, as are foreign firms (whose level
55
For the importance of engineering for economic growth, see F. Caicedo and
W. Maloney, “Engineering Growth: Innovative Capacity and Development in the
Americas,” CESifo Working Paper Series 6339, CESifo Group Munich, 2017.
396

Table 11.7 Determinants of university locations and engineers
Universities
Universities with without an
an engineering engineering
Universities school school Engineers
(1) (2) (3) (4) (5) (6) (7) (8) (9)
*** *** *** ***
Treaty ports 0.391 0.113 0.217 1.096
(0.072) (0.039) (0.065) (0.170)
Self-initiated ports 0.179** 0.108** 0.133* 0.911***
(0.082) (0.046) (0.075) (0.195)
University with engineering school (log) 0.600**
(0.298)
University without engineering school (log) 0.066
(0.200)
Foreign firms (log) 0.186* 0.100** 0.073* 0.297*** 0.151
(0.097) (0.044) (0.038) (0.104) (0.109)

Domestic firms (log) 0.047 −0.007 0.020 0.716*** 0.647***
(0.037) (0.020) (0.022) (0.058) (0.057)
Foreign banks (log) −0.001 0.058 0.177** −0.155 −0.291
(0.190) (0.079) (0.072) (0.195) (0.194)
Domestic banks (log) 0.183** 0.075 0.242*** −0.124 0.184
(0.072) (0.055) (0.042) (0.114) (0.123)
Jinshi density (log) 0.174** 0.097 0.004 −0.050 0.192** 0.073 0.346* 0.120 0.072
(0.084) (0.067) (0.046) (0.038) (0.076) (0.057) (0.198) (0.154) (0.154)
Table 11.7 (cont.)
Universities
Universities with without an
an engineering engineering
Universities school school Engineers
(1) (2) (3) (4) (5) (6) (7) (8) (9)
Shengyuan quota density (log) −0.008 0.058 0.009 0.044 −0.092 0.002 0.028 0.090 0.113
(0.100) (0.060) (0.055) (0.036) (0.091) (0.067) (0.208) (0.183) (0.180)
Provincial capital 1.066*** 0.833*** 0.402*** 0.337*** 0.747*** 0.508*** 1.885*** 1.201*** 0.770***
(0.092) (0.168) (0.051) (0.110) (0.084) (0.067) (0.218) (0.181) (0.209)
Population density (log) 0.014 0.002 −0.005 −0.009 −0.006 −0.011 0.125 −0.018 0.009

(0.034) (0.020) (0.019) (0.013) (0.031) (0.023) (0.078) (0.064) (0.062)
Provincial fixed effects Yes Yes Yes Yes Yes Yes Yes Yes Yes
Observations 255 255 255 255 255 255 255 255 255
Adjusted R2 0.610 0.721 0.352 0.567 0.461 0.701 0.602 0.769 0.777
of statistical significance has now increased from 10 percent to 5 percent). In

columns 5 and 6 we change the dependent variable to “universities without
an engineering school.” While similar results are found using treaty ports and
SIPs as the explanatory variables, this time not only are the foreign firms
significantly correlated with this dependent variable, but so too are the
banks – both foreign and domestic (which enjoy the highest level of conven-
tional statistical significance (1 percent)). This result is highly intuitive, as
banks are likely to exhibit a stronger demand for such specialists as account-
ants, lawyers, and bankers than for engineers.
We then examine the geographic distribution of the engineers’ workplace
and the kind of institutions with the strongest demand for their expertise. We
report the results in columns 7 through 9. Foremost is that a good majority of
the engineers worked in the treaty ports and SIPs (column 7); the larger
coefficient of treaty ports suggests that they provided more employment
opportunities for the engineers, again an unsurprising result (see also
Figure 11.25 for their geographic distribution).
We also find that the only demand for engineering expertise came from
firms – both foreign and domestic (column 8); the larger coefficient on the
domestic firms suggests that they had a greater demand for engineers. The
greater demand from domestic firms is further confirmed in column 9, after
controlling for the type of universities, i.e. whether a university had an
engineering school. Consistent with intuition, banks did not require the
service of engineers.
Finally, given that provinces were decreed to set up universities in their
capital city, this variable is significant throughout the regressions, as can be
visualized also in Figure 11.24. While significantly correlated with the estab-
lishment of universities, jinshi 進士 density – which measures the number
of jinshi scholars in the Ming–Qing period produced by each prefecture
normalized by its population – fails to account for the presence of those
universities with an engineering school (compare, for example, columns 1
and 3). Moreover, once we control for the presence of firms, jinshi density
becomes insignificant (columns 2 and 4). Shengyuan 生員 quota, which has
been used as a proxy for social mobility and as a corollary for political
stability,56 has no explanatory power for the founding of universities. The
same applies to population density.
56
Y. Bai and R. Jia, “Elite Recruitment and Political Stability: The Impact of the Abolition
of China’s Civil Service Exam System,” Econometrica 84.2 (2016), 677–733.
399

james kai-sing kung
Engineer
Treaty port
Self-initiated port
Provincial capital
Figure 11.25 Distribution of engineers in 1937

Resouce Committee 资源委员会 (ed.), 中国工程人名录 (A Directory of Chinese
Engineers) (Changsha, Commercial Press, 1941)
Beyond the Treaty Ports: Overseas Studies and

Political Change
The treaty ports brought not just economic change but also social and
political change, some of which was drastic beyond all expectations. The
Xinhai Revolution of 1911, for example, represents an extreme case in point,
whose unexpected occurrence brought an end to millennia-long imperial
China.
Ever since China first opened up to the West, Western knowledge and
ideas (including ideologies) spread like wildfire through the media, the
translation of books, and, as mentioned earlier, the adoption of a Western-
style education. Of the many new ideas disseminated in China, the call for
political reforms was high on the agenda, chiefly because China had fully
revealed its weaknesses in military and technological capabilities – weak-
nesses that many saw as deeply rooted in its backwardness in different arenas
vis-à-vis the Western powers.
400

Although one could learn about the West via the universities, news-
papers, and translated texts, a better way to acquire new corpuses of
knowledge firsthand and to gain educational qualifications was to study
overseas, as the political elite Zhang Zhidong strongly advocated.57
While the Qing government had already begun sending students to
study science and engineering in the United States as early as the
1870s, and later on to Europe to study military technology, the scale
was modest – only 100 or so students in total in either case. The third
and also by far the biggest wave of studying abroad did not happen until
after 1895, when China had been defeated in the First Sino-Japanese War
a year earlier. Since then, cumulatively tens of thousands of students
went to study in Japan, a trend that accelerated in 1905, when China’s
millennia-long civil service examination was eventually abolished.
Figure 11.26, which is plotted based on part of Zhou’s data on overseas
study,58 shows the number of students who studied overseas and
returned to China upon completing their studies.59 The nearly identical
patterns of the two curves suggest that the majority of students who
studied overseas did return to China.
As we shall see later, the 1911 Revolution was engineered by what may
categorically be known as the Revolutionaries or the Revolutionary Alliance,
many of whom had studied in Japan. For this reason, it is important to
understand why Japan became the most popular destination among
Chinese students. But before doing so, we would like to examine why it
became popular to study abroad in the first place. In the light of the import-
ance of Western influence, we propose several factors that we consider were
57
As the governor-general of the provinces of Hunan and Hubei, Zhang remarked that
“to study in the West for one year is better than reading Western books for five
years . . . to study in a Western school for one year is better than to study for three in
the Chinese schools.” Cited in M. Jansen, “Japan and the Chinese Revolution of 1911,” in
J.K. Fairbank and K.-C. Liu (eds.), The Cambridge History of China, vol. 11, Late Ch’ing,
1800–1911, part 2 (Cambridge, Cambridge University Press, 1980), p. 349.
58
Zhou Mian 周棉, 中國留學生大辭典 (Chinese Overseas Student Dictionary)
(Nanjing, Nanjing University Press, 1999).
59
There are altogether some 3,000 or more students in the overall sample of Zhou, 中國
留學生大辭典, but only 600 up until 1911. See Sanetō Keisha, 中国人日本留学史 (A
History of Chinese Students in Japan) (Tokyo, Kuroshio shuppan, 1970); and Li Xisuo 李
喜所, 近代中国的留学生 (Foreign-Trained Students in the History of Modern China)
(Beijing, Renmin chubanshe, 1987) for the estimates, which are cited in D. Reynolds,
China, 1898–1912: The Xinzheng Revolution and Japan (Cambridge, MA, Harvard
University Press, 1993), p. 41, according to whom “(t)he flow of Chinese students to
Japan was the most dramatic single development in relations between China and Japan
after 1898.”
401

james kai-sing kung
500
400
300
200
100
0
1840 1860 1880 1900 1920 1940
Overseas students Returnees
Figure 11.26 Number of Chinese students studying abroad and returning, a partial sample
Source: Mian Zhou 周棉, 中國留學生大辭典 (Chinese Overseas Student Dictionary)
(Nanjing, Nanjing University Press, 1999)
likely highly correlated with this particular outcome. These factors include
treaty ports and, to a lesser extent, the SIPs for their disproportionately
greater exposure to Western culture and thoughts, and more specifically
foreign firms and banks and (to a lesser extent) modern Chinese firms and
banks; an edifice of modern education ranging from primary and secondary
schools to the universities; missionary presence, and so forth. The results are
reported in Table 11.8.
We can see from column 1 that treaty ports (but not SIPs) are significantly
correlated with overseas studies, suggesting that exposure – including its
duration – to Western culture, ideas, and so forth may have increased one’s
proclivity towards studying overseas. So too does missionary presence;
individuals living in prefectures where the missionaries had set up operations
for longer were also more likely to study overseas. The same applies to the
densely populated prefectures and the provincial capitals, for the reason that
news and information about the West were likely more rapidly and thor-
oughly diffused. An interesting variable in this connection is jinshi density,
which is strongly correlated with studying overseas. This may suggest that
individuals coming from places with a strong tradition of civil service exam-
ination success were perhaps more inclined to acquire new knowledge and/
or seek modern qualifications by studying abroad.
In column 2 we replace the two treaty port variables with the firm and
bank variables (one foreign and the other domestic), and find that only firms
402

Table 11.8 Determinants of overseas studies

Studying overseas
(1) (2) (3) (4)
** *
Treaty ports 0.329 0.229 0.291*
(0.139) (0.135) (0.148)
Self-initiated ports 0.098 0.039 0.075
(0.150) (0.145) (0.157)
Domestic firms (log) 0.118**
(0.055)
Domestic banks (log) 0.047
(0.109)
Foreign firms (log) 0.169*
(0.099)
Foreign banks (log) 0.062
(0.217)
Modern universities (log) 0.381*** 0.346***
(0.135) (0.123)
Primary schools (log) 0.184*** 0.171***
(0.054) (0.052)
Secondary schools (log) 0.168* 0.102
(0.101) (0.107)
Duration of missionary presence 0.105*** 0.072* 0.090**
(0.036) (0.038) (0.045)
Railway 0.198 −0.042 0.0110 0.041
(0.159) (0.168) (0.167) (0.147)
Population density (log) 0.184*** 0.155** 0.120* 0.121**
(0.065) (0.065) (0.064) (0.050)
Jinshi density (log) 0.767*** 0.705*** 0.650*** 0.645***
(0.151) (0.150) (0.147) (0.191)
Shengyuan quota density (log) −0.403 −0.340 −0.565 −0.560
(0.833) (0.819) (0.804) (0.822)
Provincial capital 0.538*** 0.367** −0.010 −0.030
(0.172) (0.182) (0.229) (0.332)
Provincial fixed effects Yes Yes Yes Yes
Observations 239 239 239 239
Adjusted R2 0.538 0.559 0.570 0.576
are significantly correlated with studying abroad. The other variables that are
significant in column 1 continue to be significant in column 2, although
except for jinshi density their levels of statistical significance are all reduced.
We then control for the possible influence of modern education in column 3,
403

james kai-sing kung
and find that all the three measures – primary, secondary, and university – are
significantly correlated with studying overseas, with university and primary
education being much more significant than secondary schools. Of the three,
universities are the most important (the coefficient is more than twice that of
the primary schools). In column 3, provincial capital is no longer significant,
suggesting that its previous significance was likely due to the omission of the
universities. Likewise, treaty ports have also become less significant, suggest-
ing that a large part of their influence on studying abroad is likely coming
from modern education, in particular the universities. Finally, in column 4,
we put the missionary variable back in the regression, and find that secondary
schools have now become insignificant, but primary schools and universities
continue to correlate significantly with studying abroad. Summing up, mod-
ern education, missionary presence, and a strong tradition of human capital
and other unobserved aspects of treaty ports can importantly explain the
decision to study overseas. These interesting correlates notwithstanding, we
need to be cognizant of the fact that more than half of the overseas students
were sponsored by the government at various levels.
So the questions that need addressing at this point are: (1) why Japan, (2) who
went overseas and what levels of education and subjects of study did they
pursue, and (3) why did the Japanese experience turn them against China?
Japan was favored for a variety of reasons. Given the similar challenges
confronting China and Japan, the Japanese experience demonstrated the
possibility of transforming an imperial/feudal regime into a constitutional
monarchy without necessarily overthrowing the former – a reform deemed
desirable by the Manchu regime.60 Success in this matter also won Japan an
alliance with Britain. But Japan was also favored after it convincingly defeated
the Russians – a much more sizeable country – in the Russo-Japanese War, an
outcome hinting at the importance of having a superior political system.
Third, the geographic proximity of Japan easily made it a far more convincing
case than either the United States or Europe insofar as the costs of travel and
tuition were concerned, regardless of who bore the costs.61 Finally, by that
time, the Japanese had already translated most of the Western works deemed
essential to the Chinese.62
60
To learn from a number of more advanced countries about how constitutional
governance was run, in 1905 the Qing government dispatched five ministers to as
many as thirteen countries, including Japan, the United States, Britain, France,
Germany, and Russia, although it turned out that the regime favored Japan.
61
These costs were estimated at about a fifth of those of studying either in the United
States or in Europe by Zhang Zhidong.
62
See Jansen, “Japan and the Chinese Revolution of 1911”; Reynolds, China, 1898–1912.
404

While the earlier batch of students were largely sponsored by the govern-
ment at various levels, with the abolition of the traditional civil service
examination in 1905 an increasing proportion would have to become self-
sponsored as many must have felt uncertain about their career prospects and
hence the need to upgrade their qualifications.63 Still others might have gone
in the hope of acquiring knowledge that might contribute to a stronger
China.
The level of studies pursued by the self-sponsored students also varied; the
great majority of them (75 percent) attended senior secondary schools,
specialized or professional schools, and even short cram courses that lasted
for three to nine months and in some cases up to twelve months (sokuseihan
促成班). These programs allegedly catered to the Chinese students’ educa-
tional needs. Another 15 percent were enrolled in unspecified preparatory
programs, presumably involving intense training in the Japanese language. In
terms of disciplines, the vast majority of self-sponsored students were
enrolled in the liberal arts and teacher training (arts, business, and social
sciences),64 and only those who were sponsored by the government special-
ized in military studies.65 Less than 10 percent were enrolled in schools
equivalent in status to universities approved by the Japanese government;
these students studied a variety of subjects ranging from science, engineering,
and medicine to law, business, and agriculture.
The next question is why the Japanese experience turned the overseas
students against China. Essentially, in Meiji Japan a number of Western ideas
about government were no longer foreign to the overseas Chinese students,
who now could freely discuss the possibilities of future government without
being held back by traditions. Furthermore, with the positive examples of
a rising Japan and the negative daily experiences of both condescension and
discrimination fueling their sense of nationalism, these students now per-
ceived a modernized Japan as a benchmark against which to critically evalu-
ate the weaknesses of the Qing government. While some (the
constitutionalists) might wish to follow in the footsteps of the Meiji reformers
by establishing a parliament (which eventually failed),66 others (the
63
Many indeed had gone to Japan for short courses or “cram programs,” or in the words
of Reynolds, China, 1898–1912, p. 61, “quick certification for choice jobs back home.”
64
Waseda University, for example, was famous for its teacher-training program.
Reynolds, China, 1898–1912, p. 61.
65
Sanetō, Chugokujin Nihon Ryugakushi.
66
By drawing upon the expertise of a great many political elites and literati with the likes
of Zhang Jian 张謇 and Liang Qichao 梁啟超, the constitutionalists petitioned thrice
for the establishment of a national assembly after establishing the provincial assemblies,
405

james kai-sing kung
revolutionists) remained agnostic that the Manchu would be capable of continu-

ing their rule of a strengthened China – especially after witnessing its failure to
fend off the imperial Western powers that were encroaching upon China’s own
economic interest. These two major political powers changed China’s destiny
fundamentally. Between the two, the revolutionists were the ones whose goal
was to establish a Republican government via revolution. To this end, a good
number of revolutionary groups were established, most notably the Revive
China Society (Xinzhonghui 興中會, founded in 1894), the Restoration Society
(Guangfuhui 光復會, founded in 1904), and the Chinese Revolutionary Alliance
(Zhongguo tongmenghui 中國同盟會, founded in 1905 by Sun Yat-sen 孫中山).
Figure 11.27 shows the geographic distribution of these two political organiza-
tions. While they did overlap in some regions (most notably the lower Yangzi
delta region, for instance), they also occupied different parts of the country. For
instance, there were more revolutionists in the south-central part of China,
which is where the revolution was initially staged.67
Shortly after staging a revolutionary uprising in the south-central city of
Wuchang (Hubei province) on October 10, 1911, many revolts and uprisings
ensued, resulting in as many as fourteen provinces announcing independence
from the Qing state, and the subsequent announcement of the founding of
the Republic of China in January 1912.
What precisely is the connection between studying in Japan and the 1911
Revolution that overthrew the Qing dynasty? Owing to data limitations, we can
only address this question indirectly by examining how studying in Japan might
have borne upon the returnees’ choice of political organization and membership,
with implications for revolutionary outcome based on the analysis conducted in
Table 11.9.68 Specifically, the dependent variable is membership of a political
organization – constitutionalist versus revolutionist, whereas the independent
variable is the subjects in which they majored while in Japan.69 Before doing so,
however, we want to identify the general outcome of studying overseas first. As
but failed. J. Esherick, “Reconsidering 1911: Lessons of a Sudden Revolution,” Journal of

Modern Chinese History 6.1 (2012), 1–14.
67
What we are not establishing are the causal links between where the revolutionists
were located and the occurrence of the revolutions, which would require additional
work coding the geography of conflict and social networks among the leading
revolutionaries.
68
As with Table 11.8 we rely on the data provided by Zhou, 中國留學生大辭典, on the
600 or so students (out of a total of a few thousands that studied abroad) for the
analysis.
69
Ideally, we should also include a measure of whether one was sponsored by the
government or not, as that may affect a student’s political outlook and thus choice of
political organization, as well as choice of major. Unfortunately, such information is
not available from Zhou, 中國留學生大辭典.
406

Core constitutionalist
member
1–2
3–5
6–8
9–29
30–72
Core revolutionary member
1–5
6–11
12–23
24–80
81–203
Provincial capital
Treaty port
Self-initiated port
Figure 11.27 Geographic distribution of constitutionalists and revolutionists (1900–1911)

Source: Chang Yufa 張玉法, 清季的立憲團體 (The Constitutionalists in the Late Qing)
(Taipei, Institute of Modern History, Academia Sinica, 1971); Chang Yufa 張玉法, “清季
的革命團體” (The Revolutionists in the Late Qing), Bulletin of the Institute of Modern
History, Academia Sinica, 1975
shown in columns 1 and 6, studying overseas is insignificantly correlated with the

choice of joining the constitutionalists but positively and significantly correlated
with the decision to join the revolutionists.
We then examine the specific outcome of studying in Japan, by interacting
the overseas studies variable with Japan (columns 2 and 7). While column 2
shows that studying overseas is now significantly correlated with joining the
constitutionalists, its coefficient is negative, suggesting that these students
were perhaps not as politically oriented – perhaps due to the nature of their
professional training in subjects like science and engineering. Conversely,
those studying in Japan were significantly more likely to join a political
party – be it the constitutionalists (column 2) or the revolutionists (column 7) –
the much larger coefficient in column 7 suggests that these students were
distinctly more likely to join the revolutionists.
407

Table 11.9 Effect of overseas studies on political participation
Constitutionalist organizations Revolutionist organizations
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
Studying overseas −0.0001 −0.0142** −0.0140* −0.0058 −0.0174 0.2130*** 0.0021 −0.0260 0.0034 −0.0330
(0.0039) (0.0067) (0.0078) (0.0079) (0.0116) (0.0198) (0.0325) (0.0328) (0.0394) (0.0409)
Studying overseas*Japan 0.0175** 0.0177* 0.0143* 0.0178* 0.2630*** 0.2370*** 0.2140*** 0.2230***
(0.0087) (0.0092) (0.0083) (0.0102) (0.0390) (0.0401) (0.0432) (0.0442)
Studying overseas*Military −0.0006 −0.0062 0.0031 0.1060*** 0.0990*** 0.1190***
(0.0123) (0.0112) (0.0106) (0.0357) (0.0361) (0.0364)
Studying overseas*Graduate −0.0167** −0.0021 −0.0873 −0.0852
(0.0071) (0.0099) (0.0531) (0.0529)
Age −0.0038 0.0459***
(0.0023) (0.0032)
Age2 0.0001** −0.0002***

(0.0000) (0.0001)
Studying overseas*female −0.0055 0.0839
(0.0059) (0.0852)
Studying overseas*jinshi 0.2170** −0.1290
(0.0980) (0.1420)
Studying overseas*juren 0.1040 0.0982
(0.0636) (0.0993)
Studying overseas*xiucai 0.0242 0.0191
(0.0277) (0.0555)
Individual fixed effects Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
Year fixed effects Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
Clustered at individual level Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
Prefecture specific time trend Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
Observations 13777 13777 13777 13777 13777 13777 13777 13777 13777 13777
Number of people 599 599 599 599 599 599 599 599 599 599
Adjusted R2 0.490 0.491 0.491 0.450 0.484 0.584 0.595 0.597 0.598 0.659
Note. Standard errors in parentheses, * p < 0.1, ** p < 0.05, *** p < 0.01. In this analysis, a dummy indicating whether a student joined
a constitutionalist/revolutionist organization is used as the dependent variable.

james kai-sing kung
Next we consider the effect of major (field of specialization) on the choice

of political organization. We find that those who specialized in military
studies were significantly and positively correlated with joining the revolu-
tionists (column 8), but insignificantly correlated with joining the constitu-
tionalists (column 3). Given that after 1902 the Qing government strictly
prohibited all self-sponsored students from enrolling in just about any mili-
tary program, those specializing in military studies were almost certain to be
government officials who were expected to take up a military appointment
upon completing their training in Japan but instead rebelled against the
government. The finding intriguingly reflects the possible irony that these
Qing officials may have defected from the Manchu regime soon after acquir-
ing more military knowledge. In fact, officers of the New Army were heavily
recruited by the revolutionists.
What was the effect of graduate education on political participation?
Columns 4 and 9 were designed to examine this question. A graduate educa-
tion significantly dissuaded the students from joining the constitutionalists
(column 4), whereas its effect on joining the revolutionists is insignificant
(column 9). These findings are consistent with the economic intuition that
political participation comes at a higher opportunity cost for the more
educated.70 Finally, in columns 5 and 10 we add other control variables that
may also bear upon the choice of political organization. These include
overseas students’ age, gender, and whether they had achieved a traditional
qualification, with jinshi 進士 (national civil service examination) being the
highest level of achievement, followed by juren 舉人 (provincial civil service
examination) and xiucai 秀才 (county/prefectural civil service examination).
Given that all the variables of interest are included in columns 5 and 10, they
are de facto the full model and as such represent a summary finding. First of
all, column 5 shows that studying overseas in general is still significantly and
negatively correlated with joining the constitutionalists, but studying in Japan
is not, albeit with a marginal significance (10 percent). In contrast, of those
who chose the revolutionary path, studying in Japan in general and majoring
in military studies in particular were significant correlates, which is consistent
with the historical narrative that the New Army, many of whose officers were
trained in Japan, played a pivotal role in the revolution by virtue of their
access to weaponry (column 10). In terms of individual characteristics, not
surprisingly our findings suggest that those who were older tended to join the
70
See, for example, F. Campante and D. Chor, “Schooling, Political Participation, and the
Economy,” Review of Economics and Statistics 94.4 (2012), 841–59.
410

constitutionalists (column 5), whereas younger individuals often chose to

follow the path of the revolutionists (column 10). Perhaps there were not
many females involved, but gender has no effect for either type of organiza-
tion. Interestingly, those who had already achieved a jinshi or juren qualifica-
tion prior to studying in Japan were more inclined to join the political
organization established to preserve the status quo (i.e. the constitutional-
ists), whereas the same qualifications are insignificantly correlated with the
revolutionists (compare columns 5 and 10). Considering that those who had
achieved a jinshi or juren had invested tremendous human capital under the
old elite system, their preference to preserve the status quo was thus easy to
understand.
Conclusions
The nineteenth century is of unique importance to late imperial China. After
experiencing nearly two centuries of peace and economic prosperity, the
Qing tasted military defeat repeatedly within a short span of just two decades
at the hands of Great Britain – a country the size of a single Chinese province
but otherwise a military and industrial powerhouse. This series of defeats
sparked a new era for imperial China, one that was short-lived because the
vastly complicated development of events in the latter half of the nineteenth
century only culminated in the unwitting demise of the last imperial dynasty.
For some time now, economic historians have vehemently debated the
implications of the economic impact of the West, except that the lack of
data has precluded a serious analysis. Against this background, we compiled
a data set from a variety of data sources, the purpose of which was to provide
a preliminary assessment of the Western impact on the Chinese economy in
the last episode of the imperial dynasty. Others before us have analyzed the
possible impact of the treaty ports established by the foreign powers, but here
we used the micro-level data of foreign firms and banks as a unique starting
point and analyzed their impact on the growth of domestic firms and banks,
and confirmed their positive influence. In doing so we go beyond the “black-
box” approach of analyzing the treaty ports, which were likely to encompass
more than just the economic institutions of firms and banks. At the very least
we hope to have introduced a fresh approach to analyzing the economic
impact of the West.
Indeed, in addition to confirming a demonstration and persistent effect of
the foreign modern institutions (both firms and banks) on China’s own in
both the short and the long run, evidence further suggests that human capital
411

james kai-sing kung
also developed more rapidly in the treaty ports. Not only had the missionar-
ies played a significant role in erecting schools at virtually all levels, but they
also brought with them a new curriculum to China. Importantly, it was also
a time when the modern university came of age in China. And while all
provincial capitals had by imperial decree established one or more univer-
sities, their numbers were distinctly higher in the treaty ports and SIPs,
testifying to their economic importance. Specifically, both foreign and
domestic firms are significantly associated with the presence of engineers –
a specific form of human capital essential to industrialization.
In spite of these positive developments on the economic front, and partly
because it was weak to begin with, the Qing government could not survive
the havoc wreaked by those who went to study in Japan and came back with
the idea of using revolution as a platform for effectuating political change.
Inspired by the Western ideas disseminated through the modern education
system, the missionary presence, and the newspapers, those who went to
Japan for a brief stint might just have come to the realization that Qing China
had become both too stubborn and too feeble a regime for saving. Hence,
while it was Britain and Europe more generally that planted the seeds of
economic change in China by exposing it to the West, Japan arguably served
as the last straw on the camel’s back in bringing an end to imperial China. Of
course, as history unfolds, the end of the Qing only marked the beginning of
what promised to be another politically tumultuous episode for the country,
but that goes beyond our present scope.
Further Reading
Bai, Y., and J.K-S. Kung, Diffusing Useful Knowledge While Spreading God’s Message:
Protestantism and Economic Prosperity in China, 1840–1920, Journal of the European
Economic Association 13.4 (2015), 669–98.
Chang, Yufa 張玉法, “清末民初的外資工業” (Foreign Industries in Late Qing and Early
Republican China) Bulletin of the Institute of Modern History, Academia Sinica 1987,
129–249.
Chang, Yufa 張玉法, 近代中國工業發展史, 1860–1916 (A History of Modern Chinese
Industrial Development, 1860–1916) (Taipei, Gui guan tu shu you xian gong si, 1992).
Dernberger, R.F., “The Role of the Foreigner in China’s Economic Development, 1840–
1949,” in D.H. Perkins (ed.), China’s Modern Economy in Historical Perspective (Stanford,
Stanford University Press, 1975), pp. 19–47.
Du Xuncheng 杜恂誠, 民族資本主義與舊中國政府, 1840–1937 (Chinese Capitalism and
the Old Chinese Government, 1840–1937) (Shanghai, Shanghai shehui kexueyuan
chubanshe, 1991).
412

Esherick, J., “Reconsidering 1911: Lessons of a Sudden Revolution,” Journal of Modern

Chinese History 6.1 (2012), 1–14.
Feuerwerker, A., “The Foreign Presence in China,” in J.K. Fairbank (ed.), The Cambridge
History of China, vol. 12, Republican China, 1912–1949, part 1 (Cambridge, Cambridge
Jansen, M., “Japan and the Chinese Revolution of 1911,” in J.K. Fairbank and K.C. Liu
Cambridge University Press, 1980), pp. 339–74.
Jia, R., “The Legacies of Forced Freedom: China’s Treaty Ports,” Review of Economics and
Statistics 96.4 (2014), 596–608.
Perkins, D.H., “Government as an Obstacle to Industrialization: The Case of
Nineteenth-Century China,” Journal of Economic History 17.4 (1967), 478–92.
Reynolds, D., China, 1898–1912: The Xinzheng Revolution and Japan (Cambridge, MA,
Shen Diancheng 沈殿成, 中國人留學日本百年史 (A History of a Hundred Years of
Chinese Studying in Japan) (Liaoning, Liaoning jiaoyu chubanshe, 1997).
Zhang Hongxiang 張洪祥, 近代中國通商口岸與租界 (Treaty Ports and Concessions in
Chinese History) (Tianjin, Tianjin renmin chubanshe, 1993).
Zhang Pengyuan 張朋园, 中国民主政治的困境, 1909–1949 (The Predicament of
Democracy in China, 1909–1949) (Changchun, Jilin chu ban ji tuan you xian ze ren
gong si, 2008).
Zhang Jianqiu 張建俅, 清末自開商埠之研究 (1898–1911) (Research on the Self-Initiated
Ports of the Late Qing) (Taipei, Huamulan wenhua chubanshe, 2009).
413

12
Foreign Trade and Investment
Introduction
By the mid-eighteenth century, the Qing dynasty (1644–1911) was the
dominant power within Asia. Its political system and institutions of state
building were founded on structures inherited from previous Chinese
dynasties as well as on the social and cultural codes of interaction among
polities across Central Eurasia, East Asia, and Southeast Asia. Foreign trade
between China and other countries within and outside Asia was a calculated
matter of political strategy and economic gain. In the decades leading up to
the First Opium War of 1839 to 1842, China’s stance with respect to the Sino-
Western trade became increasingly at odds with British ambitions in Asia.
The growing tensions stemmed from abiding differences in the political
economy of not just two nations, but two empires. The overseas influence
of the British Empire took on a forceful new impetus with the British
Industrial Revolution, and, over the nineteenth century, technological
improvements in transport continued to power Western expansion in
global trade.
The Opium War of 1839 to 1842 was the turning point after which foreign,
and particularly Western, nations took greater control over not only China’s
international trade policy, but also important legal and economic institutions.
After 1842, Chinese ports that had previously been closed to Western traders
were forced open to trade and investment. In these so-called “treaty ports,”
tariffs on foreign imports into China were fixed at a low rate. Beyond trade,
consular offices and foreign courts were established in China, and foreign
nationals were exempt from the jurisdiction of Chinese law. The implications
Kyle Butts, Jacob Howard, Peiyuan Li, William Ridley, and Javier Andres Santiago
provided excellent research assistance. Support from the National Science Foundation
(Grants SES 0453040 and SES 1124426) is gratefully acknowledged.
414

of the semicolonial treaty port system for China’s long-run development

have been the subject of perennial interest.1
What has been less emphasized in the literature is that what we know
about China’s foreign trade increased dramatically during this period. Before
the 1800s, the quality of archival data on the quantities, prices, and types of
goods in China’s foreign trade, although substantial, is on the whole variable.
After the mid-nineteenth century, there was a sea change in what was
systematically recorded about China’s imports and exports. The reason was
not China’s new interest in foreign trade, nor improved statistical capacity
due to a rising level of development, but Western interests in China.
This chapter discusses China’s openness to foreign influence from 1800 to
1950 from the point of view of new sources of information on foreign trade
and investment that became available because of Western influence in China.
In particular, the China Maritime Customs (CMC) service was instrumental
in helping to revolutionize the system of foreign-trade statistics in China. The
records are the result of a complex yet consistent set of rules, and are of high
quality.2 Notably, the CMC data give more detail than is found even in
1
The impact of the treaty port system includes questions related to technology transfer,
legal institutions, state building, nation building, foreign policy, society and community,
and other topics. A number of historical overviews have been written by nineteenth-
century observers. W.F. Mayers, N.B. Dennys, and C. King, The Treaty Ports of China and
Japan: A Complete Guide to the Open Ports of Those Countries, Together with Peking, Yedo,
Hongkong and Macao (London, Trübner, 1867) is a compilation by British consular officers
about treaty ports and companies in China and Japan. Another comprehensive treat-
ment by a customs official in China is H.B. Morse, The International Relations of the
Chinese Empire, vol. 1 (London, Longmans, Green, and Company, 1910). See also C.S.
See, The Foreign Trade of China (New York, Columbia University Press, 1919);
Jiang Tingfu 蒋廷黻, 中國近代史 (Modern Chinese History) (Shanghai, Shanghai
guji chubanshe, 2001; first published Changsha, Shangwu, 1938); M. Greenberg, British
Trade and the Opening of China 1800–1842 (Cambridge, Cambridge University Press, 1951);
A. Feuerwerker, The Foreign Establishment in China in the Early Twentieth Century (Ann
Arbor, University of Michigan, 1976); Feuerwerker, “Economic Trends in the Late
Ch’ing Empire, 1870–1911,” in J.K. Fairbank and K.C. Liu (eds.), The Cambridge History
of China, vol. 11, Late Ch’ing, 1800–1911, part 2 (Cambridge, Cambridge University Press,
1980), pp. 1–69, for additional analysis. Recent historical treatments of the era can be
found in J.M. Downs, The Golden Ghetto: The American Commercial Community at Canton
and the Shaping of American China Policy, 1784–1844 (Hong Kong, Hong Kong University
Press, 2014); R. Bickers, and I. Jackson (eds.), Treaty Ports in Modern China: Law, Land, and
Power (London, Routledge, 2016); A. Reinhardt, Navigating Semi-colonialism: Shipping,
Sovereignty, and Nation-Building in China, 1860–1937 (Cambridge, MA, Harvard University
Press, 2018).
2
R. Murphey, The Outsiders: The Western Experience in India and China (Ann Arbor,
University of Michigan Press, 1977), pp. 213–14, stated, “the recorded figures probably
inflated the real import and export of goods by close to 100 percent,” erroneously
concluding that the data were unreliable due to double counting of what was traded. In
actuality, a high degree of internal consistency and accuracy allows us to reconstruct real
imports and exports at the port level.
415

modern-day international trade data as they capture re-exports, allowing

gross trade to be distinguished from net trade flows.3
In order to understand the opening of China in the nineteenth century, the
next section, “China’s Foreign Trade before 1839 and the Opium Wars,”
discusses the historical background of trade before 1840, and the motivations
and attitudes towards trade at the time. In the section headed “China’s
Foreign Trade during the Treaty Port Era (1842–1943),” we consider the
history of the CMC and the development of the organization during the
treaty port era (1842–1943). Drawing on recent work and original quantitative
research, we summarize recent findings based on the CMC data, especially
with respect to China’s foreign trade at the aggregate level and its compos-
ition by foreign country.4
Beyond the resource gains that arise from commodity trade, trade also
affects development by transferring knowledge with respect to new products,
institutional environments, and different legal systems.5 The section titled
“Foreign Direct Investment in China” briefly outlines the evolution of
foreign-owned firms in China and aggregate levels of foreign direct invest-
ment, both of which became more systematically recorded during this
period.
Since foreign trade often requires shipments from points of local produc-
tion to ports of export, a high degree of domestic market integration between
regions where goods are produced and the ports where the goods are
ultimately destined for export would have been important for the flow of
exports. Domestic markets in eighteenth-century China were populated by
many buyers and sellers and were relatively efficient. But how did foreign
3
See W. Keller, B. Li, and C.H. Shiue, “Shanghai’s Trade, China’s Growth: Continuity,
Recovery, and Change since the Opium War,” IMF Economic Review 2 (2013), 336–78, data
appendix, for a discussion of data quality of the CMC trade data.
4
In particular, see W. Keller, B. Li, and C.H. Shiue, “China’s Foreign Trade: Perspectives
from the Past 150 Years,” World Economy 6 (2011), 853–92; W. Keller, B. Li, and
C.H. Shiue, “The Evolution of Domestic Trade Flows When Foreign Trade Is
Liberalized: Evidence from the Chinese Maritime Customs Service,” in M. Aoki,
T. Kuran, and G. Roland (eds.), Institutions and Comparative Economic Development
(New York, Palgrave Macmillan, 2012); Keller, Li, and Shiue, “Shanghai’s Trade,
China’s Growth”; W. Keller, J.A. Santiago, and C.H. Shiue, “China’s Domestic Trade
during the Treaty-Port Era,” Explorations in Economic History 1 (2017), 26–43; and
W. Keller, and C.H. Shiue, “Capital Markets and Colonial Institutions in China,”
presentation at NBER Summer Institute, July 2020 (Cambridge, MA).
5
See W. Keller, M. Lampe, and C.H. Shiue, “International Transactions: Real Trade and
Factor Flows,” in S. Broadberry and K. Fukao (eds.), Cambridge Economic History of the
Modern World, vol. 1, 1700 to 1870 (Cambridge, Cambridge University Press, 2021), pp. 412–
37, for a survey on real trade and factor flows in the eighteenth and nineteenth centuries
from the viewpoint of global trade.
416

trade affect the domestic economy? The section titled “Quantifying Foreign
Influence in China during the Treaty Port Era” quantifies the effect of foreign
trade in China from two key perspectives: the size and distribution of its
welfare effects and the geographic scope of foreign influence on domestic
capital markets.
Finally, in the section titled “The Granular View: Chinese Commodity-
Level Trade,” we show how commodity-level trade statistics can be used to
obtain a more granular view of trade, and in particular of the role of the
extensive margin; that is, goods that newly enter foreign trade. As there were
important revisions to the manner in which China’s foreign-trade data were
collected throughout the period, we discuss new methodologies that can be
implemented to address changes in the definition of new goods. A number of
broader lessons are discussed in the concluding section.
China’s Foreign Trade before 1839

and the Opium Wars
To understand foreign trade in China prior to the Opium Wars, it is essential
to consider the motivations and preoccupations of Chinese and Western
traders and their respective governments in the period leading up to the
nineteenth century.6 Moreover, China didn’t have one foreign policy, but
multiple policies that depended on the region in question. In addition, these
policies changed over time.
Whether over land or sea, China’s borders were always porous to foreign
traders. Early on, in the Western Han (206 B C E –9 C E), China’s push into
Central Asia was instrumental in supporting the caravan trade on the famous
Silk Road. In the Tang–Song transition (755–1127) the Yangzi river valley
6
Secondary accounts on the nature of the conflicts of the Opium Wars: A. Waley, The
Opium War through Chinese Eyes (Palo Alto, Stanford University Press, 1963); H. Chang,
Commissioner Lin and the Opium War (Cambridge, MA, Harvard University Press, 1964);
I.C.Y. Hsü, The Rise of Modern China (Oxford, Oxford University Press, 1970); J. Beeching,
The Chinese Opium Wars (New York, Harcourt Brace Jovanovich, 1975); Su Zhiliang
苏智良, 中国毒品史 (A History of Drugs in China) (Shanghai, Shanghai renmin
chubanshe, 1997); R. Wakeman, Strangers at the Gate: Social Disorder in South China,
1839–1861 (Berkeley, University of California Press, 1997); P.W. Fay, The Opium War,
1840–1842: Barbarians in the Celestial Empire (Chapel Hill, University of North Carolina
Press, 1998); T. Brook and B.T. Wakabayashi (eds.), Opium Regimes: China, Britain, and
Japan, 1839–1952 (Berkeley, University of California Press, 2000); J. Spence, “Opium,”
in Spence, Chinese Roundabout (New York, Norton, 1992), pp. 228–58; J.G. Lutz,
Opening China: Karl F.A. Gützlaff and Sino-Western Relations, 1827–1852 (Grand Rapids
and Cambridge, William B. Eerdmans Publishing, 2008); J. Lovell, The Opium War:
Drugs, Dreams and the Making of Modern China (New York, Abrams Press, 2015).
417

emerged at the center of China’s economy, and with it more urbanization

and expanded domestic markets.7 Significantly, China’s cross-border com-
modity exchange permanently shifted away from the Silk Road, and
towards maritime trade. By the turn of the tenth century, Chinese mer-
chants were conducting trade over long distances with numerous foreign
countries, from Arabia and Persia to Java, Brunei, India, Japan, the
Korean peninsula, and the Philippine archipelago.8 Intra-Asian maritime
trade was active especially in the South China Sea. Over approximately
900–1300, triggered by the outward-looking policies of the Song and
Yuan, an increase in maritime trade occurred based on the export of
pepper, safflower, and spices from Southeast Asia to China, in exchange
for ceramics and metals.9 Over Central Eurasia, nomadic Kazakhs and
Mongolian tribes traded horses and furs in exchange for Chinese tea,
cloth and silks, and grain.10
Even as private, merchant-organized foreign trade routes were estab-
lished, from the point of view of the state, foreign trade was tied to foreign
diplomacy. The diplomatic terms were formalized in the framework of the
tribute system (chaogong tizhi 朝貢體制). Notably, the tributary system
encompassed different regional interests. Prior to the Opium Wars, the
Qing state organized the management of foreign relations into separate
offices: the Court of Colonial Affairs (lifanyuan 理藩院), which dealt with
Inner Asian regions, including Mongolia, Russia, and Tibet; and the Board
of Rites (libu 禮部), which handled court religious ceremonies but also the
relations with sinified tributaries. The tributaries retained their own
sovereignty, but their rulers accepted the emperor of China as the
nominal political and cultural hegemon, and they were rewarded for their
7
H. Miyakawa, “The Naitō Hypothesis and Its Effects on Japanese Studies of China,” Far
Eastern Quarterly 4 (1955), 533–52; see also chapter in R. von Glahn, The Economic History
of China: From Antiquity to the Nineteenth Century (Cambridge, Cambridge University
Press, 2016), pp. 208–17.
8
B.K.L. So and J. Su, Prosperity, Region, and Institutions in Maritime China: The South
Fukien Pattern (Cambridge, MA, Harvard University Asia Center, 2000), p. 35.
9
See G. Wade, “An Early Age of Commerce in Southeast Asia, 900–1300 C E,” Journal of
Southeast Asian Studies 2 (2009), 221–65; A. Reid, “‘An Age of Commerce’ in Southeast
Asian History,” Modern Asian Studies 1 (1990), 1–30; and Reid, “The
Seventeenth-Century Crisis in Southeast Asia,” Modern Asian Studies 4 (1990), 639–59.
Reid, “The Seventeenth-Century Crisis,” argues that these trades continued until the
mid-seventeenth century, when the expansion of the Dutch East India Company into
the region effectively ended the maritime trade boom.
10
See P.C. Perdue, China Marches West: The Qing Conquest of Central Eurasia (Cambridge,
MA, Harvard University Press, 2005), pp. 400–2, 575, for accounts related to the private
trade and co-operative official exchange of Central Eurasia.
418

loyalty.11 The Qing also recognized arms-length relationships with non-

tributary states, depending on the circumstances.12
Tributary trade – in which states ritually presented China with gifts, and
often received gifts of even greater value in return – was used to maintain and
to expand relationships with neighboring polities.13 According to Fairbank
and Fairbank and Têng, the tribute system governed the entire foreign-
relations world order of China’s empire up to the nineteenth century, and
this narrow mind-set precluded the possibility of free trade and nation-state
diplomacy based on terms of mutual equality.14
Recent research by numerous scholars has re-examined these interpret-
ations in major ways. In particular, the importance of tributary trade has been
shown to have weakened significantly over time, even if the tributary system
as such provided a way for China to promote diplomatic relations. The Ming
emperors were the last to prohibit maritime trade in favor of tributary trade,
and even they didn’t succeed fully in enforcing the ban. Moreover, compared
to Ming emperors, Qing rulers were generally more relaxed about market-
based exchange. Although Qing emperors also used maritime trade bans
(haijin 海禁), the most stringent of the bans were imposed for political ends
rather than in pursuit of any autarkic ideals. Between 1656 and 1684, the Qing
imposed a maritime ban in order to subdue the Zheng empire – a powerful
merchant organization that was loyal to the former rulers of the Ming
11
Korea, for example, was considered a loyal tributary state, and thus Ming troops were
sent in the 1590s to help Korea fight off the Japanese. Other close tributaries included
Vietnam, Siam, Laos, Burma, Cambodia, Liuqiu (Rykukyu), Luzon and Java, and the
Central Asian peoples such as the Kazakhs, Kirghiz, and Badakhshanis. In 1754, Qianlong
refers to Java as “already within the compass of Our enlightened government.”
M. Elliot, Qianlong: Son of Heaven, Man of the World (New York, Pearson, 2009), p. 126.
12
For example, the Portuguese did not accept the emperor of China as the nominal
authority but were, nonetheless, allowed to settle in Macau, establishing a private trade
center there in the mid-sixteenth century.
13
Stability on the frontiers in Central Asia could have meant a net gain from the
perspective of China’s rulers if that reduced the chance of military encounters; see
Perdue, China Marches West, pp. 402–3.
14
J.K. Fairbank, “Tributary Trade and China’s Relations with the West,” Far Eastern
Quarterly 1.2 (1942), 129–49; Fairbank, Trade and Diplomacy on the China Coast: The
Opening of the Treaty Ports, 1842–1854 (Cambridge, MA, Harvard University Press,
1969); J.K. Fairbank, and S.Y. Têng, “On the Ch’ing Tributary System,” Harvard
Journal of Asiatic Studies 2 (1941), p. 140, citing Jiang Tingfu: in the 11th and 12th centuries,
“the neo-Confucian philosophy worked, which began to dominate China, worked out
a dogma in regard to international relations, to hold sway in China right to the middle
of the nineteenth century . . . That dogma asserts that national security could only be
found in isolation and stipulates that whoever wished to enter into relations with China
must do so as China’s vassal, acknowledging the supremacy of the Chinese emperor
and obeying his commands, thus ruling out all possibility of international intercourse
on terms of equality.”
419

dynasty.15 Immediately after the Zheng empire was vanquished in 1684, the
Qing Kangxi Emperor opened all coastal ports to private trade and estab-
lished customs stations to collect taxes.16
Because of the precedents set up during Kangxi’s reign (1661–1722), which
moved the Qing further away from the traditional tributary system, private
trade frequently overshadowed the importance of tributary trade.17 The Sino-
Western trade, which first emerged out of a demand for Chinese luxury
goods – tea, silk, porcelain, furniture, art, and lacquers – continued to operate
over the eighteenth century. In addition to maritime exchange among traders
in the South China Sea, Chinese merchants sailed to Nagasaki regularly to
trade with Japanese merchants and entrepreneurs, and private Japanese
traders plied the Chinese coastal trade.18 Moreover, maritime exchange
among Asian traders in the South China Sea may have extended beyond
trade to organized private enterprises.19
Evidence of the pervasiveness of these global trading networks can be seen
in the movement of precious metals as silver from overseas surged into China
in exchange for Chinese goods.20 In the seventeenth century, China imported
around 115 tons of silver annually, approximately half of which came from
mines in Japan, and the other half from the Americas.21 Whether this exten-
sive global trade was the cause or the outcome of the divergence in living
15
In 1661, Zheng Chenggong (also known by Koxinga), successfully laid siege to the Dutch
fort, Zeelandia, located in Taiwan, and, in open defiance of the Qing, established
a separate state named Ming Eastern Capital (dongdu Mingjing 東都明京). The Zheng
family was one of the most successful merchant organizations of the period, with
annual profits from maritime trade at least one-third that of the Dutch VOC; see
Appendix 3 in X. Hang, Conflict and Commerce in Maritime East Asia: The Zheng Family
and the Shaping of the Modern World, c. 1620–1720 (Cambridge, Cambridge University
Press, 2015).
16
W. Rowe, China’s Last Empire: The Great Qing (Cambridge, MA, The Belknap Press of
Harvard University Press, 2009), pp. 136.
17
Rowe, China’s Last Empire, p. 136, concludes that tributary trade was “nil”; whereas
T. Hamashita, China, East Asia, and the Global Economy: Regional and Historical
Perspectives (London and New York, Routledge, 2008), Chapter 2, depicts tributary
trade as being intertwined with commercial trade.
18
Since the 1970s, historians have challenged the “national seclusion” view of Japan,
documenting the continuous arrival of foreign trading vessels in Japan from the
seventeenth through the nineteenth centuries. See Figure 1.2 in H. Peng, Trade
Relations between Qing China and Tokugawa Japan, 1685–1859 (Singapore, Springer, 2019);
Hamashita, China, East Asia, and the Global Economy, presents a regional maritime
history based on networks of trade that cut across national borders
19
Hamashita, China, East Asia, and the Global Economy, Chapters 3–6.
20
See Hamashita, China, East Asia, and the Global Economy, Chapter 4. G. Zhao, The Qing
Opening to the Ocean: Chinese Maritime Policies, 1684–1757 (Honolulu, University of Hawaii
Press, 2013), synthesizes additional evidence on China’s role in early globalization.
21
See von Glahn, The Economic History of China, p. 309; R. von Glahn, “Foreign Silver
Coins in the Market Culture of Nineteenth Century China,” International Journal of
420

standards between China and Europe in the twentieth century has been the
subject of long-standing debates.22 What seems clear, however, is that accu-
mulating qualitative and quantitative evidence largely overturns the perspec-
tives of an earlier literature that tended to paint China as a closed and
isolationist state before Western nations “opened” its markets.
Furthermore, not only does the recent evidence on international silver
flows point to considerable foreign-trade activity, but it also appears that the
Qing, and the earlier Ming state, paid attention to the advantages of empire
building.23 The Qianlong Emperor (1711–1799) achieved successes on this
front. At its maximal geographical extent, the total territories of the Qing
dynasty were about double the extent governed by the Ming. Around
the year 1780, the Qing state was the second-largest Chinese empire, sur-
passed only by the Mongol Empire of Kublai Khan. Neither could it be
claimed that the Chinese state had no challengers, as there were numerous
and near-continuous conflicts both from within and against nearby states.24
Instead, the Qing state had no challengers of equal stature in Asia because it
had successfully eliminated the threats from neighboring regions, as well as
the domestic rebellions from within. The military campaigns by the Qing
state over Central Asia, for example, nearly obliterated the Dzungar
Asian Studies 1 (2007), 51–78, documents that foreign coins could be found circulating
alongside domestic currencies, or as the dominating means of payment.
22
Offering a contrasting point of view to Wallerstein’s treatment of Asia as a semi-
peripheral area relative to the European core before the mid-nineteenth century, A.
G. Frank, ReOrient: Global Economy in the Asian Age (Berkeley, University of California
Press, 1998), argues that a global economy in 1400–1900 centered on Asia. E. Jones, The
European Miracle: Environments, Economies and Geopolitics in the History of Europe and Asia
(Cambridge, Cambridge University Press, 2003); and K. Pomeranz, The Great Divergence:
China, Europe, and the Making of the Modern World Economy (Princeton, Princeton
University Press, 2000), offer competing explanations and points of emphasis, with
the latter arguing that it was the relaxing of ecological constraints in Europe brought
about by access to New World resources that laid the foundations for the Great
Divergence between China and Europe. Also see R. Findlay and K. O’Rourke, Power
and Plenty: Trade, War, and the World Economy in the Second Millennium (Princeton,
Princeton University Press, 2007), who focus on the role of empire and the global
connections established through trade.
23
G. Wade, “The Zheng He Voyages: A Reassessment,” Journal of the Malaysian Branch of
the Royal Asiatic Society 1 (2005), 37–58, links the Zheng He voyages to aggressive
attempts to dominate trade routes in the Middle East and East Asia, and suggests that
the voyages constitute maritime proto-colonialism.
24
Referencing what he felt to be his top ten military achievements during his reign,
Qianlong wrote, “The ten instances of military merit include the two pacifications of
the Dzungars, the quelling of the Muslim tribes, the two annihilations of the Jinchuan
[rebels], the restoring of peace to Taiwan, and the subjugations of Burma and Vietnam;
adding the recent twin capitulations of the Gurkhas makes ten in all. Why is there any
need to include those three trivial rebellions in the inner provinces?” Elliot,
Qianlong, p. 89.
421

(Zunghar) population of 600,000.25 A combination of diplomacy and military

aggression had thus made it possible for the Qing to achieve a high level of
political, economic, and civilizational hegemony.
The stance of the Qing emperors towards the Sino-Western trade was not
inconsistent with their overall strategy of statecraft. Within China, three key
ports directed the trade coming from the West: Macau, Canton (now
Guangzhou), and Hong Kong. Starting from the year 1684, and for more than
150 years after, the Qing managed Sino-Western trade through a merchant guild,
or “Cohong” (公行). These merchants were appointed by the state to manage
the European trade. The official superintendent of maritime customs, known as
the “hoppo,” collected duties on foreign trade through the Cohong merchants.
These revenues were sent directly to the imperial household. After 1757, and until
1842, all Western trade had to be conducted from Canton. The Canton system
allowed British, Dutch, French, Austrian, Swedish, Spanish, American, and other
traders to carry out trades with a member of the Hong merchants. In practice,
this meant that traders were required to live in special quarters, in buildings
called “factories.” The factories were located outside the city along Canton
Harbor and included space for warehouses and offices.
Although Qing emperors personally gained revenues from the Sino-Western
trade, they did not seek to expand trade or diplomatic relations beyond the
Canton system. This may seem irrational, but it might be remembered that the
early and mid-Qing rulers incorporated foreign trade as a lever of power within
a very different set of institutional constraints. Both national security interests
and economic gain entered into Qing calculations of how to handle Western
traders. Rulers were concerned about the encroachment of foreigners and their
activities on domestic interests and attempted to impose regulations. Thus,
appointed Cohong merchants were responsible not only for the payment of
transit dues of foreign traders, but also for the good behavior of foreign crews, in
addition to managing the actual trade. Furthermore, Qing regulations limited
foreign merchants’ personal or diplomatic channels of communication with
Qing government officials. Outside the four-month trading season, foreigners
had to relocate away from Canton and to Macau.
From the British trader’s perspective, the Canton system was unsatisfactory.
Western traders who came to China often represented companies or syndicates
that were funded by wealthy landowners and entrepreneurs in an industrializing
Europe, or from the United States. They wanted more interaction and more
25
About 90 percent of the Dzungars were killed, died, or were taken captive; Elliot,
Qianlong, pp. 94. Qianlong targeted young and strong men for massacre in order to
destroy the Dzungars as a people; Perdue, China Marches West, p. 283.
422

representation of their interests in China. These organizations were in some

cases powerful enough to influence government politics at home.26 One prom-
inent example of the close connection between merchants’ interests and their
political activity is the British East India Company (BEIC), which operated from
1600 until 1834. The BEIC was a trading monopoly that competed with the
nationally chartered Dutch East India Company (Vereenigde Oost-Indische
Compagnie, VOC) in Asia. The expansion of foreign trade into Asia was
a central aim of the British government and the BEIC.
An incident that compellingly illustrates the different motivations of
Britain and China was Lord Macartney’s mission to China. In 1792,
Macartney was commissioned by Henry Dundas (who was president of the
board of the BEIC, and a member of Britain’s Home Ministry) to speak to the
Qing court on behalf of British interests. Specifically, Macartney was
instructed to relay to China’s emperor “the mutual benefit to be derived
from trade between the two Nations.”27 He was also to bring some products
in order to “excite at Peking a taste for many articles of English workmanship
hitherto unknown there . . . [and] turn the balance of the China trade
considerably in favour of Great Britain.”28 As imported opium was already
a growing point of discord, Macartney was also instructed that should the
subject of opium come up, and “if it should be made a positive requisition or
any article of any proposed commercial treaty, that none of that drug should
be sent by us to China, you must accede to it, rather than risk any essential
benefit by contending for a liberty in this respect.”29
On September 14, 1793, Macartney arrived in the court of Qianlong, who
was over eighty years old at the time, with numerous gifts, most of which
were European luxury items, such as German planetariums and clocks by
Vulliamy. From Qianlong’s written reply to King George I I I, we know that
Macartney successfully transmitted to the Qing court Britain’s requests,
26
R. Brenner, Merchants and Revolution: Commercial Change, Political Conflict, and London’s
Overseas Traders, 1550–1653 (Princeton, Princeton University Press, 2003), examines the
relationship between English commerce and the political activities of overseas traders
in the seventeenth century.
27
J. Chen, P. Cheng, M. Lestz, and J. Spence, The Search for Modern China: A Documentary
Collection, 3rd ed. (New York, W.W. Norton & Company, 2013), “Lord Macartney’s
Commission from Henry Dundas, 1792,” p. 80.
28
As quoted from H.B. Morse, The Chronicles of the East India Company: Trading to China
1635–1834, 5 vols. (Oxford, The Clarendon Press, 1926–1929), vol. 2, p. 215, as cited in
M. Berg, “Britain, Industry and Perceptions of China: Matthew Boulton, ‘Useful
Knowledge’ and the Macartney Embassy to China,” Journal of Global History 2 (2006),
269–88.
29
Chen et al., The Search for Modern China, “Lord Macartney’s Commission from Henry
Dundas, 1792,” p. 81.
423

which Qianlong understood well. They were: (1) to have more ports in China to
be open for purposes of trade, (2) to be able to establish a repository at the capital
in Beijing, (3) to have an island where merchants can reside and goods can be
warehoused, (4) to be able to have a place inside the city of Canton where foreign
merchants may reside, (5) to have reduced duties on merchandise, (6) to have
reduced tariffs on ships, and (7) to gain the full liberty to disseminate European
religions to Chinese subjects.30 Judging from the tone of Qianlong’s long reply to
Macartney’s entreaties, he did not view Britain as more than a presumptuous far-
away state that had overstepped proper boundaries of civilized relations.
A common interpretation of Qianlong’s dismissal of Macartney’s requests
is that it encapsulates “the Chinese policy of superior indifference to Western
things.”31 Yet a closer reading of the events suggests that, far from being
indifferent, Qianlong was an astute collector of foreign objects, as he already
possessed in his residence exactly the kinds of mechanical devices that
Macartney had brought with him on his journey to China.32 Apparently,
unbeknown to Macartney, similar objects had arrived in China through
existing channels of the Canton trade. Macartney later wrote in his journal
that on his tour of Qianlong’s pavilion, he saw
stupendous vases of jasper and agate; with the finest porcelain and japan,[33]
and with every kind of European toys and sing-songs; with spheres, orreries,
clocks and musical automatons of such exquisite workmanship, and in such
profusion, that our presents must shrink from the comparison and hide their
diminished heads.34
The Macartney mission did not fundamentally alter the way Sino-Western
trade was conducted. Corporate interests in China had neither the explicit
military backing of the state, nor, for the most part, the means to wage war
against the state. Instead, Chinese merchants typically tried to acquire official
or semiofficial roles within the state. Thus, many well-off merchants in China
sought to gain greater influence in the government by purchasing degrees or
30
Chen et al., The Search for Modern China, pp. 90–3.
31
D.S. Landes, “Why Europe and the West? Why Not China?”, Journal of Economic
Perspectives 2 (2006), 18.
32
“On the court’s side, questions concerning Macartney’s pronouncements about the
British gifts still lingered. As if to address this issue directly, the embassy was taken to
buildings filled with intricate European clocks and mechanical devices . . . The point
being made that the things Macartney had brought were in no way unique to his king’s
domain.” From J.L. Hevia, Cherishing Men from Afar: Qing Guest Ritual and the Macartney
Embassy of 1793 (Durham, NC, Duke University Press, 1995), p. 179.
33
Objects made with a dark lacquer.
34
Cited in Hevia, Cherishing Men from Afar, p. 176. See also Berg, Britain, Industry and
Perceptions of China, pp. 269–88.
424

investing in the education of their sons so that someone in the clan could gain
the ear of an official, or, better, be anointed into officialdom themselves.35
British and other Western traders felt slighted as they sought to engage
China in the new diplomatic language of equal nation-states and negotiated
benefits, while Qing emperors still considered the arrival of Western traders
on China’s shores to be nothing more significant than the understandable
desire of foreigners to partake of the benefits of China’s civilization and the
blessings of the emperor. British, American, and European merchants chafed
under the restrictions of Qing policies, but not so much as to be willing to
give up their share of the profits from the Canton trade.
The differences in the political economy in the two empires, and how each
conceptualized domestic and foreign relations, seem to be especially striking
in the years leading up to the Opium War.36 Because of the successful spread
of the British Industrial Revolution, our modern system of foreign trade and
diplomacy aligns with that of the European system, but it might be remem-
bered that Europe had itself only not that long ago – sometime between the
Treaty of Westphalia (1646–1648), the revolutionary decade of 1789 to 1799 in
France, and the Congress of Vienna (1814–1815) – settled on the nation-state
framework as a formal system of international relations among sovereign
states.37 Before that, customs, personal relationships, and family alliances
through marriage tended to play a larger role in foreign diplomatic negoti-
ations in Europe too.
From the start of the Sino-Western trade, European traders sought access
to Chinese markets not only to buy Chinese goods, but also to sell their own
wares in what they imagined to be an immense market. The problem,
however, was not simply market access, but a lack of products that ordinary
consumers in China could afford. As late as the 1830s, traders of one of the
35
These official positions were frequently the most rewarding from a socioeconomic
point of view; see C.H. Shiue, “Human Capital and Fertility in Chinese Clans,” Journal of
Economic Growth 4 (2017), 351–96.
36
The similarities and differences in the economy and political system of China and
Europe are further analyzed in detail in R.B. Wong, China Transformed: Historical Change
and the Limits of European Experience (Ithaca, NY, Cornell University Press, 1997);
Pomeranz, The Great Divergence; and R.B. Wong and J. Rosenthal, Before and beyond
Divergence: The Politics of Economic Change in China and Europe (Cambridge, MA, Harvard
University Press, 2011). On the influence of cultural change in China and Europe, see
J. Mokyr, A Culture of Growth: The Origins of the Modern Economy (Princeton, Princeton
37
For contrasts of Europe’s “Westphalian system” with East Asia’s formal hierarchy in
international relations, see D. Kang, East Asia before the West: Five Centuries of Trade and
Tribute (New York, Columbia University Press, 2010); and Hamashita, China, East Asia,
and the Global Economy.
425

dominating trading houses at the time – Jardine, Matheson & Co. – reported
that the Chinese native nankeen cotton cloth (named for Nanjing) was
superior in quality and cost compared to Manchester cotton goods.38 Thus,
even though cotton was one of the core industries that was revolutionized by
the British Industrial Revolution, it would still be some time before machine-
produced textiles could compete with the low costs of labor production in
China. By contrast, as industrialization spread from Britain to Northwest
Europe and its offshoots, wealthier classes in urban centers may have been
better able to afford foreign imports and Chinese luxury goods.39 Not only
did tea drinking become fashionable, but consumers were fascinated with
chinoiserie and other Chinese decorative goods. Chinese craftsmen and
manufacturers, for their part, also eagerly created custom-designed products
for foreign markets.
The good that tipped the trade balance was opium.40 In 1773,
140,000 pounds of opium were imported into China from India; by the
early 1820s, imports had grown tenfold.41 Opium was illegal and yet openly
smuggled, bought, and sold on the watch of Qing merchants and officials
alike. In the 1830s, 20 to 30 percent of government officials consumed opium,
and the Daoguang Emperor (r. 1820–1850) was himself an addict.42 At first
considered a foreign luxury good and a symbol of privilege and hospitality,
opium became widely used throughout Chinese society. Early debates in the
Qing court about the appropriate response to opium imports considered the
pros and cons of legalization and taxation of the drug, as opposed to strict
prohibition.43 Eventually, opium imports became a scapegoat for the failures
38
Greenberg, British Trade and the Opening of China, p. 2.
39
Average wages of urban residents in major cities of Western Europe like London were
likely trending higher than those of their counterparts in Beijing or even Suzhou.
R. Allen, J.P. Bassino, D. Ma, C. Moll-Murata, and J.L. van Zanden, “Wages, Prices,
and Living Standards in China, 1738–1925: In Comparison with Europe, Japan, and
India,” Asia in the Great Divergence, special issue of Economic History Review 64.s1 (2011),
3–38. Additional research is needed before we can be sure, but to the extent that average
wages are correlated with incomes of the wealthier classes within each region, the
trends may be similar.
40
For perspectives on the history of the consumption of opium, see F. Dikötter,
L. Laamann, and Z. Xun, “Narcotic Culture: A Social History of Drug Consumption
in China,” British Journal of Criminology 2 (2002), 317–36; also see Su, 中国毒品史.
41
Spence, Chinese Roundabout, pp. 233–5.
42
P.C. Perdue, “The First Opium War: The Anglo-Chinese War of 1839–1842,” MIT
Visualizing Cultures, 2011, https://visualizingcultures.mit.edu/opium_wars_01/ow1_
essay01.html.
43
See the arguments from proponents of legalization and taxation, as well as prohibition,
in J. Slade, Narrative of the Late Proceedings and Events in China ([Canton], Canton Register
Press, 1839), pp. 1–140.
426

of the government, social unrest, and the economic decline that character-
ized the last third of the Qing Dynasty.44
In 1839, Qing commissioner Lin Zexu was sent by the Daoguang Emperor
to end the opium problem through prohibition. Lin took the moral high
ground on the matter of opium, eventually destroying a large cargo of opium
when his entreaties to cease the opium trade were ignored.45 In response,
British traders declared property damage and quickly resorted to military
action. It was the new technology of the steam engine outfitted on British
boats, however, that determined the outcome of the Opium Wars. Finally,
European grievances about the restrictive conditions of the Canton system
could be forcefully expressed in the form of the steamships that could deftly
steer into the shallow harbor waters of Canton. British military forces took
Canton, moved up the coast and along the Yangzi river, captured Shanghai,
and eventually reached the Grand Canal, in effect threatening Peking itself.46
China quickly surrendered, agreeing to sign the Treaty of Nanking
(Nanjing) (1842), which stipulated that an indemnity had to be paid as
compensation to Britain; in addition, Hong Kong was ceded to Britain.
Beyond the initial four treaty ports (Xiamen, Fuzhou, Ningbo, and
Shanghai), additional ports were later opened to foreign trade. Trade duties
were limited to 5 percent ad valorem or less on all goods. Moreover, foreign
nationals were given the right to reside and own property in designated
treaty ports. In addition, foreigners in China would be subject to the legal
jurisdiction of their own country rather than to Chinese laws.
The issue of the legality of opium in China was hardly worth even
a mention in the Treaty of Nanking. Indeed, the coup d’état was not about
making the opium trade legal in China. The real prize was about market
access and the entry of foreign businesses into China’s economy. This
44
More recently, Z. Wang, Never Forget National Humiliation: Historical Memory in Chinese
Politics and Foreign Relations (New York, Columbia University Press, 2012), pp. 96–8,
argues that after 1991, historical revisionism shifted from class struggle to struggle with
outside forces.
45
Lin’s communication to Charles Elliot, the British superintendent of trade, in March of
1839: “While our Celestial Court has in humble submission to it ten thousand (i.e. all)
regions, and the heaven-like goodness of the great Emperor overshadows all, the nation
aforesaid (Britain) and the Americans have, by their trade at Canton during many years,
enjoyed, of all those in subjection, the largest measure of favors,” “but that they have
brought opium – that pervading poison – to this land: thus profiting themselves by the
injury of others.” Great Britain Foreign Office, Correspondence Relating to China:
Presented to Both Houses of Parliament, by Command of Her Majesty (London, T.R.
Harrison, 1840), pp. 268–9.
46
Romanization of names of locations and treaty ports in this chapter will follow that
used by the CMC in the nineteenth and twentieth centuries. Pinyin will be used in other
cases.
427

sentiment was voiced by the British plenipotentiary, Sir Henry Pottinger,

who announced after Britain’s victory over China in the First Opium War
(1840–1842) that China’s potential for trade was so vast “that all the mills of
Lancashire could not make stocking stuff sufficient for one of its provinces.”47
His overly ambitious forecast was slow to come to fruition, however.
Initially, the Qing court did not see the Treaty of Nanking as the resolution
of a grand showdown, but rather as a small concession made in order to
smooth over a conflict that could eventually be redressed in China’s favor.
For Britain, however, signed treaties between nations, compelled or not,
were all that mattered in the new era defined by contractual agreements
between nation-states. In 1844, the United States and France concluded
similar treaties with China, the Treaty of Wanghia and the Treaty of
Whampoa respectively. China’s reluctance to enforce the terms of the earlier
treaties led to the Second Opium War (1856–1860), and further treaty ratifica-
tions and most-favored-nation clauses that allowed all foreign powers oper-
ating in China to seek the same concessions. The various treaties thus
gradually opened China to international markets and ushered in an age of
ever-increasing commerce with the rest of the world until 1949, when Mao
Zedong came to power.
China’s Foreign Trade during the Treaty Port Era

(1842–1943)
The Treaty of Nanking (1842) put into motion a significant transformation of
China’s trade environment. It was followed by the Treaty of Tientsin
(Tianjin) (1858), which opened yet more treaty ports, and it also laid the
foundation for foreign-trade policies in China in essential ways. The most
important of the clauses of the Treaty of Tientsin established a system under
which trade duties would be collected under a consistent system across treaty
ports. In principle, this implied that foreign goods would be taxed only once
upon entry into China and thereafter be exempt from further duties even if
the goods were transported further inland. Among other rights granted to
foreign traders and residents, foreign vessels were permitted on the Yangzi
river and foreign merchants could also employ Chinese ships to carry their
goods. The British were officially permitted, in 1848, to establish a foreign
settlement in Shanghai. During the treaty port era, foreigners came to have
47
China Maritime Customs (CMC), Decennial Reports, Fifth Issue (1922–1931), vol. 1
(Shanghai, Statistical Department of the Inspectorate General of Customs, 2001), p. 39.
428

a much more active role in the wider economy than in earlier times, which
included the ownership of hundreds of firms and businesses, including banks
and shipyards.
The China Maritime Customs and Its Records

While the Treaty of Nanking did away with several existing elements of
China’s foreign-trade system, the Chinese customs authority initially retained
its oversight of the processing of foreign trade. However, the erosion of the
central government’s authority after the Opium War and the government’s
lasting preoccupation with the suppression of domestic uprisings (in particu-
lar the Taiping Rebellion of 1844 to 1860) meant that foreign-trade revenue
collection fell primarily into the hands of provincial and local authorities.
These local officials were ill-equipped to handle the larger volume of trade
coming in, and foreign trade was not subject to a consistent set of rules.
Rather, the payment of trade taxes was a matter of bargaining power, and the
system was rife with corruption.48
The China Maritime Customs (CMC) was founded in 1854 by the foreign
consuls in Shanghai to collect maritime-trade taxes that had been going
unpaid due to the inability of Chinese officials to collect them during the
Taiping Rebellion. Although the CMC was nominally under the jurisdiction
of China’s Foreign Office (the zongli yamen 總理衙門), which was newly
established in 1861, in practice it operated under the management of foreign
powers. In the beginning, its staff were mostly British. Later, nationals from
other Western countries joined. The top CMC position and director of its
operations was the inspector general, who worked side by side with his
Chinese counterpart, called the superintendent of customs, who oversaw
the collection of trade taxes from the so-called native trade; that is, from
Chinese-owned junks.
Early opposition to the CMC arose largely from foreign consuls who
feared that the CMC would usurp some of their powers. Foreign merchants
were also initially opposed to the CMC because now they had to deal with
customs formalities that before had been left in their entirety to Chinese
middlemen and clerks. Within only a couple of years, however, foreign
traders and entrepreneurs had come to prefer the consistent and predictable
48
Tax collection was poor even in major ports such as Shanghai. The British consul of
Shanghai estimated in one year that the loss of tariff revenue in Shanghai was at least
25 percent, and complained that “two or three sleepy menials at $5 or $6 a month” were
the sole means existing for the collection of duty, with which he was bound by the
Treaty of Nanking to co-operate. CMC, Decennial Reports, Fifth Issue, vol. 1, p. 81.
429

customs treatment by the new CMC system, and over time foreign mer-
chants were generally in favor of the foreign inspectorship system – this
smoothed the frictions between consuls and CMC officials.
Although the Chinese central government resented the loss of sovereignty
that came with the Treaty of Nanking and customs operations by the CMC,
the introduction of the CMC also substantially increased the net tariff
revenues it received.49 Local Chinese government officials likely experienced
a net decline in benefits as the CMC reduced their ability to withhold
revenues from the central government and strike deals for personal enrich-
ment. Moreover, smugglers, pirates, and adventurers saw their prospects of
gain diminished with the arrival of the CMC, especially because over time the
CMC extended its responsibilities to include antismuggling operations. Later,
the CMC also expanded its involvement into postal administration, coastal
policing, harbor and waterway management, and weather reporting.
From the point of view of the treaty powers, the establishment of the CMC
not only broadened their political influence in China but also ensured that
China would have the means to pay the indemnities imposed on it after the
First and Second Opium Wars. The information generated by this system
was so credible that China was able to put the tariff revenue down as
collateral against which it could borrow from abroad at relatively low rates
of interest. A further motive, arguably the most important, was that the
treaty powers wanted to support the expansion of commercial exchange
between China and their own economies, which necessitated a more open
and consistent Chinese system.
The CMC’s jurisdiction extended to “foreign-type” vessels, in particular
steamships, whether owned by foreigners or by Chinese, and to junks
chartered by foreigners. In addition to calculating tax revenues that were
due, the CMC was responsible for the examination of cargo, the prevention
of smuggling, and the assessment of treaty tariffs on exports, imports, and
coastal trade. The nominal tariff was fixed to yield a rate of approximately
5 percent ad valorem; however, over time the effective rate was often lower,
around 3 percent or less, due to price increases.
The number of treaty ports and customs houses expanded until there were
over forty by the year 1907. Map 12.1 displays the locations of CMC stations
49
Robert Hart, the inspector general of the CMC from 1863 to 1911 and one of the most
influential individuals in the history of the service, estimated that under the native
customs system the costs of tariff collection were larger than the customs system’s
revenues, while under the CMC at Shanghai these costs were only around 2 percent of
the revenues; CMC, Decennial Reports, Fifth Issue, vol. 1, p. 81.
430

Aigun
Aigun
Aigun
Manchouli
Manchouli
Manchouli Lahasusu
Lahasusu
Lahasusu
Sansing
Sansing
Sansing
Harbin
Harbin
Harbin Suifenho
Suifenho
Suifenho
Kirin
Kirin
Kirin
Hunchun
Hunchun
Hunchun
Newchwang
Newchwang Lungchingtsun
Newchwang Lungchingtsun
Lungchingtsun
Yellow
Yellow River
Yellow River
River
Antung
Antung
Antung
Chinwangtao
Chinwangtao
Chinwangtao Tatungkow
Tatungkow
Tatungkow
Peking
Peking
Peking Dairen
Dairen
Dairen
Tientsin
Tientsin Lungkow
Tientsin Lungkow
Lungkow Chefoo
Chefoo
Chefoo
Weihsien
Weihsien
Weihsien
Kiachow
Kiachow
Kiachow
Chinkiang
Chinkiang
Chinkiang
Nanking
Nanking
Nanking
Wanhsien
Wanhsien
Wanhsien Wuhu
Wuhu
Wuhu Soochow
Soochow
Soochow
Shasi
Shasi
Shasi Shanghai
Shanghai
Shanghai
Hankow
Hankow
Hankow Hangchow
Hangchow
Hangchow
Yangtze River Ichang
Ichang
Ichang Kiukiang
Kiukiang
Yangtze
Yangtze River
River Kiukiang Ningpo
Ningpo
Ningpo
Chungking
Chungking
Chungking
Yochow
Yochow
Yochow Wenchow
Wenchow
Wenchow
Type Changsha
Changsha
Changsha Santuao
Santuao
Santuao
Both Tengyueh Foochow
Foochow
Foochow
Samshui
Samshui
Samshui AmoyAmoy
Amoy
CMC Mengtsz
Mengtsz
Mengtsz Wuchow
Wuchow
Wuchow Canton
Canton
Canton
Swatow
Swatow
TP Nanning
Nanning Swatow
Szemao Kowloon
Lungchow Pakhoi Lappa Kongmoon
Kiungchow
Kiungchow
Kiungchow
Map 12.1 China Maritime Customs stations and treaty ports

Source: CMC customs stations given in T. Lyons, China Maritime Customs and China’s
Trade Statistics, 1859–1948 (Trumansburg, NY, Willow Creek Press, 1973); and treaty ports
(TP) with entries in the Returns of Trade of 1910
that were established over the organization’s existence.50 The CMC did not
establish customs stations in all ports, but focused on the ports that were
important for foreign trade. Generally, the more important ports were
opened relatively early, which means that even in the 1860s the CMC covered
80 to 90 percent of all foreign trade. With the opening of the CMC customs
station in Kowloon (which was located opposite Hong Kong Island and was
thus important for the trade with Hong Kong) in the year 1886, virtually all of
China’s foreign trade was covered.
The CMC’s statistical records on trade are contained primarily in the
Returns of Trade, with additional statistics and more descriptive accounts
available in the Trade Reports and Special Collections. From the start of the
CMC in 1859 to its end in 1948, records on trade were entered at least at
annual intervals. There is some daily information (e.g. for Shanghai) during
50
The map gives the list of treaty ports in the CMC Returns of Trade of 1910. There were
ninety-two treaty ports by 1917, but many were self-initiated ports.
431

some periods. Previous studies have given overviews of the institutional

features of the CMC organization and have provided broad outlines of
some of the contents of the CMC trade data.51
One of the most notable aspects of the CMC’s trade records is that they
capture trade flows that are difficult or impossible to obtain even in modern-
day advanced economies. Most data on international trade consider the
country as the unit of observation. By contrast, the CMC recorded informa-
tion not only by country, but by port. That is, the focus of observation is not
country-to-country trade, but rather country-to-port-of-entry trade, where
the port is treated independently as if it were a country. In this case, the port is
the customs area. This allows for analyses that consider international trade
flows at the intra-national level because the data effectively integrate domes-
tic trade with international trade. This unique perspective on the movement
of goods had much to do with the political circumstances at the time, when
the treaty ports of China were treated as enclaves over which certain foreign
countries had trading rights.
In order to assess trade duties in this elaborate system, the CMC’s staff
recorded the quantity and value of the goods carried. For example, in 1881,
one can find over twenty different categories of cotton goods being imported
(from velvets and velveteens to Turkey red cloths), at least ten different
varieties of woolen goods, and some seventy different sundries that included
window glass, alpaca umbrellas, needles, and dried clams. From 1875 until
around 1933, values were reported in terms of silver, also known as the
Customs tael (or haiguan liang 海關兩). Rates of exchange between the
Customs tael and the local currency existed for each port and were also
reported by the CMC.52
Some qualifications should be noted. First, the CMC improved its record
keeping over time, so that it was not until 1867 that relatively uniform
methods of accounting were put in place. Second, the statistics do not refer
51
Lyons, China Maritime Customs and China’s Trade Statistics, 1859–1948, outlines the
contents of the CMC data and also paints a detailed portrait of the tea trade at several
Chinese ports. On the institutional history of the CMC, see L. Hsiao, China’s Foreign
Trade Statistics (Cambridge, MA, Harvard University Press, 1974); H. van de Ven,
“Globalizing Chinese History,” History Compass 1 (2004), 1–5; D. Brunero, Britain’s
Imperial Cornerstone: The Chinese Maritime Customs Service, 1854–1949 (New York,
Routledge Curzon, 2004); and R. Bickers, “Revisiting the Chinese Maritime Customs
Service, 1854–1950,” Journal of Imperial and Commonwealth History 2 (2008), 221–6;
R. Nield, China’s Foreign Places: The Foreign Presence in China in the Treaty Port Era,
1840–1943 (Hong Kong, Hong Kong University Press, 2015).
52
Between 1875 and 1933, there were more variations in currency units for value – these
included the Spanish dollar, the British pound, local currency, gold units, the gold
dollar, and the Chinese dollar.
432

to the entirety of China’s trade, but only to the trade through treaty ports,
and of this trade, only that part that was carried on foreign vessels or on
Chinese ships of the foreign type (that is, steamers). At the same time, the
foreign-flag vessels included not only ships, but also those transports travel-
ing overland to Russia. Further, from the year 1901 on, the CMC also took
over the operation of Native Customs stations within twenty-five kilometers
of open ports and began to collect data on trade going through these stations
as well. The records on flows of those Chinese-produced goods were pub-
lished in separate tables.
The trade statistics are broadly consistent, both internally with other
numbers reported by the CMC and externally with foreign-partner trade
records where the data are considered to be of high quality. With the decline
of “junk” shipping, the coverage of foreign trade in the CMC data by the year
1904 was essentially 100 percent. At the same time, the CMC data collection
system underwent a number of changes, in part due to changing inter-
national practice, and in part due to structural economic changes. This is to
be expected over a long period of close to 100 years – 1859 to 1948. We will
return to this below in our discussion of China’s commodity-level trade.
China’s Overall Foreign Trade

In this section, we summarize China’s overall foreign trade. This provides
a benchmark for the more disaggregated analysis below. All data are taken
from the annual CMC reports. Figure 12.1 shows the evolution of China’s
aggregate foreign commodity trade. It is clear from the figure that the
expansion in China’s trade remained relatively stagnant from 1865 to 1885.
However, overall trade growth averaged 3.5 percent per year for imports and
2.7 percent for exports over the 1865–1900 period.
Two things are apparent. First, for the period shown, China was more
likely to have a trade deficit than a trade surplus in its commodity trade, the
difference to be covered by bullion or international debt. Second, the volume
of China’s overall foreign trade was relatively stable before the year 1885.
Afterwards, the evolution of China’s trade was reasonably well characterized
by a linear growth trend of about 5 percent per year.
Country Composition
Turning to the composition of China’s foreign trade across countries, we
analyze the nineteenth and twentieth centuries separately because several
major changes took place over this period. Table 12.1 shows China’s main
trade partners in both imports and exports between the years 1865 and 1900.
433

Imports Exports
2,000,000
1,500,000
Haekwan taels (1,000s)
1,000,000
500,000
0
1860 1870 1880 1890 1900 1910 1920 1930 1940
Figure 12.1 China’s foreign trade, 1865 to 1940
Notes. Reported are nominal values of Customs taels, the currency adopted by the CMC.
No adjustments have been made for territorial changes, for example Manchuria, which
became a client state of Japan in 1931. Shown in the graph are total imports; a fraction
between 1 and 5 percent of these were re-exported from China to other countries
The role of Hong Kong in intermediating China’s trade through re-exports

(entrepôt trade) is well known. Only a small fraction of China’s imports from
Hong Kong are produced in Hong Kong, and analogously, only a small part
of Chinese exports to Hong Kong are consumed in Hong Kong. Table 12.1
quantifies this for the nineteenth century, with around 40 percent of China’s
imports originating in Hong Kong, and nearly 30 percent of its exports
destined for Hong Kong. Because the ultimate origin and destination of
China’s trade via Hong Kong is not known for all years, the following analysis
nets out trade through Hong Kong.53
Table 12.2 presents the breakdown of China’s imports for 1900 to 1946,
showing that with the turn of the century a number of additional countries
became important in China’s trade. During the first half of the twentieth
century, Japan was the most important source of Chinese imports, followed
by the United States, while Great Britain had fallen to third place. Beyond the
53
A quantitative analysis of Hong Kong’s role as entrepôt is in Keller, Li, and Shiue,
“China’s Foreign Trade,” 853–92.
434

Table 12.1 Average trade shares, 1865–1900

Imports % Exports %
Great Britain 24.82 Hong Kong 26.94

Hong Kong 41.36 Great Britain 31.65
British India 18.23 Continental Europe 11.86
Japan 5.80 United States 11.07
United States 2.65 Russia 5.82
Continental Europe 2.31 Japan 4.93
Other countries 4.84 Other countries 7.73
level of overall trade, the types of goods imported from these countries
differed, with Great Britain and the United States exporting relatively
more machinery and other producer goods than Japan to China.
Significant amounts of imports originated from nearby sources such as
the Dutch East Indies, French Indochina, Singapore, and Australia.
Among the continental European countries, the relatively early industria-
lizers, such as Germany and Belgium, were more important than coun-
tries that industrialized later, such as Italy. Overall, while the relative
importance of trade with the British Empire had diminished, the evolu-
tion of China’s trade patterns transitioned smoothly along the foundations
laid during the nineteenth century.
It is useful to examine China’s share of world trade in comparison with
other countries. While statistics on China’s trade were meticulously recorded
by the CMC, it was only at the beginning of the twentieth century that trade
statistics for many other countries in the world became available. These
figures are given in Table 12.3.54
China accounted for about 2 percent of world trade from 1913 to 1938, with
a peak in the 1920s. As we will see below, it took a large part of the twentieth
century before it was able to capture a similar share of world trade. The value
of China’s foreign trade corresponds to about three-quarters of that of Japan
and around two-thirds of that of British India. Unsurprisingly, China’s foreign
trade during this period fell far short of that of industrialized countries such as
Great Britain, the United States, and Germany.
54
For countries other than China, see League of Nations, Statistical Yearbook of the League
of Nations (Geneva, League of Nations, 1940).
435

Table 12.2 Major sources of

Chinese imports, 1900–1946
Country %
Japan 24.51
United States 22.05
Great Britain 17.02
British India 9.65
Germany 4.19
Java (Dutch East Indies) 2.91
French Indochina 2.39
Russia (Soviet Union) 2.10
Belgium–Luxembourg 1.85
Singapore 1.56
Australia 1.25
Other countries 10.53
Note. Figures represent each country’s

share of total imports directly into
China net of imports from Hong Kong.
Table 12.3 World merchandise trade by country

Year
Country 1913 1925 1930 1938 Mean
China 1.88 2.30 1.83 1.98 2.00
Great Britain 15.24 14.90 13.44 13.90 14.37
United States 11.15 14.31 12.61 10.70 12.19
Japan 1.79 3.07 2.62 3.20 2.67
British India 3.60 3.59 2.87 2.50 3.14
Germany 13.12 8.00 9.65 9.20 9.99
Notes. Figures for China are from the CMC reports, various volumes. Numbers in the
table measure exports plus imports as a percentage share of the world total.
The Volume of Trade

This section considers the volume of foreign trade of China. For concrete-
ness, we focus on the most important port of China during this period, which
was Shanghai.55 We employ the so-called gravity equation of trade to exam-
ine Shanghai’s bilateral trade with foreign countries. Generally, this is
55
See Keller, Li, and Shiue, “Shanghai’s Trade, China’s Growth,” 336–78, for additional
analysis.
436

because the gravity equation is highly successful in explaining bilateral trade,

and it has been established that many trade theories imply a version of the
gravity equation.56 It is also of interest to see whether Shanghai’s bilateral
trade volumes during the treaty port era were unusual. The fact that trade
treaties were imposed upon China may give rise to doubts as to whether
a model of trade based on voluntary exchange can fit the data. For example, if
there were forced trade for certain bilateral partners, then this could violate
the gravity model if the trades were imposed in a way that ran counter to the
economic basis of trade. What we demonstrate is that in the case of China’s
opening, the gravity model still applies, suggesting that natural trade flows
resulted when ports were opened, with implications of gains from trade to
China and foreign partners.
The gravity equation of trade is, in its simplest form, given by
GDPiα GDPjβ
TRADEij ¼ ð1Þ
DISTijγ
where i and j are two trading economies. TRADE is either exports or imports,
GDP is gross domestic product, and DIST is shipping distance. The idea is that
bilateral commercial interaction is increasing in the size of each economy (the
GDPs) and declining as trade barriers increase (distance would be one example
of a trade barrier). In its log-linearized regression form, the equation is
0
ln TRADEij ¼ α ln GDPi þ β ln GDPj þ γ ln DIST þ Xij δ þ Eij; ð2Þ
where X refers to a set of control variables, and ε is a regression error. We

expect α, β > 0 and γ < 0. The usual signs of the estimated coefficients are α̂ > 0
and β̂ > 0 because bilateral transactions increase in the size of the trade
partners, and γ̂ < 0 because greater distance implies more resistance to trade
due to higher transport costs and other impediments. The term X includes
time-fixed effects in some specifications.
The following countries and regions are included in the analysis: contin-
ental Europe, Egypt, Hong Kong, Japan, the Philippines, Singapore,
Thailand, and the United States.57 Among these, Hong Kong and Singapore
were entrepôts, and to control for this we include an indicator variable. GDP
for Shanghai is proxied by population. Using data for the years 1869 to 1904,
56
An overview is presented in J. Anderson, “The Gravity Model,” Annual Review of
Economics 3 (2011), 133–60.
57
Results for foreign GDP and distance are similar when we include the United Kingdom.
437

Imports 1904 Exports 1904
20
20
HK CE
US
JP JP
CE US
HK
18
18
SG
Trade volume
SG
16
16
EG
PH
14
14
TH TH PH
EG
ME
12
12 14 16 18 20 12 12 14 16 18 20
Fitted values Fitted values
EG: Egypt; TH: Thailand; PH: Philippines; SG: Singapore; HK: Hong Kong; JP: Japan; US: USA; CE: Continental Europe: ME: Mexico
Figure 12.2 Predicted versus actual bilateral trade volume for Shanghai
Notes. On the horizontal axis is the predicted value of trade using values of the independent
variables in the year 1904 and the gravity equation coefficients of Table 12.4. On the vertical
axis the actual value of trade for the same years is given. Trade data from Decennial Reports,
First Issue (1882–1891), Second Issue (1892–1901), Third Issue (1902–1911), Fourth Issue (1912–1921), and
Fifth Issue (1922–1931) (Shanghai, Statistical Department of the Inspectorate General of Customs,
2001); and Shanghai Statistical Yearbooks, various volumes
we estimate a positive coefficient on foreign GDP and a negative coefficient

on distance. Furthermore, the coefficients are not far from plus one (GDP)
and minus one (distance).58
The fit of the gravity equation can be assessed by comparing actual with
predicted trade for a particular year. The results for 1904 are reported in
Figure 12.2. The diagonal in each of the graphs of Figure 12.2 denotes the forty-
five-degree line, where the prediction is equal to actual trade. We see that
bilateral trade during the treaty port era indeed follows the gravity equation.
The gravity equation in trade is typically derived for models of market econ-
omies based on voluntary exchange; the results indicate that it also explains
trade volumes at a time when colonial trade and regular trade are intertwined.59
58
This is in line with other findings by K. Head and T. Mayer, “Gravity Equations:
Workhorse, Toolkit, and Cookbook,” in G. Gopinath, E. Helpman, and K. Rogoff
(eds.), Handbook of International Economics, vol. 4 (Amsterdam, Elsevier, 2014), pp. 131–95.
59
See Keller, Li, and Shiue, “Shanghai’s Trade, China’s Growth,” 336–78, for further
analysis.
438

Foreign Direct Investment in China

With the arrival of foreign traders, consulates, firms, and residents, Chinese treaty
ports were exposed to aspects of British and Western technologies in mechaniza-
tion, transportation, and steam power, and other innovations in financial institu-
tions and banking. Two channels through which major transfers of technology
can take place between countries are capital flows, first through foreign invest-
ments, and second through foreign firms and residents. There is abundant
evidence both for direct capital flows and for the possibility of knowledge and
technological know-how to have diffused from Western countries to China.60
Capital Flows
In this and the following subsection, we highlight the fact that capital flows
increased during the nineteenth and early twentieth centuries. Remer shows
that foreign direct investment (FDI) grew considerably over time, even after
adjusting for inflation. As shown in Table 12.4, in 1902, business investments
totaled around US$503.2 million and this grew to US$1,048.5 million in 1914
and US$2,474.5 million in 1931. The primary investors were Britain, who
invested around 30 to 40 percent in each year, and Japan, who became
a large investor by 1931 with 36.9 percent of investment. Russia, on the
other hand, was a large investor in 1902 with 43.7 percent of the investment,
but was down to only 11.1 percent by 1931. To put the total FDI numbers in
context, Remer notes comparable estimates for India in 1933, with between
Table 12.4 Business investments in China by country

1902 1914 1913
US$ million % of total US$ million % of total US$ million % of total
Britain 150 29.8 400 36.9 963.4 38.9

Japan 1 0.2 210 19.4 912.8 36.9
Russia 220.1 43.7 236.5 21.8 273.2 11.1
US 17.5 3.5 42 3.9 155.1 6.3
France 29.6 5.9 60 5.5 95 3.9
Germany 85 16.9 136 12.5 75 3
Total 503.2 100 1,084.50 100 2,474.50 100
Source: C.F. Remer, Foreign Investment in China (New York, Macmillan Company,
1933), Table 13
60
See also data tabulated in Kung’s chapter in this volume.
439

Table 12.5 Geographical distribution of the direct business investments of

four countries, 1931
Direct business investments (US$ million)
Great Britain Japan Russia USA Total % of total
Shanghai 737.4 215 97.5 1,049.90 46.4

Manchuria 550.2 261.8 812 36
Rest of China (incl. 226 108.9 11.4 52.7 399 17.6
Hong Kong)
Total business 963.4 874.1 273.2 150.2 2,260.90 100
investments
Source: Remer, Foreign Investment in China, Table 11
US$2,000 million and US$3,500 million in total FDI. Similar to the trade flows,
FDI activity was concentrated at Shanghai (and to a lesser extent Manchuria).
In 1931, as seen in Table 12.5, 46.4 percent of the business FDI was based in
Shanghai.
Foreign Firms and Residents

Foreign direct investment can be seen as not just a transfer of capital, but also
a transfer of technological know-how. Furthermore, foreign-owned compan-
ies and individuals can be sources of spillovers to the local economy.
Generally, it is challenging to find quantitative evidence of technological
know-how transfers in historical contexts, but CMC statistics can provide
useful information for China, as summarized in the following.
Figure 12.3 details foreign presence over time for some of the largest treaty
powers, as measured by the number of foreign firms. Great Britain’s position
as the premier colonial power of the day is evident from the hundreds of
firms and thousands of residents present in China over the fifty years
portrayed. Other powers saw their presence grow over time, though the
number of German firms and residents declined noticeably following the
country’s defeat in World War I.
Initially, much of the foreign-owned activity was linked to trade, such as
retail and wholesale operations, banking to finance the trade, insurance to
cover trade risk, shipyards to repair ships, and railroads to provide land-based
transportation. From there it spread into other sectors of the economy.
Manufacturing and mining became important especially after the Treaty of
440

800 600
Number of foreign firms
200 400 0
1872
1877
1882
1887
1892
1897
1902
1907
1912
1917
1922
France Germany Great Britain United States
Figure 12.3 Foreign firms in China: the case of Shanghai, 1872–1921
Notes. Shown are shares of firms by foreign country
Source: Decennial Reports, various volumes
Shimonoseki (1895) established the legal right of foreigners to set up manu-

facturing firms in China.61
As the most important destination for FDI, Shanghai’s situation merits closer
examination. During the treaty port era, the number of foreign firms in Shanghai
was 152 for the year 1872 and 1,741 in the year 1921, implying an annual growth rate
of about 5 percent. These firms originated primarily from Japan and Britain, with
concerns from these countries accounting for 35 and 30 percent respectively
(Figure 12.4). The largest five sources accounted for 87 percent of all FDI into
Shanghai.
There was much heterogeneity in the nature and scope of foreign firms
operating in China.62 They included large firms such as the British Jardine,
Matheson & Co. trading firm. From its head office in Hong Kong and with
branches in every major port, it not only controlled its trade operations, but also
61
Historical treatments of the role of foreign investment in the period can be found in
C.M. Hous, Foreign Investment and Economic Development in China 1840–1937 (Cambridge,
MA, Harvard University Press, 1965), Chapter 3; and Feuerwerker, The Foreign
Establishment in China, Chapter 5.
62
See Feuerwerker, The Foreign Establishment in China, pp. 80–1.
441

Other Great
13% Britain
30%
Germany
Japan 7%
35%
United
France
States
4%
11%
Figure 12.4 Foreign firms in China by country of origin
managed other activities such as the operation of the forty-one Yangzi steamers
of its affiliate, the Indo-China Steam Navigation Company, the large Shanghai
and Hongkew Wharf Company, the Ewo Cotton Mill, and a silk filature in
Shanghai. At the other end of the spectrum was the modest retail store
Schlachterei W. Fütterer, which was the butcher for Shanghai’s German
community.
Quantifying Foreign Influence in China during the

Treaty Port Era
Integrated Statistics on Foreign and Domestic Trade
When a country shifts to opening its economy to more foreign trade, this also
affects commerce in the country’s interior. In the case of China during this
period this can be traced out because every port was treated as its own
customs area. Figure 12.5 illustrates the nature of the information contained
in the CMC reports. Figure 12.5 shows the trade flows to and from Shanghai,
for example, that were reported by the CMC. These flows are decomposed
and labeled from one to nine. The first breakdown is by type of good; flows 1
to 4 concern goods that are produced abroad (foreign goods, abbreviated F),
while flows 5 to 9 show trade in goods that are produced in China (Chinese-
produced goods, abbreviated D). Flow 1 gives the imports of goods from
Japan into Shanghai. Other imports of foreign goods into Shanghai consist of
those coming from other Chinese treaty ports; in the figure, flow 4 represents
the foreign goods reaching Shanghai via Xiamen. Once imported into
Shanghai, these foreign goods may be re-exported. The CMC data allow us
442

to distinguish between re-exports of foreign goods to foreign countries

(flow 2) and those to other treaty ports within China (flow 3).
The statistics on re-exports of foreign goods provide key information on
the extent to which foreign imports diffused throughout the country, some-
thing that has important welfare implications in the case of a large country
such as China. They also offer a direct measure of the consumption of foreign
goods in the treaty ports, as within-port foreign-goods consumption may be
obtained by subtracting re-exports from foreign imports. Separately from
foreign-goods trade, the CMC data also report trade in Chinese-produced
goods. Flow 5, for example, shows Chinese-produced goods that are exported
from Shanghai to foreign countries. These exports are direct exports in the
sense that the goods are produced in the greater Shanghai area. The direct
exports are to be distinguished from other Chinese-produced goods that are
exported abroad from Shanghai but were produced elsewhere in China
(flow 6). Both direct exports and re-exports capture major aspects of the
evolution of an economy. In particular, the size of direct exports demon-
strates the change in the production possibility of the local economy, while
the extent of re-exports sheds light on the development (and trade integra-
tion) of the hinterland as well as the capacities of the entrepôts (in this case,
Shanghai).
JAPAN
6
7
Jiujiang 1
5
8 Shanghai
1: Imports from abroad
2: Re-exports abroad (F)

CHINA 9
3: Re-exports to China (F)
2 4: Imports from China (F)
4
3
5: Exports abroad
Xiamen 6: Re-exports abroad (D)
7: Exports to China
PHILIPPINES
8: Re-exports to China (D)
9: Imports from China (D)
Figure 12.5 Export and import flows to and from Shanghai
443

Flow 7 represents Shanghai-produced goods that are exported to other

parts of China, whereas flow 8 gives the export of Chinese-produced goods
that flow from one region of China to another through Shanghai. The
information on domestic exports of Chinese-produced goods at the port
level is thus comparable to the information on foreign exports. Finally,
flow 9 shows Shanghai’s imports of Chinese goods that were produced
elsewhere in China.
Welfare Effects from Foreign Influence: China’s Domestic

Trade
This section quantifies the size and distribution of welfare effects from new
technology due to foreign influence in terms of China’s domestic trade using
a well-known trade model (that of Eaton and Kortum).63 The model captures
the Ricardian determinants of comparative advantage and differences in
relative productivities across goods, and relates them to the geography
separating the trading partners.
First, there is the technology of each trading partner, which determines the
cost at which a good can be produced in different regions, and therefore
determines which region has the lowest factory-gate production costs.
Second, there is the size of trade barriers between regions; for example,
trade barriers between regions i and j determine the trade-cost-inclusive price
in region j of a good that is produced in region i. Because trade costs increase
with geographic distance between regions, geography is the second key force
in the model. Using historical data on prices, trade flows of domestically
produced goods, and input uses, we can calibrate and solve the model in the
context of China in the treaty port era.
Figure 12.6 shows the volumes of aggregate bilateral trade, with the
thickness of each line proportional to the size of the flow. The maximum
distance between any two regions in our bilateral pairs is about 2,700
kilometers. Exports are shown in the same shade as the region’s label and
are offset towards the center of the figure. We see that Hankow exports
a large amount of its production to Shanghai, for example, while Tientsin’s
imports from Hankow are smaller than Tientsin’s imports from Shanghai.
Importantly, the trade volumes shown in Figure 12.6 are for locally produced
goods.
63
This section is based on Keller, Santiago, and Shiue, “China’s Domestic Trade,” 26–43.
See also J. Eaton and S. Kortum, “Technology, Geography, and Trade,” Econometrica
70.5 (2002), 1741–79.
444

on Newchwang
Cant
oy Tie
Am nts
in
ow
at
Sw Ki
a
oc
w
ho
ho
oc
w
Fo
Ch
efo
o
gp
o
Nin
Ich
ang
n ko w
Ha
S ha
ng
ha
i
ng
kia
K iu
u
Wuh
C h in kia n g
Figure 12.6 The size of bilateral trade between regions

Source: calculated from CMC trade data
Commodity-level trade data are employed to estimate a key model parameter

governing the strength of comparative advantage. Employing twenty-six homo-
geneous commodities that are traded between virtually all fifteen regions –
including coal, matches, and cotton yarn – price differences across ports are
used to pin down this parameter. The commodity-level trade data are also
employed to estimate trade costs between any two regions, because arbitrage
ensures that the price difference for a given good between two regions is an
upper bound on the trade costs between these two regions.64
64
See Keller, Santiago, and Shiue, “China’s Domestic Trade,” 26–43, for information on
other, related, data.
445

With key parameters and the model in hand, one can perform interesting
counterfactual experiments. One is to increase the parameter capturing port
productivity by 20 percent. This magnitude is reasonable given that customs
operation by the CMC brought with it a wide range of improvements, such as
dredging of the harbor, the construction of new lighthouses, increased
protection from pirates, and the customs process itself. Increasing the prod-
uctivity of Shanghai by this amount while leaving all other parameters
unchanged raises GDP (our measure of welfare) in Shanghai by 1.5 percent,
and general equilibrium effects lead to an increase of GDP by about a quarter
of 1 percent on average in the other regions.
Table 12.6 presents the results of reductions in trade costs, as would have
happened through the introduction of foreign steamship technology. Trade
increases by 13 percent as a result of the lower trade costs. Welfare gains,
however, are unevenly distributed across ports, and some regions, in particu-
lar Shanghai and Ningbo, experience welfare losses.
The intuition for this lies in the reallocation of production and trade.
Lower trade barriers diminish the relative importance of technology-based
Table 12.6 Lower geographic barriers and welfare

Trade barriers ↓50%
% change relative to baseline
Welfare Prices Wages
Amoy 2.58 −7.15 −2.07

Chefoo 0.17 4.25 9.36
Chinkiang 0.75 −2.90 −3.53
Foochow 2.15 0.05 4.02
Canton 1.67 −0.03 3.24
Hankow 2.01 10.23 10.30
Ichang 8.90 −26.89 −24.80
Kiaochow 1.95 −6.13 −2.79
Kiukiang 3.51 −8.00 −2.26
Newchwang 4.22 −3.48 9.96
Ningbo −31.41 15.12 −32.25
Swatow 0.27 3.22 14.44
Tientsin 3.43 4.25 8.95
Wuhu 2.38 −4.59 1.77
Shanghai −17.29 23.73 −7.76
% change in overall trade 13.14
Note. Table shows results of lowering geographic barriers by 50 percent relative to

baseline trade costs.
446

advantages. Notice that the four regions with the lowest welfare gains are
Shanghai, Ningbo, Chefoo, and Swatow. They turn out to be also the four
regions with the highest level of labor-cost-adjusted technology. For such
regions, lower trade barriers mean that they might no longer be the low-cost
source of supply in a region because with lower trade costs that region now
imports from elsewhere. As a result, welfare in the high-technology regions
might fall.
Further, Shanghai and Ningbo are relatively centrally located in China,
which means that before the reduction of trade barriers these regions had
a sizable advantage based on low transport costs compared to other regions.
In contrast, Chefoo and Swatow are located in geographically more remote
parts of China. They lose some markets as a result of the lower trade barriers
at the same time as they maintain their hold on others precisely because of
their geographic location. As a consequence, Chefoo and Swatow lose less
than Shanghai and Ningbo.
The Geographic Scope of Foreign Influence

The previous analysis has assessed the welfare effects of foreign trade and
technology for China through the lens of a specific trade model. The impact
of foreign influence in China can also be studied using a less structural but
regression-based approach. Specifically, foreign influence in China originated
from the foreign places, often treaty ports, in which the foreigners with their
firms, families, and institutions were located. While treaty ports never
accounted for more than 10 percent of China’s output, foreign influence in
the treaty ports might have generated spillovers for neighboring areas and
China’s hinterland.65 Moreover, this literature emphasizes the role of insti-
tutions for the impact of foreign countries in China.66
65
See B.K.L. So, H. Yip, T. Shiroyama, and K. Matsubara, “Modern China’s Treaty Port
Economy in Institutional Perspective,” presented at the All-University of California
Group in Economic History, February 2011 (Berkeley, CA).
66
D. Ma, “The Rise of Modern Shanghai, 1900–1936: An Institutional Perspective,” in B.K.
L. So and R.H. Myers (eds.), The Treaty Port Economy in Modern China: Empirical Studies
of Institutional Change and Economic Performance (Berkeley, University of California
Institute of East Asian Studies, 2011), pp. 33–46, discusses the expansion of Western
institutions in Shanghai in the early twentieth century; T. Shiroyama, “The Shanghai
Real Estate Market and Capital Investment, 1860–1936,” in So and Myers, The Treaty Port
Economy in Modern China, pp. 47–74, deals with institutional changes in real estate
markets that promoted economic change; and K. Chan, “The Rice and Wheat Flour
Market Economies in the Lower Yangzi, 1900–1936,” in So and Myers, The Treaty Port
Economy in Modern China, pp. 75–95, shows how the expansion of vertical integration in
grain markets was directly related to the presence of foreign technologies and eco-
nomic activity in treaty ports.
447

Connecting the anecdotal and case study evidence, Keller and Shiue
estimate the impact of foreign institutions on the level of interest rates in
China’s regional capital markets during the nineteenth century. Using vari-
ation on the location of opening treaty ports, customs stations, and foreign
consulates, which supported trade by enforcing law courts in China, they
show that foreign institutions had a positive impact by substantially lowering
regional interest rates relative to areas without foreign influence.67
Their work also quantifies the scope of foreign influence by estimating the
size of geographic spillovers of the foreign impact. Figure 12.7 shows that
a foreign consulate located up to 200 kilometers from the center of the region
leads to a lower interest rate by about 1.3 percentage points, and a similar
effect comes from an open treaty port within 200 kilometers. The results
point to a relatively strong impact through foreign consular courts, with their
foreign legal practices due to extraterritoriality, because consulates generated
significant spillovers for up to 400 kilometers (in contrast to treaty ports).
The implications for the geographic scope of foreign influence can be seen
by plotting the predicted effects from the regression on a map of China (see
Figure 12.8). While the analysis confirms that foreign influence in China was
strongest where foreigners had their strongest presence, Figure 12.8 also
indicates that by the 1890s, the geographic scope of foreign influence may
have been felt in the majority of China’s areas.
The Granular View: Chinese Commodity-Level

Trade
Quantifying New and Disappearing Goods in China’s Trade
New goods in trade matter for several reasons. For one, in the absence of
comprehensive information on supply, exports of a new good are proof of
a certain level of production capability. The arrival of new goods also
provides evidence on the level of specialization in production. On the
demand side, new goods reflect changes in consumption patterns, and they
are indicative of changes in income because demand is non-homothetic (e.g.
67
Chinese capital markets are compared with British markets for the 1770–1860 period in
W. Keller, C.H. Shiue, and X. Wang, “Capital Market Development in China and
Britain, 18th and 19th Century: Evidence from Grain Prices,” American Economic Journal:
Applied Economics 13.3 (2021), 31–64; see also W. Keller, C.H. Shiue, and X. Wang,
“Capital Markets and Grain Prices: Assessing the Storage Cost Approach,” Cliometrica
14 (2020), 367–96, for more on the estimation of comparable interest rates.
448

Point Estimate with 95% Confidence Interval 1
–1
–2
–3
km
te
km
km
km
km
km
la
0
su
0
20
20
40
40
60
60
on
0–
0–
0–
0–
in
in
rc
ith
ith
20
20
40
40
o
w
rt
in
in
in
in
po
te
rt
ith
ith
ith
ith
po
la
w
y
su
at
te
rt
te
rt
at
e
on
po
po
la
la
Tr
e
C
su
su
Tr
y
at
at
on
on
e
e
C
C
Tr
Tr
Figure 12.7 The impact of foreign influence in China: geographic effects

Source: W. Keller and C.H. Shiue, “Capital Markets and Colonial Institutions in China,”
presentation at NBER Summer Institute, July 2020 (Cambridge, MA)
the share spent on luxury goods increases with income). New goods can also
be an important source of utility (welfare) gains.68
The importance of new goods and of disappearing goods between two
years s and t, with s being earlier, can be quantified by considering the set of
goods that is available in year s, the set of goods available in year t, and the set
of goods that is available in both years s and t. Clearly, the set of goods
available in both years will typically be smaller than the set of goods that is
available in a single year.69 Now consider the value of goods available both in
years s and t evaluated at prices of year s, relative to the value of goods of year
s at prices of year s; call this expression λs. Because both of these bundles are
evaluated in year-s prices but the first set is (weakly) smaller than the second,
λs must be smaller than or equal to one (λs ≤ 1). Also, let the value of goods
68
See R. Feenstra, “New Product Varieties and the Measurement of International Prices,”
American Economic Review 1 (1994), 157–77, for more analysis.
69
The former will be smaller than the set of goods available in at least one of the years,
except when the sets of goods in s and t are the same.
449

Direct effect + spillover effect

–5 –4 –3 –2 –1 0
Map 12.2 The impact of foreign influence on local capital markets

Note. Figure gives predicted effects for the year 1890 of analysis underlying Figure 12.7;
treaty ports shown
available both in years s and t be evaluated at prices of year t, relative to the value
of goods of year t at prices of year t, be denoted by λt. For analogous reasons, also
λt must be smaller than or equal to one (λt ≤ 1).
Table 12.7 shows what the values of λt and λs tell us about the importance of
new and disappearing goods. To begin with, λt = 1 indicates that all goods that are
available in year t were also available in s (no new goods). In contrast, λt < 1
indicates that some of the goods available in year t were not available in year s.
That is, some new goods appeared by year t . The case λs = 1 indicates that all the
goods that were available in year s are also available in year t (none of the goods
disappeared). Finally, λs < 1 indicates that some of the goods available in year
s were not available in year t (some goods disappeared by period t). Thus, for any
450

earlier year s and later year t, λt provides information about new goods while λs
tells us about disappearing goods.70
To implement this in our case, ideally one would like to have the value of
trade on every item in every year – at the most disaggregated, eight-digit
level – that was ever exported in any year during the period from 1867 to 1930.
Because the information collected by the CMC covers virtually the universe
of China’s foreign trade it is well suited for an extensive-margin analysis. At
the same time, the CMC reports China’s universe of exports with
a classification that has a varying degree of disaggregation and is changing
over time. In essence, in the early years China’s exports are relatively low and
only a few relatively aggregate categories are distinguished, while in the final
years of the sample period exports are higher and many more goods are
reported at a relatively disaggregated level. The category that ensures that
CMC statistics cover the universe of exports in every year is called “Sundries,
unenumerated,” which is listed in every year. The Sundries category is not
further defined but it includes all other goods that do not fit into any of the
other categories defined in the CMC statistics in a particular year.
Our approach for studying the extensive margin of trade is based on the
fact that the CMC produced a major update of its goods classification only
every couple of years.71 In a year when there is a major increase in terms of
reported product groups, one typically observes a substantial decline in
Sundries exports. This is indicative of the fact that before this year, the
newly reported product groups were part of the Sundries category. To fix
Table 12.7 Measuring the appearance and disappearance of goods

λs = 1 λs < 1
λt = 1 • Same goods in both years • Some goods disappear

between years s and t between
years s and t
• No new goods by years t
λt < 1 • No goods disappear between s and • Some goods disappear
t years between years s and t
• New goods by year t • New goods by year t
70
For values of λs and λt below 1, it is important to keep in mind that these expressions
reflect the value, not the product count, of disappearing and new goods respectively.
71
To identify these years we employ a threshold value for the change in Sundries exports
from one year to the next; different threshold values are employed to ensure the
robustness of the analysis.
451

ideas, beginning in the year 1896, CMC statistics report the value of “Cattle,
Sheep, Goats, and Pigs” as its own export category. Presumably, this is due to
the fact that there was a sizable and sustained export of these animals, but it
would be erroneous to conclude that China did not export any cattle, sheep,
goats, or pigs before the year 1896 – export of such livestock is not a “new-
good” export of China in the year 1896. Rather, prior to that year, the value of
the exports of such livestock was recorded as part of the Sundries category.
Further, starting in the year 1910, CMC statistics report export values for
“Cattle,” “Sheep,” “Goats,” and “Pigs” separately.
An extreme approach to the new-goods margin is to assume that there are
not any new goods, only redefinitions of previously traded goods. In the
example, this means that, given that we observe positive exports of “Cattle,”
“Sheep,” “Goats,” and “Pigs” separately in the year 1930, we assume that all of
these items were also exported already in the year 1867. Our way of imple-
menting this approach to estimate disaggregated export values for 1867 (and
any later year) is to use the 1910 export values of “Cattle,” “Sheep,” “Goats,”
and “Pigs” to allocate the given export values of the single category “Cattle,
Sheep, Goats, and Pigs” during 1896 and 1909 between the four different types
of livestock. To obtain values for earlier years we use the fact that livestock
exports before 1896 are part of Sundries, which implies that livestock exports
in 1895 are equal to the change in Sundries exports between 1895 and 1896.72
We arrive at disaggregated export values for the year 1867 (and any later year)
by using the change in Sundries exports between 1895 and 1867 together with
the disaggregated livestock shares that are available for the years 1910 to 1930.
While this approach is conceptually straightforward, λt is equal to one for
every year t; that is, there are no new goods by assumption. To allow for the
possibility of new goods, we define a baseline threshold value of exports for the
new good relative to the change in Sundries exports category. Thus, for example,
if the exports of “Cattle” in 1895 that we estimate based on the change in Sundries
exports from 1895 to 1896 are below this threshold, we take “Cattle” to be
a newly exported item in the year 1896. We vary this baseline threshold to
ensure the robustness of our analysis.
72
In practice, “Cattle,” “Sheep,” “Goats,” and “Pigs” exports are not the only new
categories that are introduced in the year 1896. We account for that by assuming that
the sum of all new category exports in a year, when there are major changes, is equal to
the change in Sundries trade.
452

Mapping China’s Trade to the First International Trade

Classification
The League of Nations’ minimum list of commodities (MLC, 1935) emerged
as the first way of recording international trade data in a consistent way from
the 1931 League of Nations’ tariff nomenclature. The MLC is the precursor of
the modern-day standard international trade classification (SITC). At the time
of the MLC’s creation China’s government stated that it was not prepared to
compile statistics based on the MLC classifications. However, to be able to
consistently track China’s international trade over time and for making
international comparisons, we have matched CMC information on Chinese
exports of the years 1867 to 1930 to the MLC classification.
We create a match to the MLC classification by assigning each CMC
export to an MLC commodity number with up to eight digits for
every year from 1867 to 1930. For example, our version of the MLC classifica-
tion contains in its Chapter 8 the item “Beans and Peas,” classified with item
number 0803 (Item 03 in Chapter 08). This item is further disaggregated into
other items, such as “Beans” (080301) or “Peas” (080302). The item “Beans”
contains an even finer classification with different varieties of beans such as
“Beans, Black” (08030101) or “Beans, Green” (08030103), among others.
This detailed mapping of China’s commodity-level trade enables us to
consistently track the evolution of China’s exports starting in a period for
which, to date, there is little systematic information. It is also instrumental in
studying the changing commodity structure in China’s trade, in particular the
emergence of new goods, to which we turn now.
The Importance of New and Disappearing Goods

The sixty-four years of data, from 1867 to 1930, provide an excellent setting to
explore the introduction of new goods and the disappearance of existing goods.
We construct a 64 × 64 matrix of λs for each s and t pair, which is shown in
Figure 12.8.
For any pair of different years (s < t), λt is shown in position (t, s) to the right of
the diagonal in Figure 12.8, and λs in position (s, t) to the left of the diagonal in
Figure 12.8. In particular, in our baseline approach the (1930, 1867) element is
equal to 0.334. This indicates that the 1930 value of the goods available in both
1867 and 1930 represents about one-third of the 1930 value of all the goods
available in the year 1930. Thus many new goods appeared between 1867 and
1930, or at least the new goods had relatively high value compared to the old
453

1930 1930
1920 1920
1910 1910
1900 1900
Year 1890 1890 Year
1880 1880
1867 1867
Figure 12.8 New and disappearing goods: China’s exports
Source: authors’ computations based on trade statistics in China Maritime Customs
Service, Returns of Trade and Trade Reports (various years) (Shanghai, Statistical
Department of the Inspectorate General of Customs, 2001)
(1867) goods. This indicates that there was a substantial change in the compos-
ition of the goods exported from China over this period. Furthermore, the (1867,
1930) element in this matrix is equal to 0.999. Thus the 1867 value of the goods
available in both 1867 and 1930 represents almost 100 percent of the 1867 value of
all the goods available in that year. It means that either only a few goods
disappeared between 1867 and 1930, or those goods that disappeared had
a relatively low value. This indicates that the appearance of new goods (or
their value) tends to be more important than the disappearance of old goods in
China in the nineteenth and early twentieth centuries.
The work on mapping the commodity-level trade of China over time is still in
progress, yet the above analysis has shown that investigations of various margins
of trade are feasible with historical trade data for China. The new goods that
enter foreign trade are likely to generate implications for welfare; they are likely
to impact domestic markets and domestic firms by changing the conditions for
innovation and competition. The fact that the CMC collected detailed statistics
on commodity-level trade means that we will ultimately be able to produce
454

a complete picture of the historical evolution of China’s comparative advantage

and trade, which is rarely available in the case of other countries.
Conclusions
Today, China is one of the largest traders in the world. Unlike the emperors of the
past, China is actively engaged in and pursuing the gains from openness, be that
commodity trade or foreign direct investment. Nevertheless, China’s prominence
in global trade today can be linked to its historical past. From the seventeenth to
eighteenth centuries, China’s foreign trade and foreign diplomacy were an
inherent part of the socioeconomic and political context of Asia. The empire
had emerged as a dominant power in the region through a wide combination of
strategies that included tributary trade, colonialism, economic investment, and
military conquest. Compared to the Ming state, the Qing state had a relatively
laissez-faire stance towards the activities of merchants in domestic markets and
within the sphere of intra-Asian foreign trade. Nevertheless, Qing emperors
reserved the right to regulate trade, and in particular the Sino-Western trade.
For over a century, China was able to enforce its foreign policy on Western
traders. In 1839, a combination of conflicts over market access, opium imports,
and foreign diplomacy resulted in the Opium Wars. Over the nineteenth century,
China was forced to sign a series of treaties with foreign powers that legally
entitled Western nations and Japan to establish numerous ports of trade, customs
stations, consular cities, and other types of port cities within China.
The treaty port system originated in the effort to open up China’s markets to
Western traders, and subsequent institutions that arrived in China sought to
support that trade. Apart from the China Maritime Customs Service, pervasive
legal institutions – in the form of consular offices and formal courts – were also
established to support foreign trade. From the mid-nineteenth century until the
departure of the China Maritime Customs Service in the twentieth century, the
quantity and quality of data that were collected on Chinese international and
domestic trade increased dramatically. This chapter has documented the trends
and highlighted a few methodologies that can be used to allow researchers to gain
a window into the economy of China during this time. These data can be used to
investigate numerous additional questions about international and domestic
trade, as well as the impact of foreign trade on the welfare of China. In addition,
as the data are particularly detailed with respect to port-level information, it is
exceptionally valuable for understanding the conceptual relationships between
international trade and regional (domestic) trade more generally when a country
changes its trading regime.
455

We find that already by the 1890s foreign influence in China was strongest
where foreigners had their strongest presence, and spillover effects may have
resulted in the majority of China’s areas being affected. While we have yet to
reach a complete understanding of the global effects of foreign trade and factor
flows in the past two centuries, beyond the direct impact of trade and specializa-
tion, the most important effects during the treaty port era are likely related to
interactions with foreign institutions and technology transfer. Investigations of
the various margins of trade on welfare, innovation, and competition are prom-
ising areas of further research. More generally, it will open new ground with
respect to our understanding of China’s history of development as it relates to the
legacy of the treaty port era.
Further Reading
China Maritime Customs (CMC), Decennial Reports, First Issue (1882–1891), Second Issue
(1892–1901), Third Issue (1902–1911), Fourth Issue (1912–1921) and Fifth Issue (1922–1931)
(Shanghai, Statistical Department of the Inspectorate General of Customs, 2001).
Great Britain Foreign Office, Correspondence Relating to China: Presented to Both Houses of
Parliament, by Command of Her Majesty (London, T.R. Harrison, 1840).
Hamashita, T., China, East Asia, and the Global Economy: Regional and Historical Perspectives
(Hoboken, Routledge, 2008).
Keller, W., M. Lampe, and C.H. Shiue, “International Transactions: Real Trade and Factor
Flows,” in S. Broadberry and K. Fukao (eds.), Cambridge Economic History of the Modern
World, vol. 1, 1700 to 1870 (Cambridge, Cambridge University Press, 2021), pp. 412–37.
Keller, W., B. Li, and C.H. Shiue, “China’s Foreign Trade: Perspectives from the Past 150
Years,” World Economy 6 (2011), 853–92.
Keller, W., B. Li, and C.H. Shiue, “The Evolution of Domestic Trade Flows When Foreign
Trade Is Liberalized: Evidence from the Chinese Maritime Customs Service,” in
M. Aoki, T. Kuran, and G. Roland (eds.), Institutions and Comparative Economic
Development (Basingstoke, Palgrave Macmillan, 2012), pp. 152–72.
Keller, W., B. Li, and C.H. Shiue, “Shanghai’s Trade, China’s Growth: Continuity,
Recovery, and Change since the Opium War,” IMF Economic Review 2 (2013), 336–78.
Keller, W., J.A. Santiago, and C.H. Shiue, “China’s Domestic Trade during the Treaty-Port
Era,” Explorations in Economic History 1 (2017), 26–43.
Keller, W., and C.H. Shiue, “Capital Markets and Colonial Institutions in China,”
presentation at NBER Summer Institute, July 2020 (Cambridge, MA).
Keller, W., C.H. Shiue, and X. Wang, “Capital Market Development in China and Britain,
18th and 19th Century: Evidence from Grain Prices,” American Economic Journal: Applied
Economics 13.3 (2021), 31–64.
Morse, H.B., The Chronicles of the East India Company: Trading to China 1635–1834, 5 vols.
(Oxford, The Clarendon Press, 1926–1929).
Morse, H.B., The International Relations of the Chinese Empire, 3 vols. (London, Longmans,
Green, and Company, 1910–1918).
456

13
Transport and Communication
Infrastructure
e l i s a b e t h k ö l l
Modern means of transportation and communication along water,

rails, and roads had a profound impact on the economic and social
development of China from the mid-nineteenth century onward. After
the arrival of the steamship in the 1840s and the telegraph in the early
1860s, railroad construction began to emerge slowly at the close of the
century, followed by bus and motor traffic bringing about macadam-
ized city streets and highway expansion, with a modest level of air
traffic taking off in the 1930s. This chapter addresses the structural
changes in transportation and communication that characterized the
transition from the last decades of the Qing empire (1644–1911) through
the Republican period (1911–1949) to the early years of the People’s
Republic of China (PRC).
The focus on modern, mechanized infrastructure such as railroads or
steamships should not be interpreted as disregard of traditional trans-
portation methods via junk boats, wheelbarrows, or animal-drawn
carts which continued to offer important transportation services, espe-
cially in rural China, well into the twentieth century. By contrast, this
chapter elucidates the high level of complementarity between trad-
itional and modern transportation methods from the late nineteenth
century onward and emphasizes the pragmatic embrace of techno-
logical changes in transportation among all socioeconomic levels of
Chinese society. The chapter also pays more attention to railroads than
to other transportation methods due to the rail network’s continuing
geographical expansion and close affiliation with government agendas
under different regimes beyond the 1949 revolution. Most national
histories of railroads in Britain, the United States, India, and Japan
explore their development in the context of technological innovation
and the Industrial Revolution, the rise of capitalism, and the
457

elisabeth kö ll
emergence of the modern nation-state.1 As China’s political economy

was characterized by government patronage and the absence of capit-
alism and large-scale industrial investment before 1949, the introduc-
tion of modern transportation infrastructure particularly benefited
growth in the commercial agriculture sector while it played a much
lesser role in expanding industrial production.
Considering China’s large territorial expanse, regional socioeconomic
diversity, and decentralized, often volatile, political landscape prior to the
founding of the PRC, it should not come as a surprise that the introduction of
modern transportation infrastructure did not lead to a nationwide pattern of
economic and market integration but intensified regional economic integra-
tion instead. For example, during the Republic, passenger rail transportation
overwhelmingly consisted of passengers of modest means who traveled by
third or fourth class and generated the bulk of revenue for Chinese railroad
companies. Another important feature of the passenger market was that the
majority of the traveling public used the network primarily for short dis-
tances. As a result, the passenger business reinforced the regionalizing effects
of the freight traffic, with its powerful stationmasters and line-based logistics
companies. As this chapter will demonstrate, freight and passenger services
thus contributed to the regional rather than the nationwide orientation of
transportation infrastructure in China before 1949.
In a broader historical perspective, this chapter tries to outline the evolu-
tion of China’s transportation network, the competing business interests
among different transportation technologies and companies, and the eco-
nomic goals and regulations reflecting different local, regional, and central
government agendas – patterns that still inform infrastructure development
in contemporary China. In many ways, the development of modern trans-
portation infrastructure allows us to examine the competitive roles of the
state at various administrative levels and of private actors in the historical
process of moving people, goods, and ideas across China, as well as the
continuities and changes they imposed on Chinese economy and society.
1
R. Fogel, Railroads and American Economic Growth: Essays in Econometric History (Baltimore,
Johns Hopkins University Press, 1964); H.J. Habakkuk, American and British Technology in
the Nineteenth Century (Cambridge, Cambridge University Press, 1962); J.F. Stover, Iron
Road to the West: American Railroads in the 1850s (New York, Columbia University Press,
1978); P. O’Brien, The New Economic History of the Railways (New York, St. Martin’s Press,
1977); T.R. Gourvish, Railways and the British Economy, 1830–1914 (London, Macmillan,
1980); J. Metzer, Some Economic Aspects of Railroad Development in Tsarist Russia (New York,
Arno Press, 1977); S.J. Ericson, The Sound of the Whistle: Railroads and the State in Meiji Japan
(Cambridge, MA, Harvard Council on East Asian Studies, 1996); I.J. Kerr (ed.), Railways in
Modern India (New Delhi: Oxford University Press, 2001).
458

Transport and Communication Infrastructure
Beginnings: Postal and Telegraph Communication

Before steamships and railroads made their appearance in China, long-
distance communication in the nineteenth century was conducted within
a postal system (yizhan 驿站) that exclusively served the needs of the Qing
imperial government. Private citizens had to rely on commercial letter
agencies for sending correspondence across the country, whereas informal
messengers such as travelers or muleteers took on the function of postal
delivery in less accessible regions in the interior or rural areas. Any service
had to cope with vast distances within the empire as well as the lack of good
roads, many of them made of pounded earth, which turned into mud during
rain and were not conducive to heavy traffic. By comparison, water trans-
portation along canals and rivers was more convenient and faster, especially
with the arrival of steamships on inland waterways in the second half of the
nineteenth century. For example, letter agencies used steamboats to transfer
large mailbags whose content was sorted and delivered by carriers from local
agencies when the steamships reached the port. Some of the postal transmis-
sion of urgent Chinese government documents was taken over by telegraph
offices as the yizhan system gradually deteriorated during the last decades of
the Qing dynasty due to the neglect of maintenance of roads and equipment.2
In 1861 Sir Robert Hart, director of the Chinese Maritime Customs Service,
submitted a number of proposals to the Qing court for the organization of
a modern postal system, pointing out the potential benefits for the govern-
ment in the form of increased trade and tax revenue. However, against the
background of various domestic crises and fear of increasing foreign inter-
vention, the Qing government was not yet ready to undertake this modern-
ization project. A mix of mounted and foot courier services, government
dispatch bureaus, commercial letter agencies, foreign postal agencies ser-
vicing Westerners in the treaty ports, and the Customs Post as part of the
Maritime Customs Service offered their services until a Chinese postal system
was founded in 1896.3
As a new postal system, the Imperial Post Office (youzhengju 邮政局) was
established via extension of the Customs Post into the interior with the goal
of eventually covering the whole empire, managed by Hart under the
supervision of the Qing government. Due to the connection with the
2
Y.-W. Cheng, Postal Communication in China and Its Modernization, 1860–1896 (Cambridge,
MA, Harvard University East Asian Research Center, 1970), pp. 3–5, 44, 82–3.
3
Cheng, Postal Communication, pp. 5–6; W. Tsai, “Breaking the Ice: The Establishment of
Overland Winter Postal Routes in the Late Qing China,” Modern Asian Studies 47.6 (2013),
1749–81.
459

elisabeth kö ll
customs service, itself modeled on the British civil service system, the
Chinese post office system was modeled on the British Post Office with
appropriate modifications. Despite slow beginnings, China already had
more than a thousand post offices by 1903. After the founding of the
Republic in 1911, the Chinese Post Office became independent of the customs
service. Like many other new administrative institutions evolving in the new
Republic, the post office became part of the government’s institution-
building process in terms of structuring its organization and hierarchies,
defining responsibilities, and prescribing regulations. In 1921, comprehensive
regulations for the postal administration were publicized by the government,
and in 1935 the Ministry of Communications and Transportation introduced
the legal framework of the postal service with the announcement of a Postal
Law (youzhengfa 郵政法). Apart from its domestic postal and financial
services, the Post Office also provided an important link to transnational
Chinese communities through its overseas remittance network which main-
tained operation through crises and war until 1949.4
An important step towards modern communication involved the intro-
duction of the telegraph into China, especially as a requirement for the
construction and operation of railroads. The Great Northern Telegraph
Company, a Danish business controlled by a majority of British shareholders,
laid the underwater telegraph lines from Europe via Russia to Shanghai in
1870. Three foreign submarine cable companies, Great Northern Telegraph,
British-owned Eastern Extension Telegraph, and Commercial Pacific Cable,
an American firm, were in competition with each other and instrumental in
connecting China with other parts of Asia: by 1871 submarine cables already
connected Shanghai with Nagasaki. Recent research has focused on the
intersection between technology and politics and explored the tensions and
multilayered conflicts among different political interest groups impacting the
transfer of cable technology and its commercial use in a semicolonial
context.5
4
Cheng, Postal Communication, pp. 104–5; Qiu Runxi 仇潤喜 and Li Guangyu 李光焴
(eds.), 天津郵政史料 (Historical Materials on the Tianjin Postal Administration), vol. 1
(Beijing, Beijing hangkong xueyuan chubanshe, 1988), pp. 341–55; L.J. Harris, “The Post
Office and State Formation in Modern China, 1896–1949,” Ph.D. thesis, University of
Illinois at Urbana–Champaign, 2012.
5
E. Baark, Lightning Wires: The Telegraph and China’s Technological Modernization, 1860–1890
(Westport, CT, Greenwood Press, 1997), pp. 56–9; S. Wei, “Circuits of Power: China’s
Quest for Cable Telegraph Rights 1912–1945,” Journal of Chinese History 3 (2019), 116–18;
460

However, despite their near monopoly position in control over submarine

cables linking China to the world, Western companies did not own
a significant number of lines on land, which came to stay firmly in Chinese
hands. In December 1881, a telegraph line between Tianjin and Shanghai was
opened to traffic. The line’s main office in Tianjin came under the control of
Sheng Xuanhuai as the officially appointed manager of the new Imperial
Telegraph Administration (dianbao zongju 电报总局) established in 1882,
with a branch office in Shanghai supervised by the official Zheng
Guanying, the city’s circuit attendant. After the telegraph central office
moved from Tianjin to Shanghai in 1884, it became the core of China’s
telegraph network, which played an important role in facilitating future
railroad development. Eventually the Imperial Telegraph Administration
became the Telegraph Bureau under the control of the Ministry of Posts
and Communications in 1907.6
China’s Imperial Telegraph Administration was structured according to
the concept of government supervision and merchant management (guandu
shangban 官督商办), with the goal of revitalizing China’s statecraft, institu-
tions, and economy like many new official–private ventures founded during
China’s Self-Strengthening movement in the 1880s. According to the 1882
regulations drawn up by Sheng Xuanhuai, “The first purpose of telegraph
lines constructed in China is to transmit military messages; in addition they
serve the convenience of the merchants and populace.” However, despite the
central government’s politically driven agenda, most telegraph lines were in
fact built and maintained by provincial governments. The Qing government
thus passed on the costs for constructing and maintaining the telegraph
system to the private investors while also demanding special rates and
preferential treatment for its telegrams.7
Although the telegraph played a significant role in the communication
sector’s development, the fact that the Imperial Telegraph Administration
operated like a government monopoly without serious business competition
presented a problem. Past interpretations by historians like Albert
Feuerwerker have argued that the Telegraph Administration suffered from
administrative mismanagement, unfamiliarity with telegraphic affairs,
6
Baark, Lightning Wires, p. 166; Feuerwerker, China’s Early Industrialization, pp. 189–210;
Xiong Yuezhi 熊月之, 上海通史 (A Comprehensive History of Shanghai), vol. 4
(Shanghai, Shanghai renmin chubanshe, 1999), pp. 332–3.
7
Feuerwerker, China’s Early Industrialization, pp. 189–210, esp. p. 193; W. Yoon, “Dashed
Expectations: Limitations of the Telegraphic Service in the Late Qing,” Modern Asian
Studies 49.3 (2015), 834.
461

elisabeth kö ll
excessive bureaucratic interference, and corruption. Recent research by

Wook Yoon has added valuable operational context from the consumers’
perspective by demonstrating how various technical problems and misguided
pricing strategies presented a major obstacle to turning the telegraph into
a popular method of communication for the general population. Providing
maintenance of the telegraph lines with line guards on patrol and constant
repairs due to theft or weather damage were expensive yet necessary to
guarantee the delivery of telegrams on time. As a result, the combination of
negligent staff at work and disrepair created constant delays of telegram
delivery. At the same time, the government considered a potential solution
via the construction of submarine cables too much of an expense, and
radiotelegraphy was only introduced in the early 1900s for naval military
purposes.8
Similar to the evolving telegraph network in Japan, in the beginning
China’s telegraph communication was intended for the use of the govern-
ment and military authorities, but not for private residents. To complicate
matters for potential users of the telegraph service, no standardized system of
phoneticization or romanization was available at the time when telegraphy
started to take off in China in the 1880s, thus making the process of sending
a telegram slower and more expensive than in other language environments.
Nonofficial users of the service were also charged high fees for the transmis-
sion of telegrams while in the West technological improvements made
telegram service increasingly less expensive in the late nineteenth century.
Only after nationalization and the merging of provincial telegraph adminis-
trations with the Telegraph Bureau in 1911 did pricing become fairer for all
customers and network maintenance improve.9
Despite these considerable obstacles, the advantage of telegraphy as
a communication technology spread quite quickly in the early twentieth
century when phone and telegraph communication began to assist Chinese
commercial ventures, businesses, and affluent urban residents. Before 1900,
especially foreign-trade businesses operating from the treaty ports benefited
from the telegraph system, which allowed for more efficient communica-
tions with their agents in the field. Chinese businessmen made use of the
telegraph for commercial purposes, such as gaining up-to-date information
8
Feuerwerker, China’s Early Industrialization, pp. 205–6; Yoon, “Dashed Expectations,”
841–3.
9
K. Yasar, Electrified Voices: How the Telephone, Phonograph, and Radio Shaped Modern Japan,
1868–1945 (New York, Columbia University Press, 2018), p. 38; Yoon, “Dashed
Expectations,” 835, 854.
462

on local and regional prices for agricultural commodities. However, whereas

scholars have argued in the case of Japan that “the telegraph was put to work
in the consolidation of the nation-state and the projection of Japanese
imperial power,”10 it is much harder to make a similar case for China. As
recent historical studies have shown, the telegraph as information technology
complicated China’s connections in global politics during the late Qing
because it contributed to a certain, limited degree of internal political inte-
gration and simultaneously to destabilization due to increased military
mobility.11 In terms of advancing China’s economic growth, the expansion
and use of the telegraph contributed to the proliferation of financial and
commercial centers such as Shanghai while also linking these centers more
efficiently to the rural and urban marketplaces in the interior provinces and
regions.12
After the founding of the Republic in 1911, the Ministry of Communications
did not pay much attention to the telegraph administration due to its limited
revenue contribution and a new focus on railroad development. According to
contractual agreements with foreign companies, the Telegraph Bureau
received only 13.5 percent of the telegraph tariff for outgoing cables. Even
when the ministry received a substantial loan from the three investing foreign
companies in 1911, secured by China’s future telegraph revenue until 1930, only
10 percent of the loan was used to improve the national telegraph network.
Reflecting the ministry’s new political and economic priorities, the rest was
channeled into subsidizing the railroad sector.13 After the Nationalist Party
(Guomindang) came to power under Chiang Kai-shek in 1928, the strongly
desired reclamation of cable rights failed because the government was unable
to pay back the loans it had contracted earlier with the foreign companies, and
any existing wireless services were too unreliable to replace the cable services
at the time. Only in 1933, after the Guomindang government was able to clear
its debt with the foreign investors, were new contracts with license agreements
issued to the foreign companies. The outbreak of the Second Sino-Japanese
War in 1937 brought new challenges to the telegraph system. Control over
10
Yasar, Electrified Voices, p. 40.
11
R. Thompson, “The Wire: Progress, Paradox and Disaster in the Strategic Network of
China, 1881–1901,” Frontiers of History in China 10.3 (2015), 395–427; S. Halsey, Quest for
Power: European Imperialism and the Making of Chinese Statecraft (Cambridge, MA,
12
Zhang Zhongli 張仲禮 (ed.), 東南沿海城市與中國近代化 (Eastern and Southern
Coastal Cities and China’s Modernization) (Shanghai, Shanghai renmin chubanshe,
1996), pp. 472–3.
13
Wei, “Circuits of Power,” 119–21. Banks tied to the Communications clique also
received subsidies from the loan.
463

elisabeth kö ll
information became a strategic goal of the Japanese occupation, with censors

posted in foreign cable companies to monitor telegrams to foreign destin-
ations. In 1942 Japan took control of all three major cable companies in China
and transferred many of their cables to South Asia to assist the Japanese war
effort in the Pacific theatre.14
Steamships and Railroads in a Semicolonial Context

Steamships as new transportation infrastructure started to expand their
network along China’s coast and rivers in the early 1860s. As a result of the
1842 Treaty of Nanking (Nanjing) concluding the First Opium War (1839–
1842), steamships sailing under foreign flags were permitted to participate in
China’s coastal trade and began to serve not only the purpose of import and
export trade but also the needs of Chinese merchants in domestic trade. The
opening of treaty ports after the First and Second Opium Wars provided the
foreign steamships with an expanding network of wharves, repair stations,
warehouses, and company offices along China’s coast and the Yangzi river.
Chinese traders and merchants valued the steam-powered speed of freight
transportation, armed protection of their freight in case of piracy, and
availability of cargo insurance, which Chinese junk transportation did not
offer at the time. Domestic junk boats continued their trade on canals and
smaller rivers, while foreign steamships began to dominate the Yangzi river
trade from the 1860s onward. The trade became so profitable for Europeans
that they specifically brought ships to the China coast to participate in the
“coasting trade” which was embraced by Chinese merchants. According to
agreements by treaty port commissioners, the foreign steamships paid
Maritime Customs duties irrespective of who owned the cargo of the ship.15
The Treaty of Tientsin (Tianjin), concluding the Second Opium War
(1856–1860), evolved into an effort to negotiate issues related to the Qing
empire’s sovereignty in terms of foreign shipping and trade in China. The
1861 regulations recognized officially the coasting trade and stipulated rules
for collecting custom fees. The Yangzi river was also declared open to foreign
shipping and trade, but the treaty ensured that foreign shipping companies
were not able to sail and do business beyond the scope of the treaty ports
along the Yangzi river. Similar in function to emerging major rail hubs, treaty
14
Wei, “Circuits of Power,” 127–33.
15
A. Reinhardt, Navigating Semi-colonialism: Shipping, Sovereignty, and Nation-Building in
China, 1860–1937 (Cambridge, MA, Harvard University Asia Center, 2018), pp. 23–5;
Halsey, Quest for Power, pp. 187–9.
464

ports now became hubs for foreign and domestic shipping businesses gener-
ating forward and backward economic linkages. According to Anne
Reinhardt’s multifaceted study of Yangzi river shipping, the expansion of
the steam network benefited foreign shippers, especially British and
American, but not necessarily foreign traders doing business in China’s
domestic market. Foreign merchants were losing business to their Chinese
competitors, who, due to a combination of steam transportation, better local
market information, and lower overhead costs, were able to sell goods at
lower prices.16
The founding of the China Merchants Steamship Navigation Company
(CMSNC) in 1872 signaled the beginning of government efforts to insert itself
into the shipping business in order to secure the empire’s economic and
political interests, in particular to ensure that inland water routes would not
be completely in the hands of foreign shipping companies. The CMSNC was
able to maintain a de facto government monopoly in Chinese steam naviga-
tion and transportation in the treaty port network until 1911 because the Qing
government allowed private Chinese merchants to set up steamship com-
panies only in the late 1890s. These companies did not enjoy the government
patronage and business advantages extended to the CMSNC, but official
encouragement at least fostered the development of small shipping firms in
Chinese ownership which became a vital part of the shipping traffic on
China’s inland waterways.17
With the cost advantage of water over land transportation, sailboat traffic
did not decline due to the increased steamboat traffic. In fact, throughout the
Republican period unmechanized boat traffic increased even at major ports
where steamships and motorboats offered their services. As Thomas Rawski
has argued, before World War I I “the history of water carriage is one of
growth rather than displacement, with the use of inherited modes expanding
along with the rise of new technology.”18 According to his estimate of limited
available data, the number of motorized vessels increased steadily from just
twenty-seven in 1885 to 517 in 1900, 1,559 in 1915, 2,734 in 1924, and 3,895 in 1935,
and in those years the amount of tonnage transported increased
accordingly.19 In contrast to the railroads, which suffered serious track and
16
Reinhardt, Navigating Semi-colonialism, pp. 26–7.
17
Reinhardt, Navigating Semi-colonialism, pp. 54–5.
18
1989), p. 189.
19
Rawski, Economic Growth, p. 190, Table 4.1. The number of vessels in 1935 excludes
Manchuria.
465

elisabeth kö ll
equipment damage due to reckless warlord activities in the 1910s and early
1920s, as well as during Chiang Kai-shek’s Northern Expedition in 1926–1927,
commercial shipping only stagnated when some ships were taken over for
military purposes, but quickly recovered afterwards. Especially the number
of smaller vessels with a tonnage of less than 100 tons increased in the prewar
period, reflecting the trend of more mechanized vessels serving smaller ports
in interior provinces.20
The semicolonial origins of China’s railroads bore both similarities and
differences to railroad construction in other parts of the world in the late
nineteenth and early twentieth centuries. Railroad construction in India
began in the 1850s under the British, and European contractors built railroads
under semicolonial conditions in the independent countries of Latin America,
Southern Europe, and the Middle East. China, however, was host to more
than half a dozen imperial powers competing to dominate the construction of
this new infrastructure. The fragmentation of China’s own central political
authority since the mid-nineteenth-century Taiping Rebellion (1850–1864)
further complicated the situation as it allowed for rivalries among powerful
provincial leaders and the European powers. In short, the founding era of
Chinese railroads was dominated by intra-imperial rivalry at the Qing court,
provincial officials in disputes with local authorities and court factions, and
tensions and misunderstandings among the colonial powers and their
Chinese hosts.21
Foreign companies had long been eager to lay tracks in China, but
official opposition kept them at bay for much of the late nineteenth
century. In 1863, twenty-seven foreign firms petitioned Li Hongzhang,
then acting governor of Jiangsu province, for the right to build a line
between the administrative town of Suzhou and Shanghai, the major
trading port in the lower Yangzi region, but Li denied the petition and
even refused to forward it to the emperor. Nothing happened until 1875,
when the British trading firm of Jardine, Matheson, without officially
clearing it with the authorities in advance, constructed a ten-mile track
between Shanghai and Wusong. As the British company’s prospectus
optimistically stated, “The disposition of the local authorities is favour-
able to the scheme, and there is good reason to believe that no objection
20
Rawski, Economic Growth, p. 191; E. Köll, Railroads and the Transformation of China
21
P.A. Kuhn, “The Taiping Rebellion,” in J.K. Fairbank (ed.), The Cambridge History of
China, vol. 10, Late Ch’ing, 1800–1911, part 1 (Cambridge, Cambridge University Press,
1978), pp. 264–350; Halsey, Quest for Power.
466

will be offered to it at Peking.”22 Due to the narrow gauge the required

rolling stock was relatively inexpensive, and the company hoped to
benefit from the already considerable traffic between Shanghai and the
Wusong port. In addition, the consortium also envisioned a successful
construction as an opportunity to showcase the many advantages of
railroads and their future potential for China’s extensive inland trade.23
The train ran for more than a year, but in 1877 the Chinese government
ultimately purchased and demolished the line, fearing that foreigners
would take control of this means of transportation for imperialist military
purposes. The railroad was later moved to Taiwan, but officials there
were never able to operate it due to a lack of funds.24
Whereas only 195 miles of track had been built in China proper by 1894, the
First Sino-Japanese War altered the political calculus in favor of railroad
development. China’s loss to Japan in 1895 not only triggered efforts to
promote indigenous industrialization but also provided a new impetus to
railroad construction by weakening the government’s bargaining position
vis-à-vis foreign companies that were seeking a foothold in China. Industrial
development suddenly became a priority for the government in order to
compete with Japan and other nations on equal commercial terms. Officials-
turned-entrepreneurs like Zhang Jian (1853–1926) became advocates of rail-
roads, characterizing them as the “root of industry,” resulting in commercial
and military prosperity and, by extension, social harmony.25
The construction of the trunk line from Tianjin to Pukou, connecting the
capital with Nanjing and by extension Shanghai, grew from this political
moment and exemplifies the complex process of negotiations within multi-
national rail syndicates and with the Chinese government. After the creation
of a German sphere of interest (centered at the treaty port of Qingdao) in
Shandong province in the late nineteenth century, Germany and Great
Britain obtained a joint concession to construct a railroad line between the
city of Tianjin and the northern bank of the Yangzi river in 1898 (see
Map 13.1). The plan was to establish a transportation and communication
artery linking Tianjin, in the north of China and easily accessible from
22
“Provisional Prospectus, Woosung Road Company, Limited,” Morrison Pamphlets
P-I I I-b, no. 854, Toyo Bunko Library, n.d., n.p. The railroad’s gauge was two feet six
inches.
23
“Provisional Prospectus.”
24
H.-C. Wang, “Merchants, Mandarins, and the Railway: Institutional Failure and the
Wusong Railway, 1874–1877,” International Journal of Asian Studies 12.1 (January 2015), 50.
25
E. Köll, From Cotton Mill to Business Empire: The Emergence of Regional Enterprises in
Modern China (Cambridge, MA, Harvard University Asia Center, 2003), pp. 31–51.
467

elisabeth kö ll
Manzhouli
Qiqihar
Harbin
Jilin
Changchun
Siping
Mukden
(Shenyang)
Dandong
Shanhaiguan
Beijing Dalian
Tianjin Lϋshun
Shijiazhuang
Zhengzhou
Hankou
Cities
Railroads
1900
Completed
Proposed
Railroads in the Northeast (Manchuria)
Rivers
Map 13.1 China’s railroads, c. 1900

Map © Elisabeth Köll (cartographic design by Matthew Sisk)
Beijing, to Nanjing in the south via Pukou, a location for crossing the Yangzi
river. The British pledged to build and operate the part of the line within
Jiangsu province, the southern section, and the Germans would be in charge
of the line passing through Shandong and Zhili provinces in the northern
section. In 1899 a British–German syndicate signed an agreement with the
468

Chinese government to build the railroad.26 It took several years to negotiate

the methods of financial control, construction, equipment, and operation of
the line. Although the Boxer Uprising in 1900 delayed the project, construc-
tion of the line proceeded relatively swiftly, and both sections were com-
pleted and connected in 1912.27
Similar to early industrial enterprises and other railroad lines in China,
financing the Tianjin–Pukou Railroad proved to be extremely difficult.28 As
in the case of British India, local owners of capital initially made the rational
choice to invest in more familiar ventures, thus preferring investments in
land resources and real estate over more risky railroad investments.29 It was
not until 1905, toward the end of the Qing dynasty, that domestic investment
in industry slowly increased in a changing business environment that saw the
introduction of limited liability in China’s first Company Law, chambers of
commerce, and the first modernization efforts in the banking sector. In 1908
the financial terms for the Tianjin–Pukou railroad were finally agreed upon:
the first loan arrangement of the so-called Imperial Chinese Government 5
Percent Tientsin–Pukow Railway Loan consisted of a bond issue of £5 million
on London’s financial market, issued in two installments, with 65 percent
German capital and 35 percent British capital.30
Although the loan was oversubscribed in Europe several times
(£1.89 million was raised by the Deutsch-Asiatische Bank in Germany for
the 401 miles of the German section and £1.1 million was raised in London for
the 235 miles of the British section), the market issue also attracted some
private Chinese investors and the provincial governments of Zhili, Shandong,
Jiangsu, and Anhui, which together invested £260,000.31 Interest was to be
paid from the profits of the railroad, and the loan was guaranteed by the
26
Mi Rucheng 密汝成 (ed.), 中華民國鐵路史資料, 1912–1949 (Materials on the Railroad
History of Republican China, 1912–1949) (Beijing, Zhongguo shehui wenxian chu-
banshe, 2002).
27
Lin Cheng, The Chinese Railways: A Historical Survey (Shanghai, China United Press,
1935); Jin Jiafeng 金家鳳, 中國交通之發展及其趨向 (Developments and Trends in
China’s Transportation) (Nanjing, Zhongzhong shuju, 1937).
28
Lü Weijun 呂偉俊 et al., 山東區域現代化研究: 1840–1949 (Research on the Regional
Modernization of Shandong: 1840–1949) (Jinan, Qilu shushe, 2002), pp. 54–9.
29
I.J. Kerr, Engines of Change: The Railroads That Made India (Westport, CT, Praeger, 2007),
p. 24.
30
Cheng, Chinese Railways, p. 82; P.H.B. Kent, Railway Enterprise in China: An Account of Its
Origin and Development (London, E. Arnold, 1907), pp. 152–3; R.W. Huenemann, The
Dragon and the Iron Horse: The Economics of Railroads in China (Cambridge, MA, Harvard
University Press, 1984), p. 120. Huenemann quotes a 5.85 percent true interest rate on
the 1908 loan.
31
E.-H. Lee, China’s Quest for Railway Autonomy 1904–1911 (Singapore, Singapore University
Press, 1977), Table 4; (Santo tetsudo kaisha 山東鐵道會社), 山東鐵道ニ關スル調查
469

elisabeth kö ll
central government and secured by the income from the lijin 厘金 (com-
mercial transit tax) and internal provincial revenue from the four provinces
through which the line ran. In contrast to previous foreign-loan arrange-
ments, the Tianjin–Pukou Railroad itself was not mortgaged as collateral.
These new financial arrangements became known as Pukow Terms in loan
negotiations between different investor groups. For the first time, construc-
tion of a line was undertaken by foreign partners, but overall managerial
control lay in the hands of the Chinese government, at least on paper.32
However, while foreigners eagerly subscribed to Chinese railroad bond
issues on overseas markets, the handling of domestic subscriptions turned
into a political issue and in the case of Sichuan province led to serious political
protests that eventually became part of the 1911 revolutionary movement. As
Elisabeth Kaske has argued, “the crisis of 1911 in Sichuan was caused by a very
peculiar system of tax appropriation for the Railway Company that became
known as ‘rent shares’ (租股 zugu) and made rural landowners into share-
holders.” Whereas most other provinces tried (rather unsuccessfully) to raise
funds for local railroad ventures through commercial taxes or by approaching
urban commercial elites, Sichuan’s method of raising funds through the rent
share tax was successful, but also meant that rural elites through their
investment became drawn into the political disputes between provincial
and national interests.33 When the central government announced the nation-
alization of all the trunk lines, including the planned Sichuan–Hankou
railroad in May of 1911, the landowner–shareholders protested against the
government’s plan to offer them national bonds as compensation and chal-
lenged the provincial authorities. As Kaske shows, the issue of railroad
finance through local investors became an issue of “political participation
for the rural elites” which could not be solved during the last months of the
Qing, nor during the Republic, so that plans for building the Sichuan railroad
never materialized in pre-1949 China.34
報告: 附錄津浦鐵道借款契約 (Research Report on the Shandong Railway Company:

Appendix with the JinPu Railway Loan Agreement) (Dairen, Manshu Nichinichi
Shinbunsha, Taishō 4 [1915]); Ghassan Moazzin, “Networks of Capital: German
Bankers and the Financial Internationalisation of China (1885–1919),” Ph.D. thesis,
Cambridge University, 2017.
32
(Santo tetsudo kaisha 山東鐵道會社), 山東鐵道ニ關スル調查報告, appendix on
Pukou terms in 1915. On the use of lijin in government finance, see the chapter by
Elisabeth Kaske and May-li Lin in this volume.
33
E. Kaske, “Sichuan as a Pivot: Provincial Politics and Gentry Power in Late Qing
Railway Projects in Southwestern China,” in U. Theobald and Cao Jin (eds.),
Southwest China in a Regional and Global Perspective (c. 1600–1911) (Leiden and Boston,
Brill, 2018), p. 383.
34
Kaske, “Sichuan as a Pivot,” pp. 402–4, esp. p. 402.
470

While these events were going on in China’s interior provinces, the

emerging railroad network in Manchuria and the activities of the South
Manchurian Railway (SMR) began to assume an important role in the
historical and political trajectory of Sino-Japanese relations in the early
twentieth century. Railroad development in China’s northeastern region of
Manchuria took place in the context of foreign imperialism, often also
characterized as railroad imperialism. Russia started the construction of the
old Chinese Eastern Railway between Manchouli and Suifenhe which was
completed in 1901, adding a southern section leading to Port Arthur (Lüshun)
near Dalian two years later. After their victory over Russia in the 1904–1905
war, the Japanese obtained the right to occupy the Guandong territory on the
southern tip of the Liaodong peninsula and exert control over the line from
Port Arthur to Changchun under the new name of the South Manchurian
Railway. In order to administer the railroad and the track right of way, the
SMR Company was founded in 1906 and came to act as both a business
enterprise and an agency for the interests of the Japanese government.35
The SMR became the economic and political backbone of Japan’s imperi-
alistic ambitions in Manchuria. Japanese funding led to increased railroad
construction in Manchuria throughout the 1920s with a new branch line
eventually linking Manchuria with Korea. Some Chinese financing found its
way into railroad construction at the end of the decade but did not create any
real competition for Japanese-led rail construction. During the so-called
Mukden Incident of September 1931, the Japanese militarist government
staged a pretext to send in troops and take control of Manchuria with the
formal creation of the puppet state of Manchukuo in the following spring.
Interruptions and Consolidation during the Early

Republic
The founding of the Republic in 1911 triggered the move to nationalize China’s
railroads, but the process was slow and incomplete due to the financial and
managerial contractual arrangements with foreign syndicate partners. Over
time, improved business legislation, government policies regarding railroad
35
B.A. Elleman and S. Kotkin (eds.), Manchurian Railways and the Opening of China: An
International History (Armonk, NY and London, M.E. Sharpe, 2010). On railroad imperi-
alism, see B.A. Elleman, E. Köll, and Y.T. Matsusaka, “Introduction,” in Elleman and
Kotkin, Manchurian Railways and the Opening of China, pp. 3–9; Xie Xueshi 解學詩 and
Matsumura Takao 宋村高夫, 滿鉄與中國勞工 (The South Manchurian Railway and
Chinese Labor) (Beijing, Shehui kexue wenxian chubanshe, 2003).
471

elisabeth kö ll
transportation, and access to financial markets increased indigenous Chinese

investment during the following two decades, whereas foreign investment,
except for that from Japan, began to decline due to World War I and China’s
fragmented internal political situation. Between 1895 and 1914, a total of 3,325
miles of new track were laid in China proper, and by the mid-1930s the total
mileage supervised by the Ministry of Railways had grown to about 8,200
miles.36
It is important to note that the term “nationalization” did not imply
a sudden, radical transfer of ownership and managerial control by the
Chinese government, expropriating all foreign rail investment. For example,
in 1922 about 25 percent of Chinese railroad lines were still owned and
operated by foreigners and another 25 percent was under foreign influence
due to continuing loan agreements.37 From the Chinese government’s per-
spective, the transition toward nationalization was meant to establish the
political and economic independence of a strategic sector and to secure
national interests against foreign control. However, at the company level
the transition toward nationalization was a slow and complicated process,
producing hybrid structures that only nominally were fully fledged Chinese
rail companies.38 In financial terms, the Chinese government continued to
service the foreign loans of the former joint syndicates, and it even raised new
bond issues on the public markets of London and New York. In managerial
terms, foreigners still held many senior positions in Chinese railroad com-
panies until the late 1920s, whether they were fully government-owned or
under Chinese managerial control, due to contractual commitments tied to
the loan contracts and the lack of an open labor market for both Chinese
engineers and highly skilled workers.39
However, in the midst of the frequent administrative reorganization
during the early Republican period, the most important bureaucratic struc-
ture, which has become the hallmark of China’s railroad system, emerged:
the Railroad Management Bureau (tielu guanliju 铁路管理局), commonly
referred to as the Railroad Bureau (tieluju 铁路局). Structurally, railroad
bureaus grew out of the lines’ former head offices, but during the post-1911
transition they came under the direct jurisdiction of the Ministry for
36
Cheng, Chinese Railways, pp. 63–5.
37
“Railroad Finances,” Journal of the Association of Chinese and American Engineers 3.5 (June
1922), 22–3.
38
I use the term “railroad companies” to refer to the post-1911 nationalized railroads as
enterprises managed and owned by the government.
39
472

Communication and Transportation. It is important to note that the Railroad

Bureau structure has survived with great resilience into the twenty-first
century and still functions as a blueprint for China’s current rail administra-
tion under the same name as it did some 100 years ago.40
The regional Railroad Bureau structure emerging as a formalized system
encouraged a considerable degree of autonomy during the institution-
building process of the railroad sector in the early Republic. However, this
system developed in the context of the line-specific financial arrangements
and the physical damage to tracks and rolling stock due to warlord battles at
a time when no operational synergies existed for either the larger region or
the nation as such. It is misleading to discuss a connecting rail “network” in
the early days of the Republic even though there were a considerable number
of individual lines. The first through-transport on the five largest railroads in
eastern and central China became possible in April 1914, but many other lines
were not connected until much later. By the early 1920s the sixteen railroad
bureaus covered the existing network of major lines, with headquarters in
eleven cities throughout the country. In sum, regional political fragmenta-
tion and severely limited network connectivity engendered a regionally
divided railroad system, thereby exacerbating political efforts to centralize
rail organization and administration.41
Freight Transportation and Economic Impact

Many historical debates about the socioeconomic effects of railroads involve
the issue of their potential contribution to growth through so-called back-
ward linkages – increased demand for labor, coal, steel, or financial services –
and forward linkages in the form of the economic effects due to lower
transportation costs for agriculture and industry. Macroeconomic
studies by Ernest Liang and Thomas Rawski have made convincing argu-
ments that cities and villages along certain railroad lines profited from the
increased transportation and communication options that stimulated the
40
Shandong sheng defang shizhi biancuan weiyuanhui, 山東省地方史志編篡委員會,
山東省志: 鐵路志 (Gazetteer of Shandong Province: Railroad Gazetteer) (Jinan,
Shandong renmin chubanshe, 1993), pp. 547–56. Detailed information on each of
China’s current railroad bureaus is available through the official railroad web portal
at www.tielu.cn, accessed September 13, 2020.
41
Köll, Railroads and the Transformation of China, pp. 59–60; Yang Yonggang 楊勇剛, 中國
近代鐵路史 (A History of Railroads in Modern China) (Shanghai, Shanghai shudian
chubanshe, 1997), p. 154; Wang Xiaohua 王曉華 and Li Zhancai 李佔才, 艱難延神的
民國鐵路 (The Difficult Expansion of Railroads during the Republic) (Zhengzhou,
Henan renmin chubanshe, 1993), pp. 76–7.
473

elisabeth kö ll
commercialization of agriculture during the Republican period. According to

such interpretations, lower transaction costs resulted in higher incomes for
farmers, facilitated the integration of their products into the national market
system, and even linked some of them to international markets. The inter-
play of specific local and regional economic conditions, however, limited the
stimulation and created pockets of economic development rather than
a national trend of economic growth tied to rail infrastructure during the
Republican period.42
Although railroads attracted a considerable number of passengers when
lines or sections of lines opened, long-distance freight transport usually
generated more revenue, especially in the case of major trunk lines such as
the Tianjin–Pukou and Beijing–Hankou railroads. Although we have only
fragmentary information on the volume and revenue from freight generated
by each individual railroad line in the 1910s, data collected by the Commission
for the Unification of Chinese Railway Statistics and Accounts, under the
aegis of the Ministry of Communications and Transportation, provide helpful
insights into the capacity and economic viability of freight transportation at
the time.43
Statistics published as official records beginning in 1915 indicate the pre-
dominance of freight transportation to generate revenue for the major trunk
lines across the north China plain and the interior. In the case of the Beijing–
Hankou and the Qingdao–Jinan lines, freight transportation constituted two-
thirds of their revenue. As for the Tianjin–Pukou line, 46 percent of its
transportation revenue came from goods and 39 percent from passengers in
1915. This ratio remained fairly stable until after the Guomindang took over
the government in Nanjing in 1927, when tourist services and passenger
express trains led to increased passenger revenue. As for product types, for
42
A. Fishlow, American Railroads and the Transformation of the Ante-bellum Economy
(Cambridge, MA, Harvard University Press, 1965); Fishlow, “Internal Transportation
in the Nineteenth and Early Twentieth Centuries,” in S.L. Engerman and R.E. Gallman
(eds.), The Cambridge Economic History of the United States, vol. 2 (Cambridge, Cambridge
University Press, 2008), pp. 543–642; D. Bogart and L. Chaudhary, “Engines of Growth:
The Productivity Advance of Indian Railways, 1874–1912,” Journal of Economic History
73.2 (June 2013), 339–70; E.P. Liang, China, Railways and Agricultural Development, 1875–
1935 (Chicago, The University of Chicago Press, 1982); Rawski, Economic Growth.
43
Ministry of Communications (subsequently Ministry of Railways), 中國鐵路統計
(Statistics of Government Railways) (Beijing [subsequently Nanjing], Tiedaobu tongjichu,
1915–1936). Because of the impact of the civil war on railroad operations, reports were
published with long delays. The volume with 1925 data was published in 1929, and the
volume with 1928 data appeared in 1933. From 1926 onward each volume contained
English and Chinese versions. Beginning in 1930, the name of the series changed slightly
to 中華國有鐵路會計統計匯編 (Statistics of Chinese National Railways).
474

many years the Tianjin–Pukou railroad was the most agricultural-

commodities-oriented trunk line, with the Shanghai–Nanjing line as the
extension to the Yangzi delta and the port of Shanghai. Mineral freight
(especially coal) accounted for an even larger proportion of the revenue in
the 1930s. By 1935, freight accounted for 51 percent of the total revenue
generated by the national railroad system. Agricultural products and manu-
factured goods were also the largest categories of freight in waterborne
transportation. As Thomas Rawski has demonstrated, unmechanized junk
traffic did not decline but expanded due to increased mechanized and
motorized boat traffic.44
Shorter railroad lines with continuing competition from water-based
transportation, such as the Shanghai–Nanjing and Shanghai–Hangzhou–
Ningbo Railroads, derived most of their revenue from passenger transporta-
tion. These lines began as and remained primarily passenger lines;
agricultural products and manufactured goods presented a relatively small
proportion of the freight. For example, freight revenue in 1910 amounted to
little more than 16 percent of the passenger revenue for the Shanghai–
Nanjing Railroad.45
Chinese railroads most notably offered rapid transportation of bulk agri-
cultural commodities to new markets beyond the local and even the regional
economies. Freight statistics reveal the content and direction of commodity
flows. The combination of line location, complementary infrastructure, pre-
existing agricultural patterns, and linkages to commercial hubs and ports
conditioned the economic function of each railroad. Peanuts, tobacco, cot-
ton, and silk were among the agricultural commodities that experienced
increased demand and supply because of railroad transportation. Silk produ-
cers were early adopters of rail transportation to carry their goods to markets,
especially along the Shanghai–Nanjing line, the conduit to the silk factories in
the Yangzi delta’s treaty ports and the hinterland. The producers of silkworm
cocoons valued the speedy delivery of their precious, time-sensitive cargo to
the spinning factories in Shanghai.46
44
Ministry of Communications, 中國鐵路統計, various years; Rawski, Economic Growth,
pp. 202–8.
45
Ministry of Communications, 中國鐵路統計, various years.
46
“In Praise of the Peanut,” North-China Herald, January 13, 1911, 85; “Taianfu [Shantung],”
North-China Herald, May 20, 1911, 484; Amano Motonosuke 天野元之助, 山東省經濟
調查 (Economic Survey of Shandong Province), vol. 3, 山東農業經濟論 (Report on
Shandong’s Agricultural Economy) (Dalian: Minami Manshu tetsudo kabushiki kaisha,
1936), p. 125; “The Cocoon Season,” North-China Herald, June 24, 1910, 761.
475

elisabeth kö ll
Railroads also proved to be crucial for the business in sheep, goat, and
pig intestines. Tianjin became China’s leading collecting and exporting
center for intestines, which were transported by the Zhengding–Taiyuan
and Long–Hai railroads from the interior Sha’anxi, Henan, and Shanxi
provinces. Because the preserved intestines needed to retain some mois-
ture during transit, these commodities benefited from speedy rail trans-
port. Once they reached Tianjin, merchants specializing in intestines
further processed and repacked the consignment and shipped it to foreign
markets, with the United States as the largest consumer of Chinese animal
intestines in the early twentieth century.47
The Beijing–Hankou line presents an example of a railroad achieving
high revenue through the freight transport of mineral and agricultural
products. Output from the coal mines close to the station of Zhoukoudian
in the form of big and small coal blocks and hard coal dust was transported
by rail to markets in Beijing, Tianjin, and Dingzhou. Fifteen other mineral
products from lime to iron sand complemented a list of twenty-six agri-
cultural commodities ranging from wheat and beans, collected near the
northern end of the line, to rice, other grains, and cotton, moved from the
regions of the southern end of the line. Foodstuffs ranging from oil,
peaches, and water melons to eggs and yams, plus medicinal drugs,
tobacco, tea, hides, wool, and animal bones, as well as fertilizers and
different types of wood, were transported every year, plus a variety of
handicraft manufactured goods ranging from cotton cloth to paper and
earthenware. As the author of a 1929 review article noted, in order to
increase the crops of the farming regions along the Beijing–Hankou line,
the building of various branch railroads to the main line would be
a foremost concern “so that the transportation of goods from the place
of production to the place of consumption may become easier and
speedier.”48
For personal consumption, railroads transported larger volumes of basic
goods, such as milled flour, salt, sugar, kerosene, matches, cigarettes, and
fertilizer to the predominantly rural populations along the lines. Foreign
imports carried by rail included cotton goods and Japanese sewing cotton
(thread), Indian and Japanese cotton yarn, Japanese matches, and dried and
salted fish, as well as American kerosene. Railroads also imported more
luxurious items, such as dresses, furniture, foreign candles, glassware,
47
“Intestines for Foreign Markets,” Chinese Economic Journal 5.53 (1929), 811–16.
48
K. Loh, “Products along the Peiping–Hankow Railway,” Chinese Economic Journal 5.44
(1929), 328.
476

porcelain, and tea leaves in smaller quantities to meet the needs and tastes of
a small affluent population of consumers.49
The economic potential of railroads to stimulate production and trade in
agricultural commodities was recognized by Chinese farmers and actively
promoted by companies whose business was based on the processing of raw
agricultural products. For example, the British American Tobacco Company
(BAT) introduced and promoted cultivation of American tobacco seed along
the railroad lines in Shandong province. Under the direction of its capable
Chinese agent, a former employee of the Jiaozhou–Jinan railroad, BAT built
facilities for curing tobacco leaf and imported machinery of the highest
technological standards at six railroad stations. To encourage tobacco grow-
ing during the early 1910s, BAT extended lucrative offers to farmers, ranging
from free tobacco seed to loans of specialized equipment, credit access, and
on-the-spot cash payments for entire crops. As Sherman Cochran has shown,
these arrangements were so successful for BAT, expanding Shandong
tobacco acreage from thirty-nine acres in 1913 to 23,000 acres in 1920, that
the company saw no necessity to continue the policies afterwards.50
Coping with freight and its seasonal demands added to the existing
operational and managerial rolling-stock difficulties experienced by almost
all Chinese railroads. As foreign railroad-accounting advisers to the Chinese
government pointed out, contrary to the custom in the United States, freight
cars were not assigned by a central distribution office in the line’s manage-
ment but rather by the many dozens of individual stationmasters. The
assignment of freight cars based on local needs and particularistic interests
without sufficient attention to planning for seasonal demands for special
shipments became a negative aspect of this highly localized, stationmaster-
centric organizational model. In addition to bribery and the chronic shortage
of wagons for goods, the lack of efficient connections between lines became
yet another material constraint on freight traffic that contributed to
a predominantly local orientation.51
49
“Trade in South China: The Crisis during 1920,” North-China Herald, July 30, 1921, 354;
Qingdao shi gangwuju 青島市港務局 (ed.), 中華民國 20 年港務統計年報
(Yearbook of the Port Statistics for the 20th Year of the Republic [1931]) (Qingdao,
1931), pp. 286–9.
50
S. Cochran, Encountering Chinese Networks: Western, Japanese, and Chinese Corporations in
China, 1880–1937 (Berkeley, University of California Press, 2000), pp. 64–5; Liang, China,
Railways, pp. 116–17.
51
J.F. Baker, “Comparison of Chinese and American Railway Practices,” in J.H. Arnold,
Commercial Handbook of China (Washington, DC, Government Printing Office, 1926), pp.
127–8.
477

elisabeth kö ll
In this context railroad companies indirectly initiated the evolution of an

entire new group of logistics enterprises, the so-called transportation com-
panies (yunzhuan gongsi 运转公司). These logistics companies played an
important role as partners in transshipping regional goods and fulfilling
business functions that railroad companies were unable to manage on their
own, such as freight insurance and guaranteed deliveries. Transportation
companies quickly emerged to fill this niche, and by the mid-1910s this
logistical practice was relatively standardized. In the case of the Tianjin–
Pukou railroad, the transportation companies negotiated special agreements
whereby the line gave them exclusive lower rates on freight in return for
shouldering the burden of liability. The registration of transportation com-
panies with the Tianjin–Pukou Railroad Bureau and the rail administration
required the disclosure of company information similar to the registration of
businesses required by the 1904 Company Law. Specifically, the transporta-
tion companies had to be Chinese-owned. That Chinese transportation
companies in their agreements with individual railroad lines held a de facto
monopoly over rail shipping business drew criticism, especially from envious
foreign competitors.52
One problem every railroad line had to face was the ad valorem transit tax
or lijin, which had been collected for goods since the late nineteenth century,
most commonly at 2 percent, along the transit routes and sometimes also as
a production tax, similar to a sales tax, at the point of origin or at the final
destination. In the early years of railroad construction, the lijin was often
blamed for the relatively low utilization of railroads for freight transport
because merchants favored transport by other means to avoid the taxation.
The general manager of the Shanghai–Nanjing line complained in 1910 that
railroads lost out on business because merchants using the line had to pay the
“legitimate” lijin at the full amount, but only paid 30 percent of the author-
ized figure when sending their goods by water. Naturally, railroads tended to
lose business to rival boat traffic.53
In light of high shipping costs, including taxation, smuggling activities
along the railroad lines increased. Smuggling activities were not only actions
52
Köll, Railroads and the Transformation of China, pp. 101–11; letter, Dechamfils to Sir John
Jordan, January 17, 1919, Foreign Office archives, FO 228/2803, British National
Archives.
53
A. Feuerwerker, “Economic Trends, 1912–1949,” in J.K. Fairbank (ed.), The Cambridge
History of China, vol. 12, Republican China 1912–1949, part 1 (Cambridge, Cambridge
University Press, 1983), pp. 61–2; “The Shanghai–Nanking Railway,” North-China
Herald, January 14, 1910, 67. Rawski also cites similar incidents in Economic Growth,
pp. 232–3.
478

orchestrated to transport “illegal goods,” but one of the foremost methods of

tax avoidance integrated into the operational structure of the railroad lines
and their workforce. For example, railroad workers carried goods on their
own in order to avoid the taxes and freight charges.54 The government
supported efforts to reduce the extraneous costs on merchants shipping
goods via rail by convening the Peking Railroad Conference “to restore the
sadly abused but very great earning power of the Chinese Railways” a few
months before the Nationalists came to power in 1928. This move reflected
the railroad administration’s desire to restore state control over freight traffic
and its revenue but also to run the railroad lines more like for-profit business
entities in order “to bring back that public confidence and economy of
management which the Government Railways used to enjoy only a few
years ago both at home and abroad.”55
In the case of Manchuria, railroads provided access to the interior for
commercial farming, mining, and metallurgy enterprises. Throughout the
1930s the combination of railroads, intensive mining, and ironworks turned
Manchuria into China’s center for heavy industry, with a continued legacy in
the post-1949 period. According to propagandist Japanese sources, the heavy-
handed economic influence of Japan resulted in astonishing growth rates in
Manchuria: between 1908 and 1930, railroad construction rose by 84 percent
and soybean cultivation by 290 percent, with high growth rates in bean-cake
production and coal output from the Fushun mines and an overall increase in
foreign trade of 510 percent. In 1934, the Soviet Union succumbed to Japanese
pressure and sold the northern half of the Chinese Eastern Railway up to the
Soviet border to Japan. As a result, by 1935 all railroads in Manchuria came
under the unified management of the SMR, which operated the government
lines and SMR railroads as well as most bus lines and river transportation. In
1936, just a year before Japan’s invasion of China, the rail network in
Manchuria comprised 5,530 miles.56
54
“Opium on Railways,” North-China Herald, January 11, 1919, 70. For a comprehensive
treatment of smuggling, see P. Thai, China’s War on Smuggling: Law, Economic Life, and
the Making of the Modern State, 1842–1965 (New York, Columbia University Press, 2018).
55
“The Peking Railway Conference: List of Reforms Necessary for Efficiency,” North-
China Herald, August 13, 1927, 272.
56
Economic Conditions in Manchoukuo (Manchuria Daily News, 1940), pp. 3–4, 8; Minami
Manshū Tetsudō Kabushiki Kaisha 南滿洲鐡道株式會社, 南滿洲鐡道株式會社十
年史 (A Ten Years’ History of the South Manchurian Railway Company) (Dalian,
Minami Manshū Tetsudō Kabushiki Kaisha, 1919); Department of Foreign Affairs,
Manchoukuo Government, General Survey of Conditions in Manchoukuo (Hsinking,
1936), pp. 47–9.
479

elisabeth kö ll
With its territorial expansion and socioeconomic control backed by Japan’s

imperialist agenda, the SMR increasingly grew into a bureaucratic and
strategic enterprise that pursued reorganization with the aim of rationalizing
its railroad management and creating economic efficiencies. For example,
motor roads were considered important for the defense and industrial devel-
opment of Japanese interests and thus their management was transferred to
the SMR by the Manchukuo government. An employees’ consumption guild
was established in 1919 to provide SMR employees with access to daily
necessities at special prices. Despite Japan’s tight control over its political
and economic interests, the majority of Japanese merchants operated their
businesses directly from within the Guandong leased territory and the
heavily protected SMR rail corridor. Issues connected to personal safety,
a volatile currency situation, and relatively high operating expenses made it
preferable for Japanese merchants to stay inside the SMR zone. British,
American, and other foreign businesses were able to monopolize the petrol-
eum, tobacco, and sugar markets in Manchuria.57 However, even Chinese
merchants were attracted by the safety of the SMR zone and endured the
presence of Japanese security forces, however heavy-handed. For example, in
1920 103,000 Chinese lived and worked in the SMR zone together with 72,000
Japanese.58
As with rail lines in other parts of China, the logistics infrastructure
accompanying the tracks became equally important to sustaining the eco-
nomic impact of the rail network. In the case of the SMR, a complex system
of warehouses for soybeans, beancake, and soybean oil supported the boom
in commercial agriculture of soy and its by-products by securing efficient
onward transport of the commodities. In twelve major cities the SMR also
opened and managed bonded warehouses which allowed shippers to avoid
the inconvenience of clearing customs with their freight in the port cities
before international export.59
Passenger Transportation and Mobility

The significance of rail travel in Republican China lies less in the size, scope,
and nature of the business than in the ways it contributed to the transform-
ation of individual practices and social interactions in the public space
57
Department of Foreign Affairs, Manchoukuo Government, General Survey, p. 52.
58
Economic Conditions in Manchoukuo, p. 56; A. Kinnosuke, Manchuria: A Survey
(New York, R.M. McBride & Co., 1925), p. 86.
59
Economic Conditions in Manchoukuo, pp. 59–61.
480

associated with the railroad. Passenger services on railroads involved

a new and decidedly modern mode of engaging with the social world,
one that promoted values and behavior radically different from those
cultivated by the traditional social order. This is not to say that the
railroads dramatically changed Chinese mobility. For one thing, Chinese
had long made use of water transport along canals, rivers, and coastal
routes as well as of road transport on foot, horses, bullock carts, and other
vehicles. The high cost of rail travel relative to purchasing power, the
short average length of a rail journey, and the relatively small proportion
of the population riding the rails suggest that the advent of railroads did
not significantly shift existing mobility paradigms and practices.
Following the pattern set by freight transport, railroads assumed a new
and significant socioeconomic role for passengers via their integration
into the already existing or the newly evolving networks of transportation
(see Map 13.2).60
Passenger ticket sales were an important source of revenue for all the
nation’s railroad lines. In each year between 1915 and 1935, passenger traffic
accounted for 32 to 41 percent of revenue for all Chinese government
railroads. Although major trunk lines such as the Beijing–Hankou and the
Beijing–Mukden lines derived most of their revenue from freight transporta-
tion, they still transported a considerable number of passengers. For example,
70 percent of the revenue of the Beijing–Hankou line in 1935 came from
freight; yet in the same year the railroad also transported a total of more than
3.4 million passengers.61
For some lines, passenger traffic was the major source of revenue. Short
lines in the Yangzi and Pearl river deltas – the Canton–Kowloon, Shanghai–
Nanjing, and Shanghai–Hangzhou–Ningbo lines – connected passengers
from the densely populated hinterland to the commercial centers of
Hong Kong or Shanghai. In 1918, the Shanghai–Nanjing Railroad derived
71 percent of its revenue from passenger traffic; this figure decreased slightly
to 64 percent in 1923, rose to 72 percent in 1930, and then declined to 66 percent
in 1935, the last year of available data. The Shanghai–Hangzhou–Ningbo
Railroad followed a similar pattern, with approximately 65 percent of its
revenue originating from passenger traffic; passenger-related revenue on the
Canton–Kowloon line remained almost constant at 85 percent throughout
the 1920s and 1930s.62
60
61
Figures extrapolated from Ministry of Communications, 中國鐵路統計, various years.
62
481

elisabeth kö ll
Heihe
Manzhouli Bei’an
Jiamusi
Qiqihar
Anshan Suihua
Harbin Suifenhe
Mudanjiang
Taonan Jilin Tumen
Changchun
Siping
Mukden
(Shenyang)
Zhangjiakou Dandong
(Kalgan)Shanhaiguan
Hohhot
Beijing Dalian
Baotou
Suzhou Datogn
Datong Tanjin
i Lüshun
Shijiazhuang
Taiyuan Qingdao
Ji'nan
Lianyungang
Zhengzhou
Xuzhou
Kaifeng
Louyang Pukou
Baoji Xi'an Shanghai
Nanjing
Hangzhou Ningbo
Wuchang
Hankou
Jiujiang
Jiangshan
Jiangshan
Changsha
Nanchang
Zhuzhou
Guiyang
Shaoguan
Guangzhou
Kunming (Canton)
Sanshui Kowloon
Cities
Hekou
(Laojie)
Railroads
1935
Completed
Proposed
Railroads in the Northeast (Manchuria)
Map 13.2 China’s railroad network, c. 1935

Map © Elisabeth Köll (cartographic design by Matthew Sisk)
System-wide, passenger traffic underwent a substantial and steady increase

in volume on all national lines during the Republican era, although it
fluctuated considerably from year to year in response to adverse economic
conditions and political upheavals. According to data extrapolated from the
annual statistics collected by the Ministry of Communications, a total of
482

12.6 million passenger journeys were taken on Chinese national railroads in

1915, the first year of recorded data. This number quickly doubled to
24.6 million in 1918 and then rose to almost 40 million in 1923. During the
crucial civil war years from 1926 to 1928, the statistics include information
only on revenue figures, not on the volume of passengers. This hiatus might
be explained by the fact that during those years railroad companies were
forced to transport military personnel and soldiers without compensation.
With the consolidation of political control under the Nanjing government
after 1927, passenger travel quickly recovered. In 1932, there were 34 million
passenger journeys in total, and this number increased to almost 44 million in
1935, just before the 1937 Japanese invasion and occupation interrupted the
railroad system as well as data collection for several years.63
Although revenue from passenger transportation on most lines, especially
the major trunk lines, was secondary to revenue from freight, its distribution
among the different categories of passengers is quite surprising. Annual
statistics indicate that ticket sales for third- and fourth-class services were
by far the most important revenue stream for the passenger business of the
railroad companies. This pattern applied to traffic on all long-distance trunk
lines, including the Tianjin–Pukou and Beijing–Hankou railroads. Fourth-
class travel was no longer offered after 1932, with the exception of two short
lines connecting Shanghai with the capital in Nanjing and Hangzhou on the
West Lake, and Jinan with Qingdao. In these cases, fourth-class travel was the
preferred choice for regional commuters. With the move of the capital from
Beijing to Nanjing after 1927, the number of first-class tickets increased as
government officials and businessmen began commuting back and forth
between the two hubs. Even so, first-class travel was never financially or
numerically significant. The statistics make it clear that a long train journey
from terminal to terminal was a first-class affair for only a limited number of
passengers who had disposable income, and it never became a major revenue
source for the Chinese railroads.64
If we consider transportation companies as forward linkages resulting
from the freight business created by Chinese railroads, the establishment of
new steam-powered launch companies represented forward linkages result-
ing from the need for additional passenger transportation services. For
example, soon after completion of the Shanghai–Hangzhou railroad a new
launch company named Dexin formed a business co-operation with the line
63
Passenger journey data were based on tickets sold and included repeat journeys.
64
483

elisabeth kö ll
to facilitate passenger transport. According to their agreement, both com-

panies sold through-tickets for journeys that combined launch and rail travel
from Suzhou to Hangzhou, including free through-baggage service. The
Dexin launch took passengers on the six-and-a-half-hour boat trip via the
Grand Canal from Suzhou to Jiaxing, where they could catch the fast train
from Shanghai at 3 p.m. and arrive two hours later in Hangzhou. Commercial
launch services also ferried railroad passengers across major rivers lacking
bridges.65
Launch services to railroad terminals could be such a lucrative business
that it led occasionally to price wars between competing launch companies.
After the southern section of the Canton–Hankou railroad reached the North
river (Beijiang) in 1911, passengers continued their trip via steam launches to
Yingde and from there via a “shallow-draught stern-wheeler” to Shaozhou,
thus reducing the travel time from fourteen days to only two days. Two rival
companies competed for passengers between Yingde and the railroad head in
Guangzhou, leading to a price war and even a fist fight between the two
launch crews.66
Sometimes steam launches and railroads were competitive rather than
complementary means of transportation. In Shanghai, for example, the
Shanghai–Hangzhou Railroad attracted disapproval in 1911 for locating its
terminal too far from the center of trade and the Huangpu river. As a result,
the steam launch companies retained their edge and, on the basis of their low
prices, were still a competitor to the railroad line in 1918. Throughout the
1910s and the 1920s new launches and ferry services in their function as
complementary transportation businesses brought passengers to railroad
stations along sparse trunk line systems, often far from final passenger and
freight destinations.67
Motor Roads and Aviation

Modern roads and highways became a lynchpin in China’s infrastructure
development during the Republican period. Serious construction began in
the early years of the Republic when foreign-trained engineers advising the
65
“Soochow: A New Line to Hangchow,” North-China Herald, June 10, 1910, 618.
66
“Shiuchow [Kwangtung]: Railway Matters,” North-China Herald, May 20, 1911, 483–4.
67
“The Provincial Railways: Hangchow,” North-China Herald, September 9, 1911, 660; “The
Trade of Shanghai,” North-China Herald, August 31, 1918, 532–4; “Soochow: A New Line to
Hangchow”; “The Trade of Shanghai,” North-China Herald, August 31, 1918, 532–4.
484

Beijing Ministry of Communications lobbied for modern feeder roads as

a necessity for new railroads in operation to generate positive returns.
Although a presidential mandate with regulations for new road construction
was announced in 1919, it did not result in a concerted government effort of
creating a national highway system. Instead, Chinese and international
welfare organizations and provincial interests became the driving force
behind modern road construction.
In the treaty ports, public-works departments of the foreign concessions
undertook the construction of new, modern roads in the city. In the coun-
tryside, some road-building programs were carried out in the context of the
1920 famine to provide stricken areas with an economic stimulus through
employment options. For example, the government announced that it would
employ about a million famine victims in north China for the construction of
roads between Beijing and Shandong, Henan, and Zhili provinces at an
estimated cost of 200,000 dollars, with funds raised by additional salt and
stamp taxes and increased postal and telegraphic charges.68 Between 1923 and
1928, the China International Famine Relief Commission became strongly
involved in road-building programs, funding the construction of dykes and
roads. In addition, some provinces, with encouragement from the central
government, started independent road construction programs under the
supervision of Chinese engineers.69
Similar to the push for administrative reorganization of the rail network,
modern road construction gained serious attention as a project of national
economic and political importance once the Nationalist government under
Chiang Kai-shek took control in 1928. In the following year, the National
Highway Planning Commission, initiated and supervised by the new
Ministry of Railways, began its work and drew up projection and construc-
tion plans as well as regulations for the raising of funds through appropri-
ations from government revenue and proceeds from bond issues and special
road debentures. The regulations also specified the conscription of private
labor for national highway construction, which put a considerable burden on
the male population between the ages of eighteen and fifty residing in the
rural areas touched by the proposed highways. Most of the modern road
construction did not involve macadam or gravel but consisted of simple earth
construction. Railroad dump cars and steam rollers were the only industrial
68
“Road Making for Victims,” “New Railways for Famine Relief,” and “The Great
Famine in China: Cheap Transport of Grain,” North-China Herald, October 2, 1920, p. 22,
69
V. Smith with A. Chuh, Motor Roads in China (Washington, DC, United States
Government Printing Office, 1931), pp. 7–9.
485

elisabeth kö ll
equipment, although most of the time iron hand rollers filled with cement
and pulled by workers were the more common construction tools.70
Figures for the expansion of road mileage during the Republican period
vary, but according to a reliable source from 1931, China had about 40,000
miles of roads suitable for motor traffic at the time, including caravan routes
in Manchuria and Mongolia.71 Passenger cars amounted only to a small
number, with about 40,000 motor vehicles in all of China in the early 1930s
and one-third of the vehicles operating in Shanghai’s foreign concession. As
main users of these newly built roads, the increasing importance of bus lines
as infrastructure connecting rail stations and river docks with rural and urban
locations became evident. According to annual reports published by the
Ministry of Railways, by the early 1930s each major railroad was served by
a number of bus lines serving stations with a specific local and regional
coverage, with fixed schedules and ticket prices. The demand for bus trans-
portation of passengers and their goods in commercially vibrant market
areas along railroads such as the Shanghai–Hangzhou–Ningbo line was so
high that private motor trucks supplemented the bus system and, under the
eyes of the local authorities, offered freight transport to individual destin-
ations for hire.72
As in the case of logistics companies serving the railroads, provincial
governments often preferred to operate or grant franchises to private bus
companies for the exclusive use of certain roads, with the result that
government officials and bus lines became the main users. However,
these franchise arrangements often suffered from the problem that pro-
vincial road authorities considered the bus service a government monop-
oly and thus favored government bus lines over private companies. Even
if private lines came to a franchise agreement with a provincial govern-
ment, the conditions were often extremely harsh, including the require-
ment that private bus lines had to maintain and repair the roads on which
they operated. Similar to the relationship between railroad bureaus and
private logistics companies, the heavy-handed approach of the govern-
ment authorities vis-à-vis private companies was clearly weighted in favor
70
Smith and Chuh, Motor Roads, pp. 11, 13–14, 21–3.
71
Smith and Chuh, Motor Roads, p. 21. Thomas Rawski provides a fairly similar figure of
41,000 miles for 1931 in Economic Growth, p. 214, Table 4.8.
72
Tiedaobu nianjian bianzuan weiyuanhui 鐵道部年鑒編纂委員會, 鐵道年鑑
(Railroad Annual Statistics) (Nanjing, Tiedaobu tiedao nianjian bianzuan weiyuanhui,
1933–1936), vol. 3, 1936, pp. 1737–1750, 1744. Around 1930, Shanghai had 8,533 passenger
cars, 174 buses, 1,731 trucks, and 852 motorcycles. Smith and Chuh, Motor Roads, p. 81.
486

of government control and did not exactly encourage private business

competition.73
Compared to rail and shipping infrastructure, the development of civil
aviation in China before the beginning of the Japanese occupation in 1937 was
modest, with only thirty airplanes covering air routes of about 8,600 miles
linking major cities, especially in the eastern and southern part of China.
Recruiting and training of Chinese pilots in sufficient numbers was
a considerable challenge for the airline companies in the early 1930s, and
airlines had to readjust passenger and freight charges in order to make air
transport a viable transportation option, even for well-off customers. Similar
to railroads and shipping, commercial aviation included foreign joint ven-
tures in the beginning. In 1929 the China National Aviation Corporation
(CNAC) was founded as a Sino-American joint enterprise for passenger and
mail services in China. After reorganization, 55 percent of the capitalization
was subscribed by the Chinese government and 45 percent by the American
partner. The board consisted of nine directors, with five members appointed
by the Ministry of Communications and four by the American shareholders.
In the early 1930s three routes connecting Shanghai to Beijing, Chengdu, and
Guangzhou were the major flights, and a daily service from Shanghai to
Beijing was started in 1934.74
Passenger numbers started out small but increased relatively quickly,
although on a small scale. CNAC planes carried a total of 220 passengers in
1929, rising to 12,800 passengers in 1937. Despite the outbreak of the war,
passenger and mail transportation continued, albeit within different territor-
ial boundaries. When Chiang Kai-shek’s government moved to Chongqing as
the capital of Free China in 1938, the flight network shifted its boundaries and
began to concentrate on connecting major cities in China’s southwest,
southeast, and northwest, which were still in unoccupied territory and
accessible by air. As statistics show, throughout the war years CNAC had
rising passenger numbers, especially once government officials and refugees
moved to Chongqing.75
The Central Air Transport Corporation (CATC) was another case of Sino-
foreign collaboration. Originally founded in 1931 as the Eurasia Aviation
Corporation with the Ministry of Communications and the German
73
Smith and Chuh, Motor Roads, pp. 23–4, 113.
74
Chinese Ministry of Information (comp.), China Handbook, 1937–1945 (New York, The
Macmillan Company, 1945), pp. 233–6.
75
Chinese Ministry of Information, China Handbook, pp. 233–6; China Handbook Editorial
Board (comp.), China Handbook, 1950 (New York, Rockport Press, 1950), pp. 626–32.
487

elisabeth kö ll
Lufthansa company as partners, the enterprise became wholly Chinese-

owned after China cut its relations with Germany as a member of the
war’s Axis powers in 1941. As a government line, CATC had a much smaller
footprint, with a small number of planes, passengers, and limited service
routes focusing on Chongqing and Kunming as wartime capitals in Free
China.76
Postwar Reconstruction and Rail Expansion

In the decade following World War I I, the world’s railroad networks fell on
either side of a distinct and growing gulf. Railroads were a mature or even
declining technology in nations with democratic political systems and rela-
tively free market economies. Western Europe did not see significant add-
itions to track mileage after the war. The US rail network actually shrank by
an average of 600 miles per year between 1945 and 1954. In contrast, railroads
flourished in countries under Communist leadership. The USSR added an
average of 534 miles per year in the decade after World War I I, and post-
revolutionary Cuba boasted more railroad track per square mile in the 1960s
than any other country in the world. Throughout the Cold War era, railroads
thrived in socialist states with centralized economic planning. After the
founding of the People’s Republic of China on October 1, 1949, the adminis-
trative organization and management of the railroad system were integrated
into a core part of the new government’s political and economic agendas. As
physical entities and means of transportation, all existing railroad lines
became part of the People’s Railroad (renmin tielu 人民鐵路) system.77
At the end of the Japanese occupation and World War I I in China, about
14,500 miles of the rail network were serviceable, but tracks, rolling stock, and
maintenance facilities had suffered considerable damage. The damage due to
civil war fighting led to a further deterioration of the network so that by early
1948 only about 5,000 miles of track were operational. Especially in the south
of China, the rail network was damaged or destroyed. On the whole, the lines
within the rail network of Manchuria suffered less damage than the majority
of the railroads in China proper, partly because there the fighting between
Japanese and Russians and between Chinese Communists and Nationalists
76
Chinese Ministry of Information, China Handbook, pp. 236–8; Chia-hua Chu, The
Ministry of Communications in 1934 (Shanghai, China United Press, 1935), pp. 14–15.
77
J.N. Westwood, “Soviet Railway Development,” Soviet Studies 11.1 (July 1959), 33;
Ivan Wiesel, “Cuban Economy after the Revolution,” Acta Oeconomica 3.2 (1968),
203–20.
488

was “more highly localized than elsewhere.” Thus Manchuria’s network of

standard-gauge railroads became operational long before the remaining lines
recovered their service.78
In China’s First Five-Year Plan (1953–1957) railroads were designated to
support the country’s industrial development. Whereas the expansion and
renewal of the railroad network was not an explicit focus, the plan projected
2,500 miles of new railroad tracks as well as rail reconstruction work, includ-
ing modifying some single-track lines to become double-track. By the end of
the First Five-Year Plan, total operating mileage had increased by approxi-
mately 2,400 miles, of which 500 miles consisted of new double- or multitrack
sections. Electrification of the Chinese rail system did not take place until
1962, with a meager sixty miles of electrified lines. This situation did not
improve significantly until the beginning of the economic reforms in 1978.79
Executing huge, ambitious rail construction projects in remote areas of the
country required a large, disciplined, and dedicated workforce willing to
operate under often inhospitable and even dangerous conditions.
Throughout the 1950s the railroad system had become home for PLA
members demobilized after World War I I and the Korean War. To expedite
strategic rail infrastructure projects, in 1953 the Central Army Committee
decided to establish the Railroad Army Corps (tiedaobing 铁道兵) as
a centralized military vanguard for constructing new railroad lines in south,
central, and western China. The organizational origins of the tiedaobing can
be found in the “railroad columns” (tiedao zongdui 铁道总队) of soldiers. The
PLA mobilized these soldiers to repair sections of damaged railroad tracks in
the northeast that were of strategic importance to the advance of the
Communist troops in 1948. Renamed the Railroad Corps (tiedao bingtuan
铁道兵团) after the end of the civil war in 1949, they continued to repair and
construct railroads until they were officially integrated into the PLA as
Chinese People’s Liberation Army railroad soldiers (Zhongguo renmin jiefang-
jun tiedaobing 中國人民解放軍鐵道兵). By combining centrally directed
construction and a military presence, this strategic approach enabled the
78
N.S. Ginsburg, “China’s Railroad Network,” Geographical Review 41.3 (July 1951), 470–4,
esp. 470. On wartime damage during the Japanese occupation, see Shanghai shi
dang’anguan 上海市檔案館 (ed.), 日本在華中經濟掠奪史料, 1937–1945 (Historical
Materials on Japan’s Economic Plunder of Central China, 1937–1945) (Shanghai,
Shanghai shudian chubanshe, 2005).
79
Zhongguo tiedao xuehui 中國鐵道學會, 中國鐵路, 1949–2001 (Chinese Railroads,
1949–2001) (Beijing, Zhongguo tiedao chubanshe, 2003), p. 32. Total operating mileage
was 13,160 miles by late 1949, 14,200 miles in 1952, and 16,600 miles in 1957. Total mileage
of double- and multitrack sections amounted to 540 miles in 1949, 880 miles in 1952, and
1,400 miles in 1957.
489

elisabeth kö ll
building of ambitious railroad projects in difficult terrain, such as the 1,200-

mile-long line from Lanzhou to Urumqi, built between 1952 and 1962, and the
705-mile-long Chengdu–Kunming Railroad completed in 1970.80
Economic reconstruction and the government’s plan to introduce indus-
trialization as rapidly as possible presented major challenges but also great
opportunities for China’s railroads. On the one hand, much hard manual
work was required to repair the war-damaged network because of scarce
machinery and little modern technical and managerial expertise. The assist-
ance of Soviet experts helped improve the productivity of logistics operations
and efficiency in equipment maintenance, but the departure of the Soviets did
not lead to a decline in the rail sector. On the other hand, the desire of the
Party and the government to compensate for the lack of appropriate capital
investment by focusing on the mass mobilization of workers and any avail-
able resources meant that new railroad construction was undertaken.81
In the 1950s railroad expansion served the goals of the First Five-Year Plan
to accelerate industrialization and was aligned with PRC interests regarding
national defense. The “outsourcing” of railroad construction to the PLA in
many ways was a brilliant move because it allowed a riskier approach to the
construction of new lines, while also providing the benefits of the disciplined
work ethic of railroad soldiers and the mobility of the military. At the same
time, the experiences of former tiedaobing on construction sites also demon-
strated that certain issues related to railroad construction, such as negotiating
land acquisitions, hiring local labor, and interacting with the local population,
were not fundamentally different from the practices employed during the
Republican period. Railroad soldiers brought economic and social change to
the local communities indirectly via rail construction, but they also came to
be seen as state protectors and mediators, without their own entrenched
interests in railroad development.82
Reconstruction in the 1950s also reframed the identity of railroads as
administrative and managerial institutions and aligned them with the new
socialist vision of the Chinese state. Values such as discipline, managerial
efficiency, punctuality, and so forth, which during the Republican period
were interpreted as virtues of Western modernity and technological
80
Gao Guangwen 高光文, 鐵道兵 (Railroad Army Corps) (Beijing, Zhongguo qingnian
chubanshe, 1972), p. 29; Chen Yuanmou 陳遠謀, 昨日鐵道兵 (Yesterday’s Railroad
Soldiers) (Beijing, Zhongguo shuji chubanshe, 1994).
81
See D. Kaple, “Agents of Change: Soviet Advisers and High Stalinist Management in
China, 1949–1960,” Journal of Cold War Studies 18.1 (Winter 2016), 11.
82
490

progress, were now reinterpreted as values representing the goals of revolu-

tionary socialism. This reinterpretation and transfer of values from Western
to socialist modernity conformed with the economic goals established in the
First Five-Year Plan, but also helped create an identity for railroad workers
based on expertise, skills, and professional dedication to their work.83
The years after the founding of the PRC were characterized by a relatively
quick recovery of China’s rail network. Between 1949 and the Great Leap
Forward campaign in 1957, freight transport rose almost five times and
passenger transport three times. In terms of different infrastructure methods
in China’s circular flow of passengers and freight during the 1950s, the
numbers indicate the trend of a continuing focus on rail transportation
with an increasing expansion of the road system. As Table 13.1 indicates,
although railroad freight turnover increased rapidly in the 1950s, the ton-
miles carried by modern water transport increased even more dramatically.
Due to Cold War military activities in the Taiwan Strait, inland water
transportation developed much more rapidly than coastal shipping.
As a general trend, water navigation was clearly more important than road
transportation in terms of total freight transportation, while railroads con-
tinued to carry the lion’s share of China’s freight across the nation. The
figures for passenger transportation during the 1950s were not that different:
in 1952, 81 percent of passengers traveled by rail, 9 percent in motor vehicles,
just under 10 percent on waterways, and only 0.1 percent by air. By the end of
the First Five-Year-Plan in 1957, motor transportation for passengers had
Table 13.1 Freight turnover by modern means of transportation (in million

ton-miles)84
Year Railroads Motor vehicles Ships and barges Total
Pre-1949 peak 25,103 286 7,972 33,361

1949 11,433 155 2,678 14,266
1950 24,488 236 1,802 26,526
1952 37,382 478 6,593 44,453
1954 57,937 1,205 11,582 70,724
1955 60,987 1,566 15,186 77,739
1956 74,906 2,169 17,529 94,664
1957 83,630 2,448 21,369 107,447
83
84
V.D. Lippitt, “Development of Transportation in Communist China,” China Quarterly
27 (July–September 1966), 101–19, table adapted from Table 5, p. 113.
491

elisabeth kö ll
increased to 18 percent, with railroads at 73 percent, whereas water and air

transport maintained almost the same percentages as in 1952.85
Whereas China’s progress in the expansion of rail infrastructure was
considerable, road construction continued to lag behind those efforts for
a long time. By 1956, China had just about 38,000 miles of all-weather roads
and 56,000 miles of fair-weather roads. As one observer estimated that year,
about one quarter of the rural population at the time is not within a day’s
walk of any road . . . The communications suffice to link villages to their
market towns, but, except for the big rivers and the railways, transportation
costs over long distances are very high and discourage much movement of
grain and other heavy products.86
It would take China just another fifty years to overcome the problem and
emerge as a global leader in the construction of modern highways, bridges,
and road infrastructure in and outside China at the beginning of the twenty-
first century.
Conclusion
For a long time, historians have looked at China’s modern transportation and
communication networks emerging in the nineteenth century as
a “problem,” lagging behind Western countries in scale and speed due to
the impact of Western imperialism and semicolonialism. However, recent
studies by historians have revisited the constraints of semicolonialism in the
context of Chinese statecraft during the late Qing from a new perspective.
For example, Stephen Halsey argues that the CMSNC became “an effective
tool for mercantilist statecraft” which provided a business model for the
shipping industry by integrating logistics, insurance, and sales into its busi-
ness portfolio. China’s emerging railroad network of the late nineteenth and
early twentieth centuries presents another example of a positive develop-
ment where Chinese railroad companies as state-directed enterprises broke
the Western dominance in syndicates and slowly emerged as a technology
with important economic and social impact. Anne Reinhardt has shown how
85
Miao Qiulin 苗秋林 (ed.), 中國鐵路運輸 (China’s Rail Transportation) (Beijing,
Zhongguo tiedao chubanshe, 1994), p. 43, Tables 2-2-1, 2-2-2.
86
Report from February 16, 1956, No. 63/“S,” “Road Communications in China,” 1372/1/
56, Foreign Office files FO 371/120962, British National Archives. See also C. Meysken,
Mao’s Third Front: The Militarization of Cold War China (Cambridge, Cambridge
492

important the enactment of Chinese sovereignty was for the steam transport
network and Chinese shipping businesses.
Whether in the telegraph, shipping, or railroad sector, government-
sponsored companies enabled China’s integration into the global communi-
cations network. Especially for the pre-1949 period the question of
sovereignty played an important role in the negotiations between the
Chinese state, its officials, and foreign political and business interests which
often, expressing Chinese concerns, centered on the question of total control
over the infrastructure ventures. As we have seen, the nationalization process
and changes after the 1911 Revolution gave the Chinese state ultimate
authority while still honoring the financial obligations and managerial con-
cessions without breach of contract.87
Lack of funding for rail construction through share capital, bond issues, and
bank loans characterized China’s government responses throughout the 1880s
and 1890s, slowing down the planning, construction, and operation of railroads.
Although Sun Yat-sen envisioned the establishment of shipbuilding yards in his
1921 program as part of his industrial development scheme for the Chinese
nation, cheap labor and materials did not translate into the robust rise of
a Chinese shipbuilding industry. Along similar lines, Sun’s plea for the establish-
ment of locomotive and railroad equipment factories was based on his expect-
ation that foreign capital and expertise would help initiate these industries.88
By comparison, Meiji Japan provides a good example of a national railroad
initiative where capital in the form of private investment was scarce, but
where the government became instrumental in raising funds through gov-
ernment deposits in banks to enable the pooling of private capital. To that
end the Meiji government introduced a National Banking Act and promoted
the establishment of a modern banking sector. In contrast to China’s govern-
ment response to lacking private investment, the Japanese government
encouraged bank shareholders to fund rail construction with the promise
of government subsidies for the project. As a result, these bank shareholders
began to purchase the shares of the Nippon Railway, Japan’s first privately
held joint-stock company, which resulted in the building of 1,300 miles of
railroad in Japan through private investment as early as 1892.89 In this context,
87
Halsey, China’s Quest, pp. 211, 236; Reinhardt, Navigating Semi-colonialism, p. 296; Köll,
Railroads and the Transformation of China, p. 296.
88
Sun Yat-sen, The International Development of China (Shanghai, Commercial Press, 1920),
pp. 194–5.
89
S. Ericson, The Sound of the Whistle: Railroads and the State in Meiji Japan (Cambridge,
MA, Harvard East Asian Monographs, 1996), p. 51.
493

elisabeth kö ll
it is fair to argue that the late and slow trajectory of the rail sector in China
was at least partly due to the lack of sufficient institutional reforms that could
have brought about modern banks and legal changes for incorporated joint-
stock companies to motivate investors. However, it is also clear that the late
and initially slow trajectory of railroad development was not due to the new
technology itself or its perception and reception by the Chinese public.90
The history of China’s infrastructure development is as much a global
story as it is a Chinese story. This chapter has shown how China adopted
telegraphs, postal systems, railroads, steamships, and so on from the West as
technological and infrastructural concepts while retaining Chinese agency as
operators, managers, builders, and consumers of their services. From a global
perspective, much of China’s transportation and communication network
was financed and built under semicolonial conditions; and even after nation-
alization, Western advisers and a mix of European and North American
methods shaped the emergence of Chinese railroad and shipping manage-
ment, training, and administration. Especially in the rail sector, China
remained dependent on rolling stock and equipment imports from Great
Britain and the United States until the end of World War I I. From 1949 to the
political rift in 1961, engines, rail equipment, and technical advisers mostly
hailed from the Soviet Union. China’s rail development today, especially its
high-speed sector, still depends on technology imports, keeping alive political
debates about indigenous innovation and perceived dependency on foreign
technology.91 However, at the beginning of the twenty-first century, China’s
infrastructure development has reached a new phase as the government is
engaged in the “One Belt, One Road” initiative, with the aim of placing China
at the center of economic globalization through infrastructure expansion in
and outside China. Railroads and highways, as part of a newly envisioned Silk
Road, will complement shipping along the Maritime Silk Road in the quest to
connect China with markets, products, and people on a truly global scale.
Further Reading
Baark, E., Lightning Wires: The Telegraph and China’s Technological Modernization, 1860–1890
(Westport, CT, Greenwood Press, 1997).
90
See also H.-C. Wang, “Institutional Failure.” Wang convincingly argues for institu-
tional failure as the main reason for the Wusong railroad’s lack of success.
91
N. Ahrens, “Innovation and the Invisible Hand: China, Indigenous Innovation, and the
Role of Government Procurement,” Carnegie Papers, Asia Program, no. 114, July 2010
(Washington, DC, Carnegie Endowment for International Peace, 2010).
494

Chang Rui-te 張瑞德, 中國近代鐵路事業管理的研究, 1876–1937 (Railroads in Modern

China: Political Aspects of Railroad Administration, 1876–1937) (Taipei, Institute of
Modern History, Academia Sinica, 1991).
Chang Rui-te 張瑞德, 平漢鐵路與華北的經濟發展, 1905–1937 (The Peking–Hankow
Railroad and Economic Development in North China, 1905–1937) (Taipei, Institute of
Modern History, Academia Sinica, 1987).
Ding Xiangyong 丁賢勇, 新式交通與社會變遷：與民國浙江為中心 (New
Communications and Social Transformation: Focusing on Zhejiang Province during
the Republican Period) (Beijing, Zhongguo shehui kexue chubanshe, 2007).
Duus, P., R.H. Myers, and M.R. Peattie (eds.), The Japanese Informal Empire in China, 1895–
1937 (Princeton, NJ, Princeton University Press, 1989).
Faure, D., China and Capitalism: A History of Modern Business Enterprise in China
Feuerwerker, A., China’s Early Industrialization: Sheng Hsuan-huai (1844–1916) and Mandarin
Halsey, S.R., Quest for Power: European Imperialism and the Making of Chinese Statecraft
Huenemann, R.W., The Dragon and the Iron Horse: The Economics of Railroads in China
Kaske, E., “Sichuan as a Pivot: Provincial Politics and Gentry Power in Late Qing Railway
Projects in Southwestern China,” in U. Theobald and C. Jin (eds.), Southwest China in
a Regional and Global Perspective (c. 1600–1911) (Leiden and Boston, Brill, 2018), pp. 379–423.
Köll, E., Railroads and the Transformation of China (Cambridge, MA, Harvard University
Press, 2019).
Leung, C.-K., China, Railway Patterns and National Goals (Chicago, The University of
Chicago Press, 1980).
Liang, E.P., China, Railways and Agricultural Development, 1875–1935 (Chicago, The
University of Chicago Press, 1982).
Moazzin, G., “Sino-Foreign Business Networks: Foreign and Chinese Banks in the Chinese
Banking Sector, 1890–1911,” Modern Asian Studies 54.3 (May 2020), 970–1004.
Rawski, T.G., Economic Growth in Prewar China (Berkeley, University of California Press,
1989).
Reinhardt, A., Navigating Semi-colonialism: Shipping, Sovereignty, and Nation-Building in
China, 1860–1937 (Cambridge, MA, Harvard University Asia Center, 2018).
Shen Zhihua 沈志華, 蘇聯專家在中國, 1948–1960 (Soviet Experts in China, 1948–1960)
(Beijing, Xinhua chubanshe, 2009).
Smith, V., with A. Chuh, Motor Roads in China (Washington, DC, United States
Government Printing Office, 1931).
Thai, P., China’s War on Smuggling: Law, Economic Life, and the Making of the Modern State,
1842–1965 (New York, Columbia University Press, 2018).
Zhuang Weimin 庒維民, 近代山東市場經濟的變遷 (The Transformation of the
Market Economy in Modern Shandong) (Beijing, Zhonghua shuju, 2000).
495

14
Education and Human Capital
Introduction
China has the largest education system in the world today. It educated more
than 260 million people and employed 15 million teachers in 2015.1 Besides its
social impact, educational development has often been argued to be one of
the primary reasons behind China’s stunning economic growth after the
economic reform implemented in 1978. It is therefore of paramount import-
ance to understand how education evolved in Chinese history.
Throughout the very long educational history of China, the period that is
studied in this volume (1800–1949) is particularly important as the education
system underwent extraordinary changes and unprecedented expansion. For
over 1,300 years, the traditional Confucian teaching system had played an
essential role in shaping China’s social and political structure, economic
development, and cultural norms. From the late nineteenth century
onwards, a series of transformations occurred to shift this long-lasting trad-
itional system to a modern education system that was approximating
a Western model. This new system helped to educate millions of students,
formed modern human capital, cultivated new intellectual elites, and encour-
aged modern economic growth in China.
Despite its importance, there is very limited empirical evidence on the
measurement of education levels as well as the related concept of human
capital formation over the period studied in this volume. Some studies
show that the high demand for education generated by the civil service
examination system may arguably have led to relatively high levels of
The research leading to these results received funding from NYU-Shanghai and the
European Research Council under the European Union’s Horizon 2020 Programme
(ERC-StG 637695 – HinDI) as part of the Historical Dynamics of Industrialization in
Northwestern Europe and China c. 1800–2010: A Regional Interpretation project.
1
Y. Pan, S. Vayssettes, and E. Fordham, Education in China: A Snapshot (Paris,
Organisation for Economic Co-Operation and Development, 2016).
496

human capital.2 However, based on the limited statistics on China,

a considerable lead in mean education levels appears to have been in
existence in the West even before the nineteenth century.3 This debate in
the literature continues as a result of the paucity of data for China.
Unfortunately, reliable estimates of educational attainment prior to
1949, such as literacy rates, enrollment ratios, and average years of school-
ing, are still wanting.
This chapter surveys some major changes in the way in which people
acquired education in China from c. 1800 to 1949 and tries to provide a general
picture of the trajectory of the expansion of education based on various
sources. Human capital can be created through both formal education and
training of other kinds, such as apprenticeships. Although we occasionally
touch upon the latter, the main focus of this chapter is on the human capital
acquired through the formal education system. This chapter is organized as
follows. We will first detail the traditional Confucian teaching system, then
introduce the newly created modern education system that was built after
the turn of the twentieth century. The drastic educational expansion and
modernization also profoundly impacted China in various social and eco-
nomic areas, such as female education and the economic structure, which
will be discussed in the final two sections.
The Traditional Education System before 1905

Before it came to an end in 1905, the traditional Confucian teaching system
directly shaped China’s social and political structure, economic development,
and culture. One distinguishing characteristic of this system was that its
foremost purpose was to select state officials and social elites from the best
candidates via a series of competitive examinations.4 This means that
2
J. Baten, D. Ma, S. Morgan, and Q. Wang, “Evolution of Living Standards and Human
Capital in China in the 18th–20th Centuries: Evidence from Real Wages, Age-Heaping,
and Anthropometrics,” Explorations in Economic History 47.3 (2010), 347–59; E. Rawski,
Education and Popular Literacy in Ch’ing China (Ann Arbor, University of Michigan Press,
1979).
3
D. Cressy, “Levels of Illiteracy in England, 1530–1730,” Historical Journal 20 (1977), 1–23; P.
Gao, “Risen from the Chaos: The Emergence of Modern Education in China,” in
D. Mitch and G. Cappelli (eds.), Globalization and the Rise of Mass Education (Cham,
Palgrave Macmillan, 2019), pp. 279–309; D. Mitch, “Education and Skill of the British
Labour Force,” in R. Floud and P. Johnson (eds.), The Cambridge Economic History of
Modern Britain, vol. 1 (Cambridge, Cambridge University Press, 2004), pp. 332–56.
4
The imperial civil service examination system was implemented as early as the Tang
dynasty (618–896) and continued for more than 1,000 years until its abolition in 1905. This
system served to recruit state bureaucrats and elites from the top national talents.
497

education had been the path to upward social mobility as such success in
education could bring exceptionally high social and economic returns.5 In this
section, we briefly introduce the traditional system before 1905 first, and then
survey the quantitative evidence on educational outcomes.
The Acquisition of Education

The way in which people acquired education changed significantly over the
course of the period studied. In nineteenth-century China, education was
conducted through educational institutions that are not very familiar to
twenty-first-century readers. Under this system the educational structure
was vaguely defined, with no clear regulation on schooling levels, division
of grades, and the length of schooling.6 As shown in Figure 14.1, many types
of educational institutions with overlapping functions coexisted. This system
can be roughly divided into two levels based on different functions. The
elementary education was intended to provide basic schooling to the masses,
while the main function of high-level education was to train examinees to
prepare for the civil service examinations.
Elementary-Level Education
Unlike most modern societies, the Qing government supported high-level
education more than elementary-level education.7 The state only financed the
civil service examination together with a large number of government schools to
further train established scholars; in contrast, the responsibility to provide basic
education was mainly taken by private households and local communities. The
most common educational institutions where primary-school-age children went
to acquire basic literacy and numeracy were collectively called sishu 私塾.8 Sishu
literally means private schools, which were often single-teacher operations run
5
The successful candidates would be selected as state bureaucrats and social elites, and
they could also enjoy various privileges, including exemption from corporal punishment
and corvée (labour service), reduced tax payments, the right to wear a scholar’s robe,
and so on. Apart from official economic returns, some studies also show that Chinese
bureaucrats incurred extremely high private gains from corruption of all sorts, some-
times amounting to fourteen to twenty-two times their official salary. See, for example,
C. Chang, The Income of the Chinese Gentry (Seattle, University of Washington Press,
1962), pp. 32–43. S. Ni and P.H. Van, “High Corruption Income in Ming and Qing China,”
Journal of Development Economics 81.2 (2006), 316–36.
6
B.A. Elman, A Cultural History of Civil Examinations in Late Imperial China (Berkeley,
7
Rawski, Education and Popular Literacy, pp. 24.
8
There were many different types of sishu. A more detailed elaboration of this is outside
the scope of the present study. But, very generally, they were family sishu, lineage or
village sishu, and purely private sishu. Rawski, Education and Popular Literacy, pp. 44–53.
498

Imperial Academy
High-level education
(to prepare for imperial exams) County school/prefecture school Academies
Private popular schools (sishu),

Elementary-level education charity schools (yixue), community
(to learn basic literacy & numeracy) schools (shexue)
Figure 14.1 The school system before 1905
for profit, while some also received financial support from wealthy local families.
Public effort was not entirely absent in elementary-level education,9 and there
were two types of public schools: yixue 義學 (charity schools)10 and shexue 社學
(community schools).11 However, the number of these schools was very small.
According to Rawski, over the Qing dynasty only about 13,400 out of 40 million
school-aged boys enrolled in these publicly provided elementary schools.12
Without a clearly defined educational structure, there was no explicit
regulation on the length of schooling at every level. Chang notes that boys
started their study at quite a young age, ranging from five to seven years old,
and on average it took about three to four years to learn the basics of
literacy.13 The curriculum focused mainly on literacy and numeracy.
Commonly used textbooks were the Confucian classic texts that were con-
sidered suitable for teaching young children.14 Apart from classic texts,
practical knowledge was often incorporated in the material. For instance,
simple accounting practice and contract writing were also found to be
9
Similar common schools with one teacher and all the students in one class were also
widely found in eighteenth-century Europe and the US.
10
Yixue were established mainly by local communities to provide education for the
children from poor families.
11
Shexue were public primary schools set up by local governments mainly in rural areas;
after the 1750s they were gradually shut down.
12
Rawski, Education and Popular Literacy. Also see, for example, Ma Yong 馬鏞, 中國教育
制度通史 (A General History of Chinese Education), 5 vols. (Jinan, Shandong
Education Press, 2000).
13
C. Chang, Chinese Gentry: Studies on Their Role in Nineteenth-Century Chinese Society
(Seattle, University of Washington Press, 1955).
14
Sanzijing 三字經, baijiaxing 百家姓, qianziwen 千字文, and qianjiashi 千家詩 were
traditional textbooks for enlightened education.
499

popular teaching content.15 Sometimes even astronomy, geography, and

agronomy were introduced in the classrooms.
High-Level Education
The post-elementary education had only one purpose – training established
scholars for the imperial civil service examinations, therefore advanced
educational institutions, in various forms, functioned as training and examin-
ation centers. As shown in Figure 14.2, the examination system had
a pyramidal structure consisting of a series of tests administered at the
county, provincial, and central levels. The academic degrees associated
with each level of the examination formed a “ladder of success” in Chinese
society.
Different from the fact that private schools dominated elementary-level
education, the imperial government played an essential role in high-level
education by directly financing the examinations and setting up many gov-
ernment schools. Students usually started their studies for the examination in
privately run senior sishu where classic canons and articles were taught.16
Once the examinees passed the highest level of the entry examination, the
licensing examination (yuankao 院考),17 they obtained the degree of shen-
gyuan 生員 and were entitled to attend the publicly provided county or
prefectural school (fuxue 府學, zhouxue 州學, xianxue 縣學) to pursue their
studies.18 The most outstanding students could be further selected into the
Imperial Academy (guozijian 國子監) in Beijing.19 Apart from these
15
Li Bozhong 李伯重, “19 世紀初期華婁地區的教育產業” (Educational Institutions of
the Hualou Area in the Early Nineteenth Century), Studies in Qing History 2 (2006),
60–74.
16
Rawski, Education and Popular Literacy, pp. 31.
17
The primary entry examination was presided over by the magistrate of the county,
followed by an examination presided over by the head of the prefectural government. If
successful, the student could attend the prefectural exam (yuankao), also called “licens-
ing exam” in some studies. The prefectural exam was presided over by the highest
educational officer of the province.
18
Prefectural schools (府學) were built in prefectural seats and county schools (縣學) in
county seats. These schools were publicly funded and managed directly by the govern-
ment. They not only provided training to students; more importantly they were official
examination centers. Their main function was to hold regular mock examinations and
register further preselection examinations (keshi 科試) for each student. Only those
who passed the preselection examination were qualified to take the next-level exam –
the provincial level.
19
The Imperial Academy was established in Beijing by the central government. As the
model of teaching for other local official schools, its students had privileges such as
more funding and better teachers. The emperor himself sometimes even gave irregular
lectures there.
500

Gongshi and jinshi

( & Metropolitan exam and palace exam ( & )
Juren ( )
Provincial exam ( )
Shengyuan ( )
Licentiate exam ( )
A series of entry exams ( )
Commoners
Figure 14.2 The civil service examination system and its associated degree titles
Notes. The lowest academic degree was shengyuan 生員, which was awarded to
examinees who had passed the licentiate examination. Becoming shengyuan guaranteed
a place in a county or prefectural school that was provided by the state for examinees to
further their study. Only qualified candidates in those government schools were entitled
to participate in the next-level examination – the provincial examination, and those who
passed could obtain the degree title juren 舉人. The metropolitan examination was held in
the capital, Beijing, and juren who passed this level of examination were called gongshi
貢士. The final test was the palace examination, and the successful candidate could obtain
the highest degree – jinshi 進士.
government schools, examinees could also attend private academies (shuyuan

書院) for more training.20 These academies were privately funded but
endorsed by the state. As private educational institutions, they were more
tolerant in term of admission standard. Some academies only accepted high-
degree holders, while others also allowed examinees with weaker academic
backgrounds to attend.21
The curriculum for this stage of schooling focused entirely on the
Confucian classics. Ancient canons and classic articles were used as textbooks
to extend the training in writing and critical thinking. The curriculum
experienced slight changes over time, but throughout the Qing period it
had three main parts: (1) reading classical texts,22 (2) memorizing a shared
20
Private academies served multiple functions. First, they were libraries where Confucian
classics were kept for examinees to read. They were also centers of scholarly discussion
and research for established scholars.
21
Some academies admitted tongsheng (童生) who had only passed some low-level
examinations but obtained no official degree title yet, while others only accepted degree
holders, including shengyuan (生員) and juren (舉人).
22
Students refined their linguistic skills by reading classical articles and anthologies
written in traditional script.
501

canon,23 and (3) the ability to write elegant essays, known as eight-legged
essays.24 The strikingly narrow-focused curriculum has been criticized by
both reformers at the time and contemporary scholars. The educational
content was too distant from both modern scientific inquiry and practical
economics,25 which could discourage young talents from pursuing a wider
spectrum of knowledge and set back the development of technology and
modern science in China.26
In theory, the civil service examination was open to all males; however,
only a small proportion of people would choose to pursue high-level educa-
tion because the preparation for the examinations was time-consuming. In
addition, the length of study required to pass the examinations varied
significantly among individuals and was difficult to predict.27 On average,
the basic learning to read the classic canons, poems, and articles took
approximately six or seven years, and several more years were needed to
study eight-legged-essay writing. Most of the lower-degree holders (sheng-
yuan) enrolled in government schools at the age of twenty. Given that boys
often started elementary schooling at the ages of five to seven, it took about
fifteen years to obtain the lowest degree. Most of the top-degree holders
(jinshi) completed the final palace examination in their late thirties.28 That is
why even with high monetary return and social privilege attached to the
degrees, post-elementary education remained less attractive to people with
humble backgrounds. For instance, Wang studied 1,000 lower-degree holders
(shengyuan) in Zhili province and found that more than 50 percent of them
came from high-income families with members among the landholding
23
The compulsory readings were ancient Confucian canons, i.e. the Four Books, the Five
Classics and classic annotations of these canons by scholars. The Four Books and Five
Classics are the authoritative books of Confucianism in China. The Four Books are the
Analects 論語, the Mencius 孟子, the Great Learning 大學, and The Doctrine of the Mean
中庸. The Five Classics are the Classic of Poetry 詩經, the Book of Documents 尚書, the
Book of Rites 禮記, the Book of Changes 易經, and the Spring and Autumn Annals 春秋.
24
Elman, A Cultural History, pp. 46–93. 25 Elman, A Cultural History, pp. 53–64.
26
See, for example, G. Clark and R.C. Feenstra, Technology in the Great Divergence:
Globalization in Historical Perspective (Chicago, The University of Chicago Press, 2003),
pp. 277–322; J. Mokyr, The Gifts of Athena: Historical Origins of the Knowledge Economy
(Princeton, Princeton University Press, 2002); D.S.Landes, “Why Europe and the West?
Why Not China?”, Journal of Economic Perspectives 20 (2006), 3–22; J.Y. Lin, “The
Needham Puzzle: Why the Industrial Revolution Did Not Originate in China,”
Economic Development and Cultural Change 43.2 (1995), 269–92; D. Cantoni and
N. Yuchtman, “The Political Economy of Educational Content and Development:
Lessons from History,” Journal of Development Economics 104 (2013), 233–44.
27
See, for example, Elman, A Cultural History; Y. Xu, P. Földvári and B. van Leeuwen,
“Human Capital in Qing China: Economic Determinism or a History of Failed
Opportunities?”, MPRA Paper 43525, University Library of Munich, Germany (2013).
28
Chang, Chinese Gentry.
502

gentry, merchants, or officials. A similar composition can also be observed

when looking at the top-degree holders (jinshi).29 Consistently using personal
information about 4,035 juren, Kung and Jiang also found that only 20.27 per-
cent of their fathers were commoners who held no degree titles.30
Apprenticeship
Formal schooling is only one part of education. Vocational training that
provides job-specific skills also generates human capital. Similar to Europe
before the Industrial Revolution, vocational education in China for many
professions, especially in handicraft and service industries, was conducted
through apprenticeship.31 The learning was usually arranged in a format
where each student worked under the supervision of a master craftsman.
However, as handicraft, trade, and service sectors only occupied a small share
in the whole economy before the twentieth century,32 apprentices were
unlikely to consist of more than 4 percent of the labour force. Prior to 1900,
only a very small number of children had the opportunity to be an
apprentice.33
Unlike what we see in medieval Europe, where the relationship between
apprentices and their master was based on formal contracts,34 apprenticeships
were not strictly regulated in China and the training content was not
standardized even within the same industry. Guilds indeed attempted to
provide more regulations on apprenticeship from the nineteenth century
onwards, but the recruitment standards, apprenticeship lengths, and the cap
on the number of apprentices under one master all varied widely across
29
Wang Yaosheng 王躍生, “清代生監的人數計量及其社會構成” (The Calculation
and the Social Composition of Shengyuan and Jiansheng in Qing China), Nankai Journal 1
(1989), 40–7.
30
Q. Jiang and J.K.S. Kung, “Social Mobility in Late Imperial China: Reconsidering the
‘Ladder of Success’ Hypothesis,” Modern China, 2020, https://doi.org/10.1177/
0097700420914529.
31
Li Quanshi 李權時, 商業教育 (Commercial Education) (Shanghai, Zhonghua Book
Company, 1933).
32
Y. Xu, Z. Shi, B. van Leeuwen, Y. Ni, Z. Zhang, and Y. Ma, “Chinese National Income,
ca. 1661–1933,” Australian Economic History Review 57.3 (2017), 368–93; Y. Guo, Z. Zhang,
B. van Leeuwen, and Y. Xu, “A View of the Occupational Structure in Imperial and
Republican China (1640–1952),” Australian Economic History Review 59.2 (2019), 134–58.
33
This situation only changed after the expansion of state-owned craft workshops
(gongyiju 工艺局) in the 1900s.
34
Zhao Xueyao 趙學瑶, Yu Ping 於萍, and Lin Ruoxin 林若鑫, “中西方傳統學徒制比
較研究” (A Comparative Study of Traditional Apprenticeship in China and Western
Countries), Vocational and Technical Education 37.34 (2016), 35–40.
503

professions.35 The age of admission for apprentices was quite flexible and
children could start their training at any age between twelve and eighteen.36
There was no admission requirement on educational background for craft-
work apprentices, and the selection standards placed more emphasis on
candidates’ characteristics, such as diligence and honesty. In contrast, indus-
tries like the mercantile trade, publishing, and banking adhered to very
specific standards when recruiting their apprentices. For instance, old-style
Chinese banks set explicit rules that only candidates with at least eight years
of sishu education were considered eligible.37 The length of training also
varied greatly, from two or three years to five to seven years (the most
common duration was approximately three years).38 For example, the first-
year training in commerce was devoted to routine services in order to
develop trainees’ basic work ethics and social skills. In the second year,
training shifted towards professional skills such as accounting, letter writing,
and business etiquette. In the third year, professional training, including
negotiating skills, was deepened.39
Measuring the Expansion of Education

Existing literature remains inconclusive on the level of human capital in
China throughout the nineteenth and early twentieth centuries, because
35
Regarding regulations on the number of apprentices, the figure varies across industries.
For instance, in Changsha city, each workshop producing galvanized iron was allowed
to admit only one apprentice; while two or three apprentices per filature were allowed
for the silk industry, and five apprentices for each pawnshop. Zhong Li 李忠 and
Xiaoning Wang 王筱宁, “社會教育在底層民眾實現社會流動中扮演的角色—以
清末民國時期的學徒教育為例” (The Role of Public Education on Upward
Social Mobility of Commoners: Taking Apprenticeship in Late Qing and Republican
China as an Example), Social Sciences in Nanjing 4 (2008), 107–14. See also Qi Xia 漆侠, 宋
代經濟史 (Economic History of the Song Dynasty) (Shanghai, Shanghai renmin
chubanshe, 1987).
36
Peng Zeyi 彭澤益, 中國工商行會史料集 (Collection of Historical Archives of the
Chinese Guilds of Industry and Commerce), vol. 1 (Beijing, Zhonghua shuju, 1995).
37
With the modernization of the education system, in 1904 this requirement was changed
to the possession of a senior primary-school diploma. In 1922 it changed to the
possession of a junior secondary-school diploma. See Qiu Zhenkang 裘振康, “紹興
錢莊培植學生的一些情況” (The Situation of Student Cultivation in the Old-Style
Chinese Bank of Shaoxing), in Literature and Historical Archives of Shaoxing, 9 vols.
(Shaoxing, Shaoxingshi Wenshi Ziliao Weiyuanhui, 1990), pp. 106–9.
38
Early in the twentieth century, John Stewart Burgess investigated sixteen craft guilds in
Beijing. He found that three industries set the apprenticeship to last for three years, ten
industries called for three years and three months and in the three other industries the
duration was four, six, or seven years. See J.S. Burgess, The Guilds of Peking (New York,
39
Luo Lun 罗伦 and Fan Jinmin 范金民, “清抄本生意世事初階述略” (An
Introduction to Shengyi Shishi Chujie), The Documantat 2 (1990), 233–9.
504

direct evidence utilizing systematic data on schooling level is scarce. Without

more empirical studies directly measuring educational outcomes in China
during the period studied, the debate will continue.
Literacy Rate
The literacy rate is one of the most easily quantifiable indicators to measure
education and human capital. Perkins estimates that in the 1880s less than
50 percent of males above school age could be regarded as literate on the basis
of Maritime Customs reports.40 Similarly, Rawski argues that basic education
had already diffused nationwide, resulting in a respectably high literacy rate in
nineteenth-century China. She points out that, even though the variation across
regions could be remarkably wide, a rough guess at the literacy rate during the
late Qing might be 30 to 45 percent for males and 2 to 10 percent for females.41
Consistent with these estimates, other researchers also document that written
communications of all kinds were widely used in hiring labor, renting property,
and spreading news even in remote rural communities, which indicates a society
with relatively high literacy rates.42 If such evidence is valid, nineteenth-century
China should have a similar rate of literacy to that in late Tokugawa Japan,
where the rate for men was at 40 to 50 percent, and for women at 13 to
17 percent.43 However, compared to Europe and North America, which were
leaders in terms of the public provision of education at all levels, the literacy rate
in China lagged far behind. Before 1900, about 80 percent of Americans could be
regarded as receiving some level of education;44 and this figure was 88 percent in
Germany, 77 percent in the UK, and 70 percent in France.45
Elementary-Level Schooling
Data on school enrollments in China over our period of study are very
limited, especially for elementary-level education. As discussed previously,
40
D.H. Perkins, China’s Modern Economy in Historical Perspective (Cambridge, MA,
Harvard University Press, 1975), p. 4.
41
Rawski, Education and Popular Literacy, pp. 8–23.
42
M. Zelin, “Merchant Dispute Mediation in Twentieth-Century Zigong, Sichuan,” in
K. Bernhardt and P.C.C. Huang (eds.), Civil Law in Qing and Republican China (Stanford,
CA, Stanford University Press, 1994), pp. 249–86.
43
K. Ohkawa and H. Rosovsky, Japanese Economic Growth: Trend Acceleration in the
Twentieth Century (Stanford, CA, Stanford University Press, 1973), p. 8.
44
S.E. Black and K.L. Sokoloff, “Long-Term Trends in Schooling: The Rise and Decline of
Public Education in the United States,” in E. Hanushek and F. Welch (eds.), Handbook of
the Economics of Education 1 (San Diego, CA, North-Holland, 2006), pp. 69–105.
45
A. Green, The Rise of Education Systems in England, France and the USA (London,
Macmillan, 1990).
505

Table 14.1 Estimation of the enrollment ratio of sishu

Sishu
School-age popu- Enrollment
Year Province County lation (7–12) School Enrollment ratio (%)
1911 Zhuji 47,684 790 10,981 23.0

1909 Daishan 7,967 100 1,390 17.5
Zhejiang
1909 Fenghua 34,103 399 5,920 17.4
1907 Putuo 8,264 80 1,112 13.5
1911 Funing 29,142 227 3,155 10.8
1911 Hebei Zhaoxian 15,109 200 2,780 18.4
1911 Wangdu 7,709 71 987 12.1
1910 Anhui Lingbi 30,869 191 2,727 8.8
1911 Fujian Jianyang 15,955 200 1,000 6.3
1904 Shandong Jimo 35,624 241 4,653 13.1
Source: X. Hu, 清末中國民眾私塾就學率的考察 (A Study of Private-School

Enrollment Ratios at the End of the Qing Dynasty) Lifelong Education and Libraries 9
(2009), 29–36
most primary-school-age children attended sishu that were not public educa-
tional institutions, therefore their information was never included in the
official statistics. Table 14.1 presents some scattered county-level statistics,
indicating that basic education already had a wide coverage in nineteenth-
century China, with an average enrollment rate of approximately 14 percent.
This number matches some historical anecdotal evidence that almost all
villages had at least one or two sishu to train their children in basic literacy
and numeracy.46 But regional variation could be remarkably wide. Looking
at rich southern counties in the nineteenth century, most of the students in
the elementary-level schools were commoners, which implies that the elem-
entary-level education spread widely to all social groups in the highly com-
mercialized and prosperous Yangzi delta. In contrast, in the rural counties of
Yunnan and Jiangxi, 80 to 90 percent of the population were reported as
never receiving any level of education in the 1900s.47
46
China proper has about 1,660 counties and about 9,000 villages. Traditional private
popular schools penetrated into each village, so that there were probably more than
10,000 sishu across China.
47
Yin Mengxia 殷梦霞 and Tian Qi 田奇, 民國人口資料彙編 (Historical Data of
Population and Household Registers in the Republic of China) (Beijing, Guojia
Tushuguan chubanshe, 2009).
506

High-Level Schooling
Unlike the absence of sishu in official records, data on high-level education are
more abundant because the state directly managed these government
schools. In Table 14.2 we present the regional distribution of the two types
of high-level educational institutes, county or prefectural schools and private
academies. As shown in Table 14.2, high-level education institutions also
spread nationwide. Among the 1,660 counties in China proper,48 about
88.5 percent set up at least one government school (county or prefectural
schools) to train established scholars to prepare for the civil service examin-
ations. Private academies present a slightly different pattern. Regional dispar-
ity was higher, with 20 percent of counties having no academy, 45 percent
having fewer than three and 10 percent having more than ten.
It is impossible to estimate the enrollment ratio for high-level education,
but the number of examinees who participated in the civil service examin-
ation system offers a crude way to gain some insights on this question. There
was a long-standing quota system to regulate the number of examinees who
were allowed to enroll in county or prefectural schools.49 The original quotas,
assigned to counties, ranged from zero to twenty.50 Wealthier and strategically
more important counties were granted higher quotas. Before the Taiping
Rebellion (1851), 23,852 examinees passed the licentiate examination each
time. According to Chang, the success rate was only about 1 or 2 percent,51
48
The central plain of China includes eighteen provinces but about 80 percent of the
population. It is a term often used to express a distinction between the core and frontier
regions of China.
49
There were two kinds of quota. One was assigned directly to each county and the other
was a prefectural-level quota that was comparatively small. The figures presented in
Table 14.2 are the sum of all county-level quota within a given prefecture plus the
prefectural-level quota.
50
The Imperially Established Institutes and Laws of the Great Qing Dynasty (大清會典)
carefully documented the quota system across counties throughout the Qing period.
Whereas the quota system remained stable between 1644 and 1851, because of the
outbreak of the Taiping Rebellion (1850–1864) the Qing government increased the
quotas for many regions in order to strengthen its alliances among the local gentry
against rebel forces. After the war, the revised quota assigned in 1873 persisted until the
abolition of the exam system. See, for example, Chang, Chinese Gentry, pp. 83–92; P. Ho,
The Ladder of Success in Imperial China: Aspects of Social Mobility, 1368–1911 (New York,
51
Wang has different estimates than Chang. He argues that in the areas with large
numbers of applicants the success ratio could be even lower than 1 percent, but in the
areas with relatively few applicants, the ratio was 3 to 4 percent. See, for example,
Chang, Chinese Gentry; Wang Ligang 王立剛, “清代童試錄取率研究” (A Study of the
Admission Rate to the Preliminary Stage of the Imperial Examinations), Tribune of Social
Sciences 3 (2014), 67–76.
507

Table 14.2 Number of post-elementary schools in China proper

Government school
Prefectural County Quotas for students
Province school school Early Qing Late Qing Private academy
Anhui 12 58 1,279 1,604 312

Fujian 12 60 1,111 1,426 366
Gansu 13 69 925 931 66
Guangdong 12 89 1,359 1,768 531
Guangxi 12 66 1,024 1,122 188
Guizhou 11 71 766 767 158
Henan 13 105 1,667 1,902 404
Hubei 10 68 1,117 1,543 234
Hunan 13 71 1,262 1,675 489
Jiangsu 11 69 1,404 764 301
Jiangxi 14 79 1,360 2,130 957
Shaanxi 12 89 1,147 1,254 156
Shandong 12 107 1,870 1,993 311
Shanxi 20 101 1,528 1,612 181
Sichuan 16 150 1,486 2,017 609
Yunnan 21 80 1,369 1,372 301
Zhejiang 11 78 1,803 2,177 698
Zhili 17 133 2,711 2,751 284
Note. “China proper” is a term used to express a distinction between the core and the
frontier regions of China. More explicitly, China proper covers eighteen provinces: Anhui,
Fujian, Gansu, Henan, Hubei, Hunan, Jiangsu, Jiangxi, Guangdong, Guangxi, Guizhou,
Shaanxi, Shandong, Shanxi, Sichuan, Yunnan, Zhejiang, and Zhili. It accommodated more
than 80 percent of the population (measured by late Qing population data).
Sources: 康熙朝大清會典 (Collected Statutes of the Great Qing Dynasty); 欽定學政
全書 (The Encyclopaedia of Educational Affairs); Ji Xiaofeng 季嘯風, 中國書院辭典
(Dictionary of Academies in China) (Hangzhou, Zhejiang Jiaoyu chuban jituan, 1996)
which means the total number of candidates who participated in the licensing
examination could be as large as 1.2 million. Given that the population was
around 400 million, this crude estimate suggests that China indeed had
a “bottom-heavy” talent pool with a large number of males at least receiving
six to ten years of education.52
52
Many candidates would take the exam multiple times before they successfully obtained
a pass, therefore there was a high repeating ratio among these 1.19 million candidates.
508

The Modern Education System after 1905

From the late nineteenth century onwards, China’s defeat in a series of wars
against the West and Meiji Japan led to an increasing awareness of the
importance of Western technology in Chinese society.53 As a way to achieve
national salvation, a number of Western-influenced reforms, touching on
different aspects of society, were implemented over the course of our period
of study. This modernization movement included a call to remodel the
traditional Confucian education system. This section introduces the new
education system first, and then surveys evidence measuring the educational
outcomes of early twentieth-century China.
The New Education System

Education reforms began with cautious steps in the late nineteenth cen-
tury when a small number of foreign-language schools, military schools,
and modern teacher-training schools were built. The real milestone
occurred in 1905 when a memorandum was unexpectedly issued to abolish
the civil service examination system at all levels and a new education
system was officially implemented nationwide approximating a Western
model.54
The Formal Education System

Compared to the traditional Confucian teaching system, the new education
system had several distinguishing features. First, as shown in Figure 14.3, the
newly implemented system had a clearly defined educational structure with
explicit regulations on schooling levels, the division of grades, and the length
of schooling.55 The first schooling-level structure was a 9–4–6 division: primary
school for nine years, secondary school for four years, and tertiary school for
six years, which was modeled after the Japanese arrangement. In addition,
53
For instance, there were the two Opium Wars, the First Sino-Japanese war, the Sino-
French War, and the occupation of Beijing by eight nations after the Boxer Rebellion.
54
The education models developed in Europe, the US, and Japan from the early nine-
teenth century onwards are different in many ways, but fundamentally they are similar
and have several core characteristics: they are universal, mandatory, secular, and
academic. This education model is widely considered to be an institutionalized
model of national development around the globe. See Green, The Rise of Education
Systems, p. 297.
55
The structure was altered several times in the following decades, and Japanese influence
weakened as Western influence became more prominent. In 1912, the division became
7–4–4. On October 26, 1921, the education system was changed again to follow the
American style, which favored a 6–6–4 division.
509

Modern university
Modern secondary school

General school Vocational/teacher-training school
Modern elementary school Private popular school (sishu)
Figure 14.3 The education system after 1905
a tripartite system was introduced, with parallel tracks for general, vocational,
and teacher-training schools above the secondary education level. General
schools emphasized academic learning, vocational and technical schools pro-
vided job-specific skills for a wide arrange of trades and professions, and
teacher-training schools were intended to churn out a large number of eligible
teachers for this rapidly expanding education system. Throughout the period
under study, general education remained the dominant form of schooling.
Second, unlike the Confucian teaching system under which elementary-
level schools were mainly provided by private households and local commu-
nities, for the first time greater public effort went into primary education.56
Public funding was more supportive towards primary education while the
private sector was allowed to play a greater role in high-level education. The
proportion of public primary schools (excluding sishu) rose from 75 percent in
the 1910s to 95 percent in the 1940s, indicating the state’s strong attempt to
make at least primary education become available to the masses.57
The most far-reaching change occurred in educational content. The
Confucian teaching system only provided incentive for young talents to
study Confucian classics, which was believed to have prevented the rise of
modern science and technology in China. When the imperial government
abolished the examination in 1905, classical Chinese lost its appeal and
modern subjects were incorporated into the new curriculum. For public
56
Universal access to education that is at least partly provided by the state first emerged in
Europe and North America from the early nineteenth century onwards, while China’s
attempt to publicly provide mass education to its children started relatively late.
57
Gao, Risen from the Chaos, p. 286.
510

schools, the state drafted clear regulatory curriculum models for each level of
schooling, and the teaching content emphasized new academic learning,
technological know-how, and new ideological campaigns. As an example,
Table 14.3 presents changes in regulatory models for primary school curricu-
lums by the Ministry of Education from 1904 to 1948, and a similar pattern can
be observed in the curriculum models for secondary schools. The major
change was a substantial decline in the weight given to the Chinese classics; at
the same time new Western subjects, including arithmetic, physics, and
geography, together with new moral doctrines, e.g. nationalism and the
“Three Principles of the People,”58 were introduced to the masses and spread
through the expanding education system. The improvement in curriculum
was not confined to newly established public schools. To ensure that all types
of schools conformed to similar teaching standards, government also urged
still-existing sishu to reform their teaching content.
However, we would also like to stress that the real improvements were
less than one might have expected. The changes in curriculum bore at best
a limited resemblance to the road map provided by the state. For instance,
the government placed explicit eligibility requirements on teachers’ educa-
tion backgrounds as it was essential to appoint teachers who had the
Table 14.3 Regulatory curriculum models for primary schools (percentage)

Course 1904 1912 1922 1936 1948
Chinese and Confucian classics 35 20 20 17.5 15

Arithmetic 10 15 17.5 15 12.5
English 10 15 17.5 15 15
Art and physical education 10 10 10 15 10
Manual work 5 15 10 15 15
Social science History 20 15 15 10 15
Geography
Natural science Physics 10 10 10 12.5 17.5
Chemistry
Note. Since 1912, the Confucian classics ceased to form an independent course in
schools; what remain are some selections in the textbooks of Chinese courses
Source: Shu Xincheng 舒新城, 近代中國教育史料 (A Summary of Historical
Materials of the New Education in China) (Shanghai, Zhonghua shuju, 1928), p. 79
58
This was a political philosophy developed by Sun Yat-sen. The Three Principles
comprised nationalism, democracy, and the livelihood of the people.
511

academic competence to deliver the right training to students;59 however,

such rules were never implemented in practice. Roughly 80 percent of
primary-school teachers still failed to meet the standard set by the Ministry
of Education until the 1940s;60 and about 65 percent of sishu in 1935 continued
to teach the old Confucian content without remodeling their curriculum.61
Similarly, the quality of high-level education also failed to meet standards.
A survey of the China Vocational Education Association reports that a large
share of their graduates was far from proficient in their command of profes-
sional knowledge and skills.62
Apart from insufficient numbers of eligible teachers and a shortage of
funding, the implementation of the new system also encountered strong
resistance from local people. To a great extent, the general public showed
little interest in or understanding of this new education system, especially
during the early stages of its development. In fact, this highly Western-
influenced system was widely considered to be a severe threat to the endur-
ing cultural and social norms of local society. The Republican government
initiated a national program known as the “Temple Destruction Movement,”
which encouraged local governments to seize the assets of Buddhist and
Daoist temples in order to support the expansion of modern education.63
Such destruction of establishments symbolic of traditional culture led to
widespread resistance. The central government indeed exerted various
efforts and tried to smooth the transformation and gain more support from
the general population. For instance, graduates under the new system were
allowed to have degree titles equivalent to those under the traditional
system.64 In the same vein, the admission requirements for modern
59
For teaching posts in primary schools, only graduates of teacher-training schools (equiva-
lent to secondary school) were eligible; and for a secondary-school post, a degree from an
advanced teacher-training school (equivalent to a university) was required.
60
Chen Dongyuan 陳東原 (ed.), 第二次中國教育年鑑 (The Second Education
Yearbook) (Shanghai, Shangwu Yinshuguan, 1948).
61
Wu Jiping 吳寄萍, 改良私塾 (Improved Sishu) (Shanghai, Zhonghua shuju, 1939).
62
Wang Ji 王婕, 民國時期大學生就業研究 1912–1937 (Research on the Employment of
Tertiary-Education Students in Republican China 1912–1937) (Zhengzhou, Zhengzhou
University, 2012).
63
P.R. Katz, Religion in China and Its Modern Fate (Waltham, MA, Brandeis University
Press, 2014).
64
All students who had at least obtained a senior primary-school diploma could be
awarded such traditional scholarly titles as jinshi, juren, and gongsheng, and various
official ranks. For example, after finishing a minimum of three years of study in a college
with at least medium grades in the graduation examinations, graduates could be
awarded the academic title of juren, and they were eligible for official posts equal to
assistant magistrate. Alternatively, they could be employed by industrial and commer-
cial companies, or start their own.
512

secondary schools were tolerant and open to students from sishu that had
reformed their teaching content.
Apprentices in a Time of Change

The rise of formal vocational education presented strong competition to the
traditional apprenticeship. These two educational tracks provided similar on-the-
job training with different advantages. Formal vocational education was geared
towards modern industries and offered students formal educational credentials,
while apprenticeships had deeper roots in traditional professions, such as handi-
crafts, which continued to dominate in industry and accounted for a large part of
the economy.65 From 1900 onwards, a series of new polices were issued to
regulate apprenticeship, including training quality and the rights of
apprentices.66 In the meantime, the imperial government also established
a large number of state-owned handicraft workshops.67 Between 1902 and 1911,
757 such agencies were set up in twenty-two provinces, and each workshop could
hire dozens to hundreds of apprentices.68 The enrollment scale was further
expanded after 1911. A survey conducted in 1913 shows that 63.7 percent of
workshop owners in the machinery industry in Shanghai started their career as
apprentices.69 Even though we cannot pin down accurately the number of
apprentices participating in these training programs, it is clear that apprenticeship
continued to play a crucial role in the training of skilled workers in early
twentieth-century China.
65
For instance, 72 percent of the output value of manufacturing industry was contributed
by handicraft industries in 1933. See Wu Baosan 巫寶三, 中國國民所得 1933 (Chinese
National Income in 1933) (Beijing, Shangwu Yinshuguan, 2011).
66
In 1929, the Factory Law (gongchangfa 工廠法) issued by the national government in
Nanjing stipulated that factory owners had a duty to educate their apprentices, and in
1932 the Revised Factory Law (xiuzheng gongchangfa 修正工廠法) gave further protec-
tion to apprentices’ right to education by basically setting the final level attained during
apprenticeship equal to that of senior primary school. See Peng Nansheng 彭南生, 行
會制度的近代命運 (The Fate of the Guild System in Modern Times) (Beijing,
Renmin chubanshe, 2003).
67
These agencies include Gongyiju 工藝局, Gongyi Chuanxisuo 工藝傳習所, or
Quangongchang 勸工場.
68
Chen Shaowen 陳紹聞 and Guo Xianglin 郭庠林, 中國近代經濟簡史 (A Brief
Economic History of Modern China) (Shanghai, Shanghai renmin chubanshe, 1983).
69
Peng Zeyi 彭澤益 (ed.), 中国近代手工业史资料 1840–1949 (Historical Achievements
of the Handicraft Industry in Modern China 1840–1949) (Shanghai, Sanlian chubanshe,
1957).
513

Missionary Education
One important event that laid the foundation for China’s modern education
system was the appearance of mission schools. The presence of Christian
missionaries expanded significantly with the opening up of China following its
defeat in the Opium War (1842), and one important missionary activity was the
founding of schools. Not only did the number of mission schools expand quickly;
they also functioned as a role model of Western educational practice for Chinese
schools. For instance, they influenced their Chinese conterparts by introducing
a qualitatively new curriculum of subjects. By 1920, missionaries had established
lower primary schools in 719 counties and higher primary schools in 314 counties,
which enrolled 122,089 and 30,067 students respectively. While enrollment
figures for high schools are not available, a survey conducted in the 1920s reveals
that there were 254 secondary schools spread across 9 percent of sampled
counties.70 For tertiary education, from the late nineteenth century, a number
of Christian colleges were also built. With only three domestic state-managed
universities existing at that time,71 Christian colleges not only presented a model
of tertiary education to their Chinese counterparts but also, by educating a large
number of professionals and intellectual elites, had a long-lasting social, eco-
nomic, and political influence in twentieth-century China.
Measuring Educational Expansion

Studies on literacy rates suggest a rapid expansion of basic education over the
early twentieth century. Ch’ien finds that the literacy rate was about 70 per-
cent in the 1920s, which is a substantial jump compared to the level in the late
nineteenth century.72 Gamble and Burgess show that 84 percent of residents
in Beijing were literate.73 In urban Canton, the literacy rate was also believed
to be as high as 80 to 90 percent in the 1930s.74 However, contradictory
70
M.T. Stauffer, Christian Occupation of China: A General Survey of the Numerical Strength
and Geographical Distribution of the Christian Forces in China (Shanghai, China
Continuation Committee, 1922); Y. Bai and J.K.S. Kung, “Diffusing Knowledge While
Spreading God’s Message: Protestantism and Economic Prosperity in China, 1840–
1920,” Journal of the European Economic Association 13.4 (2015), 669–98.
71
E. Widmer and D.H. Bays, China’s Christian Colleges: Cross-cultural Connections, 1900–1950
(Stanford, CA, Stanford University Press, 2009)
72
T.S. Ch’ien, The Government and Politics of China, 1912–1949 (Stanford, CA, Stanford
University Press, 1950). pp. 13.
73
S.D. Gamble and J.S. Burgess, Peking: A Social Survey (New York, George H. Doran,
1921), p. 507.
74
J.L. Buck, Land Utilization in China: A Study of 16,786 Farms in 168 Localities, and 32,256
Farm Families in Twenty-Two Provinces in China, 1929–1933 (Chicago, The University of
Chicago Press, 1937)
514

evidence can also be found in the existing literature. For instance, in Ting
county of Zhili province, only 20 percent of the population over six years old
in the 1930s could be considered literate.75
Systematic enrollment data are available for modern education institutions
from the early twentieth century onwards, and we present enrollment rates
per 1,000 school-aged children in China for schooling at all levels in
Table 14.4. Before discussing the figures, it is important to stress that
Table 14.4 only includes modern public schools that were financed at least
in part by public funding due to the absence of sishu in official statistics. To
facilitate international comparison, a broader school-age band (five to four-
teen) is adopted for primary schooling, twelve to seventeen for secondary
schooling (twelve to fourteen for lower secondary schools and fifteen to
seventeen for higher secondary schools), and seventeen to twenty-two for
higher education.76
Looking at primary education first, enrollment ratios rose from 1.2 percent in
the early twentieth century to 12 percent in the 1930s, which is a similar level to
that of India (11.3 percent) but much lower than that of Japan (60.9 percent).77 The
substantial rise in primary schooling was interrupted by the Japanese invasion
between 1937 and 1945, which was then followed by a four-year civil war. The
enrollment ratio dropped sharply in wartime and recovered slowly, and the
prewar level was only reattained after 1947. Even though official records exclude
sishu, selective regional evidence indicates that they remained very popular
through the first half of the twentieth century. In 1924, the China Education
Department reported that the number of students attending sishu should be at
least the same as the number of students in formal primary schools. Similarly,
approximately 20,000 students were enrolled in 1,000 traditional popular schools
in Canton city, a figure higher than the number of students in formal primary
schools.78 In 1930, Zhili province reported enrollment ratios for both traditional
and modern primary schools, which also presents the same pattern of these two
types of elementary-level schools being substitutes. In 1935, a nationwide report
records that traditional popular schools accommodated 1,752,014 students while
75
S.D. Gamble and T. Hsien, A North China Rural Community (New York, Institute of
Pacific Relations, 1954), p. 185.
76
P.H. Lindert, Growing Public, vol. 1,The Story: Social Spending and Economic Growth since
the Eighteenth Century (Cambridge, Cambridge University Press, 2004); C.D. Goldin and
L.F. Katz, The Race between Education and Technology (Cambridge, MA, Harvard
University Press, 2009)
77
Lindert, Growing Public, vol. 1, pp. 91–03.
78
Tao Zhixing 陶知行 (ed.), 中國教育統計概覽 (A Summary of Chinese Education
Statistics) (Shanghai, Zhonghua Jiaoyu Gaijinshe, 1923), p. 6.
515

Table 14.4 Enrollment rates per 1,000 school-age

population, 1900–1950
China
Year Primary Secondary Tertiary
1907 12 0.7 0.07

1916 47 1.7 0.07
1922 72 2 0.05
1933 119 7.5 0.05
1949 214 20.7 0.21
Year India
1900 47 21.2 0.9
1910 65 35.1 1.3
1920 80 42.1 2.6
1930 113 65.7 3.1
1950 166.1 10.8
Year Japan
1900 507 50 5.7
1910 599 132.4 10.7
1920 602 310.7 15.9
1930 609 505.4 29.8
1950 695.1 46.5
Year United States
1900 939 106
1910 975 145
1920 924
1930 921 511
1950 978 745
Sources: for China, the enrollment data come from Yearbooks

on Education in corresponding years, and population and
age structure are from Hou Yangfang 侯楊方,中國人口
史：民國時期 (Population History of China: Republic of
China Period) (Shanghai, Fudan daxue chubanshe, 2001); for
the United States, India, and Japan the data come from P.
H. Lindert, Growing Public, vol. 2, The Story: Social Spending
and Economic Growth since the Eighteenth Century (Cambridge,
Cambridge University Press, 2004)
the number of students enrolled in modern primary schools was 15,110,119.79 In

short, sishu remained popular and functioned as a substitute for formal primary
79
Wu, 改良私塾.
516

Table 14.5 Composition of persons by education level for selected counties,

1880–1920 (percentage)
Educational attainment
Primary
schools
No Secondary Higher
Province County Year education Sishu Modern schools education
Fujian Fuqing 1880 90 10 0 0 0

1900 83 12 5 0 0
1920 82 9 9 1 0
Sichuan Chongqing 1880 78 22 0 0 0
1900 76 24 0 0 0
1920 69 27 2 1 1
Zhili Beijing 1880 89 11 0 0 0
1900 87 11 2 0 0
1920 86 10 4 1 0
Sources: Cao Ning 曹寧, 民國人口戶籍史料續編 (Population Sequel Collection in

the Period of the Republic of China) (Beijing, Guojia Tushuguan chubanshe, 2013);
Second Suburban Bureau of Beiping Bureau of Public Security 北平市警察局郊二
區分局, 1946 年 11 月戶口調查表西店和西洼村 (Household Survey Table of Xidian
and Xiwa villages, November 1946); Second Suburban Bureau of Beiping Bureau of
Public Security 北平市人民政府公安局人民政府公安局郊二分局, 戶口調查表
(Household Survey Table), 1949–4-18; First Suburban Bureau of Beiping Bureau of
Public Security 北平市警察局郊一區分局, 民国 37 年户口调查表 (Household
Survey Table, 1948)
schools until the 1940s. Hence, if we include these informal schools, China’s
educational outcome should not be as low as suggested by the official statistics
shown in Table 14.4 (also Table 14.5).
The progress of secondary education must be based on a large pool of
primary schools, which means that the early twentieth century in China does
not appear to be a time of rapid expansion. As shown in Table 14.4, in 1907
only about one in every 1,000 school-age children (i.e. 0.1 percent) attended
secondary school, and this figure was far lower than the level of India
(2 percent) and Japan (13.9 percent). Secondary-schooling expansion acceler-
ated only after the 1920s. This may be attributed to the separation of lower
secondary school from higher secondary school in 1922.80 A similar pattern
80
Issuing lower secondary-school diplomas to students who finished three years of
secondary schooling partially accounted for the significant decline in the number of
dropouts. See Zhu Youhuan 朱有瓛, 中國近代學制史料第三輯下 (Historical
517

Table 14.6 Composition of secondary schools by type

(percentage)
Year General Teacher-training Vocational
1912 60.1 30.4 9.5

1916 67.8 20.6 8.7
1922 49.9 35.1 16.8
1930 62.6 28.3 9.1
1937 65.4 19.2 15.4
1941 73.3 14.5 12.2
1945 73.5 15.2 11.4
Source: Chen Dongyuan 陳東原 (ed.), 第二次中國教育年鑑 (The

Secondary Education Yearbook) (Shanghai, Shangwu Yinshuguan,
1948), p. 1428
can be seen when looking at tertiary education,81 the scale of which remained
minimal throughout the period of study, with enrollment ratios below
0.1 percent before the 1940s. To put these figures into perspective, the
enrollment ratio was 4.05 percent in Japan at this time, and in India 0.48 per-
cent (i.e. more than four times the level for China).82 The regional variation
was significantly wide, though, with 60 percent of universities being concen-
trated in metropolises like Beijing and Shanghai.
Apart from general education, vocational and teacher-training education also
expanded quickly under the new system. As shown in Table 14.6, general schools
still dominated in secondary education, while about 30 to 40 percent of students
chose vocational and teacher-training tracks. The total number of students in
vocational schools in 1912 was around 32,000; this figure rose to 50,000 in the mid-
1930s. One thing worth stressing is that teacher-training schools were particularly
important, especially in the early stages, because training eligible teachers was
crucial for the new education system. Furthermore, these teacher-training
schools also played an essential role in female education, since teaching was
regarded as one of the few acceptable career options for females.
Materials of the Modern Educational System in China), vol. 3, part 2 (Shanghai,
Huadong Shifan daxue chubanshe, 1992).
81
Tertiary education includes universities (four-year program) and colleges (usually
institutions that offered a two-year program).
82
One thing worth stressing is that the Japanese invasion did not greatly impact tertiary
education. Most of the universities, located in big cities, were able to take advantage of
the relative safety of the international settlements and French concessions in these
cities. Meanwhile, the universities located in occupied China did not cease to operate;
they merely fled south and continued to function throughout the war.
518

Education and Socioeconomic Changes

As discussed, the education system in China underwent unprecedented
expansion and modernization between 1800 and 1950, which affected the
society profoundly in various ways. In this section, we briefly discuss three of
the main socioeconomic changes, namely economic modernization, female
education, and the rise of new intellectual elites.
Educational Development and Economic Modernization

There is a growing body of work that studies the association between
the introduction of Western-influenced new institutions and economic
development in China in the late nineteenth and early twentieth
centuries.83 However, empirical studies directly examining the economic
and social impact of this educational movement remain limited.84 Did the
rise of modern education contribute to the development of modern industry
and economic growth in China? How did the new education system affect
social mobility?
Some studies specifically look at the impact of the abolishment of the civil
service examination. The civil service examination system did not incentivize
young talents to study practical knowledge, which was an institutional
obstacle to the rise of modern technology. The rapid expansion of modern
schools with a modern curriculum could cultivate new intellectual elites and
form professional human capital for various modern industries in China.
Such changes could greatly promote modernization and industrialization in
late nineteenth- and twentieth-century China. For instance, as presented in
Table 14.7, the number of engineers in China expanded rapidly after the birth
cohorts of the 1890s. Bai also finds that misallocation of talents largely existed
under the civil service examination system, therefore regions with better
talent pools established significantly more modern firms after the abolish-
ment of the old examination system.85
83
See D. Ma, “Why Japan, Not China, Was the First to Develop in East Asia: Lessons from
Sericulture, 1850–1937,” Economic Development and Cultural Change 52.2 (2004), 369–94;
D. Ma, “Economic Growth in the Lower Yangzi Region of China in 1911–1937:
A Quantitative and Historical Analysis,” Journal of Economic History 68.2 (2008), 355–92;
R. Jia, “The Legacies of Forced Freedom: China’s Treaty Ports,” Review of Economics and
Statistics 96.4 (2014), 596–608.
84
There are no census or individual-level survey data that include questions about both
income and education for this period.
85
Y. Bai, “Farewell to Confucianism: The Modernizing Effect of Dismantling China’s
Imperial Examination System,” Journal of Development Economics 141 (2019), 102382.
519

Table 14.7 Number of engineers by birth cohort

Birth cohort 1870s 1880s 1890s 1900s 1910s
Number of engineers 10 270 1,206 2,503 3,640
Source: Design Board of National Defense 國防設計委員會, 中國工程名人錄

(Who’s Who of Chinese Engineers) (Chongqing, Shangwu Yinshuguan, 1941)
Conversely, the rise of modern industries and commerce further stimulated

the demand for modern education from the late nineteenth century onwards.
New talents were in short supply especially in those emerging sectors, such as
the railroad, the telegraph, steamships, the mining industry, law, and modern
manufacturing. Utilizing a railroad company’s employee records, Yuchtman
finds that the wage premium of modern education was higher than that of
traditional education, indicating that the number of newly educated elites did
not meet market demand.86 Another example is that the Tianjin Technology
and Business University reported that their graduates, especially those that
majored in engineering and business, had very bright career prospects in the
1930s. Graduates with business majors mainly landed good posts in the modern
banking industry and the public sector, while engineering graduates had even
more diverse career choices, ranging from construction to rail transport, water
conservancy, civil engineering, and municipal administration.87
However, we cannot be overly optimistic about the extent of modern
education’s contribution to economic development. First, as shown in
Table 14.4, only a very selective group of people received post-elementary
education. Liang et al. analyzed the family background of 110,000 university
students during the 1906–1952 period and found that more than 60 percent of
them had parents with high-income occupations and came from large cities.88
Another problem lies in the pronounced mismatch between students’ choice
of major and the labour market’s needs. The rapid expansion of high-level
education did not meet the substantial rising demand for modern human
86
N. Yuchtman, “Teaching to the Tests: An Economic Analysis of Traditional and
Modern Education in Late Imperial and Republican China,” Explorations in Economic
History, 63 (2017), 70–90.
87
Chen Yanzhong 陳炎仲, “工學院之過去未來” (The Past and the Future of Tianjin
Technology and Business University), Gongshang Xuesheng 1.4 (1937), 6.
88
Liang Chen 梁晨, Dong Hao 董浩, Ren Yunzhu 任韻竹, and Li Zhongqing 李中清,
“江山代有才人出—中國教育精英的來源與轉變 1865–2014” (Social Transformation
and Elitist Education: Changes in the Social and Geographical Origins of China’s
Educated Elites 1865–2014 ), Sociological Research 3 (2017), 48–70.
520

capital in China partly because science and engineering remained unpopular

among students. In the 1930s, over 70 percent of university graduates chose to
major in the arts and social science, even though these majors did not
guarantee bright career prospects.89 As Zhu Jiaye, the minister of education
in Republican China, once pointed out, unemployment was widespread
among university graduates. On the contrary, experts with specialized skills
or professional knowledge, especially in the sciences and engineering, were in
short supply.90 According to an official report of Shanxi province, the total
number of graduates from colleges and universities was 8,905 in the three
decades after 1905, and about half of them (4,700) were out of work.91 As
a response, the central government even began to issue regulations to restrict
the number of students majoring in the arts from 1934, with the effect that the
share of students choosing science subjects rose to 59.4 percent in 1939.
Furthermore, the efficiency of the labor market at the time was also ques-
tionable since a large proportion of students never had the opportunity to
apply the knowledge and skills that they had obtained in their own jobs. For
example, the National Federation of Education in 1917 reported that some
graduates trained in textile schools ended up as teachers in primary schools
and those that attended agricultural schools worked as general administrative
assistants.92
Rising Gender Equality

Besides economic modernization, another social improvement under the
modern education system was the rising gender-neutrality in education.
Gender inequality in education is shared by many developing countries,
and in Chinese history females had always been excluded from formal
education. Not all women were illiterate, at least among those from gentry
families. Girls could receive home education in literature, female ethics,
and skills that would make them more competitive in the
marriage market. However, even this type of training was not available
for girls in the lower social classes in undeveloped areas. For example,
around 1900 less than 1 percent of school-age girls in Yugan county (Jiangxi
province) and Chenggong county (Yunnan province) had ever attended
89
Wang Jie 王婕, 民國時期大學生就業研究 1912–1937 (Research on the Employment
of Tertiary-Education Students in Republican China 1912–1937) (Zhengzhou,
Zhengzhou University, 2012).
90
Wang, 民國時期大學生就業研究.
91
“太原通訊” (Taiyuan Communication), Takungpao, September 24, 1934.
92
Zhu, 中國近代學制史料第三輯下.
521

sishu.93 Allowing women the right to formal education after the turn of the
twentieth century profoundly changed women’s social position and the
landscape of the labor market.
As noted, mission schools were the pioneers in educating women in China.
In 1844, Miss Aldersey established the first mission schools specifically for girls
in Ningbo, and a great number of mission schools exclusively for girls were
opened in the following few decades.94 In 1907, the Ministry of Education
promulgated statutes which paved the way for the establishment of a great
number of primary schools and teacher-training schools exclusively for girls;95
and from 1912 fully functional mixed-gender schools were also allowed. The
share of female students attending senior primary school in 1923 was 6.34 per-
cent and by 1930 the figure had more than doubled, reaching 15 percent.96
More systematic records are presented in Table 14.8, showing that the gender
gap in secondary-school enrollment was pronouncedly narrowed. After 1916,
the number of female students attending secondary school began to rise at an
unanticipated pace, such that the female–male student ratio fell from one to
seventy-nine in 1912 to one to seven in 1929 and continued to fall, finally
reaching one to three in 1946. Looking at tertiary education, only 887 female
students attended universities or colleges in 1922, accounting for 2.5 percent of
all students, and this figure rose to 15.2 percent in 1936.97
Two patterns are worth highlighting. First, the private sector played an
important role in promoting equal gender opportunities in education. By
1907, 428 schools exclusively for girls had been established, of which more
than half were private. Second, another growth engine came from the
teacher-training schools.98 The tuition fee for teacher-training schools was
substantially less than that for general education, while another attractive
feature was that after graduation a job was guaranteed for every student,
93
Yin and Tian, 民國人口資料彙編; Yugan county, Shangrao city, Jiangxi province,
Fifth township, Fourth district, Household registration books, c. 1947.
94
L.-D. Djung, A History of Democratic Education in Modern China (Shanghai, The
Commercial Press, 1934).
95
Wang Zhongtian 汪忠天, “國內大學及專門學校畢業生就業狀況的一個調查” (A
Survey of the Employment Status of Graduates of Domestic Universities and Colleges),
Chinese Educational Circles 22.6 (1934), 49–97.
96
Li Huaxing 李華興, 民國教育史 (The History of Education through the Republic of
China) (Shanghai, Shanghai Jiaoyu chubanshe 1997), p. 729.
97
Yu Qingtang 俞慶棠, “三十五年來中國女子之教育” (Female Education in China
during the Past Thirty-Five Years), in Zhuang Yu 莊俞 (ed.), 最近三十五年之中國
教育 (Education in China during the Past Thirty-Five Years) (Shanghai, Shangwu
Yinshuguan, 1931), pp. 175–214.
98
The proportion of females in general secondary schools before 1930 did not exceed
4 percent, but girls accounted for about 18 percent of normal-school students (equiva-
lent to secondary school). See Tao, 中國教育統計概覽, p. 4.
522

Table 14.8 Gender composition in secondary schools (1912–1946)

Female students Male students
Year Number Proportion Number Proportion
1912 677 0.01 51,423 0.99

1913 470 0.01 57,510 0.99
1914 956 0.01 66,298 0.99
1915 948 0.01 68,822 0.99
1916 724 0.01 60,200 0.99
1922 3,249 0.03 100,136 0.97
1929 33,073 0.13 215,595 0.87
1930 59,939 0.15 337,009 0.85
1933 73,667 0.18 342,281 0.82
1946 379,087 0.26 1,106,060 0.75
Source: Education Committee 教育審委員會 (eds.), 第一次中國教育

年鑑 (The First Chinese Education Yearbook) (n.l., Ministry of
Education Press, 1934)
including females. Together with medical care and journalism, teaching was
one of the most popular career choices for the “new women” throughout the
first half of the twentieth century.99
The development of female education did not, however, suggest an
equivalent increase in female participation in the labor market. Education
could be considered a new fashion or accessory rather than a means to
economic independence for women. Therefore many female students, espe-
cially those with high-level education, went back to family life after gradu-
ation. During the mid-1930s, a survey in urban Canton showed that only
16.4 percent of the female population was employed.100
The Rise of New Intellectual Elites

The upper tail of human capital has been found to be an essential driving
force for modernization and industrialization in Europe.101 From the late
nineteenth century onwards, an emerging new intellectual elite group was
99
To attract more students, those who enrolled in teacher-training schools were exempt
from paying tuition fees, while their living expenses were also covered by public
funding; in recompense, after graduation they had to serve at least three years in
a local primary school. See Li, 民國教育史, pp. 512–13.
100
Jin Zhonghua 金仲華, 婦女問題 (Women’s Issues) (Shanghai, Shangwu chubanshe,
1936).
101
M.P. Squicciarini and N. Voigtländer, “Human Capital and Industrialization: Evidence
from the Age of Enlightenment,” Quarterly Journal of Economics 130.4 (2015), 1825–83.
523

formed in China, including students returning from overseas together with

graduates from major domestic universities. These knowledge elites exerted
a profound impact on Chinese society over the course of the following
decades.
Following the educational reform, there was an ever-growing number of
Chinese who left home to study overseas. According to Wang, around
100,000 students went abroad during the first half of the twentieth century
to study, and the majority of them chose to come back to China after
graduation.102 Among all destinations, Japan was the most popular because
of its geographic proximity and cultural similarity, and also because it was
much less expensive compared to the US and Europe. Between 1900 and 1911,
90 percent of the 20,000 Chinese studying overseas went to Japan.103
These foreign-educated scholars directly helped to professionalize this
newly established education system in China, especially for tertiary
education.104 Over the period of study, a very large proportion of faculties
and administrators working in major Chinese universities received their
education abroad.105 For example, in 1937, sixty-nine of a total of ninety-
four professors in Tsinghua University had studied in the US.106 These
scholars brought back not only knowledge but also new ideas to reform
existing educational institutions. One important case is that Peking
102
Wang Qisheng 王奇生, “民國時期歸國留學生的出路” (The Employment of
Foreign-Educated Students during the Republican Era in Chinese History), Journal of
Republican Era Studies 3 (1994), 12–14.
103
L. Yao, “The Chinese Overseas Students: An Overview of the Flows Change,” 12th
Biennial Conference of the Australian Population Association (2004), 15–17.
104
Zhou Mian 周棉, “留學生群體與民國時期新式教育體制的建立” (The Group of
Students from Overseas and the Establishment of a Modern Education System in the
Period of Republican China), Zhejiang Academic Journal 5 (2012), 59–68.
105
Using Beijing University as an example, Cai Yuanpei, the first chancellor, studied at the
University of Leipzig for four years from 1907. His successors, Jiang Menglin and Hu
Shi, were returnees from the USA. Jiang stayed in the USA for nearly ten yeas and got
his doctoral degree at Columbia University. Hu went to Cornell University in 1910 and
finally defended his Ph.D. thesis at Columbia University. Chancellor Mei Yiqi of
Tsinghua University, who shouldered the position for seventeen years; President
Zhu Kezhen of Zhejiang University; President He Lian of Nankai University;
Presidents Liu Tingfang and Mei Yibao of Yenching University; President Yang
Yongqing of Soochow University; President Li Zhaohuan of Shanghai Jiaotong
University; Presidents Gu Yuxiu and Wu Youxun of National Central University;
and others, were all outstanding students returned from overseas.
106
Tsinghua University was funded by the “Gengzi Indemnity,” also known as the “Boxer
Indemnity.” The same funding source also supported many outstanding students
studying abroad. See Tsinghua University, 清華同學錄 (Schoolmate Records of
Tsinghua University) (Beijing, Tsinghua University, 1937); Yi Chu Wang, Chinese
Intellectuals and the West, 1872–1949 (Chapel Hill, University of North Carolina Press,
1966).
524

University, led by Cai Yuanpei, who came back from Germany, quickly grew
into an independent academic institute with a democratic management
system and the spirit of “academic freedom and inclusiveness,” and continues
to be one of the best universities in China to this day.107
The influence of these new knowledge elites was not confined to class-
rooms. New ideologies, including democracy, nationalism, and communism,
were introduced and spread to a much wider group of people outside the
universities through public talks. Some of the auditors, including Mao
Zedong, Hua Luogeng, and Shen Congwen, were strongly inspired and
continued to shape the course of Chinese history. For instance, the New
Culture Movement, which was a movement to lead a revolt against trad-
itional Confucianism, has been considered to be an important watershed,
marking a break between tradition and modernity in China; the leaders and
followers of this movement also clustered in major universities, such as
Peking, Tsinghua, and Fudan Universities.108 Apart from educational institu-
tions, these new knowledge elites also played an active role in other centers
of literature and intellectual activity, such as publishing houses, journals, and
literary societies throughout the first half of the twentieth century.
Lastly, this elite group with modern education also had a strong impact in the
political arena, as many were recruited for government positions at all levels. For
example, 22.5 percent of the staff working in the central government in 1916 had
received education aboard.109 In the Ministry of Agriculture, Industry, and
107
Research association of studies on Cai Yuanpei, 蔡元培全集 (Collected Works of Cai
Yuanpei), 4 vols. (Hangzhou, Zhejiang Jiaoyu chuban jituan, 1997). Founded in 1898,
Peking University was originally known as the Imperial University of Peking. The
admissions, the recruitment of teachers, and the designation of graduates were all
controlled by the central government. It also served as the highest administration for
education at its founding. As a result, it was unavoidably affected by bureaucratic
ideology and practices. The main purpose of students who came here was a promising
political career. When Cai took the position of president, he emphasized learning and
research as the most important tasks of both teachers and students. Academic author-
ity was highly esteemed and became dominant instead of administrative power. He
invited many leading scholars full of enthusiasm to be teachers to join his team, while
those teachers who did not aspire to academic work were gradually removed. In this
way, Peking University became so attractive to those who were really interested in
academia that many of the best scholars in that era chose to work there. See
Feng Youlan 馮友蘭, “我所認識的蔡孑民先生” (My Thoughts on Mr. Cai Jiemin),
People’s Daily Overseas Edition, January 9, 1988.
108
Zhou Mian 周棉, “論留學生群體作用於民國社會發展的諸種互動關係”
(Impacts of the Group of Foreign-Educated Students on Social Development during
the Republican Period of China), Zhejiang Academic Journal 5 (2015), 51–61.
109
Tang Yueliang 唐悅良, “青年會與留學生之關係” (The Relationship between
Qingnianhui and Returning Students from Overseas), Eastern Miscellany 14.9 (1917),
196–8.
525

Commerce, 49.8 percent of employees had foreign tertiary-education certificates,

and this figure was about 40 percent in the Ministry of Transport and the
Ministry of Justice.110 Another long-term legacy of these new knowledge elites
in the Chinese political arena was the creation of several political movements
from the late nineteenth century. The earliest revolutionary organizations that
tried to overthrow the Qing dynasty were founded outside China, and a majority
of their early members were overseas students. About 90 percent of the earliest
participants of the Tongmenghui (Alliance Society) were students who had
returned from abroad. Similarly, after the foundation of Republican China, in
the major political parties, Nationalist Party and the Communist Party, more
than half of top leaders were foreign-educated students.111
Conclusion
A rich literature documents the important role of education and human capital
for economic development in the modern world. China has a very long educa-
tional history, but the rapid expansion of mass education and modernization in
the curriculum started relatively late, only after the early twentieth century. Over
the period studied in this volume (from 1800), drastic changes occurred in the
education system, and this chapter has attempted to survey several milestones.
The traditional Confucian teaching system had been operating in China for
over a thousand years. Under this system, state bureaucrats and social elites were
recruited using a merit-based examination. The provision of education was
mainly taken on by private parties and the Confucian classics were used as
teaching content. The major transformation in the education field during the
period of study was the shift from this long-lasting traditional Confucian teaching
system to a modern one that approximated a Western model. This change was
a slow and difficult process that started in the 1860s, and generated the following
three important improvements: public provision of mass education, curriculum
reforms, and an increase in educational opportunities for girls. The new system
educated millions of students, formed professional human capital, cultivated new
intellectual elites, and encouraged modern economic growth in China.
110
Tang, “青年會與留學生之關係,” 196–8.
111
C. Hsueh, Huang Hsing and the Chinese Revolution (Stanford, Stanford University Press,
1961). M. Zhou, “留学生与中国同盟会的创建” (Returned Students from Abroad and
the Establishment of Tongmenghui), Journal of Tsinghua University (Philosophy and
Social Sciences) 23.4 (2008), 46–54. Zhou Mian, “論留學生群體作用於民國社會發展
的諸種互動關係”; Wang Xiaoting 王效挺 and Huang Wenyi 黃文一, 戰鬥在北大
的中國共產黨人: 1920–1949 (CPC Members in Peking University 1920–1949) (Beijing,
Peking University Press, 1991).
526

Even though the traditional education system discussed in this chapter

ended more than a century ago, its legacy remains in China to this day. For
example, the institutional heritage of the imperial examination system can
still be found in national college entrance examinations, and admission for
major universities still uses regional enrollment quotas just like the civil
service examinations. Besides China, ninety-one out of 156 countries world-
wide still recruited employees in the public sector via competitive examin-
ations in 2015. Another important impact of the long history of education in
China is the persistent culture of placing a high value on success in education.
Chinese continue to emphasize human capital and education as an effective
ladder for upward social mobility today. Despite the lack of public invest-
ment, Chinese parents are willing to invest heavily in their children’s
education.112 Empirical evidence confirms that the historical regional patterns
in educational achievement throughout history are associated with modern
educational outcomes through the transmission of people’s attitudes towards
education by successive generations.113 The persistence of this cultural trans-
mission was partly interrupted during the Mao era (1949–1977), during which
the merit of education was largely denounced and deadly attacks on scholars
and intellectuals were prevalent.114 The widespread educational breakdown
indeed had a negative impact on later educational outcomes, but the culture
of valuing education was only weakened shortly and continued to show
strong persistence.115
112
China’s public investment in human capital has been small in comparison with nations
at similar levels of economic development, and its geographical dispersion has been
large. J.J. Heckman, “China’s Human Capital Investment,” China Economic Review 16.1
(2005), 50–70. This means that private investment plays an essential role in human
capital formation even today. China, at all levels of government together, spent about
2.5 percent of GDP on schooling in the year 2000. In the US, this figure was 5.4 percent.
In South Korea, it was 3.7 percent.
113
T. Chen, J. Kung, and C. Ma, “Long live Keju! The Persistent Effects of China’s Civil
Examination System,” Economic Journal 130.631 (2020), 2030–64.
114
Chinese education has undergone major shifts between 1949 and 1977 in response to general
swings in Chinese policy. The interruption was particularly catastrophic during the
Cultural Revolution (1966–1976), when most schools in urban China were heavily dis-
rupted. All the tertiary-level institutions ceased regular operation with no teaching carried
out and no new students were admitted. Not only educational opportunities were
disrupted; education quality was also compromised. Education became a Party device to
cultivate political ideology rather than provide academic training. China’s educational
system only returned to normalcy after 1977 when the Cultural Revolution ended.
115
X. Meng, and R.G. Gregory, “The Impact of Interrupted Education on Subsequent
Educational Attainment: A Cost of the Chinese Cultural Revolution,” Economic
Development and Cultural Change 50.4 (2002), 935–59; J. Zhang, P.W. Liu, and L. Yung,
“The Cultural Revolution and Returns to Schooling in China: Estimates Based on
Twins,” Journal of Development Economics 84.2 (2007), 631–9; Chen et al., “Long Live
Keju!”.
527

Further Reading
Baten, J., D. Ma, S. Morgan, and Q. Wang, “Evolution of Living Standards and Human
Capital in China in the 18th–20th Centuries: Evidence from Real Wages, Age-Heaping,
and Anthropometrics,” Explorations in Economic History 47.3 (2010), 347–59.
Borthwick, S., Education and Social Change in China: The Beginnings of the Modern Era
(Stanford, CA, Hoover Institution Press, 1983).
Chen, T., J.K.S. Kung, and C. Ma, “Long Live Keju! The Persistent Effects of China’s Civil
Examination System,” Economic Journal 130.631 (2020), 2030–64.
Deng Hongbo 鄧洪波, 中國書院史 (History of Traditional Chinese Academia)
(Shanghai, Dongfang chuban zhongxin, 2004).
Elman, B.A., A Cultural History of Civil Examinations in Late Imperial China (Berkeley,
Gao, P., “Risen from the Chaos: The Emergence of Modern Education in China,” in
D. Mitch and G. Cappelli (eds.), Globalization and the Rise of Mass Education (Cham,
Palgrave Macmillan, 2019), pp. 279–309.
Ho, P., The Ladder of Success in Imperial China: Aspects of Social Mobility, 1368–1911
Jin Linxiang 金林祥, 中國教育制度通史 (A General History of Chinese Education), 6
vols. (Jinan, Shandong Education Press, 2000).
Ma Yong 馬鏞, 中國教育制度通史 (A General History of Chinese Education), 5 vols.
(Jinan, Shandong Education Press, 2000).
Perkins, D.H., China’s Modern Economy in Historical Perspective (Cambridge, MA, Harvard
Rawski, E., Education and Popular Literacy in Ch’ing China (Ann Arbor, University of
Michigan Press, 1979).
Wang, M., B. van Leeuwen, and J. Li, Education in China, ca. 1840–Present (Leiden, Brill,
2020).
Wang, Y.C., Chinese Intellectuals and the West, 1872–1949 (Chapel Hill, University of North
Carolina Press, 1966).
Zhang Jie 张杰, 清代科舉家族 (Keju Hereditary Families in the Qing Dynasty) (Beijing,
Social Sciences Academic Press, 2003).
528

part II
1950 TO THE PRESENT

15
The Origin of China’s Communist
Institutions
chenggang xu
Introduction
In the nineteenth century, the Chinese Empire – the longest-lasting empire in
human history – was the largest economy on earth with a decent per capita
GDP level. But it shrank rapidly after its collapse. Since the founding of the
PRC in 1949, China had been one of the poorest economies in the world until
the post-Mao reform, which has enjoyed high growth for three decades.1 But
a sustained slowing down since 2009 reminds us of the trend of the Soviet
economy since the mid-1970s.2
Understanding the nature of the Chinese institutions and their drastic
changes, particularly since the rise of the Chinese Communist Party (CCP),
is critically important for making sense of the Chinese economy in the past
seven decades. This is because China’s institutions have been distinctive in
the world both historically and contemporarily. After decades of unsuccessful
Republican revolution efforts, in 1949 the CCP seized power by armed force,
and transplanted Soviet-type totalitarian institutions into China. In this
I thank Patrick Bolton and Debin Ma for very helpful comments, and Nancy Hearst for
copyediting. I have benefited greatly from comments in workshops at Utrecht, CKGSB,
HUJI (Jerusalem), and the Coase Institute (Tel Aviv and Warsaw), and comments in my
public lectures at Stanford, CUHK, Corvinus, FMSH-Paris, Sinica-Taiwan, WINIR-Hong
Kong, and AEA-ACES. I acknowledge support from the CKGSB and the hospitality of
Corvinus University of Budapest, Imperial College London, and the LSE.
1
A. Maddison, The World Economy: Historical Statistics (Paris and Washington, DC,
Organisation for Economic Co-operation and Development, 2003); C. Xu, “The
Fundamental Institutions of China’s Reform and Development,” Journal of Economic
Literature 49.4 (2011), 1076–1151.
2
W. Chen, X. Chen, C.-T. Hsieh, and Z. Song, “A Forensic Examination of China’s
National Accounts,” Brookings Institution paper on economic activity, conference draft,
March 2019. The Soviet economy started to slow down steadily from the time when its
per capita GDP reached one-third of the US level. Maddison, The World Economy. In
comparison, Chinese per capita GDP (in purchasing-power parity) today is about one-
quarter of that of the US. World Bank (2021), at https://data.worldbank.org/indicator/
NY.GDP.PCAP.PP.KD.
531

chenggang xu
chapter, totalitarianism is defined (or described) as a modern party-state

institution, whereby in all respects the party totally controls (1) the
state, (2) the armed forces, (3) the economy, (4) the media, and (5) the
ideology.3
Importantly, the Chinese did not completely stick to a Soviet totalitarian
model, like other Eastern Bloc nations. Instead, taking the totalitarian insti-
tutions as their root, the Chinese institutions have been further evolving,
until today. Two waves of campaigns led by Mao – the Great Leap Forward
(GLF) (1958–1960) and the Cultural Revolution (CR) (1966–1976) – changed
Chinese institutions into what I call regionally decentralized totalitarianism
(RDT), in which some of the features of Chinese imperial institutions
reappeared prominently while fundamental features of totalitarianism were
kept. Under China’s RDT regime, the CCP, as the sole political party,
monopolizes all political power and controls the most important personnel
matters in the country, including the enormous party-state bureaucracy that
penetrates every level of the entire society. It is because of this combination
of an extremely high degree of political centralization and a high degree of
administrative decentralization that Chinese institutions are unique in the
world. Yet administrative and economic issues are highly decentralized to
party-state agents at the regional levels.
The RDT institutions were the institutional basis for the post-Mao
reforms.4 From the 1980s, under its RDT regime, Chinese institutions created
the powerful mechanisms, i.e. regional competition and local experimenta-
tion, that were responsible for the success of the early post-Mao reforms,
particularly the creation and growth of the private sector within a totalitarian
regime. With the changes during the early stages of the reform process, the
Chinese RDT institutions gradually evolved towards what I called regionally
decentralized authoritarian (RDA) institutions.5 In comparison, authoritar-
ianism is a less extreme type of autocracy in which limited pluralism is
3
C.J. Friedrich and Z.K. Brzezinski, Totalitarian Dictatorship and Autocracy (Cambridge,
MA, Harvard University Press, 1956). Regarding victims of the two major types of
totalitarian regimes worldwide, “Communist regimes have victimized approximately
100 million people in contrast to the approximately 25 million victims of the Nazis.”
S. Courtois, “Introduction,” in S. Courtois, N. Werth, J.-L. Panne, A. Paczkowski,
K. Bartosek, and J.-L. Margolin, The Black Book of Communism: Crimes, Terror,
Repression (Cambridge, MA, Harvard University Press, 2015), p. 15.
4
In the media the post-Mao reform is often refered to as the Deng Xiaoping reform. But
I will explain later in this chapter that this popular description is controversial, or errs in
many important historical facts.
5
C. Xu, “The Fundamental Institutions of China’s Reform and Development.”
532

The Origin of China’s Communist Institutions
allowed, economically and ideologically. But the hardcore CCP leaders

would not tolerate the evolution from totalitarianism towards authoritarian-
ism, so this trend was aborted, and we are observing a return to RDT since
2012. It is because of this combination of an extremely high degree of political
centralization and a high degree of administrative decentralization that
Chinese RDT/RDA institutions are unique in the world.
How and why have Chinese institutions evolved in the ways we observe?
What are their impacts on the economy? What are their origins? This chapter
will explain how Chinese RDT/RDA institutions originated from Chinese
imperial institutions and Soviet totalitarian institutions. These explanations
are a base for understanding later institutional changes which led to devas-
tating disasters from the late 1950s to the mid-1970s, spectacular growth at
earlier stages of the reform, and grave problems in recent years. Due to the
space limitations of this chapter, my discussions will focus on key arguments
and evidence, many of which are stylized facts abstracted from details, with
some reference to my book in progress, Institutional Genes: A Comparative
Analysis of the Origin of Chinese Institutions.
Institutional Genes: An Analytical Concept

Analyzing institutional evolution is a huge intellectual challenge. To facilitate
our discussions, let me first briefly lay out my conceptual framework. When
examining long historical processes of institutional change, one can
observe some regularity whereby certain basic institutional elements appear
repeatedly, even when there are important regime changes. More import-
antly, some of these repeatedly reproduced institutional elements have
deep impacts on the long-term trajectories of further institutional changes.
I call these repeatedly reproduced institutional elements “institutional
genes.”6
The institutional gene is defined as the basic institutional element that
determines the players’ incentives. They are repeatedly self-produced, repro-
duced, and evolved with the change of institutions over a long historical
process. Institutional changes are endogenous processes in that, given the
existing institutions and other constraints, they are created through strategic
interactions among players (referring to all individuals in a society). The
reproduction of institutional genes in a changing environment is caused by
6
C. Xu, “Institutions and Institutional Genes,” in Xu, Institutional Genes: A Comparative
Analysis of the Origin of Chinese Institutions (forthcoming from Cambridge University
Press).
533

chenggang xu
players’ selection of certain institutional elements from existing institutional

genes out of self-interest. This sheds new light on the “path dependence”
nature of institutional change, a popular concept in the literature of economic
history.7 We explain the processes of institutional evolution by identifying
the reproduction and evolution of the institutional genes. This helps us
understand the mechanisms of institutional evolution over and over again
during historical processes, and their consequences.
For understanding institutional genes and their evolution, the following
concepts are essential. Among institutions, those with rules that are followed
regularly (including under threat and coercion) are stable.8 An institution is
regarded as an incentive-compatible institution if the incentives that the
institution provides (including social norms) are compatible with the major-
ity of players (including rulers and the ruled) in the institution: i.e. the players
chose to follow the rules of the institution. Here, incentives consist of
material and nonmaterial rewards and punishments, which can include
imprisonment, torture, and killing. Thus coercive rules can also be incentive-
compatible as long as individuals choose to follow under the threat of the
rules.
Similarly, incentive-compatible (IC) transformation is defined as
being when the incentives of the key players are consistent with the
rules established by the transformation. Thus an IC transformation is
more likely to be stable, whereas a non-IC transformation is more
likely to be unstable, as the incentives of a majority of key participants
are violated.
Institutional Genes in the Chinese Empire and the

RDT/RDA Regime
To establish a foundation for later analysis of the origins of China’s institu-
tions in the remainder of this chapter, this section briefly illustrates how the
institutional genes of the Chinese empire evolved and reproduced over the
two millennia of the history of the Chinese empire, and outlines the institu-
tional genes of contemporary Chinese RDT/RDA regimes. Figure 15.1
depicts the institutional genes of the Chinese empire.
In Figure 15.1 there are three basic institutional elements (blocks), reflect-
ing the strong complementary relationship of these blocks. I call this gene
7
D.C. North, Institutions, Institutional Change, and Economic Performance (Cambridge,
8
Here being stable is a neutral description which does not bear any normative meaning.
534

Imperial junxian
system
Political control
Central–local
governance
Judicial system
Imperial exam system

Imperial land system
Personnel control,
Economic/legal
ideological
foundation
legitimacy
Figure 15.1 The institutional genes of the Chinese empire: an institutional trinity
map an institutional trinity.9 The central block is the bureaucracy, led by

the emperor and the imperial court, the imperial junxian 郡县制 system.
This system emerged during the Spring and Autumn period and became
the dominant governance institution of the Chinese empire from 220 B C
until the collapse of the Chinese empire in 1911.10 The junxian bureaucracy
was composed of multiple levels, from the top – the imperial court and the
central government – to the bottom – the county governments. The
judicial system was completely integrated into the executive bureaucracy,
and the emperor was the supreme judge of the empire.
The lower-left block of the institutional trinity is the imperial land
system, featuring the ultimate imperial land property rights, which were
the economic and legal foundation of the empire. This building block was
established simultaneously with the creation of the junxian system. The
emperor possessed ultimate rights to all the land in the entire empire.11
Imperial land rights are indispensable for the survival and stability of the
Chinese empire as they eliminated the economic and legal foundations for
9
C. Xu, “Institutional Genes of the Chinese Empire,” in Xu, Institutional Genes.
10
Zhou Zhenhe 周振鹤 and Li Xiaojie 李晓杰, 中国行政区划通史 (The History of
Chinese Administrative Divisions) (Shanghai, Fudan daxue chubanshe, 2009), p. 1.
11
Beginning in the Song dynasty, landlords and peasants were allowed to trade their land
rights on the market. But there were strict restrictions on the amount of land that could
be purchased and accumulated. No one was allowed to possess an amount of land that
would be sufficient to enable him to challenge the county government. Moreover, the
emperor was entitled to repossess any land he needed or desired, or if the landlord was
found to be disloyal or noncompliant with the throne.
535

chenggang xu
the landed nobility; they also served to prevent anyone from accumulating
landed power through the purchase of land.12
The lower-right block of the institutional trinity is personnel control and
the imperial examination system. The emperor controlled all bureaucrats
through selection, appointment, and promotion/demotion, among other
things. By controlling the selection of candidates for the bureaucracy, the
imperial examination system deprived everyone except the emperor of
inherited powers. At the same time, by designing the content of the imperial
examination, the emperor and the imperial court used it as an ideological
control mechanism to maintain the legitimacy of their rule.
After the imperial examination system was created, many contradictions
appeared between the aforementioned central and the lower-left blocks of the
institutional trinity, as revealed by the unstable nature of the Qin–Han imperial
institutions, when high officials in the bureaucracy could accumulate power by
inheriting bureaucratic titles across generations. Despite the lack of de jure
ownership of the land they managed, some high officials became de facto
nobility, with their accumulated power. The creation and growing power of
such de facto seigneurs gradually eroded and challenged the central authorities,
eventually leading to the disintegration of the empire. The Chinese empire was
not stabilized until the imperial examination system was established (during
the Sui dynasty, 581 A D) and perfected (in the Tang dynasty, 618 A D, and the
Song dynasty, 960 A D).13 This stable and consolidated imperial examination
institution lasted until the early twentieth century, making the Chinese empire
distinctively different from other empires in the world.
To control the territory of what had been one of the largest empires in
human history is an immense challenge. Figure 15.2 depicts the institutional
genes of the administrative governance structure of the empire, i.e. the
imperial junxian system, which was fully established during the Qin empire.
After evolving for hundreds of years, this system was codified and institu-
tionalized in the Sui Code, then essentially copied into the Tang Code, and
then largely continued in the administrative laws of the later dynasties of the
Chinese empire until the eventual dissolution of the system.14
12
C. Xu, “Property Rights and Sovereignty under the Chinese Empire,” in Xu, Institutional
Genes.
13
Zhang Xiqing 张希清, Mao Peiqi 毛佩琦, and Li Shiyu 李世愉, 中国科举制通史
(General History of the Chinese Imperial Exam System) (Shanghai, Shanghai renmin
chubanshe).
14
Zhou and Li, 中国行政区划通史.
536

Emperor–royal court
Functional control Territorial control
Personnel Finance Education Military Justice Manufacture Province A Province B Province C
Personnel Finance Education Military Justice Manufacture
Figure 15.2 The institutional genes of the imperial junxian system
The top ruler of the junxian bureaucracy was the emperor, who governed
all local governments via the imperial court. All bureaucrats in the subna-
tional governments – at the provincial, prefectural, and county levels – were
appointed and managed by the imperial court, and all of the appointed
bureaucrats had to retire after completion of their tenures, i.e. by the rule
that no one could pass power on to descendants. Another key feature of the
junxian system was its decentralization of concrete administrative matters
whereby subnational governments played central roles in local administra-
tive matters. A good illustration of this feature is how administrative func-
tions were co-ordinated at each level of government. At the national level, all
administrative tasks were divided into six functions: personnel, finance,
education, military, justice, and manufacturing/construction (li 吏, hu 户,
li 礼, bing 兵, xing 刑, gong 工) and managed by six ministries (liu bu 六部).
Replicating this structure, at each subnational, provincial, prefectural, and
county level, the local government controlled the six administrative func-
tions, through six offices (liufang 六房), e.g. at the county level.15
Importantly, it was the responsibility of the head of the local government
to co-ordinate all the administrative functions within a jurisdiction. In
15
For concrete information on stereotypical bureaucratic offices at the county level and
the office layout in a typical county government, which impacted contemporary
institutions, see Bai Hua 柏桦, 明代州縣政治體制研究 (Prefectural- and County-
Level Political Systems during the Ming Dynasty) (Beijing, Zhongguo shehui kexue
chubanshe, 2003); Miao Quanji 缪全吉, 明代胥吏 (Local Officials during the Ming
Dynasty) (Taipei, Jiaxin shuini gongsi wenhua jijinhui, 1969).
537

chenggang xu
contrast, although a minister had the same bureaucratic rank as a provincial

governor, in general he would not make direct decisions regarding the
functionalities of the subnational localities. This combination of political
and personnel centralization and administrative decentralization represented
a compromise between top-down control and implementation efficiency.16
As an illustration of the persistence of institutional genes, the institutional
elements depicted above endured through dynasties and rebellions and have
been replicated again and again not only in those dynasties established by
rebellious elites or created by rebellious peasants who overthrew the preced-
ing dynasty, but also in the dynasties established by the Mongols or the
Manchus. Moreover, such institutional elements of the Chinese empire were
reproduced even years after the collapse of the empire. When China’s
institutions were transformed from those of Soviet totalitarianism to those
of RDT, one of the key institutional elements re-established was the junxian
structure illustrated by Figure 15.2.
Before turning to explanations of why totalitarianism has prevailed in China,
let us now compare the institutional genes of the past and present. Although
the Chinese empire was dissolved more than a century ago and the CCP has
always declared its ideology to be anti-feudal (which is the CCP’s way of
labeling the Chinese imperial system), similarities between the institutional
genes of the Chinese empire and those of the Chinese RDT/RDA regime can
be seen clearly. Parallel to Figure 15.1, Figure 15.3 depicts the governance
structure of the Chinese RDT/RDA institutional trinity at an abstract level.17
The seemingly superficial similarity between the institutional genes of the past
and those of today provides deep insights, which will be explained below.
Similar to the structure of the institutional genes of the Chinese empire,
the institutional trinity of the RDT/RDA regime comprises three basic
institutional blocks: the party-state bureaucracy, party-state ownership of
land/finance, and party-state control over personnel and ideology.
Moreover, the judicial system is also fully integrated into the top-down party-
state bureaucracy in the central block.
Akin to the Chinese empire, the lower-left institutional block is
the economic, legal, and power foundation of the CCP and the
RDT/RDA regime.18 It consists of complete state ownership of
16
Xu, “Institutional Genes of the Chinese Empire.”
17
C. Xu, “A Full-Fledged Totalitarian Regime in China,” in Xu, Institutional Genes.
18
The imperial land system of the Chinese empire is an essential part of its institutional
genes, upon which totalitarianism based on state ownership is developed. C. Xu,
“Property Rights as Institutional Genes,” in Xu, Institutional Genes.
538

Central–local top-down bureaucracy

Decentralized administration/resources to the
local judicial system within the bureaucracy
State control of land

and financial Party control of
resources personnel and
Economic foundation ideology
of RDA
Figure 15.3 The institutional trinity of the RDT/RDA system
land;19 overwhelming state ownership of financial resources and other

assets, including most of the banks in China and the majority of
companies listed on the Chinese stock markets; and monopolization of
all strategic sectors in the Chinese economy. Similar to the imperial
institution, the lower-right institutional block – control by the central
party of national personnel and ideology – is the key element to
guarantee ultimate control by the central authority over all government
levels and all individuals.
Parallel to Figure 15.2, Figure 15.4 depicts a highly stylized governance
structure of the central–local party-state bureaucracy of an RDT/RDA
regime. This simplified figure shows only two levels in the hierarchy. In
reality, the Chinese government consists of a multilevel hierarchy that, below
the central level at the top, features three levels of subnational governments:
the provincial level, the municipal level (or the prefectural level), and the
county level. Most administrative, economic, and public-service functions are
carried out by the subnational governments. Each region is self-contained,
and each subnational government controls all major functions, such as
personnel,20 finance, industry, agriculture, and so forth within its respective
19
The concept of ownership here is defined as the residual control rights of ultimate
control rights. O. Hart, Firms, Contracts and Financial Structure (Oxford, Clarendon
Press, 1995). According to the PRC State Constitution, the state has ultimate control
rights over all land in the country, including “collectively owned” arable land, since the
collectives must surrender their ownership to the state either when requested by the
state or when their land is to be used for nonagricultural commercial purposes.
20
Highly centralized control over nationwide personnel matters is implemented by
a nested structure. The central authority directly controls personnel matters at the
539

chenggang xu
Central govt
Central adm function Territorial control
Personnel Finance Agriculture Industry Province A Province B Province C
Personnel Finance Agriculture Industry
Figure 15.4 Stylized governance structure of China’s RDT/RDA central–local regime
jurisdiction. The national government is relatively hands-off in most concrete

operations of the national economy, whereas the subnational governments
are deeply involved in the economies within their respective jurisdictions.21
Comparing this structure with Figure 15.2, one can easily identify the related
institutional genes in the imperial junxian system.
However, the Chinese empire and the RDT/RDA regime are categorically
different in one crucial respect, the Leninist party, which was implanted from the
Soviet Union to China and has ruled China totally for more than seven decades.
The Party as a modern organization is ubiquitous, controlling the whole fabric of
society from top to bottom. The next two sections explain briefly how totalitar-
ian institutions were created in the world, and how they were introduced into
China.
provincial level, whereas the provincial authorities directly control personnel matters
at the city levels within their respective jurisdictions. Finally, the city-level authorities
directly control county-level personnel matters within their respective jurisdictions. To
guarantee that the central authority is able to control all important personnel matters,
all key party-state positions are subject to rotation.
21
Y. Qian and C. Xu, “Why China’s Economic Reforms Differ: The M-Form Hierarchy
and the Entry/Expansion of the Non-state Sector,” Economics of Transition 1.2 (June
1993), 135–70; Xu, “The Fundamental Institutions of China’s Reform.”
540

The Origins of the Institutional Genes of

Totalitarianism
It is well documented that the CCP was created by the Comintern, which was
a Russian agency founded and led by Lenin almost immediately after the
totalitarian regime was created in Russia. The CCP developed in China very
rapidly. Only one decade after the creation of the CCP, the Chinese Soviet
Republic was created (in 1931), when the Baltic states were not part of the
Soviet Union yet. In less than two decades the CCP had taken over
the entirety of China by military victory, and in 1949 established the
People’s Republic of China. However, the very foundations of the totalitarian
regime – the ideology of Marxism–Leninism – and the organization of the
Leninist party were completely foreign to the Chinese. For these foreign
exogenous factors to play such a fundamental role in changing China so
enormously, there must have been deep endogenous reasons. In the current
section and the next, I look at shared critical institutional elements of China
and Russia. Thus the creation of totalitarianism in Russia and its transplant-
ation to China are not coincidental.
The first modern totalitarian regime in the world, Bolshevism, emerged
from the following three institutional genes: (1) the Christian Orthodox
church, which had almost completely penetrated Russian society and con-
trolled Russian spiritual life; (2) the Russian imperial institutions, which
almost completely controlled secular society, making the constitutional
reforms more difficult; and (3) political terrorist organizations or secret and
violent elite organizations, which completely controlled their members
through discipline and terror.22
Both leaders of the Marxist movement and leading critics of Communism
pointed out that Communism and violent Communist movements originate
from Christianity,23 and naturally the Christian church was essential for the
creation of the first totalitarianism. For centuries nearly all Russians were
members of the Russian Orthodox church. Russians’ belief in and loyalty to
22
See Xu, “Institutional Genes upon Which Totalitarianism Is Born,” in Xu, Institutional
Genes.
23
F. Engels (1895), “A Contribution to the History of Primitive Christianity,” pub-
lished online by the Socialist Labor Party of America (www.slp.org), February
2007, at www.scribd.com/document/48277975/A-Contribution-to-the-History-Of-
Primitive-Christianity-Frederick-Engels; K. Kautsky, Communism in Central Europe
in the Time of the Reformation (Nabu Press, 7 January 2010); L. von Mises, Socialism
(New Haven, Yale University Press, 1981; first published 1951). Kautsky and Mises
discussed the bloody violence of the first communist city-state in Münster
(Germany) created by radical Anabaptists in 1534.
541

chenggang xu
the Orthodox church were deep and passionate. Moreover, the church was
a pillar that supported the empire. Since Peter the Great, the czars had
intertwined the religious order with the government to increase their
power and to implement political reform, to use the church to rule the
empire.
By borrowing some key elements from the church, the Bolsheviks trans-
formed the revolutionary party into a political church, and Marxism–
Leninism is the political religion of the church.24 This political church incited
and organized the masses to initiate and implement revolutionary actions. In
addition to mobilizing the mass movement, the Bolshevik political church
was essential to justify or legitimize the totalitarian regime.25
The second institutional gene of Bolshevism was the Russian imperial
institutions. The power of the czars was much greater than that of Western
European absolutists. The nobility, entrepreneurs, and merchants in Russia
had been very weak since the emergence of the Russian empire, compared
with their counterparts in Western Europe. The traditional Russian “parlia-
ment,” the boyar Duma, which represented the nobility, was essentially
a bureaucracy subordinate to the czar since Ivan I V. Even this was not
allowed to last when Peter the Great abolished the boyar Duma
completely.26 The essential institutional elements of the czarist empire not
only made the constitutional reforms difficult, but also prepared the institu-
tional foundation for the forthcoming totalitarian regime.27
The third institutional gene of Bolshevism consisted of the terrorist
organizations associated with the populist movement.28 The first significant
such organization was the Decembrists, a violent secretive organization
composed of radical young military officials who, in the early nineteenth
century, attempted to establish constitutionalism by launching terrorist
campaigns.
A leading populist terrorist organization, Land and Liberty, directly influ-
enced the birth and organization of Bolshevism. This party later split into several
movements, among them are a proto-Marxist organization led by Georgi
24
C. Xu, “Christianity as an Institutional Origin of Totalitarianism,” in Xu, Institutional
Genes. The similarities and differences among Christian denominations and their
implications for totalitarianism are also discussed there.
25
C. Xu, “The Creation of the First Full-Fledged Totalitarian Regime,” in Xu, Institutional
Genes.
26
J.K. Sowards, Makers of the Western Tradition, 5th ed. (New York, St. Martin’s Press,
1991), vol. 2, p. 29.
27
Xu, “The Creation of the First Full-Fledged Totalitarian Regime.”
28
L.H. Haimson, The Russian Marxists & the Origins of Bolshevism (Cambridge, MA,
542

Plekhanov and a purely terrorist organization called People’s Will.29 Plekhanov

was the first self-claimed Russian Marxist and was the founder of the social-
democratic movement in Russia, the predecessor of the Bolshevik movement.
Lenin, the founder of Bolshevism, was deeply influenced both by Plekhanov
and by his own elder brother, who was a local leader of People’s Will respon-
sible for a failed attempt to assassinate the czar and was later executed.
Deeply rooted in Russian society, these three institutional genes were
particularly appealing to radical revolutionaries, radical intellectuals or intel-
ligentsia, and the masses at the bottom of society. The shared features of the
czarist imperial institutions with totalitarianism rendered the transformation
of Russia into a totalitarian state acceptable or even preferable among many
radical revolutionaries, proletarians, and soldiers.
Introducing Totalitarianism into China

Totalitarianism is foreign to the Chinese. When the CCP was established in
1921, the number of Chinese who knew constitutionalism was far more than
those who knew Marxism or Bolshevism. The CCP was only a small branch of
the Comintern, which was quite unknown nationally and internationally at
that time. But why did Bolshevism grow so fast in China and eventually
become the dominant force there? Our explanation is in the roles of institu-
tional genes. Indeed, China shares two of the three essential institutional genes
of Bolshevism – an imperial institution and secretive organizations. The
Chinese empire had established a more centralized and more sophisticated
institution than its Russian counterpart. Arguably, popular support for imperial
institutions among Chinese, for many of whom it might have been subcon-
scious as it was the only known order, was even stronger than in Russia.
Moreover, China had a long history of secretive rebellious organizations. One
example is the triads that partnered with Sun Yat-sen’s revolution. These were
powerful mafia-type organizations established in the eighteenth century.30
But China would not have been able to create a native totalitarian regime
on its own as a critical institutional gene was missing from Chinese soil: the
Christian church,31 the ideology, and ideology-centered institution. Without
29
According to its leader, Lev Tikhomirov, the party’s ambition was to organize a coup
d’état to seize power. D. Offord, The Russian Revolutionary Movement in the 1880s
(Cambridge, Cambridge University Press, 1986), pp. 26, 47–9.
30
C. Xu, “China: From Constitutional Reform to Bolshevism,” in Xu, Institutional Genes.
31
Although Western missionaries had established hundreds of churches in many Chinese
cities and towns beginning in the sixteenth century, the empire did not allow these
churches to become influential in society.
543

chenggang xu
the church and without many cultural connections with the West, not only
did Chinese institutions and Chinese culture make it impossible to create
a totalitarian ideology in China, but also no Chinese had ever even heard of
Marxism until the early twentieth century. Marxism was only brought back
to China with students who had been studying in Japan, and their knowledge
of Marxism was extremely shallow.32 Thus the mission of the Comintern was
pivotal for the creation of the Marxist totalitarian regime in China.33
The Comintern was set up by Lenin in 1919 to wage a global Communist
revolution, the success of which, according to Marxism–Leninism, was
a necessary precondition for the survival of the Bolshevik Revolution. In
1920, Grigori Voitinsky, head of the Far Eastern Bureau of the Comintern,
traveled to China to establish a Communist Party there. He contacted Li
Dazhao, Chen Duxiu, and others, who later became the founders and first
leaders of the CCP. Several months later, together they created the
Comintern China Branch, the Chinese Communist Party. Then, in 1921,
another Comintern representative, Maring (real name Henk Sneevliet) was
sent to China directly by Lenin to chair the First Chinese Communist Party
Congress.34 The Constitution of the CCP (passed at the Second Chinese
Communist Party Congress, in July 1922) declared that the CCP was
a branch of the Comintern. The CCP’s subordination to the Comintern
remained until the dissolution of the Comintern in 1943.35
The reorganization of the Guomindang (KMT) and the creation of
a KMT–CCP coalition sponsored by the Comintern was a vital step to foster
the newborn CCP. In 1923 Adolf Joffe, a representative of Lenin, signed the
Sun–Joffe Manifesto.36 Thereafter, the Comintern provided large-scale mili-
tary, financial, and personnel aid as well as training and advisers to the KMT,
and created the Huangpu Military Academy in Guangzhou and Sun Yat-sen
University in Moscow (officially called Sun Yat-sen Communist University of
the Toilers of China) to train KMT and CCP officials.
At the 1924 First National Congress of the KMT, Sun announced that the
KMT would be reorganized along the lines of the Bolshevik Party and
32
Xu Liangying 许良英 and Wang Laidi 王来棣, 民主的历史 (History of Democracy)
(Beijing, Law Press, 2015).
33
T. Saich, The Chinese Communist Party during the Era of the Comintern (1919–1943) (n.l.,
CreateSpace Independent Publishing Platform, 2014).
34
B. Lazitch with M.M. Drachkovitch (eds.), Biographical Dictionary of the Comintern, new,
rvsd and exp. ed. (Stanford, CA, Hoover Institution Press, 1986), p. 436.
35
Xu, “China: From Constitutional Reform to Bolshevism.”
36
W.L. Tung, The Political Institutions of Modern China (Dordrecht, Springer, 1968), p. 92.
544

China’s revolution would “learn from Russia” (yi e weishi 以俄为师). But the
long-run goal of the Comintern was to prepare the CCP to erode the power
of the KMT in the near future.37 Except for Sun Yat-sen, all the top KMT
leaders, including Chiang Kai-shek, Liao Zhongkai, Chiang Ching-kuo, and
so on, had trained in Russia or had spent lengthy visits in Russia. The
Comintern’s support for the KMT was maintained until 1927.
Under the leadership of the Comintern, in 1931 the CCP created the
Chinese Soviet Republic, China’s first totalitarian regime, with Mao
Zedong elected as chairman, even though he was not yet a top CCP leader.
To some extent, the Chinese Soviet Republic at that time was similar to the
Ukrainian Soviet Socialist Republic that loosely became part of the USSR. In
retrospect, the Chinese Soviet Republic was a prototype for the future PRC,
with a constitution that was a simplified version of the 1924 Soviet
Constitution.
In 1938, under the command of Comintern head Georgi Dimitrov, Mao
Zedong became the top leader of the CCP at the Sixth Plenary Session of the
Sixth Central Committee of the CCP, a position he would hold until his death
in 1976.38 Because Dimitrov’s instructions were communicated orally to
Wang Jiaxiang, the CCP envoy to the Comintern, without any witnesses,
the credibility of this order is still contested by some Russian historians.
Inspired by Stalin and the Comintern, Mao consolidated his power by
launching the Yan’an Rectification Campaign in 1942. From ideology to
strategy, the Yan’an Rectification Campaign was essentially a Chinese version
of the Stalinist Great Purge. The key reading of the Yan’an Rectification
Campaign was Stalin’s History of the Communist Party of the Soviet Union
(Bolsheviks): Short Course; the primary strategy of the Yan’an Rectification
Campaign was a Stalinist purge backed by terror, and the goal was to
establish the absolute power of the leader by building a personality cult.
Following Stalin’s steps after the 1936 Great Purge, “Mao Zedong
Thought,” coined by Wang Jiaxiang in 1943, was forcefully promoted in the
CCP by Liu Shaoqi (a secretary of the Central Secretariat and vice chairman
of the People’s Revolutionary Military Council at that time, later the presi-
dent of China, purged to death during the Cultural Revolution),39 and
37
38
He Fang, 党史笔记: 从遵义会议到延安整风 (Notes on the History of the CCP:
From the Zunyi Meeting to the Yan’an Rectification Movement) (Hong Kong, Liwen
chubanshe, 2005); Yang Kuisong, 毛泽东与莫斯科的恩恩怨怨 (Mao Zedong’s
Resentment of Moscow), 3rd ed. (Nanchang, Jiangxi renmin chubanshe, 2005).
39
H. Gao, 1930–1945 How the Red Sun Rose: The Origins and Development of the Yan’an
Rectification Movement (Hong Kong, Chinese University Press, 2018).
545

chenggang xu
became the core ideology of the CCP and entered the CCP Constitution at
the Seventh CCP National Congress in 1945. Although the Yan’an
Rectification Campaign was a duplication of Stalin’s Great Purge and
Wang’s invention was an imitation of Kaganovich’s creation of Stalinism in
the mid-1930s, the ideology of the great leader created in the process took on
a life and a soul of its own such that since then the CCP has become a fully
fledged Bolshevik institution in its own right. Thenceforth, the Bolshevik
institutional genes became part of China’s institutional genes, dominating
China’s institutional evolution until today.
If Bolshevism, the first fully fledged totalitarian regime in human history,
was so repressive, how was it possible for the Bolshevik Revolution to be
incentive-compatible with its wide-ranging participants, from social elites to
the vast masses of society?
Part of the answer is the Communist ideology, which has appealed to
intellectual elites and the masses at the bottom of any highly unequal society.
Yet, ironically, all totalitarian regimes are extremely unequal institutions.
The other part of the answer is the conspiratory strategy common to secret-
ive organizations. The Bolsheviks divided the grand revolutionary goal into
stages, promising different goals at each stage. This strategy would make
each stage incentive-compatible to the masses, and overall inconsistency
between stages would be handled in other ways, often by violence. One
example is Lenin’s promise of a nationwide election to be held within three
months immediately after the forced overthrow of the Russian Provisional
Government, i.e. the October Revolution. He said the Bolsheviks would step
down if they did not win a majority of the votes.40 But after losing the
election, the Bolsheviks declared that the election was invalid and arrested
opposition-party leaders, thus triggering a multi-year civil war.
Learning from the Russian Bolsheviks, the CCP designed their ingenious
strategy to change the incentive-incompatible grand revolution into a stage-
wise incentive-compatible revolution by making different and often contra-
dictory promises over time. At each stage, the CCP issued appealing promises
to participants, similar to what Lenin did in 1917–1918. The CCP has system-
atically breached its promises. To deal with the contradiction between
assurances made at various stages of the revolution and reality, they use
censorship, brainwashing, and suppression.
40
V.I. Lenin, “Reply to Questions from Peasants,” November 18, 1918, in Lenin, Collected
Works, vol. 26 (Moscow, Progress Publishers, 1971), pp. 300–1.
546

For example, total nationalization, which was the goal of the Bolshevik
Revolution, would not be incentive-compatible for the Chinese peasant
masses.41 To make the revolution incentive-compatible for the peasants,
from the 1940s the CCP promised them ownership of private land and
publicly announced that there would be no sovietization. However, after
the CCP took and consolidated power, and penetrated every village and
neighborhood and every corner of society, nationalization and collectiviza-
tion began to be ruthlessly implemented. In the process, the social status of
a handful of CCP village leaders was elevated. Most peasants hesitated to join
the collectives, and after they did they later regretted it, but it was already too
late for them to withdraw. Ultimately, almost all of them were forced to give
up all their private assets. Only at that time did peasants find that to survive
they had no choice but to join the collectives under the total control of the
CCP.42
Similarly, for the seizure of power to be incentive-compatible, the CCP
also promised democracy, freedom, and constitutionalism to the national
capitalists and intellectuals. Consequently, a large number of liberal intellec-
tuals joined the CCP.43 To win the hearts of enlightened Chinese, in 1946 the
CCP worked closely with the KMT and other parties to draft the Constitution
of the Republic of China. In the process, the CCP delegates made numerous
sensible motions to implement constitutional principles, including proactive
protection of human rights measures, strengthening of checks and balances,
and so forth. And most of those were incorporated into the final version of
the Constitution (which, interestingly, remains the Constitution in Taiwan to
this day).44
However, after taking power by military force in 1949, the CCP immedi-
ately withdrew its promises to implement constitutional principles in its
provisional constitution – the Common Program of the Chinese People’s
Political Consultative Conference – promulgated in September 1949. The
most important basic constitutional principles promised by the CCP, also
written into the Republic of China Constitution that the CCP had supported,
41
A Bolshevik revolution is a proletarian mass movement. With a weak urban proletarian
force, the CCP was forced to organize and lead peasant rebellions, and seized power by
peasant force. In many respects, the CCP revolution shared similarities with the peasant
rebellions that had been an integral part of the repeated cyclical dynastic revolutions of
the Chinese empire.
42
Xu, “China: From Constitutional Reform to Bolshevism”; Xu, “A Full-Fledged
Totalitarian Regime in China.”
43
44
Li Bingnan 李炳南, 政治协商会议与国共谈判 (The Political Consultative
Conference and the KMT–CCP Negotiations) (Taipei, Yongye chuban gongsi, 1993).
547

chenggang xu
such as a separation of powers, constitutional constraints on the power of the

executive, a multiparty system, and so forth, all disappeared from the
Common Program.45 Article 1 of the Common Program declares that the
People’s Republic of China “carries out the people’s democratic dictator-
ship.” Article 15 states, “The organs of state power at all levels shall practice
democratic centralism.” It should be noted that the essence of the so-called
“democratic centralism” is centralism alone and it is one of the basic Leninist
principles established for the totalitarian Bolshevik party.
Creating a Fully Fledged Classic Totalitarian

Regime
The Soviet Republic of China established in 1931 was a prototype totalitarian
regime and the basis for later developments in the People’s Republic of
China. The creation of a fully fledged totalitarian regime began after the
CCP took power. Stalinist institutions were fully transplanted into China.
The first step, even before transplanting the Soviet model, was political
centralization because previously all CCP-controlled territories had been
governed as a kind of federation whereby each territory had its own banking
system, legal system, and powerful leaders, such as Gao Gang in the north-
east, Deng Xiaoping in the southwest, Xi Zhongxun in the northwest, and so
forth. Beginning in 1950, all these powerful regional leaders were moved to
the central government and “promoted” as national leaders.46 Consequently,
regional challenges and constraints on central authority were greatly weak-
ened. Such centralization was well accepted by top CCP leaders as it was part
of the institutional genes of the Chinese empire following the imperial
pattern of enthroning the emperors.
The transplanting of the central planning institutions from the Soviet
Union, including state ownership, bureaucratic resource allocation, and
bureaucratic management, was one of the most critical elements in the
creation of the Chinese totalitarian regime, perhaps second only to party
building. A popular official slogan at that time was, “Today’s Soviet Union is
tomorrow’s China.”47 In the early 1950s, the Soviets transferred to China 156
huge projects covering all sectors of manufacturing. Compared with the
capital, equipment, technology, and management of modern state-owned
45
Common Program of the Chinese People’s Political Consultative Conference, at www
.lawinfochina.com/display.aspx?id=abb13dba42840de7bdfb&lib=law (accessed April 28,
2020).
46
Xu, “A Full-Fledged Totalitarian Regime in China.” 47 People’s Daily, 15 October 1951.
548

firms associated with these projects, the far more important essence of these
projects was the concrete implementation of central planning. Such large-
scale and comprehensive aid gave China a chance to duplicate everything
from the Soviet Union from the top to the bottom, from ideology to
propaganda, from law to rule, from central planning to management, from
technology to skills. The central planning apparatus was created as
a duplicate of the Soviet system. In this system, almost all state assets,
including the 156 Soviet-aided projects, were directly controlled by special-
ized central ministries, thus greatly strengthening the power of the central
authority of the CCP. With this unprecedented Soviet aid and several
campaigns which will be explained below, by the second half of the 1950s
China had already established a fairly complete classic totalitarian regime.48
The PRC Constitution was even drafted under repeated pressure from
Stalin. Advised by Soviet experts, the CCP drafted and passed the first PRC
Constitution in 1954.49 Unlike the Soviet Constitution, the Chinese
Constitution recognized the peasants’ rights to private land and the property
rights of the owners of private firms. Again, this was part of the CCP’s
strategy of maintaining its united front, composed primarily of peasants,
capitalists, and intellectuals, in the CCP’s seizure of power.
However, immediately after passage of the Constitution, the CCP publicly
announced that the Constitution would be transitional. Less than one year
after the passing of the PRC Constitution, collectivization and nationalization
began to sweep across the party-state, and the Constitution was de facto
abandoned.50
Concerning the CCP’s promises in the late 1940s that the PRC would be
governed by a coalition of the CCP and democratic parties, the 1954
Constitution breached this promise by drastically decreasing the power of
the democratic parties to participate in the governance of the PRC. The
Chinese People’s Political Consultative Conference degenerated from an
acting congress to a rubber-stamp forum. Most national-level positions held
by non-CCP members were abolished; all vice premiers were CCP members,
48
R. MacFarquhar and J.K. Fairbank (eds.), The People’s Republic, part 1, Emergence of
Revolutionary China, 1949–1965 (Cambridge, Cambridge University Press, 1987).
49
Mao did not want to have any constitution in the PRC as it would be an unnecessary
constraint on his dictatorial rule. But Stalin insisted that the CCP must implement
a constitution for the PRC to be considered a “normal” nation. Zhang Ming 张鸣, “1954
年宪法是怎么来的: 从 ‘共同纲领’到 1954 年宪法” (Where the 1954 Constitution
Came From: From the “Common Program” to the 1954 Constitution) 炎黄春秋
(History of the Chinese People), 10 (2014), 28–33, at www.yhcqw.com/30/9628.html
(accessed April 30, 2020).
50
Xu, “A Full-Fledged Totalitarian Regime in China.”
549

chenggang xu
and only a very few ministers were members of the democratic parties. Over
time, full-scale sovietization and the CCP’s totalitarian control over firms,
NGOs, and universities created strong discontent among intellectuals and
democrats. Their grievances began to mount and spread rapidly.51
Triggered by Nikita Khrushchev’s secret speech at the Twentieth Party
Congress of the Communist Party of the Soviet Union (CPSU) and the
Hungarian Revolution, both in 1956, the CCP launched a campaign to “let
a hundred flowers bloom and let a hundred schools contend.” Some years
later, Mao described this campaign as a plot to lead the snake from out of his
hole. Responding to this CCP initiative and unaware of the true intention of
the Party, most democratic leaders, who only a few years earlier had been
allies of the CCP, openly demanded that the CCP fulfill its promise to create
a coalition government. Hundreds of thousands of intellectuals openly criti-
cized the constitutional and administrative shortcomings of the totalitarian
party-state. But within only several months, all of those who had dared to
speak out were purged as “rightists,” i.e. political enemies. Many of them
were sent to labor camps or prisons; the luckier ones were placed under de
facto house arrest.52 In this “anti-rightist campaign,” as it was officially called,
550,000 intellectuals, including more than 30,000 professors, were purged as
rightists. These numbers should be placed in context. In 1956 China had
a total of fewer than 240,000 engineers and there were even fewer intellec-
tuals in business, finance, science, and the humanities.53 Many of the so-called
“rightists” were college students. Thereafter, constitutionalism was elimin-
ated from Chinese college curricula and replaced by a so-called “politics
course,” which essentially was a course on Party doctrine.
The significance of the “anti-rightist campaign” is comparable to that of the
“Yan’an Rectification.” The Yan’an Rectification established a complete and
independent totalitarian party, whereas after the anti-rightist campaign totali-
tarian rule was fully established over the entire country. Every person,
regardless of whether or not she was a CCP member, had to strictly follow
the Party line and the Party leader. The prohibition against dissident views
cut off any channels of outside information and destroyed any possible checks
and balances, thereby nurturing the conditions for future changes, such as the
51
R. MacFarquhar, The Origins of the Cultural Revolution: The Coming of the Cataclysm 1961–
1966 (New York, Columbia University Press, 1997).
52
Xu, “A Full-Fledged Totalitarian Regime in China.”
53
L. Orleans, Professional Manpower and Education in Communist China (Washington, DC:
National Science Foundation, 1961), pp. 68–9, 74–5. See also the chapter by Perkins in
this volume.
550

Great Leap Forward and the Cultural Revolution, and consequent

catastrophes.
The widespread and punitively coercive suppression after the anti-rightist
campaign established a foundation for a nationwide personality cult of the
Party leader. Liu Shaoqi, the president of China, declared that Party members
should become “tame tools (驯服工具)” or even “screws.”54 Students grow-
ing up in this environment were ready to follow any command whatsoever
issued by the great and powerful leader. They had no clue about the basic
rights of citizens, constitutions, laws, checks and balances, or even their own
rights and interests. When a replica of the Soviet totalitarian regime was
eventually established in China, totalitarian institutional genes also became
part of Chinese institutional genes. Nevertheless, China soon deviated from
the Soviet model. The classic totalitarian regime was transformed into
a regionally decentralized totalitarian (RDT) state.
The Creation of Regionally Decentralized

Totalitarianism (RDT): The Great Leap Forward
and the Cultural Revolution
As noted above, before 1949 the CCP-controlled territories (i.e. the liberated
areas) were governed by a quasi-federal structure. These regions enjoyed
substantial autonomous power, and local forces were an essential part of the
CCP power base. Political centralization after 1950 was more or less antici-
pated and tolerated by most CCP leaders as this had occurred whenever
a new dynasty was established in Chinese history. However, full-scale imple-
mentation of the Soviet model would have allowed the central ministries to
take over all the resources and powers from the regional governments.
Sovietization would thus have caused resentment among regional
officials.55 In 1956, Mao addressed the issue of the central–local relationship
in his speech entitled “On the Ten Major Relationships” (lun shida guanxi
论十大关系), revealing Mao’s rethink of the institutional details of
a totalitarian regime. Soon his thoughts were implemented through two
waves of campaigns, the GLF and the CR, and totalitarianism with Chinese
characteristics was born.56
54
Liu Shaoqi, “同北京日报社编辑的谈话” (A Conversation with the Editor of the
Beijing Daily) (June 8, 1958), at www.marxists.org/chinese/liushaoqi/1967/112.htm.
55
C. Xu, “Regionally Decentralized Totalitarianism (RDT),” in Xu, Institutional Genes.
56
Xu, “Regionally Decentralized Totalitarianism.”
551

chenggang xu
In 1958, after the anti-rightist campaign, which had further established Mao
as the supreme and unchallenged leader, Mao launched the GLF, thus
drastically moving China in a direction away from a classic totalitarian
regime. The first step in this campaign was to further centralize Mao’s
political power by weakening the fragile remaining checks and balances
within the Party and by suppressing other top leaders who held views that
were slightly different from those of Mao. At the Second Session of the Eighth
Congress of the CCP Central Committee in May 1958, Mao sharply con-
demned Zhou Enlai and forced him, as well as several vice premiers who
were responsible for central planning, such as Chen Yun, Li Xiannian, Bo
Yibo, and others, to make self-criticisms. All the party officials who supported
Mao and criticized Zhou became more powerful or were promoted, such as
Liu Shaoqi, Lin Biao, and Ke Qingshi (Party secretary of Shanghai).57
The series of institutional changes that began with the GLF and was
completed during the CR (to be discussed below) created what I call
a regionally decentralized totalitarian (RDT) regime, which featured the
coexistence of, on the one hand, totally centralized control of society in
ideology, politics, and even personal lives by the Party, and, on the other,
decentralization in administration and management. The most important
institutional change was the reintroduction of some of the institutional genes
of the Chinese empire into the totalitarian institutional genes.58 From an
orthodox view of totalitarianism, the GLF and the RDT institutions created
by the CCP were heretical to Marxism–Leninism. Thus the Soviet Union and
other Eastern Bloc Communist countries criticized the GLF sharply and
comprehensively.59
Although Mao had almost absolute power in the CCP, shaking up the
established totalitarianism had to be revolutionary as the latter had strong
backing from the whole international Communist movement led by Moscow
and the nested interests of some top CCP leaders. Indeed, an important part
of the efforts was officially called the Cultural Revolution, and the basic
principle that Mao emphasized was a continuous revolution. The mechan-
isms of both the GLF and the CR were fanatic mass movements, and they
featured fierce competition among regional forces, including regional
governments.
57
Xin Ziling 辛子陵, 紅太陽的隕落: 千秋功罪毛澤東 (The Fall of the Red Sun: The
Sins of Mao Zedong) (Hong Kong, Shu zuo fang, 2007).
58
59
552

During the GLF, local governments competed over grain output per unit of
land, over steel output, over promises of output, over the creation of novel
Communist institutions (experimentation), and so forth. An essential part of
the institutional foundation for regional competition was created at the begin-
ning of the GLF campaign, a drastic decentralization to regional governments,
with full-scale state ownership (or control) of all production assets.60 It is
important to point out that, since both were created from the same RDT
institution, at an abstract level, the mechanism of the regional competition in
the GLF is similar to that during the post-Mao reforms. Of course, as the party
lines were different there were some key differences between the GLF and the
post-Mao reforms. The GLF aimed to eliminate all markets, thus the competi-
tion was only over quantitative targets. Without independent channels to
check the veracity of the claims made by local governments, these quantitative
targets could be easily manipulated. In comparison, the post-Mao reforms
attempted to re-establish markets, and the target of the competition was GDP
growth or comprehensive market activities that could be verified independ-
ently, e.g. through random-sampling market surveys.
New institutions were created by trial and error through regional experi-
mentation in both the GLF and the CR, which will be discussed below.
Regional officials were incentivized to compete over the creation of institu-
tions that would accelerate the transition to communism. The most promin-
ent institution that was created during this campaign was the People’s
Commune (PC). The PC emerged in Chayashan town, Suiping county,
Henan province, on April 20, 1958. The local party officials called this new
institution the “Chayashan Satellite People’s Commune” to commemorate
the 1871 Paris Commune and the first Soviet satellite, sputnik I, which had
been launched some six months earlier. After revision by Mao, the charter of
the PC was published in the party’s theoretical journal, Red Flag, to promote
the implementation of communes throughout the country. In late 1958, all
rural communities were required to organize such communes and all peas-
ants were required to join them. Under the leadership of the CCP, all Chinese
peasants had “joined” a commune by the end of that year.61
Mao regarded the PC as the social foundation of the regime,62 and
he highlighted its features with two keywords, “large and public” (yida ergong
一大二公). “Large” refers to the scale of a commune, about 20,000 to 30,000
peasants in each commune, and the scope of a commune, which included
60
61
62
At the peak of the GLF, many urban communities were also organized into PCs.
553

chenggang xu
industry, agriculture, commerce, education, and the militia (gong nong shang
xue bing 工农商学兵). “Public” refers to complete public ownership, i.e.
elimination of all private property rights, and dominance by the CCP, which
controlled all “publicly owned” assets in the nation. Every commune created
tens or hundreds of commune–brigade industrial enterprises, the predeces-
sors of the reform-era township and village enterprises (TVEs).63 Thus each
PC was a self-contained social unit.
A totalitarian society composed of tens of thousands of self-contained PCs
made China distinctively different from the Soviet Union, which consisted of
highly specialized large-scale enterprises. At the same time, the hard-core
totalitarian institutions in the CCP were further consolidated, such as the cult
of personality, the absolute power of the leader, the absolute control of the
party, and so on. China’s transformation from classic totalitarianism to an
RDT regime is not coincidental as it involved essential institutional genes
inherited from imperial China. Mao made this point clear at the Central
Committee’s Beidaihe meeting at the beginning of the 1958 GLF. He
described himself as a combination of Karl Marx and Qin Shihuang. His so-
called “Marx” referred to the imported totalitarianism, whereas his so-called
“Qin Shihuang” referred to the imperial institutions.64 A decade later during
the CR Mao argued more explicitly that institutions of the Qin dynasty (Qin
youzai 秦犹在) continued to exist in China to the present.65
Replacing central planning by regional competition and forcing peasants to
work in communes during the GLF consequently not only destroyed infor-
mation about cost, quantity, and quality in all sectors of the Chinese econ-
omy, but also forced the peasants to hide food as government procurements
squeezed their rations and their seeds, such that their survival was threat-
ened. The chaos and disincentives led to the largest man-made famine in
human history, with the death of 30 million.66
Due to the great famine, the GLF was prematurely aborted and the newly
created RDT institution was still primitive. The second wave of pushing
towards the RDT, the CR, started in 1966 and lasted until 1976. On the one
hand, the CR thrust the centralization of politics, ideology, and personnel
63
64
65
Wang Nianyi 王年一, 大动乱的年代 (A Time of Great Upheaval) (Beijing, Henan
People’s Press, 1988), p. 470.
66
This estimation is by Kung, in this volume. For narratives of the great famine, which
occasioned 36 million deaths, see Yang Jisheng 杨继绳, 墓碑：中国六十年代大饥荒
纪实 (Tombstone: A Record of China’s Great Famine in the 1960s) (Hong Kong, Tiandi
tushu youxian gongsi, 2008).
554

matters even further, to an unprecedented level. All power was concentrated

in the hands of the top leader, Mao, and his lieutenants. On the other hand, as
an essential part of the CR, most central ministries were entirely closed down
and almost all centrally controlled assets were delegated to the regional
governments.67 The extreme centralization of politics and personnel matters
and the frenzied decentralization of administrative powers were highly
complementary to each other. The decentralized administrative powers
weakened the de facto powers of the central administrators because they
could challenge the top leader in technical respects as they were endowed
with indispensable resources. In contrast, the regional officials had no chance
to influence the central leaders as their powers were thinly distributed. Thus
the weaker the central ministries were, the more powerful the supreme
leader became.
At the peak of the CR, more than 98 percent of central government-
controlled assets were handed over to the regional governments. Except
the People’s Liberation Army (PLA) and those making nuclear weapons and
ballistic missiles, almost all state-owned enterprises (SOEs) were controlled
by the regional governments. The number of centrally controlled SOEs
dropped from 10,533 in 1965 to 142 in 1970. Most central commissions and
ministries, including the Central Planning Commission, the Central
Economic Commission, the State Statistical Bureau, and so forth, were left
with no functions. Many ministries, such as the ministries of Metallurgy,
Coal, Commerce, and others, were permanently abandoned.68 Nevertheless,
because an RDT structure was already in place and self-contained regional
economies had already been operational since the GLF, and the RDT regime
was consolidated and enhanced, during the CR there was no great famine and
the economy did not completely collapse. Associated with the administrative
decentralization, complementary to the frenzied cult of Mao’s personality,
the major driving incentive mechanism of the CR was regional competition
at every stage: the Red Guard movements; the Seizing-Power campaigns; the
agricultural Learn from Dazhai campaign; the Five Small Industries (FSI)
campaign, and so on.
Similar to the GLF, which created tens of thousands self-contained PCs as
a foundation for a primitive RDT regime, the CR created thousands
of self-contained counties as a foundation for a consolidated and industrial-
ized RDT regime. To make all counties autarkic in terms of metallurgy,
67
MacFarquhar, The Origins of the Cultural Revolution.
68
555

chenggang xu
energy, machinery, construction, construction materials, and chemistry,

a nationwide FSI campaign was launched in 1970, whereby each county
was required to establish its own SOEs in five sectors, namely steelmaking,
coal mining, machinery, cement, and chemical fertilizers. By the end of the
CR, a substantial proportion of Chinese counties had become self-sufficient.69
At the cost of 30 million lives in the GLF, and arguably even higher human
costs in the CR,70 the RDT was fully consolidated and codified by the state
Constitution in 1975. As will be explained in the next section, ironically the
RDT is the institutional foundation of the once successful post-Mao reform.
The Evolution of Regionally Decentralized

Authoritarianism during the Post-Mao Reforms
The CR era is one of the darkest periods in human history. The devastation of
the CR awakened the majority of CCP elites as the legitimacy of the CCP was
deeply shaken, thus paving the way for change in the CCP after Mao’s death.
In its earlier stages, the post-Mao reforms induced an unintentional insti-
tutional change towards RDA.71 The market replaced administrative plan-
ning in most areas of the economy; private property rights in production
emerged and become the largest sector of the national economy; limited
ideological pluralism and NGOs were somewhat allowed and accepted,
although censorship still prevailed and tolerance was always contested,
sometimes violently. The private and individualistic institutional elements
grew fast, eroding RDT institutional genes. However, not surprisingly,
limited liberalization and pluralism were not incentive-compatible with
some powerful groups in the old regime, who regard the remaining RDT
institutional genes as the foundation of their power and began to roll back the
trend towards RDA from 2012. Sharing the same kind of institution with pre-
1989 Soviet Eastern Europe, the CCP’s negative reaction towards liberalism
could be expected. The puzzling question is why China succeeded in creating
69
70
According to a report by the Central Committee of the CCP published in the 1980s,
more than a million people were killed and more than 10 million were injured or
disabled, plus more than 113 million were politically persecuted during the CR. Cited in
Yang Jisheng 杨继绳, “道路·理论·制度—我对文化大革命的思考” (The Course,
Theory, and Institutions: My Reflection on the Cultural Revolution), Jiyi 记忆
(Memory), November 30, 2013, 2–23.
71
C. Xu, “Institutional Evolution in the Post-Mao Era: Regionally Decentralized
Authoritarianism (RDA),” in Xu, Institutional Genes.
556

a private sector in the early stages of reform. This question is addressed by

examining the institutional genes inherited from the GLF and the CR.
Concerning both content and timing, the starting point of the post-Mao
reforms was the ending point of the CR. The leaders of the coup d’état in 1976,
which occurred several weeks after the death of Mao, arrested Mao’s wife
Jiang Qing and her lieutenants and removed those who insisted on continu-
ing the CR from both the central leadership and the subnational levels.
Consequently, a campaign was launched to transform the Party line from
class struggle to economic development.72 Changes in the Party line were
associated with personnel changes at all levels of the party-state hierarchy.
Those who had seized power during the CR were replaced by party-state
officials who had been purged at the same time. Importantly, many of those
who were purged during the CR were de facto political dissidents during the
Mao era as they vehemently disagreed with the Party line and were keen to
introduce change. Such a systematic political change paved the way for the
coming decades of reform.
With changed leadership, CCP central leaders forged a new consensus on
the following major issues: (1) the monopolistic political power of the CCP
must not be challenged, i.e. maintaining the essence of the RDT regime
unchanged;73 (2) the Party line has changed and the essence of socialism
should be interpreted as economic development, which is the least contro-
versial objective among the competing powerful factions and infighting
ideologies; and (3) the Mao type of personalistic leadership should be replaced
by a consensus-based collective leadership.74 These principles were docu-
mented in the communiqué of the Third Plenum of the Eleventh CCP
Central Committee in December 1978,75 which became the official manifesto
for political, ideological, and economic change, whereas it emphasized main-
taining socialism, particularly insisting that state and collective ownership
72
Before Deng Xiaoping returned to power in late 1978, the major changes were led by
Hu Yaobang, at the time general secretary of the CCP. J. Hu, “Hu Yaobang Selected the
Breakpoint for the Reform,” Kaifang 开放 (Kaifang magazine) 4 (2008), 66–8; Hu,
“What Is ‘Reform and Opening Up’? When Did It Occur?” Zhengming 争鸣
(Zhengming magazine) 4 (2009), pp. 66–70.
73
The following argument by Deng depicts the goal of the CCP clearly: “to build
socialism it is necessary to develop the productive forces . . . Not until . . . we have
reached the level of the moderately developed countries shall we be able to say that we
have really built socialism and to declare convincingly that it is superior to capitalism.
We are advancing towards that goal.” Deng Xiaoping, “To Uphold Socialism We Must
Eliminate Poverty,” April 26, 2987, in Selected Works of Deng Xiaoping, vol. 3, 1982–1992
(Beijing, Foreign Languages Press, 1994), p. 223.
74
Xu, “Institutional Evolution in the Post-Mao Era.”
75
See http://cpc.people.com.cn/GB/64162/64168/64563/65371/4441902.html.
557

chenggang xu
must not be touched. Consequently, “Four Modernizations” became the

slogan of the Third Plenum, and “Reform and Opening Up” became the
slogan after the 1987 Thirteenth Party Congress (when Zhao Ziyang was
the CCP secretary general). Make no mistake, the change in the objective of
the CCP from class struggle to economic development was always meant to
be fully consistent with Communist ideology and to serve the survival of the
CCP regime.76
As China was still suffering from the consequences of the CR in the late
1970s, the post-Mao reforms began by following the Eastern Bloc nations due
to shared similar institutions and objectives. However, the reforms in all
Eastern Bloc nations ran into deep problems caused by a totalitarian bureau-
cracy. China was no exception.
A totalitarian regime controls all power and resources in society through
the Leninist party in a top-down manner with a long chain of command. The
entire society and the national economy are ruled by millions of party-state
bureaucrats, who enjoy great benefits from their powers. Moreover, super-
iors have to rely on their subordinates for information to evaluate their
subordinates, but subordinates with better local information have no incen-
tive to report truthfully. Their vested interests are major obstacles to any
reform that challenges existing institutions. Not only will they not take any
initiative to attempt such reforms, but also they will find excuses not to
implement reforms in their respective jurisdictions, regardless of the Party
line or the dictates of their superiors. However, implementing any reforms at
least has to rely on subordinates. In reality, very often the reforms have to
rely further on their initiatives as they are better informed regarding local
information (à la Hayek). Thus a solution to the incentive problem in the
party-state bureaucratic hierarchy is the key to determining what reforms are
76
The survival of the CCP regime is exactly the reason Deng and his lieutenants argued in
cracking down on the Tiananmen demonstration. Moreover, ideologically, according
to Marx, one respect in which socialism is better than capitalism is in its higher capacity
to advance “productive forces.” Thus, in order to prove the validity of the Communist
Party’s doctrine, it is necessary to deliver a higher growth rate than the capitalist
economies. For this reason, most Communist leaders in the Soviet Union and
Eastern Europe, even including Mao in the 1950s, attempted to grow their economies,
although such attempts all eventually ended in failure. After the collapse of the Soviet
Union and the Eastern Bloc, the top CCP leaders believed that continued economic
development was crucial for the survival of the regime. For example, Tian Jiyun, a vice
premier in the 1990s, attributed the collapse of the Soviet Union and the Eastern Bloc to
their decades of failure to improve productivity. Du Mingming and Qingquan Xu,
“田纪云谈 1992 年中央党校讲话” (Tian Jiyun on his 1992 Speech at the Central Party
School), Yanhuang Chunqiu 3 (2009), at www.yhcqw.com/11/4679.html (accessed
May 22, 2020).
558

feasible, who will implement the reforms, and how the reforms will be
carried out.
Given the failure to find such a solution, intrinsic resistance to institutional
reforms by party-state bureaucrats in the Eastern Bloc countries led to the failure
of two decades of reform attempts and ultimately to the total collapse of their
totalitarian regimes. In contrast, in the early stages of Chinese reform, the private
sector emerged and grew fast, which drove China’s growth thereafter. A key
observation for understanding why China differs is that all the reforms related to
property rights in China were not designed by the central Party or state. Instead,
these were experiments at local levels, and initially they were not even permit-
ted. Under anticapitalist laws and rules, local governments initiated the experi-
ments and assumed high risks, as recognition, formal rules, and/or the
legalization of reforms related to property rights almost always occurred after
the fact was established, and risks were taken.77 What motivated local party-state
bureaucrats to take such risks?
The mechanism which drove many local party-state bureaucrats to engage
with experiments relating to property rights in the first two decades of the
reforms was tournament-like regional competitions over the GDP growth
rate launched by the central authority. Promotions of officials in subnational
governments were linked to their relative performance vis-à-vis the perform-
ance of officials in other regions. This provided high-powered incentives to
local bureaucrats.78
The central authority focuses on GDP growth as long as the Party is in power.
But to fulfill this goal is not easy, as shown in the lessons from the Eastern Bloc.
Thus regional governments were encouraged to find ways to develop faster than
other regions. Under this mechanism, to succeed in the regional competition,
many regional bureaucrats experimented with privatization, either partially or
wholly, and indirectly or directly, even when private property rights were illegal.
Only later were successful methods promoted or copied nationwide. The most
prominent such examples include land reform (the “household responsibility
system”), the special economic zones (SEZs) (protecting foreign private property
rights for Chinese land), the TVEs and their later privatization,79 the privatization
of SOEs (starting in Zhucheng, Shandong province), and, most importantly, the
77
Xu, “The Fundamental Institutions of China’s Reform.”
78
Y. Qian and C. Xu, “Why China’s Economic Reforms Differ: The M-Form Hierarchy
and the Entry/Expansion of the Non-state Sector,” Economics of Transition 1.2 (June
1993), 135–70.
79
M.L. Weitzman and C. Xu, “Township–Village Enterprises as Vaguely Defined
Cooperatives,” Journal of Comparative Economics 18.2 (1994), 121–45.
559

chenggang xu
rapid development of de novo private firms, among other things. The large-scale
change in property rights was the foundation for the institutional change towards
RDA. In 2004 the changed regime was codified in the PRC Constitution where
private property rights are recognized.80
However, regional tournament-like competition requires strong condi-
tions that do not always provide the desirable high-powered incentives in
the long run. That is why Chinese practices during the post-Mao reforms
appear unusual in comparison with the Eastern Bloc. The following is the set
of conditions for tournament-like regional competition to be an effective
incentive mechanism for national policies:
1 There must be a top-down hierarchical bureaucracy that effectively con-
trols the appointment, supervision, evaluation, and execution of all subor-
dinate-level bureaucrats.
2 All, or the majority of, subordinate bureaucracies must consist of self-
contained structures.
3 The government focuses on only one well-defined and measurable
objective.
4 The government’s disregard for all other objectives does not result in
serious consequences.
Here, 1 and 2 are institutional conditions; whereas 3 and 4 concern the
nature of competition targets. All totalitarian and authoritarian regimes,
including those in China and in the Eastern Bloc nations, satisfy condition
1. However, only Chinese RDT and RDA satisfy condition 2; whereas classic
totalitarian institutions in the Eastern Bloc violate this condition.81 These
conditions are also helpful for understanding why regional competition was
an essential part of the incentive mechanisms of the GLF and the CR after
administrative powers and economic resources were systematically decen-
tralized after 1958.
Indeed, at the beginning of the post-Mao era, when the party line focused
on economic development, and China was so desperately poor that people
were more willing to make sacrifices in other aspects in order to improve
their income, conditions 3 and 4 are satisfied such that regional competition
80
Xu, “The Fundamental Institutions of China’s Reform.”
81
With the implicit assumptions 1 and 4, Maskin, Qian, and Xu provide a theory and
preliminary evidence showing that the Chinese M-form (condition 2) can provide high-
powered incentives for economic growth (i.e. condition 3); whereas by violating
condition 2, the Soviet U-form will not be able to provide high-powered incentives
for economic growth. E. Maskin, Y. Qian, and C. Xu, “Incentives, Information, and
Organizational Forms,” Review of Economic Studies 67.2 (2000), 359–78.
560

targeting GDP growth was effective during the early stages of the Chinese
reforms.82 Moreover, GDP as a comprehensive indicator of total market
activities is well defined and well measured, and also it can be verified
independently. Thus setting the GDP growth rate as the objective of local
government competition categorically differentiates the consequence of the
post-Mao reforms from that of regional competition during the GLF and the
CR, although institutions were akin and strategies were also similar.
However, the role of any government regardless of the type of institution
must always involve multiple dimensions. Associated with growth sustained
over three decades and the authoritarian nature of governance, conditions 3
and 4 were violated substantially when China was no longer poor and when
government-driven growth was associated with deep socioeconomic prob-
lems (e.g. land-grabbing local governments forced citizens in their jurisdic-
tions to relocate by demolishing their homes, qiang chaiqian 强拆迁), such as
social stability, inequality, degradation of the environment, corruption, and
so forth. Even worse, tournament-type regional competition with multiple
targets led to a race to the bottom for some targets. Thus regional competi-
tion as a solution for RDT/RDA bureaucrats, and associated growth per-
formance, can only be transitional. The unprecedented fiscal stimuli of more
than 1 trillion RMB spent during the global financial crisis pushed growth up
temporarily, then it appeared to steadily slow down, with problems of
overcapacity and overleveraging. Consequently, regional competition was
abandoned.83 But under the RDA institutions, there are no alternative effect-
ive solutions.84
Facing increasing troubles in the economy, calls for reforming the state
sector, for better protections of property rights and human rights, and for
further reform in general were strong and popular. Given that the private
sector employed more than 90 percent of the labor force in China, these
reform calls are incentive-compatible with the majority of Chinese.
82
H. Li and L.-A. Zhou, “Political Turnover and Economic Performance: The Incentive
Role of Personnel Control in China,” Journal of Public Economics 89.9–10 (September
2005), 1743–62. After Maskin, Qian, and Xu, “Incentives, Information, and
Organizational Forms,” and Li and Zhou, “Political Turnover and Economic
Performance,” a sizable empirical literature in economics and political science provides
systematic evidence that Chinese tournament-like regional competition is effective
when the government’s only objective is growth.
83
C. Xu, “The Rise and Fall of the RDA,” in Xu, Institutional Genes.
84
There is no general optimum incentive solution for a bureaucracy when it has multiple
objectives. B. Holmstrom and P. Milgrom, “Multi-task Principal–Agent Analyses:
Incentive Contracts, Asset Ownership, and Job Design,” Journal of Law, Economics,
and Organization 7 (1991), 24–52.
561

chenggang xu
However, events recently have gone in the opposite direction. This change is
related to, but is more than, the change of the CCP leadership in 2012. These
changes and policies against the popular demand and expectations need to be
backed by strong coercive power.
This strong power consists of elements of the RDT institutional genes,85
which have been eroded by privatization and by the transformation towards
RDA. Indeed, from the beginning of the post-Mao reform, the ultimate
purpose of the CCP is to sustain the totalitarian regime both politically and
economically. This is manifested in Deng Xiaoping’s Four Cardinal Principles
of modernization announced in 1979 (upholding the socialist path, upholding
the people’s democratic dictatorship, upholding the leadership of the CCP,
and upholding Mao Zedong Thought and Marxism–Leninism),86 and is
evident in his decision to crack down on the Tiananmen demonstrations in
1989 and to purge the reform-minded CCP secretary generals Hu Yaobang
and Zhao Ziyang sequentially in 1986 and 1989.87 The propaganda and
ideological leading figures and agencies, state banks, large SOEs, and the
State-Owned Assets Supervision and Administration Commission of the State
Council (SASAC) have always been fighting against privatization and against
amendments to the CCP Constitution in 2002 and the State Constitution in
2004. One argument which they have emphasized is that state assets are the
foundation of Party rule.88
Since 2013 all private firms and NGOs, including foreign firms and organ-
izations, are required to set up CCP branches within the firm and organiza-
tion. “Everything must be led by the Party” is being enforced everywhere.
Discussions of constitutionalism and judicial independence are prohibited.
“Upholding and strengthening the Party’s absolute leadership in political and
legal affairs” becomes the rule above the law.89 Criticism of or even dissent
from top leaders is punished. Government media must follow the party
85
The most important elements include: (1) the party/state which controls the society,
the court, the legislature, and the armed forces; (2) the state sector of the economy,
including the financial sector, land, and SOEs; and (3) the subnational level party/state
bureaucracy (the RDT structure).
86
“邓小平：坚持四项基本原则” (Deng Xiaoping: Upholding the Four Fundamental
Principles), at https://baike.baidu.com/reference/280112/6b4atU-GeIeXMCJwbyQm
Cc4QXielL0TgnCZmNtYbbGoZ5sCwZZTYNwOhXn48UuE5NTm3yW7qXJ6Mw-q_
hCSaPeHBr51NSIGlMJIpPP7HGOY5-verWOhtuw.
87
Zhao Ziyang, Prisoner of the State: The Secret Journal of Premier Zhao Ziyang (New York,
Simon & Schuster, 2009).
88
Xu, “The Rise and Fall of the RDA.”
89
“中国共产党政法工作条例” (Regulations on the Political and Legal Work of the
Communist Party of China), at www.gov.cn/zhengce/2019-01/18/content_5359135
.htm.
562

(dangmei xing dang 党媒姓党). And the State Constitution has changed,
allowing the state presidency to become for life. All bureaucrats, particularly
subnational bureaucrats, are evaluated foremost on their loyalty to the top
leader. Their morals have dropped drastically as their main goal now is not to
make punishable mistakes.90
Conclusion
Enlightened Chinese elites have launched reforms and revolutions aimed
to establish a constitutional republic since the late nineteenth century, and
these endeavors are still unfulfilled to this very day. On the contrary,
China has implanted Bolshevism and further created an RDT regime
with deep local roots. The narrative of this chapter has explained how
the institutional genes inherited from the Chinese empire impeded consti-
tutional reforms, and instead nurtured Bolshevism in China with deep
roots, and further localized it. After totalitarianism prevailed in China, the
institutional genes were transformed into new forms. The institutional
genes of today’s RDT regime appear to be mutations of their counterparts
in the Chinese empire, like grafting institutional genes of the Chinese
empire onto the genes of totalitarianism. The persistence of the institu-
tional genes implies the difficulties of changing China’s fundamental
institutions. Indeed, from Deng Xiaoping’s “upholding the leadership of
the CCP” in his Four Cardinal Principles to Xi Jinping’s re-emphasizing
Mao’s words that “the Party is the leader of everything,”91 from the
cracking down in Tiananmen Square in 1989 to the suppression in
Hong Kong in 2020, and from the continuous anti-peaceful-evolution
efforts of the CCP from the 1950s to this day, the consistency in basic
principles among the CCP leadership, and their tenacious resistance to
constitutional reform, are evident.
Although institutional genes of the old regimes are persistent, institutional
genes can mutate in diverging directions. Contrary to mainland China, sharing
the same historical institutional genes, institutions in Taiwan have evolved into
those of a full constitutional democracy. If in China the private sector, including
NGOs and communal organizations (formal and informal), becomes the domin-
ant sector in society and comes to enjoy full autonomy; if the judiciary becomes
90
Xu, “The Rise and Fall of the RDA.”
91
“The Party is the leader of everything, from the Party, the government, the army,
the people, the school, the East, the West, the North, the South, and the Center” (党政
军民学、东西南北中，党是领导一切的), said Mao in 1962, as Xi repeated in 2018.
563

chenggang xu
independent in protecting property rights and human rights; if a sufficient share

of the population is enlightened about their basic rights and takes collective
action to protect the their own rights and those of others, then new institutional
genes will breed, and change. Under that situation, institutions in mainland
China could eventually converge with what Taiwan has achieved.
Further Reading
Brandt, L., D. Ma, and T. Rawski, “From Divergence to Convergence: Re-evaluating the
History behind China’s Economic Boom,” Journal of Economic Literature 52.1 (March
2014), 45–123.
Courtois, S., N. Werth, J.-L. Panne, A. Paczkowski, K. Bartosek, and J.-L. Margolin, The
Black Book of Communism: Crimes, Terror, Repression (Cambridge, MA, Harvard
Friedrich, C.J., and Z.K. Brzezinski, Totalitarian Dictatorship and Autocracy (Cambridge,
MA, Harvard University Press, 1956).
Gao, H., 1930–1945 How the Red Sun Rose: The Origins and Development of the Yan’an
Rectification Movement (Hong Kong, Chinese University Press, 2018).
Li, H., and L.-A. Zhou, “Political Turnover and Economic Performance: The Incentive Role of
Personnel Control in China,” Journal of Public Economics 89.9–10 (September 2005), 1743–62.
MacFarquhar, R., The Origins of the Cultural Revolution: The Coming of the Cataclysm 1961–
1966 (New York, Columbia University Press, 1997).
MacFarquhar, R., and J.K. Fairbank (eds.), The People’s Republic, part 1, Emergence of
Revolutionary China, 1949–1965 (Cambridge, Cambridge University Press, 1987).
Maskin, E., Y. Qian, and C. Xu, “Incentives, Information, and Organizational Forms,”
Review of Economic Studies 67.2 (2000), 359–78.
Qian, Y., and C. Xu, “Why China’s Economic Reforms Differ: The M-Form Hierarchy and the
Entry/Expansion of the Non-state Sector,” Economics of Transition 1.2 (June 1993), 135–70.
Wang Nianyi 王年一, 大动乱的年代 (A Time of Great Upheaval) (Beijing, Henan
People’s Press, 1988).
Xin Ziling 辛子陵, 紅太陽的隕落: 千秋功罪毛澤東 (The Fall of the Red Sun: The Sins
of Mao Zedong) (Hong Kong, Shu zuo fang, 2007).
Xu C., “The Fundamental Institutions of China’s Reform and Development,” Journal of
Economic Literature 49.4 (2011), 1076–1151.
Xu, C., Institutional Genes: A Comparative Analysis of the Origin of Chinese Institutions
(forthcoming from Cambridge University Press).
Xu Liangying 许良英 and Wang Laidi 王来棣, 民主的历史 (History of Democracy)
(Beijing, Law Press, 2015).
Yang Jisheng 杨继绳, 墓碑：中国六十年代大饥荒纪实 (Tombstone: A Record of
China’s Great Famine in the 1960s) (Hong Kong, Tiandi tushu youxian gongsi, 2008).
Zhao, Ziyang, Prisoner of the State: The Secret Journal of Premier Zhao Ziyang (New York,
564

16
China’s Struggle with the Soviet Growth
Model, 1949–1978
dwight h. perkins
When the Communist Party of China announced a new government on

October 1, 1949, the economy that government inherited was in shambles.
China had been at war for over twelve years and much of the infrastructure
of the country had been destroyed or badly damaged and prices were rising at
51 percent per month or 13,000 percent per year. The Guomindang govern-
ment fleeing to Taiwan took much of the country’s foreign-exchange and
gold reserves with them, along with many of the managers of the banks and
industrial firms. Inflation and war left many of the businesses that stayed
barely able to function even when their managers and technicians did
not flee.
The government that inherited this shambles had very little experience
managing a modern urban economy. Their economic management experi-
ence was largely confined to overseeing some of the poorest regions of the
rural economy, notably the area around Yan’an in the remote northwest. The
one exception was in the northeast where the Soviet defeat of Japanese forces
there in 1945 helped make it possible for the Communist forces to gain
control of a major region with large cities and modern industry. Future
economic leaders, notably Chen Yun, did gain experience there that was to
help them later. For the most part, however, those charged with implement-
ing economic policies in the region often had little formal education and their
practical experience mainly involved military actions or farming. The
Communist base areas, however, did have real economic problems that
had to be dealt with.1 To maintain the support of the people in their base
areas, their efforts focused on the redistribution of land from richer landlords
This chapter has benefited greatly from the comments of its discussant, Xu Chenggang, at
the Utrecht Conference.
1
For an in-depth analysis of the economic policies of the Yan’an period, see P. Schran, “On
the Yenan Origins of Current Economic Policies,” in D.H. Perkins (ed.), China’s Modern
Economy in Historical Perspective (Stanford, Stanford University Press, 1975), pp. 279–302.
565

dwight h. perkins
to poorer farmers, but the base areas also had to produce enough food and
other necessities to maintain both an acceptable standard of living for the
general population and a surplus to feed, clothe, and provide ammunition to
the army. Furthermore, this had to be done in an area that was blockaded by
both the Guomindang and Japanese armies, severely limiting the ability of
the Communist-held regions to trade for necessities. The Communist armies
could not survive for long in these circumstances using the tactics of the
typical warlord army that met its needs mainly by pillaging the areas it
controlled.
Given this background, it might have been expected that the new govern-
ment would rely as much as possible on those remaining from the previous
government who had experience managing the economic institutions of the
modern urban sectors of the economy. To a limited degree that was the case,
but it lasted for a comparatively brief period. The main tasks during 1949–1952
were to achieve as rapidly as possible a restoration of production in the
enterprises that still existed and to end the hyperinflation. The latter was
achieved by replacing the old currency with the new renminbi currency and
then maintaining a government budget where revenue and expenditures
were more or less in balance. Rapid recovery in production and hence
taxation, together with the end of most fighting, also helped make
a balanced budget feasible.
During these first years (1949–1952) the government completed the process
of land reform that it had begun during wartime. Landlords were deprived of
their land without compensation and the land was distributed to the peasants,
as much as possible to the poorer peasants who were the mainstays of
support for the revolution. Land to the tiller at this stage did not involve
formation of co-operatives or collective farming except for a few experimen-
tal areas where it was done on a volunteer basis. The campaign did, however,
involve considerable violence, with large numbers of former landlords killed
or sent off to do hard labor. After 1952 experiments with co-operative forms
accelerated, but there were debates within the government about how fast to
push collectivization. Those debates were decisively ended when Mao
Zedong ordered that virtually all of agriculture be formed into producer co-
operatives on an involuntary basis where necessary in the winter of 1955–1956.
The government also early on began to establish direct control over key
areas of the modern urban economy. Steps were taken almost immediately to
nationalize the banking system and most of the transportation system. Under
the Guomindang many key industries were run by the state and these
simply changed management from one government to the next. The
566

China’s Struggle with the Soviet Growth Model
Japanese-owned companies in the northeast and elsewhere were also trans-

ferred to Chinese government ownership. Even private Chinese-owned
enterprises where management had stayed in China gradually lost most of
their independence and were converted into joint state–private enterprises
where the state was clearly dominant.
The transformation from private enterprise to joint state–private owner-
ship began first in 1951 and 1952 during the war in Korea with the “Three Anti”
and “Five Anti” political campaigns that were designed, among other pur-
poses, to intimidate private-sector owners and managers into following the
government’s directives. These campaigns were followed by a variety of
regulations designed to eliminate what the government perceived as undesir-
able private-sector practices, such as speculation. By 1955 it is unlikely that
there was much, if any, resistance to the state effectively taking over these
firms, converting former owners and managers into what in effect were
employees. Foreign trade during this process was taken away from the
producing enterprises and was put into the hands of state corporations.
Only they could hold foreign exchange with which to purchase imports.
The Centrally Planned Command Economy

By 1955–1956 most of the institutions of what had become a Soviet-type
economic system were in place. The first Soviet-style five-year plan theoret-
ically covered the years from 1953 to 1957, although it was not actually
published until July 1955. Annual plans, however, were made and imple-
mented as early as 1953. Soviet advisers were stationed in key positions
overseeing the economy to help China complete the process of creating
a Soviet-type economic system. The regulations governing the various
economic institutions were often little more than translations of the relevant
Soviet regulations.
The Soviet-type system that China created had the following
characteristics:
1 Except for a few state-managed farms, all agriculture was collectivized into
what initially were called agricultural producer co-operatives that theoret-
ically were managed by leaders elected by the members, but were in fact
appointed by the Chinese Communist Party and were expected to follow
orders from the Party and government. Land was held by the co-operative
as a whole although individual farm families were allowed to have small
private plots on which they could grow vegetables and raise pigs. Family
567

dwight h. perkins
incomes within the co-operative were based on the number of “work

points” earned by family members where the value of each point was
determined by the total income of the co-operative minus certain collect-
ive expenses and investments.
2 Most commercial banks were abolished and the People’s Bank of China
became a monobank that performed the functions of both a central bank
and a commercial bank. Enterprises were expected to deposit virtually all
the money they received in this monobank.
3 Industrial decisions of what to produce and with what inputs were deter-
mined by a central plan that developed targets for all industrial products
and inputs and then passed those targets down through the government
bureaucracy that broke those national targets down into targets for each
individual enterprise.
4 Commerce was in state enterprise hands for everything except small
handicraft and subsidiary agricultural products that could be sold on
small informal markets. Essential consumer products such as grain were
rationed for urban residents and prices were set by the government. Farms
received marketing quotas that they were expected to meet and farm
purchase prices were set by the state. Industrial firms turned over their
output to state commercial enterprises which then distributed them in
accordance with plan targets at state-set prices – prices that had little
relation to the supply and demand for the products.
That was the formal structure of state control of virtually the entire econ-
omy. The only areas by 1956 still mainly controlled by markets were small
private plots and informal rural markets.
How this system was supposed to work and to a degree did work can best
be understood by looking at its operation within the industrial sector. The
planners drew up a detailed plan every five years giving a broad outline of
what the economy was expected to produce, sector by sector. The real
operational plan, however, was an annual plan that drew up what was
intended to be a consistent set of output and input targets for each industry.
That was then broken down into output and input targets for each industrial
enterprise. Enterprises were legally required to implement these targets, but
law in the Chinese (and Soviet) case was not an effective means for enforcing
the plan targets.
Enforcement instead relied first on the fact that enterprises received the
inputs required from state distribution enterprises that were supposed to
make these allocations in accordance with the central plan. Supplementary
568

inputs could not be obtained on the market because no such market existed.
Enterprises could informally trade excess inputs for inputs they were short of,
but only on a limited basis and prices played no role in these trades. Further
backing up enforcement of the plan was the fact that the central bank, the
People’s Bank of China, was charged with checking to make sure that
enterprise deposits and loan funds were only spent on items authorized by
the plan.
Most profits of enterprises were turned over to the government, with
small exceptions for research efforts and the like authorized in the plan.
Investment in new plant capacity was funded out of the government budget
and carried out according to the central plan by separate enterprises that on
completion turned the investment over to the producing enterprise.
This centrally planned command system is capable in theory of producing
a consistent set of targets that provide the individual enterprises with the
inputs they need to meet the plan production targets for their enterprise. For
this to happen in practice, however, the central planners have to collect
enormous amounts of information about the specific conditions in each
industry and each enterprise. Much of this information must be obtained
from the enterprises themselves and those enterprises have an incentive to
overstate the inputs needed to fulfill their output targets. Thus the system has
a built-in tendency to use excessive inputs in production. To prevent this, the
central planners typically adjust the information they get from the industries
and enterprises, reducing the inputs requested or increasing the output
targets for a given level of inputs. By this method they hope to pressure the
enterprises to use the inputs more efficiently.
Even in an economic system where all enterprises have sophisticated
means for collecting the required information, therefore, there are going to
be input and output targets that will be inaccurate. One result in all systems
of this type is that enterprises accumulate much larger inventories than
would typically be the case when distribution and production are governed
by market forces.
The Command System in Operation in 1956–1957

When the centrally planned command system was more or less fully imple-
mented in 1955 and 1956, China’s industrial economy was a far cry from the
conditions ideally required for this system to operate efficiently. To begin
with, the Chinese Communist Party leadership required that virtually all
positions of authority be held either by members of the Party or by others
569

dwight h. perkins
with loyalty to the Party. The major sources of such people were institutions
such as the Eighth Route Army that had won the revolution. But whatever
their skill in warfare, few members of these core Party-supporting institutions
had much experience with a modern industrial economy. Most were origin-
ally from the countryside and had little formal education.
A survey in 1955 found that only 5.73 percent of the top leadership in
industrial enterprises had a university-level education or its equivalent. The
situation was better with respect to the engineering and technical personnel
in the enterprises, where 56.03 percent had a university education or its
equivalent.2 But China in 1955 had 125,000 industrial enterprises, of which
perhaps 30,000 were classified as medium- or large-scale. Medium- or large-
scale firms, however averaged less than 300 workers per enterprise. The other
90,000-plus firms averaged around fifteen workers each.3 The engineers
available in China by 1956 were drawn from roughly 170,000 who graduated
after 1948 together with the roughly 70,000 graduates educated before 1949,
many of whom did not remain in China.4 One survey in 1955 indicated that
there were only 32,000 engineers in China in that year, although one suspects
that the definition of what constituted an engineer in this survey differed
from the definition of an engineering graduate. The number of graduates
before and after 1949 in “economics and finance” and in “science” was much
smaller. Furthermore, the quality of many of these graduates was not
particularly high. Many had fewer than four years of training and the training,
at least after 1949, tended to be narrow and geared to immediate job
requirements. Engineers, scientists, and economists, of course, also played
an important role in many areas other than industry. It is not likely, therefore,
that even the 30,000 large and medium firms averaged more than three or
four university graduates each, and most of the small firms had none.
Given this situation, it is not surprising that the quality of the central
planning effort was not very high. The relationship between plan targets and
actual production for a limited number of high-priority industries is
presented in Table 16.1.
2
Tongji gongzuo tongxun 统计工作通讯 (Statistical Work Report), “The Breakdown and
Organization of the Number of Workers and Employees in Our Country in 1955,” 新华
半月刊 (Xinhua Semi-monthly), January 25, 1957, 89.
3
These figures were derived from various official Chinese sources in D.H. Perkins, Market
Control and Planning in Communist China (Cambridge, MA, Harvard University Press,
1966), Table 9, p. 109.
4
These data were derived from the graduation and enrollment rates for engineering and
other graduates of universities and equivalent institutions before and after 1949 in
L. Orleans, Professional Manpower and Education in Communist China (Washington, DC,
National Science Foundation, 1961), pp. 68–9, 74–5.
570

Table 16.1 Annual plan completion record (% of plan target)

1953 1954 1955 1956 1957
Steel 143.2 164.9 96.9 169.5

Electricity 112.9 134.2 161.4 121.1
Cement 210.8 208.7 105 112.7
Coal 197.8 108.1 284.2
Petroleum 103.9 163.7 83.8 87.5
Cotton yarn 148.1 193.5 113.3 97.6
Paper 785.7 185.7 135.3
Sugar 38.8 32.3
Cigarettes 30.1 36.2
Source: the data on plan targets and their realization come mainly from the reports of
ministers such as Bo Yibo and Li Fujun; see Dwight H. Perkins, “Industrial Planning
and Management,” in Alexander Eckstein, Walter Galenson, and Ta-chung Liu (eds.),
Economic Trends in Communist China (Chicago, Aldine Publishing, 1968), Table 3, p. 611
That the government was able to implement central planning at this early
stage of development was no doubt made easier by the fact that only certain key
sectors followed through on the full plan process. The number of plan targets
and input–output coefficients required was only a tiny fraction of what was
required for Soviet Union plans. Many other sectors received some plan guid-
ance but most of these small enterprises used simple machinery to produce farm
tools or other household goods or processed raw materials such as oil seeds
available locally. In these cases there was not much need to plan and co-ordinate
their inputs and allocate them through the state distribution system. But where
co-ordination of inputs and outputs with other industries was necessary, getting
the input and output plan targets was made especially difficult by the fact that
costs and the vintages of machinery in firms in many industries varied widely.
The Eighth Party Congress of the Chinese Communist Party in
September 1956 was the occasion for the Party and government to modify
this Soviet-type centrally planned command system to better fit the reality of
the Chinese economy. The key speech was that by Chen Yun, who became
minister of commerce in the fall of 1956 but more importantly was the leading
economics figure in the Party.5 In that speech Chen recognized the backward
and small-scale nature of much of the industrial and commercial sectors. For
5
This speech, “New Issues since the Basic Completion of the Socialist Transformation,” is
available in N. Lardy and K. Lieberthal, Chen Yun’s Strategy for China’s Development:
A Non-Maoist Alternative (Armonk, NY, M.E. Sharpe, 1983), pp. 7–22.
571

dwight h. perkins
the smaller-scale enterprises he recognized that many did not require co-
ordination through a national plan or distribution through state distribution
enterprises. Many probably did not even keep the kind of records that the
central plan required, nor did they have the capacity to do so. Much of the
statistical information collected from these smaller firms “turned out to be
useless.” By forcing all products to be purchased and sold by the state system,
there was the real prospect that incentives would be generated that led to the
neglect of many smaller products. Prices for these smaller items needed to
reflect market conditions, and that was best accomplished by allowing
producers to sell directly through local markets.
Large-scale enterprises and essential products, however, were to be strictly
controlled by the plan and distributed through the state system. Chen,
however, cautioned against the practice of some officials of simply combin-
ing a number of small enterprises in order to create larger enterprises.
Artificial mergers of this sort where no economies of scale were present
would undermine efficiency, not enhance it. Chen Yun was a supporter of
planning but only for the most important and largest industrial sectors where
he believed planning could be implemented efficiently. Given the primitive
state of much of China’s industry, it was far better to allow much of the rest of
the economy to continue to be co-ordinated by markets. Establishing the
centrally planned command system as the leading component of the econ-
omy but using market forces to supplement the plan, particularly in the
distribution of many consumer products, was to remain Chen Yun’s basic
position throughout his career, including into the first decade of the post-1978
reform era. By the 1980s, however, the nature of China’s industrial economy
was very different from what it had been in the 1950s. In 1956 Chen’s views
placed him among the most “liberal” or pragmatic economic leaders who
best understood the role that market forces could play, whereas by the 1980s
Chen came to be seen as a “conservative” blocking moves toward a full
market economy.
The state also took over internal management of the enterprises. A central
issue in this area was the relationship between the plant manager, who in
principle controlled daily production decisions and was more of a technocrat,
and the enterprise Party secretary, who, also in principle, was supposed to
ensure that the enterprise carried out the goals of the Party. The division of
authority between these two figures would be a central issue throughout the
pre-1978 period. During periods when major political campaigns dominated
Chinese society, Party and political campaign goals typically dominated
572

technocratic considerations, but 1956 and early 1957 were a period when
pragmatic management considerations played a central role.
The state also replaced worker and manager wages set by market forces
with wages set by the state. Urban worker wages were based on an eight-
grade system where each grade in the mid-1950s was roughly a bit over
20 percent higher than the previous grade, with the top grade being roughly
three times the lowest grade. There were other similar systems for manage-
ment and technical personnel. This system, like central planning in other
areas, was largely in place throughout the country by 1955 and 1956 and was
patterned after similar wage structures in the Soviet Union. China also relied
heavily on piece-work wages in 1955–1957. The main role of wages was to give
workers incentive to work hard and develop their skills. In addition to wages,
workers received a number of other benefits in kind. Housing and medical
insurance, for example, were supplied by the enterprise. Higher-ranking
personnel might receive larger apartments and other perquisites, such as
access to an automobile for a limited few at the top of the enterprise.
Wages in China played little role in the allocation of labor. In this respect
the situation in China was very different from that in the Soviet Union.
Where the Soviet Union had a shortage of labor, China had a large surplus of
unskilled labor based mainly in the countryside but ready to migrate to the
cities when allowed. In the Soviet Union labor was largely free to move from
one job to the next and migration from rural areas to the cities was actively
encouraged in part through wage policies. In China, in contrast, skilled labor,
particularly engineers and other highly skilled personnel, was directly allo-
cated to jobs by the state. Changing employment from one enterprise to
another was difficult and could mainly only be done with the permission of
the enterprise one was leaving. An informal system developed that was still in
place in the 1970s where a person wishing to transfer employment in one city
to another city would attempt to trade his position with a worker in a similar
position in the desired new location by posting a description of his position
and his contact information on a local billboard or telephone pole.
The issue with unskilled workers was different. There the problem was
that there were too many more than willing to migrate to jobs in the cities. As
early as 1955 the government began taking steps to limit migration from the
rural areas to the cities and to remove people from the cities who were not
employed. The 1956 wage reform exacerbated the migration problem by
raising the bottom wage as well as higher wages to levels that increased the
incentive to migrate to the cities. In early 1958, in order to discourage
migration, the government lowered the wage at the bottom end of the
573

dwight h. perkins
eight-grade system, making it roughly equal to the income of rural workers in

each area plus a cost-of-living differential.
The situation in rural areas with respect to central planning was funda-
mentally different from that in the cities. The government did draw up plans
for agriculture specifying such things as the sown area to be planted in a given
crop and the level of production expected. There were also targets for the
amount of the crop that was to be sold to the state purchasing enterprises.
Planning for the 110 million farm families that existed in the 1950s was clearly
impossible but the organization of these families into a million agricultural
producer co-operatives made some degree of central direction seem feasible
at least to some. However, there was enormous variation in the quality and
quantity of land between co-operatives and the suitability of that land for
particular crops. Adding to differences in land quality (and the many dimen-
sions of quality that included location and topography) were variations in
weather across regions and within regions from one day to the next. Then
there was the fact that some farmers were more skilled or worked harder
than others. To top things off there were even less reliable data on these
conditions than was the case in small-scale industrial firms.
A centrally planned command system in agriculture, therefore, was never
really implemented. There were efforts in the direction of such a system on
the part of some cadres but again it was the voice of Chen Yun in 1956 and no
doubt earlier that injected basic reason into the guidance given to agriculture.
State direction with respect to crop production mainly took the form of
efforts to establish quotas for the state purchase of a portion of the major
crops produced. These quotas were backed up with price policies designed to
ensure that co-operatives were sufficiently rewarded for their sales to make
the sales at least to some degree voluntary. For large numbers of farm
products such as vegetables and hogs, however, even that level of state
guidance was impossible. Given the lack of refrigeration both in the country-
side and in the cities, vegetables had to get from farm to market in a day or
risk arriving spoiled. Hogs required personal care in the Chinese context that
could only be appropriately managed by individuals or families, not even by
the co-operative leadership and certainly not by a central planner. For these
products families had their small private plots and local markets on which to
sell their produce.
In China, unlike in the Soviet Union, the agricultural surplus available to
the cities was not very large and the growing rural population (at 2 percent
per year in the 1950s) meant that increasing production was essential. There
was a genuine belief that the agricultural producer co-operatives would be
574

a means for accelerating growth in agricultural output as well as the surplus

available to the cities. The Chinese Communist Party may not have had
many members with experience in modern urban industries, but it did have
experience in the rural areas and with issues of agricultural production. Mao
Zedong himself, unlike some of the other top leaders, grew up in a rural
setting in Hunan province.
The logic behind the belief that co-operatives could be a vehicle for
accelerating production growth rested on several basic ideas. One was the
traditional practice of mobilizing rural labor to build rural infrastructure,
notably to expand and improve irrigation systems. In the nineteenth century
and much earlier, large-scale rural works could be built almost entirely by
mobilizing rural labor to move large amounts of dirt and rock to form
irrigation ponds, canals, and roads. Much of this mobilization in earlier
times involved a degree of coercion because the largest beneficiaries of
these efforts were the owners of land, but it was the tenants on this land, at
least in the southern half of the country, that did most of the work. With the
formation of co-operatives the increase in income resulting from construc-
tion by mobilized labor went to the whole co-operative rather than to
landlords and rich peasants and that in turn raised the value of each work
point earned. The incentive to participate in rural construction efforts was
thus enhanced and the need for coercion by the co-operative leadership
reduced.
The other main economic reason why the co-operative form was seen as
a vehicle for increasing production started from the view that peasant farmers
tended to be very conservative when it came to using new technology and
new plant varieties. There was a good reason for this conservatism.
Experimentation with a new technology that failed, leading the farmer’s
crop to fail, could mean that the farmer’s family had nothing to eat until
the next year’s harvest. This is a problem throughout farming in low-income
countries and it also limited the willingness of farmers in these countries to
rely heavily on the market. Most farm families thus tried to raise enough
grain to feed their families even if their land would normally produce more
income by planting the fields with a crop such as cotton all of which would be
sold on the market. With co-operatives, the leadership cadres could be
ordered to carry out the use of new technologies that central planners and
agricultural research labs believed would improve crop yields.
There were other barriers to increasing production that some in China
believed co-operatives could help overcome. With family-based farming,
plots of land tended to be small and often got even smaller when the death
575

dwight h. perkins
of the family head led the land to be further divided among the male heirs.
Potentially offsetting all of these advantages, of course, was the fact that
pooling of land and rewarding individuals and families with work points
provided a weaker incentive to work hard and with skill than did family-
owned farms where reward for effort and skill were directly related to the
income the family received without any intermediate process such as the
work point system. In addition, decision making, at least for major crops and
collective construction activities, moved from the household head to the
leadership of the co-operatives. This might have improved decision making
over such things as labor allocation if the senior cadres had been more skilled
managers and more knowledgeable about the land and the crops than most
household heads. Selection of the senior cadres, however, was based on
criteria that placed a heavy emphasis on loyalty to the Communist Party.
These were potential benefits, but were they actually realized in practice?
The agricultural production growth rate, whether measured as gross value of
output or as value added, was much the same in 1956–1957 (4.7 percent to
4.4 percent, 3.8 percent to 3.9 percent) and in the years immediately preceding
the nationwide establishment of agricultural producer co-operatives. This in
itself is a considerable achievement given the kind of disruption that often
occurs with a fundamental restructuring of ownership and management of
a rural economy. The number of hectares affected by natural disasters was
also somewhat higher in 1957 than in earlier years.
Data published in the 1950s indicate that there was a large increase in the
area under irrigation in 1956 and 1957, implying that labor mobilization had
made effective use of rural surplus labor. From 1953 through 1955, irrigated
acreage increased by an average of 1.25 million hectares per year. In 1956,
according to these data, the irrigated acreage jumped by 7.9 million hectares,
and in 1957 it rose by another 2.9 million hectares.6 According to data
published in the 1950s, the total area under irrigation was 34.7 million hectares
in 1957, whereas data published in the early 1980s after the long statistical
blackout estimated that total irrigated acreage in 1957 had been 27.3 million
hectares, 7.4 million hectares lower than the earlier estimate.7 The total area
in the country irrigated by power (for pumps) in that year was only 1.2 million
hectares. It is likely that the estimates of irrigated acreage in 1957 were from
6
N.R. Chen, Chinese Economic Statistics (Chicago, Aldine, 1967), p. 289. The original data
came from Guojia tongjiju 国家统计局 (National Statistical Office), Ten Great Years,
p. 130.
7
Guojia tongjiju 国家统计局 (National Statistical Office), Statistical Yearbook of China 1981
(Hong Kong, Information and Agency, 1982), p. 185.
576

reports by local officials based on varying criteria. These optimistic reports

may have had an influence on the senior leadership and on Mao Zedong in
particular in their decisions with respect to agriculture in 1958.
The situation with respect to improvements in technology may also have
been subject to local biases. The area planted to improved grain crop
varieties, for example, rose from 20.6 percent of the total sown acreage in
grain in 1955 to 36.4 percent in 1956 and 55.2 percent in 1957, an increase of
42.3 million hectares. Grain production in 1957, however, only rose by
6 percent over 1955, suggesting that a liberal definition was given to what
constituted an improved grain variety. A more concrete example of techno-
logical change that failed to achieve expected gains is the country’s experi-
ence with the double-wheeled double-bladed plow. These plows had worked
well in a number of areas in 1954 and 1955 and the authorities concluded that
every well-equipped co-operative should have a supply of these plows. At
considerable cost to other priorities, they raised the plow production quota
from 400,000 a year to at first 5 million plows, then cut that target back to
2 million plows. It turned out, however, that the plows were completely
unsuitable for large areas of the country. By 1956, 1.2 million plows had
actually been produced but only 800,000 of those had been sold and 40 to
50 percent of those sold were not being used.8
Finally there is the issue of how macroeconomic policies were managed in
China once the Soviet-type economic system was in place. Under the
centrally planned command system, achievement of price stability, balance-
of-payments equilibrium, and employment growth was handled very differ-
ently from the methods used in a market economy.
The major macroeconomic issue in China in the 1950s was price stability.
Employment growth, meaning urban employment growth, was largely
governed by the level and type of investment, and that was the responsibility
of the State Planning Commission. It was the State Planning Commission
that determined the level and sectoral allocation of that investment using
administrative (as contrasted to market) mechanisms. The level of invest-
ment outside the agricultural sector was largely determined by the excess of
government revenue over its recurrent expenditures (household savings in
the 1950s were small). That excess was in turn determined mostly by the
profits and taxes paid by state enterprises (82 percent of revenue in 1956).
Those state enterprise profits and taxes were determined by the prices from
8
The account of the double-wheeled double-bladed plow appeared first in “Why There Is
a Stopping of Production and an Obstruction of Sales of the Double-Wheeled Double-
Bladed Plows,” 计划经济 (Economic Planning), September 23, 1956, 1–4.
577

dwight h. perkins
their sales minus the prices of their inputs (including wages). These industrial
input and output prices were set by the state, with output prices deliberately
set at a high level to generate more revenue.
Prices played no role in the allocation of inputs or in the State Planning
Commission decisions concerning output targets.9 Enterprise demand for
inputs often exceeded supply but there was little pressure on input prices
because enterprises could not use money to purchase additional inputs.
There was no market on which to buy these inputs. If an enterprise desired
more of a given input, it could lobby the State Planning Commission to give
the enterprise a larger allocation or it could trade some of its excess inputs
with another enterprise that had a surplus of the needed input. When an
input was allocated to the enterprise, the enterprise did have to pay money
for that input, but if the enterprise did not have the necessary funds, the
central bank, the People’s Bank of China, would lend the necessary funds
without charging interest. The financial aspects of these transactions were
mainly accounting devices having little role in determining what was pro-
duced. The central bank was thus a passive supplier of funds in support of the
central plan. If the plan called for a particular allocation, the central bank had
to finance it when needed. This passive role for the central bank would
periodically become an issue in the pre-1978 period and was a source of
inflation in the post-1978 reform period. It was also a major element in what
has been called the “soft budget constraint.”10 State enterprises did not face
hard budgets that forced financial discipline. That discipline was supposed to
be provided by the state planners.
Maintaining balance-of-payments equilibria was controlled by the central
planners in an even more straightforward way that eliminated most market
forces. Enterprises were not allowed to deal directly with foreign customers.
Enterprise output destined for foreign markets and purchase orders for
foreign inputs were turned over to state trading companies that negotiated
with the foreign suppliers. These negotiations could be with individual
trading companies, as occurred at the annual Canton Trade Fairs, or
could be part of country-to-country or country-to-trading-group
9
The State Planning Commission did, to a limited degree, have to take the level of
demand for consumer goods into account in setting prices. However, consumer prices
could be and were made higher by limiting the amount of such goods available and the
decision on the level of output of consumer goods was controlled directly by the State
Planning Commission.
10
The “soft budget constraint” term was originally used by János Kornai and refers to the
fact that state enterprises in a centrally planned command system typically faced few
pressures to use inputs efficiently.
578

negotiations. The state trading companies, when dealing with individual

companies for the most part, would buy and sell at international prices.
The state-set prices used to sell to or buy from the domestic producing
enterprise typically had no relation to the international price.
There were also few capital movements into or out of China that the
government had to deal with. There was no stock market in China and hence
no portfolio investment. Foreign direct investment was prohibited except for
a few projects supplied by the Soviet Union, and the trading enterprises and
domestic enterprises could not borrow from abroad. Thus China had
a modest level of foreign trade (in 1956 and 1957 exports plus imports equaled
10.4 percent of GDP as contrasted to 29.8 percent in 1990 after a decade or
more of open trade and 40 percent in 2000). The domestic economy, how-
ever, was largely cut off from any influence on prices or production from the
international economy. The exchange rate for renminbi could be set at any
level the planners desired since it had no influence on trade. The central
planners could plan the domestic economy without paying much attention to
international economic forces except for when they set export and import
targets for particular enterprises.
The restriction on foreign direct investment, together with the tight
control of foreign trade that involved little effort to promote exports,
meant that China had to follow one other characteristic of the Soviet-type
economic system: the emphasis in the industrial sector needed to be focused
on producer-goods industry if the country wanted to grow. Even in the First
Five-Year Plan period (1953–1957), export earnings were only 5 percent of
GDP and were an even lower 4 percent in the Fourth Five-Year Plan period
(1971–1975). An investment rate of 30 percent would thus have to supply
domestic investment goods for 25 of that 30 percent. A strategy of producing
a large surplus of textiles and other consumer goods and then exporting them
for the large share of the investment goods required was not possible. This
relationship between industrial strategy and limited foreign trade was well
recognized by Soviet theorists, notably Yevgeny Preobrazhensky, and was
presumably understood by Chinese planners.
In this system achieving price stability became mainly a matter of ensuring
that the incomes of the households did not exceed the availability of con-
sumer goods in both rural and urban areas. In rural areas this meant ensuring
that there were enough industrial goods produced to soak up income earned
from the sale of grain and other crops and rural handicrafts. In urban areas the
challenge was to keep wage growth more or less equal to the growth of
industrial consumer goods plus the supply of food products from the
579

dwight h. perkins
countryside. The central bank had no real role in this process. Price stability
was mainly in the hands of the central planners. As pointed out above, the
task of the central planners was further simplified by the fact that China had
a surplus of labor that eliminated pressure on unskilled wages and through
that on prices. Enterprises did not compete for skilled labor through wage
increases either, because skilled labor was not free to move to alternative
employment.
The Transition from Pragmatic Planning to the

Great Leap Forward
The first six or more years of Communist Party rule had transformed a rural
peasant economy with a veneer of modern industry and services relying on
market forces to an economy dominated by collective agriculture and a state-
directed urban economy that rejected market forces in favor of state
administrative directives. It was a radical change that involved considerable
violence and economic disruption, but the Chinese economy continued to
recover and grow throughout. The period from mid-1956 through the middle
of 1957 was in sense a period of consolidation and adjustments to the new
kind of economic system that had been created, represented by Chen Yun’s
modifications of central planning discussed above. Mao Zedong’s speech on
the “Ten Major Relationships” at the Eighth Party Congress also appears to
have accepted a more pragmatic approach to the country’s political and
economic challenges.11
This period of roughly a year was also accompanied in mid-1956 by
a similar movement in the political sphere, the “Hundred Flowers” move-
ment (“let a hundred flowers bloom, let a hundred schools of thought
contend”). This movement may in part have been motivated by a desire
for constructive criticism, but Mao Zedong was disturbed by Khrushchev’s
secret speech in 1956 denouncing Stalin, and the Hundred Flowers campaign
could also be used to “draw snakes out of their holes,” as Mao later said. The
Hundred Flowers campaign did elicit a torrent of criticism of the government
and the Party in early 1957. By the middle of 1957 the Hundred Flowers
movement had given way to the Party Anti-Rightist rectification campaign
11
This speech would be republished during the reform period after 1978 as an effort to
show that Mao’s views were not necessarily opposed to reform efforts of the kind then
underway. Mao Zedong, “On the Ten Major Relationships,” April 25, 1956, in Selected
Works of Mao Zedong, vol. 5 (Beijing, People’s Publishing House, 1977), pp. 272–6.
580

that involved purging many of the critics, and that rectification campaign
created the political environment that would lead to the Great Leap Forward.
There were elements of the earliest phases of the Great Leap that did not
have the irrational political campaign aspects of a few months later. The
economic reform measures introduced at the Third Plenum of the Eighth
Party Congress, also introduced by Chen Yun, called among other things for
the transfer of many industrial, commercial, and financial functions from the
central government in Beijing to lower government levels. The idea was that
this would give local units more flexibility to take into account local condi-
tions and exercise their own creativity.12 In a different political context, this
transfer could be seen as a continuation of the adjustments to the centrally
planned command system that began in 1956. Beijing was in no position to
understand the great diversity of conditions that existed in different areas of
China and it lacked the information needed to make suitable adjustments to
local conditions. Many of the goods and services produced by industry also
did not need to have their inputs and outputs co-ordinated at the national
level.
In the actual context following the purges of the Anti-Rightist campaign,
this decentralization transferred wide authority over the economy to cadres
who, whatever their personal beliefs, were going to demonstrate that they
were not “rightist.” They would try their best to do whatever Mao wanted
them to do. Mao himself, who had apparently been inspired by Khrushchev’s
goal of surpassing the United States in fifteen years, called for China to
surpass the United Kingdom in fifteen years. That later became passing the
United Kingdom in three years and the United States in ten years. The
question then became one of how to implement this ambitious goal. Mao,
whatever his other strengths and weaknesses, was a master at mobilizing
large numbers of followers in enthusiastic political campaigns, and the Great
Leap became a political campaign on an unprecedented scale.
Formally the leading economic group of the Party still headed by Chen
Yun and including all of the senior economics ministers brought a semblance
of planned decentralization to the process. Thus in April of 1958 the State
Council directed all enterprises belonging to ministries, except for a handful
of particularly important ones, to be transferred down, often far down, to the
various levels of local government. In 1958 only 13.8 percent of industrial
output was under central control, down from 39.7 percent in 1957. In
12
This discussion of the early phases of the Great Leap Forward is based to a large degree
on J. Wu, Understanding and Interpreting Chinese Economic Reforms (Mason, OH,
Thomson/South-Western, 2005), p. 44–7.
581

dwight h. perkins
September 1957 planning was to a large degree transferred to local govern-

ments and those local governments were given considerable leeway to make
their own plan targets. The number of categories of materials and equipment
planned by the State Planning Commission was reduced from 530 in 1957 to
132 by 1959. Of particular significance, local governments were authorized in
July 1958 to make their own investment plans.13
In the campaign spirit of the time the leading economic group of the Party
by the middle of 1958 had become a figurehead and the local governments
and enterprises were implementing a massive expansion in investment and
employing millions of new state enterprise workers. Total investment in the
three years from 1958 to 1960 averaged over 33 billion renminbi (RMB)
per year, as contrasted to RMB 14.3 billion in 1957. The number of state
enterprise workers rose from 24.5 million at the end of 1957 to 45.3 million by
the end of 1958 and 59.7 million by the end of 1960. The urban population by
1960 was 30 million above the 99.5 million people of 1957.14
This was what was happening in what had been the more organized and
planned part of the industrial economy. Mao also got the idea that China’s
labor surplus in both rural and urban areas could be mobilized to accelerate
the growth of the country’s iron and steel industry. Tens of thousands of
small-scale iron and steel furnaces suddenly, with Mao’s blessing, appeared
throughout the nation. These furnaces could not produce iron or steel from
iron ore, but they could melt down scrap iron and steel and make simple tools
using locally made molds. On a modest scale this might have been a useful
addition to the making of farm tools, for example. There were a few such
furnaces that later made their way to the New Territories in Hong Kong and
made simple parts for such things as flush toilets, and they were able to make
a profit in capitalist Hong Kong.15 On the massive scale that actually occurred,
it ended up diverting resources to highly unproductive uses. There was not
enough scrap metal to feed these small furnaces so local cadres began taking
perfectly good iron and steel instruments and melting them down in order to
show their enthusiasm for meeting their inflated targets.
None of this massive expansion in investment and personnel was co-
ordinated to make sure that what was being produced would be useable or
that the necessary inputs would be available. By the latter half of 1958 there
was no longer much, if any, co-ordination by the planners. The enthusiasm
13
Wu, Understanding and Interpreting Chinese Economic Reforms, pp. 46–7.
14
Wu, Understanding and Interpreting Chinese Economic Reforms, p. 50.
15
This statement is based on personal visits to one of these small-scale furnaces by the
author in 1961–1962.
582

and loyalty of the local cadres was supposed to offset the abandonment of
planning, but political enthusiasm is not a mechanism for co-ordinating
inputs and outputs. The result throughout industry was chaos. The statistical
authorities and the local cadres reporting statistics on their achievements
made the situation even worse. The statistics reported soon had no relation
to reality.
The situation in rural areas was comparable. The underlying logic of what
became the rural people’s commune movement was not wholly irrational.
The agricultural producer co-operatives of 1956–1957 solved the problem of
relating local construction projects carried out by co-operative labor to who
received the benefits from these efforts. But the problem still existed at
a larger level. If a construction project required labor from five or ten co-
operatives, but the project disproportionally benefited one or two of those
co-operatives, then the incentive problem for the co-operatives that did not
benefit remained. This problem could be solved by creating a much larger
collective unit.
The rural people’s commune was invented locally in Henan province. In
August 1958 Mao Zedong visited the experimental Qiliying Rural People’s
Commune in northern Henan where a number of co-operatives had been
formed into a single production unit and such vestiges of a market economy
as private plots were abolished. Mao said that the experimental commune
was “good” and that was apparently sufficient for the Party and government
to launch a campaign to convert all of the co-operatives into communes.
That started a competitive process among the country’s rural cadres to form
communes. The process of consolidation was completed before the year was
over. Where the 700,000 agricultural producer co-operatives had averaged
around 200 families each, the 26,000 communes averaged from 4,000 to 5,000
families each. In principle, management of crop production was to remain
with a subunit of the commune, the brigade, which was roughly comparable
to the previous co-operative. In practice rural cadres in general and the
commune leadership in particular were under enormous pressure to produce
dramatic results not only in crop production, but also in rural construction.
By the winter of 1958–1959 rural China was engulfed in the largest movement
to physically reconstruct the countryside in the history of the world.
Hundreds of millions of farmers and their families moved massive amounts
of dirt and rock to build dams and irrigation systems, to smooth uneven crop
land, and much else.
The demands on the time of all rural adults for this construction work
inevitably interfered with work on crops and other farm products. Vegetables
583

dwight h. perkins
and hogs were the first to decline sharply because these were produced on the
private plots and sold on the market and both private plots and these markets
had been abolished. The commune leadership, much like the national leadership,
was focused on producing more grain and more cotton. In the words of Mao,
grain was to be taken as the “key link” in agriculture (and steel was the key link in
industry). There were adjustments as early as 1959 to restore limited private plots
in an effort to support vegetable production. Decentralization of statistical
reporting to the local level, combined with enormous political pressures on
local cadres to show dramatic results, however, led to massive falsification of
data. As a result, few in the leadership had any notion of the crisis that was
building in the countryside (or in industry). The relatively good weather and
resulting good harvest of 1958 further obscured the emerging crisis. Mao himself,
for example, in mid-1958 was thinking about what to do with the large grain
surplus.16 China exported 41.6 million tons of grain in 1959, nearly 18 million tons
above the annual average grain exports of 1955–1958,17 despite the fact that grain
production in 1959 was falling by 28 million tons (Table 16.2). Purchases of grain
from farmers rose from an average of 44.6 million tons a year in 1955–1957 to
51.8 million tons in 1958 and 64.1 million tons in 1959.18
There was an attempt by Minister of Defense Peng Dehuai at the Eighth
Plenum of the Eighth Party Congress in mid-1959 held on Lushan to recognize
the magnitude of the disaster that was emerging. Mao purged him for his efforts
and that ended any real resistance to Mao’s desire to keep the Great Leap
Forward going well into 1960 with only minor adjustments. Data reconstructed
by the National Statistical Office many years later during the post-1978 reform
period make clear that the 1959 grain harvest involved a sharp drop of 14 percent
and the 1960 harvest was another 15 percent below that of 1959 for a two-year fall
in output of 27 percent (Table 16.1). The fall in subsidiary farm products such as
vegetables and pork was undoubtedly much larger but the data needed to
estimate their decline probably do not exist.
The decline in industry lagged behind that in agriculture. Substantial parts
of the initial increases in industrial output in 1958 and 1959 were real in that
they reflected the fact that China had begun a number of large industrial
investment projects in the mid-1950s, many with Soviet assistance, that were
destined to go into production in those years. That said, there was enormous
16
Wu, Understanding and Interpreting Chinese Economic Reforms, p. 50.
17
National Statistical Office, Statistical Yearbook of China 1981, p. 372.
18
Guojia tongjiju 国家统计局 (National Statistical Office), 历史统计资料汇编
(Compilation of Historical Statistics) 1949–1989 (Mason, OH, China Statistics Press,
2005), p. 26.
584

Table 16.2 Impact of the Great Leap Forward on the economy

1956 1957 1958 1959 1960 1961 1962 1963
GDP (index 1952 = 100) 148.1 155.6 188.6 205.3 204.6 148.7 140.4 154.7
Primary value added (index 117 120.6 121.1 101.9 85.2 86.5 90.4 100.6
1952 = 100)
Gross value of agricultural out- 121.9 128.7 131.9 113.9 99.5 97.1 103.2 115.2
put (index 1952 = 100))
Grain (million tons) 192.7 195 197.7 169.7 143.8 136.5 154.4 170
Cotton million tons) 1.445 1.64 1.969 1.709 1.063 0.803 0.75 1.2
Oil seeds (million tons) 5.086 4.196 4.77 4.104 1.941 1.814 2.003 2.458
Secondary value added (index 227.3 245.5 375.4 472.3 498.6 288.8 257.8 295.2
1952 = 100)
Steel (million tons) 4.47 5.35 8 13.87 18.66 8.7 6.67 7.62
Cement (million tons) 6.39 6.86 9.3 12.27 15.65 6.21 6 8.06
Electric power (billion kWh) 16.6 19.3 27.5 42.3 59.4 48 45.8 49
Cloth (billon meters) 5.77 5.05 6.46 7.57 5.45 3.11 2.53 3.34
Exports (billion US$) 1.65 1.6 1.98 2.26 1.86 1.49 1.49 1.65
Retail prices (index 1952 = 100) 119.5 121.3 121.6 122.7 126.5 147 152.6 143.6
Tax revenue (billion RMB) 14.09 15.49 18.74 20.47 20.37 15.89 16.21 16.43
Source: Guojia tongjiju 国家统计局 (National Statistical Office), 新中国六十年

(New China’s Sixty Years) (Beijing, China Statistics Press, 2009), various years
waste of industrial resources in 1958 and chaotic construction of new and

expanded enterprises. In 1960 the Soviet Union also abruptly withdrew the
1,400 specialists and technicians that it had in China, some of whom were
probably supporting the new industrial projects. This removed one more
influence on the side of rational planning and management and helped
precipitate the sharp decline in industry that occurred in 1961.
The full nature of the crisis and the famine that resulted is the subject of
a separate chapter in this volume and so will not be discussed further here. By
1960 the economy was in shambles, famine was beginning that was to lead to
tens of millions of premature deaths, and rule by the Communist Party was
threatened, or so it was widely perceived.
From Recovery to the Cultural Revolution,

1961–1964
The main task from late 1960 and for the next few years was to bring order to
the economy and rapid recovery. Mao largely withdrew from direct involve-
ment in this recovery effort and returned to his focus on “class struggle” – the
585

dwight h. perkins
struggle to inculcate values in the Chinese people that would prevent the
restoration of bourgeois values. President and second in command Liu
Shaoqi took overall charge of the recovery. Chen Yun was restored to
a leading position in the design of economic matters. Mao would later
claim that both Liu and Deng Xiaoping were ignoring him during this period.
Mao’s major concern was political but that concern did involve him in
defending the Great Leap as having accomplished more good than bad.
That view contrasted with that of many senior people in the party who
had a much more negative view. Class struggle also meant making sure that
the efforts at recovery did not go too far and undermine Mao’s view of what
constituted a true socialist society and economy. A true socialist economy, in
his view, did not have much, if any, place for either markets or material
incentives. In 1961 through early 1963, however, there was a brief interlude in
which a more open debate was encouraged and in which the prominent
economist Sun Yefang, for example, advocated for substantial enterprise
autonomy that presumably would have required a reduced role for central
planning and an increased role for market forces. That period of open debate
ended in 1963 when the Socialist Education Movement got underway.19
The most urgent task in 1961 was to restore agricultural production and
end the famine. To that end the massive rural construction effort was scaled
back. The communes, with their three-tier structure (commune, brigade, and
production team), continued to exist but the management of crop production
was passed down and would eventually reside at the level of the production
team until the late 1970s. In the early 1960s, however, there were also areas
that applied what was labeled the “responsibility system” and that often
involved turning production over to individual families. The “household
responsibility system” was to reappear again in the first years of the reform
period (1979–1983), at which time it led to the total elimination of the
commune system, replacing it throughout the country with household-
based production. It is unlikely that there was any such intent in the early
1960s, however, given Mao’s opposition to anything that ended socialism in
the countryside. It was probably mainly an attempt to let the poorest rural
farmers do whatever they thought was necessary to survive. The responsi-
bility system of the early 1960s, however, became one of the principal charges
19
This brief discussion of the political background during the recovery years draws on the
much more complete and nuanced discussion in A. Walder, China under Mao:
A Revolution Derailed (Cambridge, MA, Harvard University Press, 2015); and
R. MacFarquhar and M. Schoenhals, Mao’s Last Revolution (Cambridge, MA, Harvard
586

brought against Liu Shaoqi and Deng Xiaoping in the first phases of the
Cultural Revolution.
Private plots and limited free markets, as already indicated, had begun to
be restored even earlier. To relieve pressure on farm families further, the
government began to import substantial quantities of grain (mainly wheat)
from abroad. Prior to 1961 China had exported over 20 million tons of grain
a year and it still exported 10 million tons in the famine year of 1961. There
were almost no imports of grain prior to 1961. In 1961, however, the govern-
ment imported 5.8 million tons and kept importing grain on average at that
level into the early 1970s (at which point grain imports rose even further).20
These imports were mainly destined for consumption by urban residents, but
they made it possible to reduce the purchase and tax quotas for farmers by
roughly 10 percent.21 In actuality, grain purchases averaged 35.3 million tons
in 1961–1963, down 21 percent from the 1955–1957 average.22 “Means of
subsistence,” which included grain but also other necessities for consump-
tion, rose from a minuscule 5 to 8 percent of total imports in 1953 through
1960 to 38 percent in 1961 and an average of 44 percent in 1962–1964.23
Recovery in agriculture, however, was not rapid. Production of major
crops was still well below the level of the pre-Great Leap period (1955–1957).
The gross value of agricultural output has been estimated to have recovered
to the level of 1955, although it is not clear how that figure was arrived at
given the performance of major crops. Population, however, had also risen
by 12.5 percent over that time so per capita output was still down from the
pre-Leap level even if one accepts the gross-value estimate.
Serious malnutrition resulting in famine ended but that may have had as
much to do with restoration of order in the transportation and statistical
reporting systems. Famine in China throughout its history had more to do
with the distribution of food, mainly grain, than it did with the level of
production. Production failures in the past were often confined to small
areas. When the imperial government was well run, the emperor regularly
received reports on grain prices. Local grain price spikes were a good indica-
tion of a crop failure and the government could use its “ever-normal granar-
ies” to supply the stricken area. China in 1959–1961 was an example of
20
21
Grain purchase plus tax quotas in 1955–1956 and 1957–1958 averaged 62.5 million tons
a year, although some of the agricultural tax quota was not paid in kind in grain.
Perkins, Market Control and Planning in Communist China, p. 44.
22
National Statistical Office, 历史统计资料汇编, p. 26.
23
587

dwight h. perkins
a government where a modern version of this system had ceased to exist. The
information reaching the top was falsified on a grand scale and the transpor-
tation system was busy with other priorities. The statistical system, along
with a number of other government functions, however, was recentralized in
1962, including banking and finance, making it possible to gain an under-
standing of the food disaster and to use the transportation system to restore
transport to the worst-affected areas.
Restoration of the industrial sector was somewhat more rapid than in
agriculture. Basically the industrial planning and management system was
restored to the system as it had existed in 1957 but with the decentralization
started in 1958 still intact. A large share of industry was to be planned and
inputs and outputs co-ordinated at provincial, county, and even lower levels,
and that practice lasted until the first major industrial reform effort of the
post-1978-reform period (which began in 1984). Only industrial sectors of
critical importance were planned and co-ordinated at the national level. The
official statistics for the early 1960s indicate that industry not only recovered
to pre-Great Leap levels, but output by 1966 was more than double the level
of 1957. This is not surprising because much new capacity was installed in
Chinese industry both in the years just preceding the Great Leap and during
the first phases of the Great Leap itself.
Employment in industry, however, did not recover to pre-Great Leap
levels. Employment in urban areas in 1958–1960 nearly doubled from
32 million to 61 million workers and employees. The restrictions that had
been designed prior to the Great Leap to ensure that rural-to-urban migration
would be related to planned employment targets completely broke down and
the urban population rose from 99.5 million in 1957 to 130.7 million in 1960. In
the context of 1961–1965, this increase in employment was no longer needed
as industry and urban services output fell sharply.
Rather than continue to support this inflated urban population, the govern-
ment instead took steps to reduce that population as rapidly as possible.
Employment in urban areas fell by 15 million by 1962 and 14 million were forced
to return to the countryside by that year. The policies used to make this transfer
back to the countryside remained in force until the early 1980s. They were
draconian. To begin with, all households were registered as either urban or rural
residents, the so-called hukou system. It was illegal to reside in urban areas unless
one had an urban hukou. Enforcement of this restriction involved several
elements. Enterprises wishing to hire additional workers were required to
work through government employment agencies and the employment agencies
were required to draw from the registered urban population for the most part.
588

Rural residents wanting to migrate had to first get the permission of their
commune’s leadership in order to leave legally. If they managed to get to an
urban area, they had to apply to the local Public Security Bureau for a stay of
over twenty-four hours. There were local residents throughout the cities charged
with reporting any strangers appearing to be staying without permission. If the
rural-to-urban migrant still managed to stay, that migrant faced the problem of
how to get sufficient food and other necessities. Most necessities, including grain
and edible oils, were rationed and could only be obtained with ration coupons no
matter how much money one had.
This system in effect created a two-class system that persisted in China into the
twenty-first century. Its impact in the 1960s through to 1978 was to hold China’s
urban population far below what it would likely have been if people had been
free to migrate. The data in Figure 16.1 show that both urban population and
urban employment peaked in 1960 and fell sharply in 1962. They both stayed low
in 1963 and then began rising in 1964. The urban population had a particularly
large increase in 1964 but this jump was probably an upward revision by the
statistical authorities resulting from the 1964 population census (the first census
since 1953). The population and many other statistics in 1958–1963 must have been
educated guesses by the statistical authorities given the chaos of 1958–1961, not
least in the statistical offices themselves.
From 1964 on, urban employment grew at a rate similar to the fairly rapid
growth of 1953–1957 (5.0 percent 1965–1978 versus 5.2 percent 1953–1957), but
annual urban population growth fell to 2.1 percent as compared to 6.8 percent
during the more liberal rural-to-urban migration period of 1953–1957. Urban
population growth in the 1960s and 1970s came almost entirely from the natural
increase of urban residents. Migration from the countryside was negligible. The
high growth in urban employment in the 1960s and 1970s was made possible
mainly by the employment of urban women. In 1957 women made up only
13.5 percent of workers and employees (those employed and receiving wages
mostly in urban areas) but in 1979–1980 they made up 48.4 percent of the
industrial workforce and over half of most of the services workforce.24
24
The figures presented here for 1979–1980 are not entirely comparable to those for 1957
but they do clearly support the view that the employment of urban women during the
period of tight restrictions on migration from the countryside was very large. See
N.-R. Chen, Chinese Economic Statistics: A Handbook for Mainland China (Chicago, Aldine,
1967), p. 473 (the original source was National Statistical Office, Ten Great Years); and
National Statistical Office, Statistical Yearbook of China 1981, pp. 99–103. From 1965
through 1978 the number of women workers and staff in state enterprises (all enter-
prises were either state- or collective-owned) rose from 7.86 million to 21.26 million
(Guojia tongjiju 国家统计局 (National Statistical Office), 中国劳动工资统计资料
1948–1945, (Statistics on Chinese Labour and Wages, 1948–1945), p. 32) and kept on rising
589

dwight h. perkins
350
300
250
200
Urban population
150
Urban employment
100
50
0
1949
1952
1955
1958
1961
1964
1967
1970
1973
1976
1979
1982
1985
1988
Figure 16.1 Urban population and employment growth 1952–1990 (millions)
Source: National Statistical Office, 新中国六十年, pp. 608–9
Another component of the Great Leap aftermath is that the government

for a brief time lost control of prices. The official figures for the rise in retail
prices in 1958 through 1964 are in Table 16.3. Again, given the collapse of the
statistical systems, these are probably best-guess estimates of the statistical
authorities. Because of the hyperinflation that helped bring the Communist
Party to power in China, both the population and the Party have been very
sensitive to increases in prices that most developing countries would consider
to be quite modest. Even as late as 1988–1989 an annual average increase in
consumer prices of 18 percent contributed to the political instability of those
years. In 1961 an increase of 16 percent in the midst of the famine then
underway must have contributed to the view that China was in a crisis and
the actual rate may have been higher. As indicated earlier, China’s main
method for controlling inflation was to control the urban wage bill and the
amount and prices of state-purchased agricultural crops. The crop failures
sharply reduced the availability on the market of agricultural products and
the chaos in industry reduced the availability of industrial consumer goods.
Average wages had fallen as early as 1958 presumably because the large
in the 1980s and early 1990s (in all urban units), peaking at 58.9 million in 1995, but then
began to decline after 1997 to only 41.6 million in 2002 (National Statistical Office, China
Statistical Yearbook, 2003, p. 139).
590

Table 16.3 Wages and the urban cost of living (1957–1964)

Urban cost of living Average urban money wage
index (previous year = 100) (yuan per year)
1957 571
1958 100.2 470
1959 100.9 430
1960 103.1 409
1961 116.2 380
1962 103.8 405
1963 94.1 371
1964 95.3 358
Source: National Statistical Office, 新中国六十年, pp. 610, 625
increase in the labor force occurred mostly at the low-skill end, but wages
stayed down even after most of these unskilled migrants had been returned
to the countryside. Wages stayed down through 1964 and then began to rise,
but average nominal urban wages did not pass the level of 1957 until 1980,
twenty-three years later. Consumer prices during 1965–1978, however, only
rose at 0.2 percent per year, so the increase in urban wages of 2.5 percent
a year over that same period was real even if it did not restore average wages
to the 1957 level.
Clearly the government in 1962–1964 gave price stability a high priority and
inflation was brought to an end quickly. Price stability, however, was bought
at the cost of a decline in real average wages of 18 percent (from 1960 to 1962)
and a far larger decline than that if one takes into account that the average
wage in 1960 included 15 million low-skilled workers who had been returned
to the countryside by 1962.
Economic Organization and Policies 1965–1978

With economic recovery achieved, China in 1965 drew up a Third Five-Year
Plan (1966–1970). The regulations developed earlier to decentralize the plan-
ning and management of industry remained in place. The staff of the
planning and implementation agencies, however, had lost most of their
authority and so the plan was mostly on paper. Still, from that point on
until well after the beginning of the reform period in 1978, the formal
planning and management system for industry remained much the same.
In 1971–1975 there was a Fourth Five-Year Plan. As in the 1950s there were
591

dwight h. perkins
annual plans that governed what was produced and distributed but did so
within the broader guidelines of the five-year plans. That was the way the
system was designed, but in reality the operation of this Soviet-type com-
mand system was modified in major ways by two other movements. The first
of these was the Great Proletarian Cultural Revolution initiated by Mao in
1966 that among other things caused the collapse of the agencies responsible
for planning. The second was the Third Front industrial development pro-
gram created at much the same time to deal with the perceived growing
military threat from both the United States and the Soviet Union.
The Cultural Revolution, unlike the Great Leap Forward, was initiated
with political or “class struggle” ideological objectives. There was no specific
economic content to these objectives and the movement did not make any
major policy or organizational changes in the economic sphere. The move-
ment, however, did have a major impact on the economy. To begin with, the
first years, particularly 1967–1969, involved rising conflict between different
factions of students and workers that at times resulted in pitched battles with
military-style weapons. There was also the commandeering of passenger
trains by students, “Red Guards,” so that they could travel around the
country fomenting “revolution.” Those in positions of authority, including
industrial plant managers, were attacked. Even high-priority military enter-
prises were not immune to Red Guard actions.25 Anyone with a background
that included international experience was particularly vulnerable. Unlike in
the Great Leap Forward, however, the countryside, although not entirely
immune, was less affected by the disruption.26 Crops were planted and
harvested and the production teams remained in charge of managing farm
output. Private farm plots on which to raise vegetables continued to exist.
The disruption of the Red Guards was gradually brought to an end when
the army was charged with the task of restoring order and ending the “civil
war” between factions. The values inculcated during this first phase of the
Cultural Revolution, however, did not disappear. There was a strong anti-
foreign element that had the effect until after 1976 of limiting efforts to
promote foreign trade and import industrial complete plants and compo-
nents. Powerful anti-intellectual currents continued and the role of experts
was denigrated, inhibiting research. In propaganda, peasants were said to
25
To give only one example, at the factory making jet engines for fighter planes in Xi’an,
Red Guards locked up the senior management and the top engineers in a room in
a factory for nearly a year (personal communication from one of the engineers held
there).
26
Macfarquhar and Schoenhals, Mao’s Last Revolution, pp. 268–72.
592

have insight on technical matters superior to that of trained specialists.

Perhaps most importantly, much of the leadership of government and
industry was simply removed from position and sent to the countryside or
to May 7 schools designed to re-educate leadership cadres. Many of these
people were not restored to positions of influence until the early to late 1970s.
Others died or committed suicide. There were still people in government
who drew up plans and workers and foremen who carried out their daily
duties, but there were few in charge who had the authority to make signifi-
cant changes or develop new approaches.
The principal exception to this lack of leadership in the economic sphere
was what was called the Third Front.27 The Third Front was a major effort to
relocate industry to the interior so that it would be less vulnerable to attack. It
also involved an attempt to make the major economic regions self-sufficient
so that if one was put out of commission the rest could still function. The
decision to launch the Third Front was Mao’s, but it reflected the very real
possibility that China might be drawn into a major war with the United States
or the Soviet Union or both.
The major American escalation in the Vietnam War began almost imme-
diately after the November 1964 presidential election in the United States.
The sustained aerial bombardment of Vietnam by the United States began in
March of 1965. A US invasion of northern Vietnam was not seriously con-
sidered in part because of the fear that it would involve the United States in
another war with China, as in Korea, but the Chinese could not be confident
that that restraint would continue. China at the same time had broken its
alliance with the Soviet Union and relations between the two countries
steadily worsened, with China accusing the Soviet Union of restoring capit-
alism and the Soviet Union replying in kind. There was a major increase in
Soviet military deployment along the border with China that in 1969 would
break out into open conflict on an island in the Ussuri river.
Relocating the center of industry into the interior of the country was an
enormously expensive undertaking. In 1963–1965, 38.2 percent of capital
construction investment was allocated to the Third Front, with southwestern
provinces receiving a particularly large share. Capital construction invest-
ment in Sichuan rose from 5 to 13 or 15 percent of national investment, and
that in Guizhou rose from 1 or 2 percent to 5 percent.28 The level of Third
27
This discussion of the Third Front is based largely on B. Naughton, “The Third Front:
Defense Industrialization in the Chinese Interior,” China Quarterly, September 1988,
351–86.
28
Naughton, “The Third Front,” 362, 365.
593

dwight h. perkins
Front investment nationwide in the Third Five-Year Plan (1966–1970) rose

sharply to 52.7 percent of the national total and involved most of the interior
provinces. In addition to this industrial investment, Mao called on the entire
nation to “dig tunnels deep and store grain everywhere,” and elaborate
tunnel systems were dug under many and possibly nearly all cities.29
The Third Front effort was cut back sharply in 1972, although Third Front
investment in the Fourth Five-Year Plan (1971–1975) was still 41.1 percent of
the nationwide total. The reason for the change was almost certainly the
opening up of relations with the United States. Following Henry Kissinger’s
visit to China in 1971 and the historic visit of President Nixon in February 1972
that established a working relationship between the two countries, there was
no reason to defend against a possible American bombing campaign,
although the perceived threat from the Soviet Union continued.
One further distortion brought about by the Cultural Revolution in its
early phases belongs in any discussion of the impact of that movement on the
economy. That was the chaos in the universities and secondary schools that
concluded in 1968 with the universities being closed and most of the students,
together with many high-school students, sent to the countryside. The
universities were not reopened until 1972 and enrollment was still only
a fraction of what it had been earlier.30 In that year the total number of
students officially enrolled in institutions of higher learning was 134,000,
about half the level of 1956–1960. The number of such students enrolled
each year did not pass a million until 1997.31 Furthermore, selection of higher-
education students when the schools did reopen was not based on examin-
ations but mainly on the perceived ideological posture of the potential
student. Entrance examinations were not reinstated until 1978.
The number of upper secondary-school students enrolled each year, in
contrast, officially only dropped sharply in 1966 and 1967, although one
doubts that much useful education was going on through 1968 and 1969. By
1970, the number of secondary-school students enrolled each year was four or
more times the level of the late 1950s and that number kept climbing rapidly
29
In 1979 when accompanying Senator Henry M. Jackson we were taken on a tour of
Hohhot, the capital of Inner Mongolia. Somewhere in the middle of the city what
looked like garage doors were opened and we drove down into the tunnel system and
continued for about seven kilometers, passing side tunnels branching out in various
directions. We ended up on the edge of hills outside the city. Hohhot was considered at
the time to be on a likely invasion route of Soviet forces.
30
For a further discussion see Walder, China under Mao, pp. 267–70.
31
The enrollment figures are from Guojia tongjiju 国家统计局 (National Statistical
Office), 新中国六十年 (New China’s Sixty Years) (Beijing, China Statistics Press,
2009), pp. 674–5.
594

through 1977. Political interference, however, must have affected the quality
of education negatively at the secondary level as well.
Any estimate of the impact on the economy of the school closings plus the
chaos and highly politicized environment when they were open is not really
possible. Many talented individuals were never able to reach their full
potential. That was particularly true of those who reached the appropriate
age for entrance when the schools were closed and many were not able to
make up the loss later. According to the 1964 census, in that year there were
114 million people aged between seven and twelve and 70 million aged
between thirteen and seventeen. Those were the age cohorts most impacted,
but if enrollments had stayed at the levels of 1957–1960 (8.6 million total
students in both lower and higher secondary schools on average in those
years and 718,000 total students in institutions of higher learning), fewer than
15 percent would have gone on to secondary school and perhaps only
1 percent would have gone on to university. It would be decades of reform
and rapid economic growth before most urban residents and a large but still
minority share of rural residents would graduate from upper secondary
school.
To complete the economic record of China prior to the Third Plenum of
the Party Congress in December 1978 that marked the beginning of radical
reform, the post-chaos period of 1973–1978 needs to be divided into two parts.
Phase one involved Mao’s acceptance of the need for order and the return of
Deng Xiaoping as a principal manager of day-to-day affairs, first in foreign
affairs, then the military, and finally in the economy. He was brought back to
a position of power by Mao but he mainly worked closely with Zhou Enlai in
this period. Zhou, however, was dying of cancer and not strong enough to
actively play a day-to-day management role.32
The second phase was preceded in 1976 by, first, the death of Zhou, then
the removal of Deng from office, and finally the death of Mao, followed by
the arrest of the radical Cultural Revolution group, the “Gang of Four.” In the
background to this political drama was the Tangshan earthquake that killed
hundreds of thousands and had much of the population of Beijing, among
32
See, for example, the discussion of E. Vogel, Deng Xiaoping and the Transformation of
China (Cambridge, MA, Harvard University Press, 2011), pp. 74–119. In July 1974
I accompanied Senator Henry M. Jackson to China on a visit that included a five-
hour discussion with Deng. Zhou Enlai was said to be unavailable but the Chinese
government people finally relented and allowed Jackson and his wife (but none of his
aides) to spend a few minutes with Zhou in what was described to his aides by Helen
Jackson, who had familiarity with such matters, as basically a hospital room in
Zhongnanhai full of oxygen tanks, among other medical paraphernalia.
595

dwight h. perkins
other places, living outdoors in fear of the aftershocks. With the end of the
radical “leftists,” there was a return to a more or less orderly centrally
planned command system, but one unconstrained by attacks from the left
on such issues as whether it was appropriate to import foreign technology.
The economy in 1973–1975 was run in a reasonably orderly centrally
planned way, although political campaigns continued to have a prominent
role. Exports began to grow from US$2.26 billion in 1970 (the same amount as
in 1959) to US$5.82 billion in 1973 and US$7.26 billion in 1975. Imports that had
shifted back toward producer goods in the Third Five-Year Plan continued at
that level in the fourth plan. Because of the rise in exports, the shift back to
imports of producer goods meant that their total rose 3.3 times from
US$1.9 billion in 1970 to US$6.2 billion in 1975.33 Particularly when this
increase in foreign-made producer goods involved such things as complete
chemical fertilizer plants, it was a source of political attacks from the Cultural
Revolution radicals. However, while those attacks may have inhibited some
trade deals, they did not stop many of them. Industrial growth accelerated to
9 percent per year in the fourth plan from 7 percent a year in the third plan,
and conceivably the greater import of foreign producer goods played some
role in this rise.
The year 1976 was mainly one of political upheaval, but the economy was
affected and GDP fell by 1.6 percent as compared to a rise of 8.7 percent
the year before. Of far greater importance for economic policy, as well as
everything else, was the arrest of the Gang of Four and the end of radical
criticism from the left. Deng Xiaoping was restored to a position of influence
over the economy but he was not yet in overall charge as was to be the case
by late 1978. Mao’s designated successor, Hua Guofeng, still held the top
leadership posts. The one policy change, and it was a major one, was to end
the criticism of those wishing to import technology from abroad. Enterprises
in 1977 were instead encouraged to go abroad for advanced technology and
they reacted with enthusiasm. One international bank estimated that Chinese
enterprises in 1977 and 1978 signed letters of intent to purchase goods from
abroad worth around US$600 billion.34 With exports in the two years of 1977–
1978 of only US$17.3 billion, imports of such magnitude were out of the
question and most of these letters of intent had to be torn up.
Offshore oil exploration was becoming an important possibility in the late
1970s as well and in the new atmosphere the government began negotiating
33
These estimates were derived from data in National Statistical Office, Statistical
Yearbook of China 1981, pp. 357–8.
34
Verbal communication.
596

with foreign oil companies to participate in exploration and production –

something that would have been impossible before 1976. GDP growth of
7.6 percent in 1977 involved an element of recovery from the disruptions of
1976, but the official 11.7 percent growth rate in 1978 was robust. At that point,
the Third Plenum of the Eleventh Central Committee of the Communist
Party late 1978, with Deng Xiaoping in charge, set China on a very different
path.
An Evaluation of Economic Performance 1952–1978

An evaluation of China’s overall economic performance during the pre-
1978 period is best done taking the period as a whole. We have corrected
the official GDP figures for this period for the distortion caused by the very
high state-set industrial prices. The results of this calculation are presented in
Table 16.4.
At 5 percent overall and 3.1 percent per capita, the published official GDP
growth rates for 1958–1978, one would have cause to wonder why the
Chinese government was concerned about the poor performance of the
economy at all. Per capita GDP would have nearly doubled. In fact, even
at market prices, per capita GDP increased by 50 percent, a lower but still
nonnegligible figure. As the discussion that follows indicates, however, actual
incomes and consumption of the population grew much more slowly than
that.
Table 16.4 Price distortions in GDP growth estimates

Official Chinese GDP growth rates Chinese GDP growth rates
(early-year state-set prices) (year 2000 market prices)
1953–1957 9.25 6.32

1958–1961 −1.13 −4.51
1962–1965 9.54 9.94
1966–1969 4.01 3.33
1970–1975 5.9 7.2
1958–1976 5 3.9
Note. The methodology used to recalculate GDP growth in year 2000 market prices is
entirely based on official Chinese data. For more on the methodology see D.H.
Perkins and T.G. Rawski, “Forecasting China’s Economic Growth to 2025,” in
L. Brandt and T.G. Rawski (eds.), China’s Great Economic Transformation
(New York, Cambridge University Press, 2008), pp. 834–5.
597

dwight h. perkins
Table 16.5 Sources of growth (1952–1990)

Period GDP Fixed capital Raw labor Education-enhanced TFP growth
1952–1957 6.5 1.9 1.2 1.7 4.7

1958–1965 2.4 5.2 1.5 2.1 −1
1966–1978 4.9 7.7 2.4 3.1 −0.2
1979–1985 9.7 9.2 3.4 4.5 3.2
1986–1990 7.7 6.9 2.5 2.9 3.1
Source: this is a modified version of Perkins and Rawski, “Forecasting China’s

Economic Growth to 2025,” Table 20.2, p. 839
An analysis of what was driving this GDP growth in the pre-1979 period can
usefully start with an analysis of the sources of growth during these years. In 1952–
1957 Chinese economic growth enjoyed a high-growth spurt. The investment
rate had begun to rise but the growth rate of the capital stock was still very low.
Growth was mostly (72 percent) explained not by the growth in labor and capital
inputs, but by total factor productivity. It is not entirely clear why the growth rate
of productivity in this period was so high given that the period included the
potentially disruptive effects of the collectivization of agriculture and the state
takeover of industry. The most likely explanation is that high growth was largely
a result of the comparative stability of this period after years of civil war and
hyperinflation. Growth was more a matter of bringing existing agricultural and
industrial units up to full capacity than of an increase in new capacity.
The years 1958 through 1965 began with the massive mobilization of rural
labor and the general chaos of the Great Leap Forward, followed by the
famine and then recovery. Investment in modern plant and equipment also
accelerated and that largely explains why GDP in 1965 was 21 percent higher
than in 1957. The efficiency with which that plant and equipment were used,
however, was undermined by the politics and chaos of the early part of that
period and there is little doubt that that explains the decline in total factor
productivity of capital and labor over the entire eight years.35
There was no total factor productivity growth in 1966–1978 either. GDP
growth was mainly explained by the 7.7 percent annual average growth of capital
throughout the period. The chaos of 1967–1968 is part of the reason for the
slightly negative TFP estimate. The general inefficiency of Soviet-type centrally
35
If the periodization was broken down further, it would show that there was
a particularly sharp decline in TFP in the first years of the period followed by a rise
that brought average TFP growth up to minus 1 percent.
598

planned economic systems, combined with the obvious inefficiency of Third

Front investment, however, are probably the main reasons for the lack of
productivity growth. The chaos of the early years of the Cultural Revolution
was more like a series of worker strikes temporarily closing down or slowing
production, with production quickly recovering when the strikes ended. Sorting
out and quantitatively estimating the precise contributions of each of these
components of inefficient growth, however, is far beyond what is possible in
this chapter. Finally, what separated the performance of the first decade of reform
after 1978 from what occurred before 1978 was not the slight acceleration in the
growth rate of the capital stock. It was the large jump in total factor productivity.
The performance of GDP is one element of any appraisal of the pre-
1979 period. What actually happened to the standard of living of the
Chinese people in this period is arguably even more important and GDP
growth does not necessarily translate into better living conditions for the
majority of a country’s population. We will begin with the rural population
since that made up 87.5 percent of the total population, falling to 82 percent in
the 1960s and staying at that level until 1978.
Several indicators related to rural incomes are presented in Table 16.6. The
share of per capita agricultural value added (all wages and profits from farming,
forestry, and fisheries) in GDP rose slightly during the First Five-Year Plan, fell
sharply during the famine years, and then rose back to the level of 1957 but no
further until the return to household agriculture after 1978. Essentially the same
story holds true for the output of grain, the main source of calories in the Chinese
diet.
The outlier index is the National Statistical Office’s estimate of rural per capita
consumption, which indicates that the decline in per capita consumption in the
famine years was only 6.4 percent and the rise in per capita consumption by the
1970s was 50 or more percent above 1952 (or 35 percent above 1957). The estimated
modest decline in the famine years does not appear to be plausibly related to the
widespread malnutrition and death that occurred in those years. The growth in
consumption after 1965 could be partly explained by the rise in income from rural
industries that were actively promoted in the 1970s and mainly employed rural
labor, but only partly. These industries only used 10 percent of the rural labor
force in 1980, and probably fewer workers earlier, and their per capita incomes are
not likely to have been much higher than the average wages from farming in the
regions where they were located.36 The most plausible conclusion is that the per
36
This 10 percent share includes employees in commune, brigade, and team enterprises
(roughly 30 million workers). In the 1970s, China included income from brigade and
team enterprises in gross value of agricultural output and the contribution of brigade
599

dwight h. perkins
Table 16.6 Per capita rural economic performance

Grain output
Primary value added (per capita) Rural consumption
1952 100 100 100

1957 110.9 109.6 116.8
1961 81.9 78.9 93.6
1965 105.4 100.4 125.2
1969 102.1 97.3 134.2
1975 112.1 114.4 150.4
1978 108.3 118.4 157.6
1984 162.6 155.6 272.2
1990 194 162.8 339.4
Source: National Statistical Office, 新中国六十年, pp. 608, 614, 637
capita consumption estimates are biased upward and give an overly favorable
view of conditions in the countryside after 1958. This bias is not really surprising
since reliable rural consumption figures require carefully managed sample sur-
veys of rural areas and there is no evidence that surveys of that sort were carefully
carried out in the 1958–1978 period as they have been in the post-1978 reform years.
Rural incomes, therefore, were nearly stagnant and did not grow by more
than 10 percent or so (mainly due to rural small-scale enterprises) after 1957.37
The situation in urban areas, in contrast, was different. Wages of workers and
employees did not increase at all. The figures for nominal average wages for
both state and urban collective enterprises are in Table 16.7. Wages in current
prices peaked in 1957 and then fell, not recovering to the 1957 level until after
1978. Urban consumer prices rose significantly after 1957, peaking in 1962 and
then falling, but not back to the level of 1957. From 1965 through 1976
consumer prices were unchanged, rising by a small amount in 1977 and
1978. Thus the nominal drop in wages in the early 1960s was reinforced by
and team enterprises to gross agricultural output in 1978 was 11.7 percent, up from
around 2 or 3 percent in 1965. The total wage bill from commune, brigade, and team
enterprise in 1980 was equivalent to only 8 percent of agricultural value added. The data
on rural enterprises are from National Statistical Office, Statistical Yearbook of China 1981,
pp. 137, 195.
37
The National Statistical Office estimated the rise in nominal rural net income per capita
to be seventy-three yuan in 1957, 107.2 yuan in 1965, and 117.1 yuan in 1977, but did not
attempt to deflate these figures by the price increases in the early 1960s. Since prices rose
hardly at all between 1965 and 1977, however, these nominal estimates are likely to be
close to the real estimates. National Statistical Office, 新中国六十年, p. 627.
600

Table 16.7 Urban employment and consumption estimates

Urban population Urban consumption
Urban employees* growth Urban wages per capita
1952 100 100 445 100

1957 128.9 138.9 624 131.7
1960 246.1 182.5 511 113.7
1962 182.5 162.8 551 101.9
1965 206.6 182.1 590 144.2
1970 253.9 201.4 561 158.3
1975 330.7 223.8 580 194.2
1978 382.7 240.8 615 217.6
1984 491.9 335.3 974 276.7
1990 685.5 421.5 2140 415.1
* Urban employees includes collective as well as state and private enterprises.

Source: National Statistical Office, 新中国六十年, pp. 608–10, 616
the price increase, but for the rest of the period real wages and nominal wages
did not rise or fall.
What, then, explains rising urban household consumption per capita in 1965
through 1978? The primary answer, as discussed earlier, is that urban enter-
prises, blocked from bringing in migrants from the countryside, turned instead
to women urban residents. The share of women in the urban labor force rose
from under 15 percent of the total urban workforce to over 40 percent. In
absolute numbers that was an increase of well over 30 million women, starting
from a small base of 2 million urban employed women in 1955.38 The average
urban household in 1981 with 4.24 members included 2.39 employed persons.
I have not found estimates for the 1950s but the number of employed persons
per household must have risen by nearly one person. Still, from 1965 through
1978 urban consumption per capita increased at a rate of only 1.8 percent
a year, far below the rates that would be achieved after 1978.
In addition, what urban areas were able to consume depended on what the
government enterprises were willing to produce, and the centrally planned
command economy did not often make decisions based on consumer prefer-
ences. In probably the most egregious example, the amount of urban living space
per capita was allowed to decline to 4.4 square meters per capita in 1978.39
Priority in construction and industry was not focused on consumer goods.
38
Chen, Chinese Economic Statistics, p. 482.
39
601

dwight h. perkins
Heavy industry (roughly similar to the producer goods industry) rose from
35.6 percent of industrial gross value output in 1952 to 57.3 percent in 1978.
China’s planners were following the closed-economy strategy of
Preobrazhensky throughout the pre-reform era. Machines were produced in
order to make more machines that someday in the future would make more
consumer goods.
With respect to the distribution of income, China’s image in this period
was of a country that was one of the most egalitarian in the world. The land
reform that began before 1949 and continued through the first years of
Communist Party rule did significantly reduce income inequality. Land was
given to the actual tillers, many of them former tenants paying rent to
landlords.40 The landlords were stripped of their land without compensation
and often with the loss of their life. Urban private enterprises were in effect
taken over by the state also without compensation to the private owners.
Wealthy individuals who did not flee with their wealth lost most or all of it.
Thus there is little doubt that inequality in income had fallen sharply by the
mid-1950s.
The impression held by some visitors to China that this decline in inequal-
ity largely eliminated differences in income, however, was false. Within rural
areas the work point system used to determine individual and hence family
incomes in the co-operatives and communes was based on the amount of
work each member did. That in turn led to some inequality between families
with more than one adult worker. The major rural differences in income,
however, were due to differences in the amount of co-operative or commune
land per capita, the quality of that land, and its location (near a city meant
much higher incomes for the members). For the most part the government
did not attempt to reduce regional differences in inequality by moving farm
families from poor regions to richer regions or vice versa. Such a move, if it
had been tried, would probably have been politically explosive. Subsidies to
the poorer communes were also limited until many years after the beginning
of the reform period.
Urban inequality was driven mainly by government-set wage systems. The
basic eight-grade system for most urban workers had a wage differential
between each grade of about 20 percent, with the top grade roughly three
times the bottom grade. There were also other systems for technical and
management personnel and government officials, among others. The
40
The one quantitative study I am aware of that estimated the impact of land reform on
income distribution was by C. Robert Roll.
602

average monthly wage in 1957 was a bit over RMB 50 and wages of top
professionals could be RMB 200 or more a month. The differential between
the top and bottom wages in enterprises could be as much as sevenfold or
more.41 These wage systems came under attack during the Cultural
Revolution but, however they may have been modified at that time, the
system in the 1970s was much the same as had existed when it was set up in
1956–1957. Political attacks were not just against the wage structure but also
against anything that appeared to the “leftists” to be emphasizing material
incentives. Thus piece-rate wages were sometimes attacked and abolished,
managerial bonuses were used sparingly or not at all, and individuals in the
1960s and 1970s could go for long periods without promotions.
There were in essence no property incomes in urban China from 1957 until
well after 1978. The closest approximation to property incomes was the
special privileges of the elite. The top government and enterprise officials
had larger and better apartments and access to automobiles, and in some
cases could command staff to provide them with services. At the very top,
individuals such as Mao, Jiang Qing, and a few others could command
airplanes and private automobiles. This top-level elite was a very narrow
one, however. To give one a general idea of its size, the number of passenger
automobiles in the country was 34,400 in 1957 and had only risen to 173,000 by
1975.
There were no surveys of inequality during 1958–1978, or at least none that
have become publicly available, so it is not possible to measure in any precise
way the Gini coefficient or some other measure of inequality for that period.
There was, however, a major survey done in 1988 early in the reform period
that has been analyzed in depth by a number of scholars.42 This survey is not
ideal for looking back at inequality before 1978. Still, 1988 was well before the
rising role of private enterprises, the privatization of many of the township
and village enterprises, the privatization of housing, and a level of rural-to-
urban migration where most of those registered as being in rural areas
between the ages of eighteen and forty were in fact working in the cities.
Those changes would lead to an explosion in inequality, but in 1988 the level
of inequality was probably only modestly above the level of the 1970s.
41
See C. Howe, Wage Patterns and Wage Policy in Modern China, 1919–1972 (Cambridge,
42
For analyses based on this survey, see K. Griffin and R. Zhao (eds.), The Distribution of
Income in China (New York, St. Martin’s Press, 1993). In addition to the chapter by the
editors, see also the chapters by Carl Riskin, Li Shi, John Knight, Song Lina, Azizur
Rahman Khan, and Terry McKinley.
603

dwight h. perkins
The one element of inequality that can be measured for 1952 through 1978
is the difference between average urban and rural incomes. This difference is
a major source of inequality in all countries and that was and is the case in
China. Two alternative estimates of the urban–rural per capita consumption
ratio and its changes over time are presented in Table 16.8. The “official” ratio
is based on the National Statistical Office’s estimates. The “estimate” ratio
was obtained by dividing the official urban per capita income by rural per
capita income, assuming that rural income in real terms was better repre-
sented by per capita agricultural value added than by the official estimates for
rural per capita consumption in this period. This latter procedure may
overstate the ratio in 1975 and 1978 because there was some commune
industrial income that is not included in household income in those two
years, but that income was not likely to have been much, if at all, more than
10 percent of total rural income.
If one accepts the “estimated ratio” as the more accurate, it is clear that the
rural–urban consumption ratio rose rapidly from 1965 to 1978. The combin-
ation of slow agricultural growth with tight restriction on the migration of
rural people out of the countryside was the main cause of this rise.
China’s overall economic performance during 1958–1978 therefore
involved modest GDP growth, little increase in per capita consumption in
the rural areas, and 2 percent a year growth in urban areas. It is not surprising
that China’s leadership in 1978 saw this overall performance as inadequate,
and that view was reinforced by the obvious superior economic performance
of several of China’s neighbors.
Table 16.8 Urban–rural per capita

consumption ratio
Official Estimated
1952 2.37 2.37

1957 2.71 2.81
1961 2.85 3.6
1965 2.49 3.14
1969 2.41 3.44
1975 2.68 3.96
1978 2.93 4.59
1984 2.15 3.89
1990 2.85 4.66
Sources: see text
604

The post-1978 reform period is often described as a rejection of the Soviet-

type centrally planned command system described in this chapter in favor of
a market economy and capitalism – in effect a complete transformation of
China’s economic model. There is considerable truth in this characterization,
but there were also elements of the pre-1978 economic system that persisted
for some time after 1978 and in a few cases were still present during
the second decade of the twenty-first century. That story, however, is
covered in other chapters in this volume.
Further Reading
Donnithorne, A., China’s Economic System (New York, Praeger, 1967).
Eckstein, A., W. Galenson, and T.-C. Liu (eds.), Economic Trends in Communist China
(Chicago, Aldine, 1968).
Griffin, K., and R. Zhao (eds.), The Distribution of Income in China (New York, St. Martin’s
Press, 1993).
MacFarquhar, R., and M. Schoenhals, Mao’s Last Revolution (Cambridge, MA, Harvard
Naughton, B., “The Third Front: Defense Industrialization in the Chinese Interior,” China
Quarterly, September 1988, 351–86.
Perkins, D.H., Market Control and Planning in Communist China (Cambridge, MA, Harvard
Walder, A., China under Mao: A Revolution Derailed (Cambridge, MA, Harvard University
Press, 2015).
Wu, J., Understanding and Interpreting Chinese Economic Reforms (Mason, OH, Thomson/
South-Western, 2005).
605

17
Living Standards in Maoist China
chris bramall
Introduction
Support for the Chinese Communist Party (CCP) in the late 1940s owed much
to its willingness to “stand up” against external interference in Chinese affairs,
whether by Japan, the USA, or any other colonial power. This nationalist
agenda may have been the decisive factor in its victory during the civil war, and
it continues to be a key driver of popular support for CCP rule.
Nevertheless, the CCP’s social agenda – that, on assuming power, it would
reduce poverty and inequality by promoting growth and by redistributing assets,
particularly farm land – was also important in garnering support.1 Precisely for
that reason, much of the scholarship on the Maoist era has examined the degree
to which this “promise of the revolution” had been fulfilled by the time of Mao’s
death. By the end of the 1980s, the settled conclusion was that this second CCP
promise was largely unfulfilled. To be sure, the regime’s achievements in terms
of social indicators – especially life expectancy and morbidity – were recognized
to be extraordinary, the mortality crisis of 1958–1962 notwithstanding. However,
it was all but universally believed that the Party had failed to raise material living
standards – and that this failure was both an indictment of Chinese socialism and
a powerful explanation of the policy changes which occurred after 1976–1978.
Little of significance has been published on these issues in recent years.2 An
I am indebted to Tom Rawski, Debin Ma, and Xu Chenggang for their comments on an
earlier version of this chapter.
1
Xu Chenggang has suggested to me that the CCP never intended to keep these promises.
However, even accepting this point, it seems reasonable to judge the CCP’s record
against generally accepted measures of “development” rather than in terms of whether it
achieved its own self-defined objectives.
2
Some of the more influential older pieces are N. Lardy, “Consumption and Living
Standards in China, 1978–83,” China Quarterly 100 (1984), 849–65; World Bank, China:
Socialist Economic Development, vol. 1, The Economy, Statistical System and Basic Data
(Washington, DC, World Bank, 1983); C. Riskin, China’s Political Economy (Oxford,
Oxford University Press, 1987); and C. Bramall, In Praise of Maoist Economic Planning
(Oxford, Clarendon Press, 1993). A fine recent summary of China’s development is
606

indication of the scholarly malaise is that, for something as pivotal as food

consumption, we are reliant on the estimates made by Piazza in the early
1980s. His work was pioneering, but the sources available to him at that time
were limited. A far richer range of Chinese materials is now available, and the
time is therefore ripe for a re-evaluation of Maoist living standards.
This chapter takes up the challenge of re-evaluation. It begins by looking at
estimates of per capita GDP and per capita income. However, our GDP
estimates for Maoist China are based on fragile statistical foundations. In any
case, per capita GDP is a poor guide to consumption. The same is true of data
on per capita income. Accordingly, this chapter looks directly at trends in
food consumption, the key component of material living standards. Then we
consider broader measures of living standards, principally life expectancy.
The sixth section expands the discussion to cover distributional questions.
The contribution made below is original in two respects. First, it offers
new estimates of food consumption by using the abundant materials in
Chinese which are now available. These allow us to construct more accurate
food balance estimates of consumption (based on production data) than was
previously possible. Second, it presents new estimates of rural income
inequality, especially for the period immediately after land reform, based
on more detailed information from the 1954 rural income survey.
Trends in Per Capita GDP

In thinking about the Maoist record on per capita GDP, it is logical to
compare the late 1970s with the immediate prerevolutionary period. In
practice, because of the distorting effects of both the war with Japan and all-
out civil war between the CCP and the Guomindang in the late 1940s, the best
comparison is with 1934–1936, when the Nanjing regime was relatively well
established and before the Japanese invasion. The most consistent series on
long-run per capita GDP is that estimated by Angus Maddison.3 This shows
average per capita GDP rising from US$570 in 1934–1936 to US$886 for the
offered in L. Brandt, D. Ma, and T. Rawski, “From Divergence to Convergence:

Reevaluating the History behind China’s Economic Boom,” Journal of Economic
Literature 52.1 (2014), 45–123. One recent paper is L.-L. Chen and J. Devereux, “The
Iron Rice Bowl: Chinese Living Standards 1952–1978,” Comparative Economic Studies 59
(2017), 261–310. However, this does little more than rehearse the findings of the older
literature; it makes scant attempt to interrogate that literature’s underlying assumptions
or to engage with Chinese sources.
3
A. Maddison, “World Population, GDP and Per Capita GDP, 1–2003 A D” (2010), at www
.ggdc.net/Maddison (accessed March 25, 2016).
607

chris bramall
1974–1978 period, a long-run real-growth rate of around 1 percent per year

(1990 prices). However, although the Maddison data are widely used, the
limitations of his estimates for the 1930s are equally widely recognized.4 The
main problem is that the underlying yield and cultivated-area data are
unreliable; because agriculture was the largest sector, this compromises the
GDP estimates as a whole. The estimation of per capita GDP is rendered
even harder by a lack of reliable population data; China’s first modern and
relatively reliable population census occurred only in 1953, and its results
cannot easily be back-projected because of the impact of war.
We are on safer ground in comparing the early Maoist era with the 1970s
because the establishment of the State Statistical Bureau (国家统计局,
hereafter SSB) in the 1950s led to more systematic data collection. Even so,
these estimates are still subject to considerable error. The data for the early
1950s were largely based on the old system of data collection used by the
Guomindang’s National Agricultural Research Bureau, which underreported
cultivated area. Underreporting was also pervasive in the 1970s, both by
collective farms (to avoid procurements) and by the private sector (widely
viewed as a “capitalist remnant” during the Cultural Revolution). Putting
these caveats aside, the SSB data suggest that per capita growth for 1952–1978
was 2 percent per year. A 2 percent per capita GDP growth rate is not a bad
achievement and certainly better than China’s record during the “Great
Divergence” of the Qing era.5 It would have been better still but for shocks
inflicted by the Great Famine (1958–1962) and the early years of the Cultural
Revolution (1966–1968).6
Trends in Per Capita Income

The real issue, however, is whether disposable income and consumption rose
at the same rate as GDP. We know that the investment share was substan-
tially higher by the late 1970s than it had been in the 1930s or the 1950s;
a comparison of the estimates made by Liu and Yeh for the 1930s, and the
4
K. Deng and P. O’Brien, “China’s GDP Per Capita from the Han Dynasty to Communist
Times,” LSE Economic History Working Paper 229 (2016).
5
The Qing data are sufficiently unreliable that we cannot be sure that growth accelerated;
see Deng and O’Brien, “China’s GDP.”
6
This is not the place for a discussion of the causes of slow growth. Most scholars would,
of course, blame the system of Soviet-style central planning (systemic failures), but the
impact of China’s high level of military spending (especially on the program of Third
Front industrialization, which focused mainly on inhospitable parts of western China)
also played an important role. See Bramall, In Praise of Maoist Economic Planning.
608

World Bank for the 1970s, suggests that the share of gross capital formation
rose from about 6 percent of GDP in the 1930s to 35 percent for 1974–1978.
This rise in the investment share implies the possibility of a consumption
squeeze. However, the outcome depends on whether the investment was
productive. After all, the logic of the Maoist/Soviet approach to industrial-
ization – set out in the classic Feldman model of the 1920s – was that raising
the investment share in the short run, and using that higher rate of invest-
ment to expand the supply of producer goods, would lead to a dividend in
terms of higher per capita consumption in the long run. We therefore need to
measure consumption trends directly, and not just rely on GDP per capita as
a proxy.
Rural Incomes
One way of evaluating consumption trends is to use the national surveys of
rural income and expenditure. The surveys conducted during the 1950s were
quite reliable. The first national survey by the SSB in 1954 covered 14,334
households.7 The 1957 national survey sampled 17,378 households (Table 17.2),
These surveys covered most provincial-level units. For example, the con-
sumption part of the 1954 survey included households from twenty-one
provinces; Yunnan, Jiangxi, and Tibet were the only provinces omitted.
Later surveys were considerably smaller; the 1962 survey sampled 4,658
households and the 1965 one included 11,632 households. These SSB all-
China surveys ceased after 1965, and resumed only in 1976 (though some
provinces, such as Hubei, conducted their own surveys during the 1970s).
The surveys of the mid-1970s were smaller than those of the 1950s. That of
1976, covering 3,646 households, was but a shadow of the survey of 1954. The
1977 and 1978 national surveys of 6,089 and 6,095 households across twenty
provincial-level units were little bigger. Only in 1979 was the 10,000-
household mark breached, and the 1954 survey total was not exceeded until
1980. The small scale of the national surveys of the late 1970s meant that
provincial subsamples were often very small indeed. Admittedly, some
provinces were well represented in the national survey; 750 Hubei house-
holds were included in the 1979 survey.8 But the overrepresentation of Hubei
7
The number of households included in the national surveys is given in Table 17.2 below.
For a useful English-language discussion, see Y. Matsuda, “Survey Systems and Sampling
Designs of Chinese Household Surveys, 1952–1987,” Developing Economies 28.3 (1990),
329–52.
8
Hubei tongjiju 湖北统计局, 农民家庭经济调查资料 1974–1982 (Rural Household
Economy Survey Materials 1974–1982) (Wuhan, Hubei tongjiju, 1983), p. 27.
609

chris bramall
meant the underrepresentation of other provinces. Even in the 15,914-

household survey of 1980, provincial sampling was uneven. The Henan
sample, for example, had increased from 270 households in 1978 to 1,296 in
1980, with thirty-six counties included instead of nine, and the Sichuan sample
included 2,181 households. Moreover, the small samples for Ningxia and
Shanghai (200 and 160 households respectively) reflected their small popula-
tions. However, some populous provinces were grossly underrepresented.
For example, the Shaanxi survey included only 328 households, and the
Hunan survey just 327 households in 1980.9 There were other problems
with the SSB surveys of the 1970s. Although the SSB used the comprehensive
data collected by the Ministry of Agriculture for every collective in selecting
production teams for inclusion, the production teams sampled during the
mid-1970s were not randomly chosen but “typical examples” of rich, mid-
dling, and poor teams. By requiring the inclusion of a rich team in every three
sampled from a population that was largely poor, the typical example
methodology oversampled rich teams. Within teams, however, poor house-
holds were undersampled because they lacked the language and arithmetic
skills needed to record annual income and expenditure. The overinclusion of
rich teams, and the omission of many poor households, generated an upward
bias to survey per capita income in 1977, and especially during 1978 when
Deng Xiaoping assumed control. This suited the politics of the late 1970s; it
was highly desirable that the surveys demonstrated the success of the post-
Mao policy changes.
What the surveys did do was to reveal a national increase in per capita
rural income from sixty-four yuan in 1954 to 133 yuan in 1978, marginally less
than 3 percent per annum (Table 17.1). The rise occurred in poor provinces
such as Guizhou, Sichuan, and Anhui; middle-ranking provinces like Jiangsu;
and affluent provinces such as Liaoning. The real rate of increase was about
2 percent per year – smaller than the nominal rate but nevertheless significant
enough.10
9
The number of households surveyed by province in 1980 is listed in Henan tongjiju 河
南统计局, 河南省农村经济调查统计资料 1982 (Statistical Materials from a Survey
of Henan’s Rural Economy, 1974–1982) (Zhengzhou, Henan tongji chubanshe, 1982),
pp. 567–9.
10
I use the implicit deflator from the State Statistical Bureau’s real personal expenditure
(xiaofei 消费) series to estimate real income growth here. Real rural incomes were
affected by increases in the price of urban goods sold in rural areas. However, the trend
in agricultural procurement prices also mattered. To be sure, rising procurement prices
improved the internal terms of trade, which benefited surplus-producing collectives
and is captured in the nominal income data. But a rising proportion of procured
commodities was resold to the rural sector; by 1978, around 46 percent of expenditure
610

Table 17.1 Per capita net peasant income, 1954–1978 (current yuan)
China Guizhou Anhui Sichuan Henan Jiangsu Liaoning
1954 64 51 61 56 66 71 105
1956 73 65 65 66 68 86 n/a
1957 73 64 75 68 65 86 109
1962 99 113 131 121 77* 107 n/a
1965 107 110 109 106 74 114 125
1975 113** 90 n/a 95** n/a n/a 128
1978 133 109 113 127 105 155 185
*1963, **1976
Note. These data show a significant increase in nominal income between 1957 and 1962
but the real increase was negligible because of inflation during the Great Leap Forward.
Sources: Nongyebu 农业部, 农业经济资料 1949–1983 (Materials on the Agricultural
Economy) (Beijing, Nongmuyuye bu, 1983), p. 523; Zhonggong Sichuan sheng wei
yanjiushi 中共四川省委研究室, 四川省情 (Conditions in Sichuan) (Chengdu,
Sichuan renmin chubanshe, 1984), p. 226; Guojia tongjiju 国家统计局 (National
Statistical Office), 各省自治区直辖市农民收入消费调查研究资料汇编
(Summary Research Materials from Rural Income and Expenditure Surveys in
China’s Provinces) (Beijing, Zhongguo tongji chubanshe, 1985), pp. 139, 223, 349, 434–
6; Liaoning tongjiju 辽宁统计局, 辽宁省农村发展五十年 (Fifty Years of Rural
Development in Liaoning) (Shenyang, Liaoning tongji chubanshe, 1999), p. 367;
Jiangsu tongjiju 江苏统计局, 江苏农村经济五十年 (Fifty Years of Jiangsu’s Rural
Economy) (Beijing, Zhongguo tongji chubanshe, 2000), pp. 153–4; Anhui sheng
renmin zhengfu 安徽省人民政府, 安徽五十年统计资料 (Statistical Materials on
Fifty Years of Anhui) (Beijing, Zhongguo tongji chubanshe, 1999), p. 739
Alternative income data were collected by the Ministry of Agriculture (nongye

bu 农业部, hereafter MOA) – see Table 17.2. The MOA had access to
a comprehensive reporting system: every collective was required to report
collective income and output data to the ministry. By estimating income
generated in non-collective activities (notably household sidelines), the MOA
was able to derive estimates of overall per capita income. However, the
MOA grossly undersampled sideline activity and hence total income. In 1978,
the MOA figures in Table 17.2 imply sideline income of just seventeen yuan,
was on commodities and (an imputed) 47 percent on own-produced goods, compared

with 31 percent and 63 percent respectively in 1952. See Guojia tongjiju 国家统计局
(National Statistical Office), 中国国内生产直核算历史资料 1952–1995 (Historical
GDP Data on China, 1952–1995) (Beijing, Dongbei caijing daxue chubanshe, 1997), p.
44. The rise in agricultural procurement prices by 59 percent between 1954 and 1978
therefore had a significant impact on rural households, making them significant market
purchasers of agricultural goods – most obviously, grain-deficit households.
611

chris bramall
Table 17.2 Alternative estimates of national per capita peasant income (yuan)
Total net income per Of which: net collective distributed
head income per head SSB sample size
SSB MOA SSB MOA (households)
1954 64 n/a 2 n/a 14,334

1956 73 52 45 43 14,172
1957 73 52 43 41 17,378
1962 99 56 52 46 4,658
1963 101 57 55 46 9,724
1964 102 61 55 48 12,095
1965 107 66 63 52 11,632
1976 113 80 78 63 3,646
1977 117 81 76 65 6,089
1978 134 91 89 74 6,095
1979 160 103 102 83 10,282
1980 191 104 108 86 15,914
1981 223 117 116 101 18,529
Notes. (1) Net income here is gross income minus production costs. (2) Data are for
nominal per capita income. (3) The first SSB survey was conducted in 1954 and the first
reporting by collectives to the MOA was in 1956. Both surveys ceased during the late
1960s.
Sources: Nongyebu, 农业经济资料, pp. 516–17, 521, 523
whereas the SSB survey put the figure at forty-five yuan.11 This SSB figure is
more plausible; we know that decollectivization had begun in some areas by
1978, and that the restrictions on household activities imposed during the
Cultural Revolution (on, for example, chicken and pig breeding) had been
largely removed.
The key issue, however, is not the absolute level of income but the trend. The
surveys may have been biased in the late 1970s, but it is not obvious that these
biases changed much over time. Political imperatives during the 1950s – a wish to
demonstrate the success of land reform and collectivization – led to a bias
towards success. As for the 1970s, the political pressures were to exaggerate the
damaging impact of the Cultural Revolution, when the SSB itself had been closed
down, and its members rusticated, and to show that the Chinese economy had
performed better after the arrest of the “Gang of Four” (October 1976). These
pressures imparted some upward bias to the data for 1977–1978.
11
These numbers are the difference between total and collective income in Table 17.2.
612

In short, therefore, the incentive to overstate success was present both in

1954–1957 and after 1976. The income trend therefore ought to be relatively
unbiased. And this trend was clearly upwards. In fact, Table 17.2 shows that
all five measures of per capita rural income rose significantly between the
mid-1950s and the late 1970s.
Urban Income
Survey data for urban incomes are more scarce, probably because the
collection of urban data was a lower priority for the CCP leadership: the
urban population was comparatively small (18 percent of the total in 1978),
and more affluent than the rural population because of higher urban wages
and substantial subsidies.
Nevertheless, we do have large-scale survey results for 1956, 1957, and 1958,
covering over 5,000 households.12 There are also some surveys for individual
cities. One such is that for Shanghai in the mid-1950s, which compares
consumption in 1956 with 1929–1930.13 Post-1957 urban data are more scarce,
though we do have a national survey covering 3,537 households for 1964.14
There are also some provincial data.15 However, large-scale national sample
surveys of consumption ceased between 1964 and 1980.
The urban survey results (Table 17.3) show a growth rate of around
3 percent per annum between 1952 and 1980, in line with rural income
growth.16 The urban retail price index rose by 0.5 percent per year, so real
urban income growth was closer to 2.5 percent per year – about 0.5 percent
12
These surveys necessarily included few households from any individual province. The
1957 Henan survey, for example, included 365 households drawn exclusively from the
large cities of Zhengzhou, Luoyang, and Kaifeng. See Henan tongjiju 河南统计局, 河
南省职工家庭生活调查资料汇编 (Summary Materials from a Survey of Living
Standards in Worker Households in Henan) (Zhengzhou, Henan tongjij chubanshe,
1982).
13
Guojia tongjiju 国家统计局 (National Statistical Office), “1927 年以来上海市工人生
活水平的变化情况” (Changes in the Living Standards of Shanghai Workers since
1927), 经济研究资料 (Economic Research Materials) 16.7 (1957), 1–6. The data show,
for example, that pork consumption was eight kilograms per person per annum, similar
to the national estimate – therefore suggesting that the national survey was skewed
towards more prosperous cities like Shanghai.
14
National Statistical Office, 六五年计划, p. 15.
15
Jiangxi tongjiju 江西统计局, 江西统计年鉴 1982 (Jiangxi Statistical Yearbook 1982)
(Nanchang, Jiangxi tongji chubanshe, 1983), p. 249; Henan sheng renmin zhengfu
diaocha yanjiushi 河南省人民政府调查研究室, 河南生情 (Conditions in Henan)
(Zhengzhou, Henan renmin chubanshe, 1987), p. 623; Liaoning tongjiju 辽宁统计局,
辽宁城市人民生活史料 (Historical Materials on Urban Living Standards in Liaoning)
(Shenyang, no publisher, 1986), pp. 145–6.
16
These exclude subsidies and therefore understate true growth; see also the discussion of
the urban–rural gap below.
613

chris bramall
Table 17.3 Selected per capita urban incomes, 1943–1980 (current yuan)
China Liaoning Henan Beijing Guangdong Wuhan
1952 191 137 126

1956 242 220
1957 254 236 216 251 151 210
1958 253 232 202
1962 231 229* 221 300
1965 243* 233 239 252 243 240
1978 494 365 402 402**
1980 514 508 444 501 462 538
* 1964, **1979.
Sources: Liaoning tongjiju 辽宁统计局, 辽宁城市人民生活史料 (Historical
Materials on Urban Living Standards in Liaoning) (Shenyang, no publisher, 1986),
pp. 145–6; Guojia tongjiju 国家统计局 (National Statistical Office), 六五年计划我国
城镇居民家庭收支调查资料 (Survey Materials on the Income and Expenditure of
Urban Households during the Sixth Five-Year Plan) (Beijing, Zhongguo tongji
chubanshe, 1988), p. 15; Guojia tongjiju 国家统计局 (National Statistical Office),
全国职工家庭生活调查资料 1956, 1958, 1980 (National Survey Materials on
Worker Living Standards 1956, 1958, 1980) (Beijing, Zhongguo tongji chubanshe, 1956,
1958, 1980); Henan sheng renmin zhengfu diaocha yanjiushi 河南省人民政府调查
研究室, 河南生情 (Conditions in Henan) (Zhengzhou, Henan renmin chubanshe,
1987), pp. 621–2; Wuhan tongjiju 武汉统计局, 武汉统计史志资料 (Statistical
Material Record for Wuhan) (Wuhan, Wuhan tongji chubanshe, 1988), p. 942; Beijing
tongjiju 北京统计局, 北京五十年 (Beijing Fifty Years) (Beijing, Zhongguo tongji
chubanshe, 1999), p. 373; Guangdong tongjiju 广东统计局, 广东统计年鉴 1990
(Guangdong Statistical Yearbook 1990) (Beijing, Zhongguo tongji chubanshe, 1990),
pp. 436–7. These Guangdong figures are for living expenditure and therefore lower
than for per capita income
above rural income growth. Much of the rise occurred after the mid-1960s
and the main driver was rising participation rates, especially by women; the
average wage increased very little. The regional pattern of increase shown is,
at least in terms of absolute magnitudes, largely as one would expect.
Wuhan, for example, developed rapidly in the heavy-industry-oriented late
Maoist era and income growth reflects that industrialization. The same is
true of Liaoning. By contrast, Henan’s cities were smaller and not focal points
for heavy industrialization. The same is broadly true of Guangdong. Urban
incomes rose in Beijing at about the same rate as the national average, not
perhaps surprisingly given that it was not a center of heavy industry.
614

Trends in Food Consumption

The data from surveys on per capita income and expenditure are broadly
consistent, and tell a story of rising material living standards. Nevertheless,
because of the sample biases discussed above, it is useful to have corrobor-
ation from other indicators. Perhaps the most useful alternative approach is
to look at food consumption because it is a key metric for any poor country,
and because it can be estimated from annual production – instead of sample
surveys – using the food balance sheet method pioneered by the UN’s Food
and Agriculture Organization (FAO).17 It therefore serves as an independent
check on the income survey data.
Previous Estimates
Official output data suggest reasonable levels of food consumption in the
1930s and 1940s; one estimate by the FAO put daily food consumption at 2,226
kilocalories per day for 1931–1937.18 However, this seemingly plausible figure
is suspect. First, the population data used were gross underestimates in the
order of 100 million (as China’s first proper census, that of 1953, was to reveal).
Second, the FAO estimate relied upon the Crop Reports compiled by the
National Agricultural Research Bureau (NARB). These reports are problem-
atic because they systematically underreported sown area and, for 1931–1937,
excluded both Manchuria and ethnic minority regions in the west where the
NARB was unable to collect data.19 The exclusion of Manchuria was espe-
cially significant, both because it had a large population and because
Manchurian calorie consumption was high: Shen, using Japanese-collected
data, put Manchurian consumption at 2,557 kilocalories per day during 1935–
1938.20 Much scholarly effort has been expended on correcting these biases.
Buck’s large-scale Land Utilization survey of the early 1930s put food con-
sumption levels at 2,537 kilocalories per day.21 Liu and Yeh came up with
a national estimate of 2,130 kilocalories for 1933, of which 1,936 kilocalories
17
The food balance approach estimates consumption as the residual after the subtraction
from crop production of net exports, stock increases, waste, processing losses, and that
part of the crop used for seed, animal feed, and industrial inputs.
18
T.H. Shen, Agricultural Resources of China (Ithaca, NY, Cornell University Press, 1951),
p. 381.
19
One of the most comprehensive compilations of the Crop Report data is that by
Xu Daofu 许道夫, 中国近代农业生产与贸易统计资料 (Agricultural and Trade
Statistical Materials for Modern China) (Shanghai, Shanghai renmin chubanshe, 1983).
20
Shen, Agricultural Resources of China, p. 383.
21
J.L. Buck, Land Utilization in China (Oxford, Oxford University Press, 1937); T.-C. Liu
and K.-C. Yeh, The Economy of the Chinese Mainland (Princeton, Princeton University
Press, 1965), p. 31.
615

chris bramall
came from grain crops.22 However, Buck’s estimate is too high because his
team oversampled relatively prosperous villages and Liu and Yeh’s otherwise
careful estimates are flawed because of their partial reliance on Buck’s work.
I have therefore recalculated food consumption, and estimate that it averaged
2,191 kilocalories per day for 1931–1937.23 It is purely coincidental that it is so
similar to the FAO figure.
There are several series for post-1949 (Table 17.4).24 The FAO series is most
pessimistic: average consumption fell from 2,126 to 1,940 calories between
1946–1948 and 1975–1978. The MOA series is almost as pessimistic, and
Piazza’s series also shows little improvement, especially if the high figure
for 1978 is disregarded.25 Together, these estimates suggest that the Maoist
record was poor.26
None of the estimates in Table 17.4 are reliable. The FAO series assumes
very high feed grain use for pigs during the 1970s but Chinese pig feeding
actually relied upon household scraps and scavenging, and made only modest
use of grain. The Ministry of Agriculture data are available only for grain,
pork, sugar, and vegetable oils (though these did account for over 90 percent
of calories consumed), so we need to infer total consumption using additional
data from other sources. And whilst Piazza’s estimates are certainly superior
to the others, he had only limited access to Chinese sources and therefore
made questionable assumptions for grain used for seed and animal feed, as
well as milling loss.27
22
Liu and Yeh, The Economy of the Chinese Mainland, pp. 29–30. This post-1949 Chinese
definition of grain includes potatoes, soybeans, and pulses.
23
Details of my estimation methods are available on request.
24
It makes sense to focus on four-year averages because of the impact of weather, an issue
discussed further below. I have chosen to omit 1952 and 1953 because there are more
questions about the data for these years – when the SSB had been barely established –
than for the mid- to late 1950s. According to Liu and Yeh, The Economy of the Chinese
Mainland, pp. 43–55, all the pre-1956 data are suspect. That may be true, but the change
between the 1950s and 1970 is not sensitive to the inclusion of the data for 1954 and 1955.
25
Piazza’s early estimates were more optimistic; see A. Piazza, “Trends in Food and
Nutrient Availability in China, 1958–1981,” World Bank Staff Working Paper No. 607
(1983), 47–73. However, Piazza discarded his 1983 series when producing his 1986 book;
see A. Piazza, Food Consumption and Nutritional Status in the PRC (Boulder, CO,
Westview Press, 1986). This was because his 1983 series assumed implausibly high use
of grain for animal feed in the 1950s – which lowers consumption for the 1950s and
produces a marked upward trend. His 1986 series uses a more appropriate assumption
about feed.
26
Early World Bank estimates were also pessimistic. These put calorie consumption at
2,024 kilocalories in 1957 and 2,044 kilocalories in 1977, before surging to 2,441 kilocalor-
ies in 1979 as result of the early effects of reform. See World Bank, China: Socialist
Economic Development, pp. 32, 118.
27
For example, Piazza assumes that Chinese citizens consumed white rice; he therefore
uses a milling/polishing loss rate of 33 percent in his calculations for both the 1950s and
616

Table 17.4 Trends in food consumption during the Maoist era (calories per
capita per day)
Piazza 1986 MOA FAO
1952 2,083 2,280 2,126*

1953 2,048 2,266 n/a
1954 2,041 2,260 n/a
1955 2,232 2,273 n/a
1956 2,326 2,358 n/a
1957 2,217 2,336 n/a
1975 2,266 2,217 1,909
1976 2,235 2,210 1,875
1977 2,233 2,225 1,914
1978 2,413 2,280 2,062
1954–1957 average 2,204 2,307 2,126*
1975–1978 average 2,287 2,233 1,940
Change from 1954–1957 to 1975–1978 +83 –74 –186
* Average of 1946–1947 and 1947–1948; as with other countries, there are no FAO
time series data for pre-1961. Data refer to the whole population.
Sources: FAO, The State of Food and Agriculture 1948 (Washington, DC, FAO, 1948),
p. 49; Nongyebu 农业部, 中国农村经济统计大全 (Statistics on the Chinese Rural
Economy) (Beijing, Nongyebu, 1989), p. 58; A. Piazza, Food Consumption and
Nutritional Status in the PRC (Boulder, CO, Westview Press, 1986), p. 77
However, the biggest problem is that previous estimates uncritically

accepted official production data. These data are problematic because they
underreport production (manchan 瞒产) on private plots during the 1970s. To
be sure, grain was not the only crop grown on private land; Gao’s villagers (in
Jiangxi) focused upon green vegetables.28 However, grain production usually
featured prominently on private plots. The households studied by Potter and
Potter in Guangdong, for example, grew sweet potatoes, soybeans, and rice
1970s. However, Hinton, relying upon his experience in the Chinese countryside,
rejected Piazza’s assumption, arguing that consumption of brown rice was common-
place, especially in lean years. W. Hinton, Through a Glass Darkly (New York, Monthly
Review Press, 2006), p. 252. This critique is supported by the evidence provided by the
big 1954 survey and in the 1959 Agricultural Handbook; see Nongyebu 农业部, 农业经济
资料手册 (Handbook of Materials on the Agricultural Economy) (Beijing, Nongye
chubanshe, 1959). Both sources suggest that rice was lightly polished (these sources put
the milling loss at between 24 and 29 percent of harvested weight). These are seemingly
arcane technical issues, but the food balance methodology is such that overall con-
sumption is sensitive to assumptions made about rice milling because it was the key
grain crop.
28
M. Gao, Gao Village (London, Hurst and Co., 1999).
617

chris bramall
as well as a range of vegetables, and Thaxton’s villagers in Henan focused on

sweet potatoes. And private-sector grain output was far from insignificant.29
Soya and sweet potatoes were both classified as grain by China’s State
Statistical Bureau and, although private plots typically accounted for just 5
to 7 percent of cultivated area in the 1970s, their output share was consider-
ably higher. This was because high procurement quotas at low prices for
collectively produced grain induced a higher degree of labor intensity on
private plots. Evasion of procurements was difficult.30 However, private plots
were not subject to procurement quotas and therefore substitution effects
occurred between collective and private activity. Labor intensity fell on
collectives, but the time and energy allocated to private plots rose
significantly.
Private-sector activity of itself did not guarantee underreporting; private
production could in principle still have been measured. The reporting prob-
lem arose because repeated Maoist attempts to outlaw private-sector activity
caused widespread evasion. Take sweet potatoes. They were consumed
directly in significant quantities but, more importantly, they were crucial
pig fodder. Households were therefore determined to grow them, come
what may. Moreover, sweet potato production was easy to hide. Potatoes
were quite literally an underground crop; they were grown underground and
stored in underground cellars. As a result, private-sector production
remained significant during the 1960s and 1970s, just as it had during the
famine years. Many local CCP officials even connived in this because collective
pig breeding was generally a failure due to disease, lack of care, and the
impossibility of collecting household food waste at a collective level.
Allowing household pig rearing was therefore essential.
As to the scale of underreporting, official MOA national data put private-
sector grain production at 6 million tonnes, around 2 percent of the 1978
total.31 But other evidence suggests that it was higher. For one thing, the
share of household income in total net rural income was no less than
27 percent in 1978 according to the SSB income survey. To be sure, some of
this income came from pigs and poultry, but the production of sweet
29
S. Potter and J. Potter, China’s Peasants (Cambridge, Cambridge University Press, 1990);
R. Thaxton, Catastrophe and Contention in Rural China (Cambridge, Cambridge
30
It required collusion between collective farm workers and collective leaders, as well as
inadequate higher-level monitoring, and appears to have occurred significantly only
during the famine years.
31
Nongyebu, 农业经济资料手册, pp. 143, 240, 519. I calculate this as a residual after
subtracting collective and state farm production from the total.
618

potatoes was an especially important component of income. For another

thing, the share of private plots in total cultivated area was around 6 percent
in 1978. If the private share in grain output was equal to this area share, it
would imply true private-plot grain production of 18 million tonnes. If
productivity on private plots was double that on collective land – not
implausible given very high labour inputs – true output would be close to
36 million tonnes. We also have direct SSB estimates of private-sector grain
production. These put the private-sector total at 26 million tonnes for 1979,
and 28 million tonnes in 1980.32 Finally, some micro-studies (Thaxton’s work
in northern Henan, for example) suggest that privately produced grain
contributed about 25 percent of grain consumption in the late Maoist era.
If, as Thaxton discovered, private plots produced fifty kilograms per capita
per annum in Henan – the official Henan data put the figure at only twenty-
eight kilograms – that would imply a national total of nearly 40 million
tonnes.33 In short, multiple sources reveal significant underreporting of
private grain production in the late 1970s. Understating output in the official
publications served the political purpose of the 1980s of blackening the Maoist
record and lauding the achievements of the Dengist regime.
Widespread underreporting during the 1970s also helps to explain the
otherwise implausible increase in grain output reported in official Chinese
sources between 1978 and 1984. If official sources are to be believed, grain
output rose from 305 million tonnes to 407 million tonnes during 1978–1984,
a rise of 4.7 percent per annum. By contrast, annual growth between 1965 and
1978, from a much lower base, was only 3.3 percent. The usual explanation is
that there was indeed a one-off surge in production during 1978–1984 and that
it occurred primarily because of agricultural decollectivization. But if the
impact of decollectivization was so critical, it is hard to see why production
grew at a very similar rate between 1977 and 1982 in counties which had
decollectivized, and counties (the majority) which had not.34 The best explan-
ation of the apparent output surge is simply that a process of reintermedia-
tion occurred. This occurred because incentives for underreporting of output
diminished due to rising procurement prices, and because controls on pri-
vate-sector activity – both the type of production and the scale of private
plots – were gradually lifted after 1976. By the mid-1980s, underreporting had
largely ceased.
32
Renmin Ribao, June 16, 1981. 33 Thaxton, Catastrophe and Contention, pp. 273–4.
34
C. Bramall, “Origins of the Agricultural ‘Miracle’: Some Evidence from Sichuan,” China
Quarterly 143 (1995), 731–55.
619

chris bramall
Revised Estimates of Food Consumption

In the light of the underreporting and methodological problems that afflict
existing food balance estimates, I have re-estimated consumption as follows.35
For the 1950s, I accept the official production data but use the results of the
SSB’s 1954 income and expenditure survey for seed and animal feed.36 I also
assume that rice was lightly polished, in contrast to Piazza’s assumption that
white rice consumption was the norm. For the 1970s, a number of Chinese
sources give estimates for seed and feed grain, which I use. I assume a milling
figure of 70 percent for rice; the FAO/Piazza 67 percent milling rate implies
that consumption of highly polished rice was normal in rural areas, which
seems implausible.37 My estimates for the 1950s and 1970s also accept official
data on changes in the grain stock. However, I use revised estimates of grain
production for the 1970s. Although there is no definitive answer to the extent
of underreporting, the various sources mentioned above suggest that it was
about 10 percent of official output, the equivalent of 30 million tonnes of
grain. I therefore raise annual output for 1975–1978 by that amount.38
Putting these various assumptions together, and using the food balance
approach to estimate consumption, generates the revised estimates shown in
Table 17.5. I have included my estimates for 1931–1937 for comparative purposes
but, because the figure is very similar to that for the 1950s, I focus on the change
between 1954–1957 and 1975–1978.39 As Table 17.5 shows, my estimates are
virtually identical to those of Piazza for the 1950s; our assumptions are quite
different but produce a similar result, mainly because my assumption of less rice
polishing is offset by assuming higher use of grain for feed purposes. For the
1970s, however, my consumption estimate is around 4 percent higher. It reflects
my assumption of output underreporting, which is only partly offset by my
35
Full details of the methodology and sources are available from the author.
36
Guojia tongjiju 国家统计局 (National Statistical Office), 1954 年全国农家收支调查
资料 (Materials on the 1954 National Peasant Household Income and Expenditure
Survey) (Beijing, Zhongguo tongji chubanshe, 1956).
37
Most scholars are inconsistent in assuming a low milling rate. They argue that late
Maoist living standards were low – yet simultaneously assume that the rural population
consumed highly polished white rice. The evidence suggests that the rice milling rate
fell from 71.5 percent in 1954 to 70 percent in 1978, implying a modest rise in rural
consumption during the Maoist era – but that relatively coarse rice was still the rural
consumption norm in the late 1970s.
38
In other words, grain output on private plots was about 36 million tonnes, compared
with the 6 million tonnes officially reported in 1978. This suggests that the yield was
around double that on collectively farmed land, but it is somewhat below Thaxton’s
40 million tonnes.
39
I do not attempt to derive complete time series data because we lack reliable data on
feed use, milling rates, and so on during the famine years.
620

Table 17.5 Contrasting estimates of food consumption, 1954–1978 (calories

per head per day)
Piazza 1986 My revised estimates
1954 2,041 2,084

1955 2,232 2,128
1956 2,326 2,352
1957 2,217 2,260
1975 2,266 2,367
1976 2,235 2,343
1977 2,233 2,342
1978 2,413 2,478
Average, 1931–1937 n/a 2,191
Average, 1954–1957 2,204 2,206
Average, 1975–1978 2,287 2,382
Change from 1954–1957 to 1975–1978 +83 +176
Sources: Table 17.4 above and text
assumption of higher livestock feed.40 These revisions imply an overall rise in

daily per capita consumption of about 180 kilocalories (8 percent) by the late
1970s.
My revised consumption series, showing a significant rise between the
1950s and 1970s, is more plausible than other estimates because the underlying
assumptions are more solidly based on survey evidence that has become
increasingly available. My series also accords with the urban and rural
consumption surveys; the rural surveys, for example, show rice consumption
rising from fifty-two kilograms per head in 1954–1957 to fifty-eight kilograms
in 1977–1978, and meat consumption rising by 20 percent over the same
period.41 My estimates also fit with the data on rising heights and weights,
which provide an alternative measure of nutritional intake. The best source
here is the China Health and Nutrition Survey (CHNS), which covered both
urban and rural areas. The survey has been repeated over a number of years
and collected data by year of birth and height during survey years. The 2006
40
I put grain use for animal feed at 33 million tonnes in 1978, about 9.5 percent of available
grain. By contrast, Piazza’s figure is 27 million tonnes. Piazza, Food Consumption and
Nutritional Status in the PRC, p. 190. Both our estimates are far below the FAO figure of
70 million tonnes.
41
Nongyebu, 农业经济资料手册, pp. 548–9. The production of fine grains (wheat and
rice) rose from 56 percent of grain output in 1954–1957 to 61 percent in 1975–1978 – hardly
the sign of an economy in crisis. Nongyebu 农业部, 中国农村经济统计大全
(Statistics on the Chinese Rural Economy) (Beijing, Nongyebu, 1989), pp. 146–53.
621

chris bramall
version of the survey shows a rise of around four centimeters of height

between men and women born in the 1930s and those born in the mid-
1970s. The data on average weight show a comparable upward trajectory.42
Evaluating China’s Consumption Record

The increase in food consumption during the Maoist era is impressive enough at
face value. It becomes even more so when properly contextualized. For one
thing, the rise indicates that China had ended the chronic food insecurity which
had been such a feature of the pre-1949 era, and which reappeared with devastat-
ing force during the Great Famine. And China had not only achieved basic food
security: it had done so without any significant recourse to food imports. Second,
food consumption levels in China by the 1970s were at least on a par with those of
many other land-poor countries during the early days of their successful industri-
alization. For example, recent estimates by Broadberry et al. put British food
consumption at only around 2,170 kilocalories between 1750 and 1800, and little
better than 2,100 kilocalories even in the 1850s.43 Consumption levels in fast-
industrializing Meiji Japan were also rather low. Few question the notion of
a British Industrial Revolution (or indeed a Japanese one) despite the stagnation
of food consumption in the early decades of industrialization. China managed to
industrialize and to raise consumption levels during the Maoist years. Third,
Chinese food consumption levels would have been higher still in the late 1970s
but for a run of unusually poor weather. I noted earlier that the apparent rise in
output between 1978 and 1984 is exaggerated by underreporting in 1978, but
equally serious is the distorted perception induced by weather fluctuations. One
of the most interesting features of the period is that the weather – as proxied by
rainfall – was considerably worse in the late 1970s than in 1982–1984, the years
during which decollectivization was in full swing. If we look at patterns of rainfall
in nine of China’s key grain-producing provinces, a normal year saw a deviation
from 1951–2000 average rainfall of 17 percent.44 During 1976–1978, however, the
actual deviation recorded by the sixty-two weather stations was five percentage
points higher than normal (Table 17.6). By contrast, the deviation in 1982–1984
42
CHNS, China Health and Nutrition Survey (2006), at www.cpc.unc.edu/projects/china/
about/proj_desc (accessed July 4, 2012).
43
S. Broadberry, B. Campbell, A. Klein, M. Overton, and B. van Leeuwen, British Economic
Growth 1270–1870 (Cambridge, Cambridge University Press, 2015), p. 289.
44
In arithmetic terms, the deviation would be zero in a normal year, but I treat flooding
and drought symmetrically by assuming that both were equally bad for output:
flooding in one region only offsets drought in another region in arithmetic terms.
This explains why a normal deviation was 17 percent; in any given year, some stations
recorded drought and others recorded flooding.
622

Table 17.6 Rainfall deviations in 1976–1978 and 1982–1984 from the 1951–2000
norm (percentage points)
Actual deviation Normal deviation Difference
1976 16 17 −1
1977 18 17 1
1978 22 17 5
Total deviation, 1976–1978 56 51 5
1982 13 17 −4
1983 19 17 2
1984 12 17 −5
Total deviation, 1982–1984 45 51 −6
Notes. (1) Data on rainfall are for sixty-two weather stations located in nine
provinces – Anhui, Guangdong (including Hainan), Heilongjiang, Henan, Hubei,
Hunan, Jiangsu, Sichuan (including Chongqing), and Shandong – which together
supplied nearly 60 percent of grain output in the late 1970s. (2) The national deviation
is calculated by averaging the deviation for each of the sixty-two weather stations.
Source: Data on the sixty-two stations are derived from the 160 weather station
database at http://dss.ucar.edu/datasets/ds578.1/data/ (accessed May 9, 2009)
was six percentage points less than normal. In other words, the anecdotal
accounts suggesting a string of excellent harvests during 1982–1984 are con-
firmed by the rainfall data, whereas the years between 1976 and 1978 saw more
droughts and flooding than normal. In terms of single years, the twenty-five
percentage point deviation recorded in 1954 (when there was extensive flooding
along the Yangzi) was by some way the worst of the Maoist era.45 Other bad
years were 1956, 1963, and – of most interest for our purposes – 1978. By
contrast, 1984 was the second-best year of the entire period, being second
only to 1952 (when the deviation was just 11 points). It is therefore little wonder
that output apparently shot up between 1978 and 1984; 1978 was a bad year and
1984 a very good year in terms of the weather.
To be sure, it is important not to exaggerate the improved level of food
consumption. Even at close to 2,500 kilocalories per day, China’s level of food
consumption in the late 1970s was little more than adequate for a population
45
Rainfall patterns during the Great Leap Forward were not particularly unusual. It is true
that there was poor weather in the provinces of Sichuan, Henan, and Anhui, and that all
three suffered severe famine. In principle, however, the central government could have
reduced procurements in the three, and transferred grain to them from provinces
where the weather remained good (especially in the northeast). Such transfers would
not have offset the fall in production caused by policy failure, but they would have been
enough to mitigate the impact of poor weather.
623

chris bramall
predominantly engaged in hard manual labour. Moreover, most kilocalories

were provided by plant products (primarily grain and vegetables), and very
few from meat or fish. And consumption levels in the 1970s were not so very
much higher than they had been in the 1930s. Nevertheless, a positive
appraisal of the Maoist record is in order on the basis of these new estimates.
The old view that food consumption levels stagnated after the 1950s requires
modification in the light of these new findings.
The Mortality Record

The Maoist record on life expectancy is even better than for material living
standards (Table 17.7). The population censuses of 1953 and 1964 were impres-
sively complete by the standards of poor countries. Nevertheless, they tended
to underreport mortality (especially infant mortality) and therefore official
estimates of life expectancy for the 1950s are too high; compare them with the
Banister and World Bank estimates in Table 17.7. By the late 1970s, the
underreporting problem was much less because of the careful 1973–1975
mortality survey, which collected data on the causes of over 15 million deaths
during the three-year period but reached as far west as Tibet and Xinjiang.46
The results of the 1982 census are therefore very reliable and, when combined
with adjusted estimates for the 1950s, suggest a true life expectancy increase
of around fifteen years at birth between 1957 and 1978, almost double the
official increase. The infant mortality data exhibit very similar trends. This is
hardly surprising because infant mortality was the main reason for low life
expectancy in the 1950s. Nevertheless, the improvement is striking. Banister’s
reconstruction shows the infant mortality rate in 1978 as being about
a quarter of what it was in the early 1950s, and the official data show that it
was only around a third of its 1952 value.
China’s remarkable mortality record contrasts sharply with that of India.47
This is a useful comparison. The countries are of similar size, they broke free
of colonial influences at approximately the same time, and they entered the
postcolonial era with similar levels of development; China’s life expectancy
46
The survey covered twenty-four of China’s provincial-level jurisdictions; it omitted
Gansu, Hubei, Shandong, Guangxi, and Guangdong. See Rong Shoude 戎寿得,
Li Junyao 黎均耀, Gao Runquan 高润泉, Dai Xudong 戴旭栋, Cao Dexian 曹德贤,
Li Guangyi 李广义, and Zhou Youshang 周有尚, “我国 1973–1975 年居民平均期望
寿命统计分析” (Statistical Analysis of Mortality and Life Expectancy in China in 1973–
1975), in Beijing jingji xueyuan 北京经济学院, 中国人口科学理论 (A Symposium on
Chinese Population Science) (Beijing, Zhongguo xueshu chubanshe, 1981), pp. 49–58.
47
A. Sen, “Food and Freedom,” World Development 17.6 (1989), 769–81.
624

Table 17.7 Life expectancy and infant mortality in China, 1952–1981

Life expectancy (years at birth) Infant mortality (per 1,000 live births)
Banister World Bank Official Banister Official
1952 133
1953 40 175 122
1954 42 164 112
1955 45 154 105
1956 47 143 106
1957 50 60 132 92
1958 46 146 119
1959 43 160 135
1960 25 43 284 121
1961 38 44 183 108
1962 53 44 89 73
1963 55 46 62 87 73
1964 57 47 86 79
1965 58 49 84 73
1966 59 51 83 64
1967 59 53 82 67
1968 60 55 81 55
1969 61 57 76 57
1970 61 59 70 51
1971 62 60 65 51
1972 63 61 60 52
1973 63 62 56 51
1974 63 63 52 50
1975 64 64 66 49 50
1976 64 64 45 45
1977 65 65 41 41
1978 65 66 68 37 41
1979 65 66 39 40
1980 65 67 42 43
1981 65 67 68 44 38
Notes. Official life expectancy data are given in life tables for the years shown. The
1957 and 1963 official estimates are based on very small rural samples, and therefore
overstate the true level of life expectancy.
Sources: Judith Banister, China’s Changing Population (Stanford, Stanford University
Press, 1987), p. 116; World Bank, World Development Indicators 2017, at https://data
.worldbank.org/data-catalog/world-development-indicators (accessed July 10, 2017);
Zhongguo shehui kexueyuan renkou yanjiusuo 中国社会科学院人口研究所, 中国
人口年鉴 1991 (Chinese Population Yearbook) (Beijing, Jingji guanli chubanshe,
1991), p. 537; Zhongguo renkou qingbao 中国人口情报中心, 中国人口资料手册
(Handbook on Chinese Population Materials) (Beijing, Jingji guanli chubanshe, 1986),
pp. 475–516
625

chris bramall
was higher in the early 1950s (forty years at birth compared with thirty-seven
years), but the difference was small. Despite these similarities, the life
expectancy gap between China and India had grown to over ten years (sixty-
six years compared with fifty-six years) by the early 1980s. In accounting
terms, the “demographic gains” from this improvement more than offset the
impact of the famine deaths that China experienced.
The reasons for China’s success are relatively well known.48 Part of it was
down to public-health campaigns. Examples include a reduction in syphilis-
related deaths (itself reflecting the campaign against prostitution and the
increased use of penicillin), a ban on opium use, and the anti-schistosomiasis
campaign implemented by draining and clearing waterways. More important
for the mortality reduction, especially amongst infants, was the creation of
a relatively simple but extensive system of rural health care which centered
on immunization programs. At the center of the program were the “bare-
foot” doctors – paramedics perfectly capable of carrying out inoculation and
giving health care advice. Although it has been fashionable in the last ten
years to decry the achievements of this group – and their predecessors
(strictly, barefoot doctors only existed after 1968) – their efforts were remark-
ably successful. Their key roles were in terms of the training of midwives,
which was instrumental in the massive decline in the infant mortality rate,
and inoculation against traditional diseases such as smallpox, diphtheria, and
tuberculosis. By contrast, successive Indian governments provided high-
quality health care for high-income earners in urban centers but neglected
rural health care.49
Income Distribution
How can we square Maoist China’s modest record on per capita GDP and
food consumption with its outstanding achievements in terms of life expect-
ancy? Part of the answer lies in the relatively low-cost but effective health care
48
For a discussion of mortality and health care, see J. Banister, China’s Changing Population
(Stanford, Stanford University Press, 1987), pp. 78–120; J. Banister and S.H. Preston,
“Mortality in China,” Population and Development Review 7.1 (1981) 98–110; Gao, Gao
Village, Chapter 6; J. Banister and K. Hill, “Mortality in China 1964–2000,” Population
Studies 58.1 (2004), 55–75; K.S. Babiarz, K. Eggleston, G. Miller, and Q. Zhang, “An
Exploration of China’s Mortality Decline under Mao: A Provincial Analysis, 1950–80,”
Population Studies 69.91 (2015), 39–56; R. Alvarez-Klee, “China: The Development of the
Health System during the Maoist Period (1949–76),” Business History 61.3 (2019), 518–37.
49
See J. Drèze and A. Sen, An Uncertain Glory: India and Its Contradictions (London, Allen
Lane, 2013), Chapter 6.
626

system noted in the previous section. The rest probably lies in the more equal
distribution of income and consumption that was the norm during the post-
1949 era.50
Rural Inequality
Previous work on rural inequality has usually compared Roll’s estimate of
inequality for 1952 with the data published for the late 1970s when the SSB
resumed data collection after the Cultural Revolution.51 However, we now have
data for other years, notably the results of the 1954 SSB income survey mentioned
earlier. There are still many gaps – I have not, for example, seen any data for 1965,
which was the year of a relatively large SSB income survey. Nevertheless, we can
paint a more accurate picture of inequality than has hitherto been possible.
Inequality in rural China dropped sharply between the mid-1930s and 1949 as
a result of land reform and CCP-led rent and interest rate reduction campaigns in
regions “liberated” early (Figure 17.1). Inequality fell further once land reform
was complete: the Gini for per capita income was just 0.18 in 1954. Further
reductions occurred as a result of collectivization, which eliminated the rich-
peasant economy and its associated sideline activities; by 1956–1957, the rural Gini
was just 0.1, barely a quarter of what it had been in the mid-1930s.
The trajectory of rural inequality between 1957 and 1965 was probably
upwards. Inequality may have declined further during the late 1950s as a result
of the Great Leap Forward. However, the impact of the famine was to usher
in a period of liberalization of the household economy, and even decollecti-
vization in some provinces (notably Anhui). Evidence for Hebei, for example,
shows that the share of sidelines rose from just 2.5 percent in 1957 to 30 percent
in 1963, and it was still at 23 percent in 1965. In Anhui, the share rose from
29 percent in 1956 to 56 percent in 1962.52 Of course, the extent of liberaliza-
tion varied across provinces in the early 1960s and decollectivization was
50
Considerations of space preclude a discussion of broader aspects of inequality, such as
inequalities in power and human rights. The record of the Maoist regime was dismal in
most of these areas. The central issue, as Sen, “Food and Freedom,” rightly notes, is
whether the lack of these freedoms was offset by the positive impact of food security (at
least after 1962), and rising life expectancy, for the Chinese poor. The case can certainly
be made that Maoist China’s record, for all its failures, was far better than that of India
precisely because of its achievements in terms of basic needs.
51
C. Roll, The Distribution of Rural Incomes in China (London, Garland, 1980); I. Adelman
and D. Sunding, “Economic Policy and Income Distribution in China,” Journal of
Comparative Economics 11.3 (1987), 444–61.
52
Guojia tongjiju 国家统计局 (National Statistical Office), 各省自治区直辖市农民收
入消费调查研究资料汇编 (Summary Research Materials from Rural Income and
Expenditure Surveys in China’s Provinces) (Beijing, Zhongguo tongji chubanshe,
1985), pp. 183, 237.
627

chris bramall
0.50
0.45
0.40
0.38
0.35
0.30
0.25
0.23
0.20
0.18
0.15 0.21
0.10 0.10
0.05 0.10
0.00
1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980
Figure 17.1 Gini coefficients for rural per capita income, 1934–1978
Note. The trend shown here is my interpretation of the data. I have used the Brandt–Sands
estimate for per capita rural incomes for the 1930s.
Sources: L. Brandt and B. Sands, “Land Concentration and Income Distribution in
Republican China,” in T. Rawski and L. Li (eds.), Chinese History in Economic Perspective
(Berkeley, University of California Press, 1992), p. 202; C. Roll, The Distribution of Rural
Incomes in China (London, Garland, 1980); Guojia tongjiju 国家统计局 (National
Statistical Office), 1954 年全国农家收支调查资料 (Materials on the 1954 National
Peasant Household Income and Expenditure Survey) (Beijing, Zhongguo tongji
chubanshe, 1956); Nongyebu 农业部, 农业经济资料手册 (Handbook of Materials on the
Agricultural Economy) (Beijing, Nongye chubanshe, 1959), pp. 92–3; Guojia tongjiju 国家
统计局 (National Statistical Office), 中国农村居民调查年鉴 2006 (Chinese Rural
Household Survey Yearbook 2006) (Beijing, Zhongguo tongji chubanshe, 2006), p. 34
swiftly halted. Nevertheless, although the rise is difficult to quantify, rural

inequality was higher by 1965 than it had been in 1957.
We know least about rural inequality between 1965 and 1978. It is likely
that inequality fell back to its late 1950s levels during the late 1960s as
collective farming was reimposed in renegade provinces, as sidelines were
suppressed and because the relatively egalitarian need-based Dazhai work
point allocation model was used. After the fall of Lin Biao in 1971, however,
controls were relaxed and work point allocation was increasingly driven by
work done. As a result, the share of household income in total income may
have risen; Dikötter makes much of this, portraying the revival of the private
sector as the engine of late Maoist growth.53 There is some support for this: in
Guangdong, the sideline share rose from 34 percent of total income in 1974 to
53
F. Dikötter, The Cultural Revolution (London, Bloomsbury, 2016).
628

40 percent by 1980.54 But Guangdong was probably exceptional. In

Yunnan the sideline share barely changed between 1966 and 1976,
hovering at around 34 percent. Even in Anhui, which decollectivized
in the late 1970s, the sideline share was only 24 percent in 1978, well
below the 37 percent recorded in 1966. Moreover, it is possible that, if
we (controversially) assume that poor households focused more on
sideline production than richer households because of desperation,
rising household sideline income may have reduced inequalities. In
the case of rural Hubei, for which we have data on 750 households,
the rural Gini seemingly declined from 0.22 in 1974 to 0.19 in 1978.55
But whatever the trend in the 1970s, there is little doubt that collective
farming and restrictions on the private sector constrained inequality.
The national survey results for 1978, for example, reveal a Gini for
rural per capita income in that year of just 0.21, somewhat higher than
twenty years earlier but nevertheless low by Chinese historical
standards.
It is, of course, true that all these data on income inequality are
suspect. The 1934 estimate probably understates inequality because the
National Land Commission survey excluded much of southwestern
China where landlordism was a particular problem. Second, many of
the post-1949 surveys were colored by powerful ideological impera-
tives. For example, the 1954 survey covered over 72,000 peasants and
disaggregated income and consumption by province and by class,
which provides 115 data points. However, one of its purposes was to
show the advantages of co-operation; for that reason, the usual class
categories of poor peasant, middle peasant, rich peasant, and landlord
were supplemented with that of co-operative members. Third, we
know little about the surveys of 1956 and 1957. Fourth, the surveys
of the late 1970s and early 1980s were small-scale and tended to under-
sample both rich and poor households.56
For all that, we can be relatively confident about the trend: there was
a sharp fall in inequality in the early 1950s, further falls in the mid-1950s, and
54
Guangdong tongjiju 广东统计局, 广东省国民经济和社会发展统计资料 1949–88:
农业 (Statistical Materials on the Development of Guangdong’s Economy and Society
1949–88: Agriculture Volume) (Beijing, Zhongguo tongji chubanshe, 1989), pp. 78–9.
55
National Statistical Office, 各省自治区直辖市农民收入消费调查研究资料汇编,
pp. 12–13, 152 and 183.
56
C. Bramall, “The Quality of China’s Household Income Surveys,” China Quarterly 167
(2001), 689–705.
629

chris bramall
then a gradual rise from the early 1970s onwards. The catalysts for the long-
run decline were undoubtedly land reform, collective ownership of land,
and the suppression of private-sector production, moneylending, and
commerce.
There is not much evidence that the decline in rural inequality was driven
by a narrowing of income gaps between provinces. To be sure, the data for
nominal per capita net peasant income do show a decline (Table 17.8). If we
compare 1954 (the year of the first large-scale post-1949 income survey) with
1978 (the year in which large-scale income surveys resumed after the Cultural
Revolution), the coefficient of variation for provincial-level jurisdictions
declined significantly from 0.36 to 0.29. If we exclude the metropolitan
centers of Beijing, Tianjin, and Shanghai, the decline is larger still (from
0.35 to 0.21).
However, this apparent decline in spatial inequality is misleading. For one
thing, we need to adjust for differences in provincial inflation rates. If we
proxy income by per capita personal expenditure, and employ the deflators
used by the SSB, the apparent fall in the coefficient of variation (CV) is in fact
reversed – the CV rises from 0.28 to 0.32 between 1952–1957 and 1973–1978. If
we exclude Beijing, Shanghai, and Tianjin, the CV for real per capita personal
Table 17.8 Coefficients of variation (CVs) for provincial per capita net
peasant income, 1954 and 1978 (current yuan)
All provincial-level jurisdictions Excluding metropolitan centers
1954 1978 1954 1978
Mean 79.1 143.7 73.8 134.2

SD 28.5 41.7 25.5 27.6
CV 0.36 0.29 0.35 0.21
Note. “Metropolitan” refers to Beijing, Tianjin, and Shanghai.

Sources: I estimate the 1954 data figures from the 1954 income and expenditure survey
by subtracting production costs from gross income; see SSB, 1954 调查, pp. 155, 185,
218. Data for Yunnan, Ningxia, Beijing, and Shanghai are averages of the 1952 and 1957
values given in National Statistical Office, 各省自治区直辖市农民收入消费调查
研究资料汇编. We lack data on Qinghai and Tibet, Guangdong includes Hainan,
and Sichuan includes Chongqing. The 1978 data come from Guojia tongjiju (National
Statistical Office), 各省自治区直辖市农民收入消费调查研究资料汇编
(Compilation of Survey Materials on Rural Income and Expenditure by Province)
(Beijing, Zhongguo tongji chubanshe, 1985).
630

expenditure does decline, but rather modestly (from 0.28 to 0.24).57 Second,
the measure of inequality matters. If we calculate the Gini coefficients for per
capita expenditure for 1954 and 1978 rather than the CV, the Gini rises from
0.126 to 0.137. Even if we exclude the metropolitan centers, the Gini coeffi-
cient barely declines (0.125 in 1954 and 0.121 in 1978). Third, the 1978 data are
unreliable because the SSB samples were biased; as discussed earlier, the
income estimates for 1978 are almost certainly too high. In fairness, however,
the upward biases in the 1978 survey were probably also present in the 1954
survey, and therefore the reported trend in provincial incomes may still
adequately capture the reality.
On balance, the safest conclusion is that spatial inequality at the provincial
level changed little between the 1950s and the 1970s. It was the decline in
intra-local inequality – due to land reform, collectivization, and the suppres-
sion of private-sector activity – which drove the decline in the overall rural
Gini coefficient between 1952 and 1978.
Rural Poverty
Asset redistribution, income growth, and the partial allocation of work points
on the basis of need did much to reduce poverty. A sign of China’s success
comes from the experience of provinces where food consumption levels were
low in the 1950s. In Anhui, for example, calorie consumption rose from 1,967
kilocalories in 1957 to 2,333 kilocalories in 1978. For Sichuan, the rise was from
1,976 kilocalories in 1957 to 2,835 kilocalories in 1978. Jilin reported a rise from
1,970 in 1965 to 2,281 in 1978. The Fujian data show an increase from 1,800
kilocalories in 1954 and 1957 to 2,229 in 1978. And Hubei survey data record
a rise from 1,855 kilocalories in 1954 to 2,222 kilocalories in 1978.58
A rare complete data series for the very poor province of Guizhou also
shows a clear rise in grain and meat consumption between 1938 and 1978
(Figure 17.2). Meat consumption, for example, rose from around five kilo-
grams per head to around 7.5 kilograms per head. Such survey estimates of
consumption may well be more reliable than food balance estimates because
they are not affected by the pervasive problem of output underreporting.
57
I use the SSB deflators for per capita peasant expenditure because, flawed or not, it is
better to use them than rely on current price data. As noted earlier, personal expend-
iture data are derived by the SSB from the national accounts.
58
National Statistical Office, 各省自治区直辖市农民收入消费调查研究资料汇编,
pp. 44, 174, 229, 340; Liu Hongkang 刘洪康, 中国人口—四川分册 (China’s
Population – Sichuan) (Beijing, Zhongguo caizheng jingji chubanshe, 1988), p. 369.
631

chris bramall
600 50
Grain consumption 45
500
40
35
400
Grain (jin)
Meat (jin)
30
300 25
20
200
15
10
100
5
Meat consumption
0 0
1946
1948
1958
1960
1962
1970
1972
1974
1938
1940
1942
1944
1950
1952
1954
1956
1964
1966
1968
1976
1978
Figure 17.2 Per capita rural consumption of meat and grain in Guizhou, 1938–1978
Note. 1 jin = 0.5 kilograms.
Source: National Statistical Office, 各省自治区直辖市农民收入消费调查研究资料
汇编, pp. 421–2
Moreover, the post-1949 trends look plausible enough, not least the dramatic
decline in consumption during the Great Famine.
Nevertheless, and despite these long-run improvements in income and
consumption, absolute poverty persisted in the Chinese countryside. In 1978,
for example, there were 377 counties across China where distributed collective
income was less than fifty yuan, well below the national average of eighty-nine
yuan.59 Of the 377 counties, many were located in geographically disadvantaged
western China; there were fifty-eight in Guizhou, twenty-eight in Yunnan,
twenty-four in Shaanxi, and twenty-seven in Gansu. Yet, as already observed
for Henan, many poor counties were also located on the north China plain.
There were forty-six very poor counties in Shandong, seventeen in Hebei,
fifteen in Anhui, and, reflecting its low level of grain production, forty-five in
Henan.60 Moreover, the county data hide pools of poverty. Even within
relatively rich provinces, income levels were low in many locations. Twenty-
nine percent of Liaoning’s 87,744 production teams had a per capita distributed
income of less than fifty yuan in 1974, and that percentage rose to 31 percent in
59
Nongyebu, 农业经济资料手册, p. 523. Total net per capita income, which includes
sidelines and private-plot production, was 134 yuan.
60
Nongyebu 农业部, 全国农村人民公社收银分配统计资料 1956–80 (Statistical
Materials on Income Distribution in China’s Communes, 1956–80) (Beijing, Nongyebu,
1981), pp. 142–3.
632

1976. Even in 1978, when the harvest was better, incomes in 19 percent of teams
were still below the fifty-yuan mark.61 This does not mean that there were poor
counties; per capita income did not fall below fifty yuan in any Liaoning county
in 1978. But there were pockets of poverty within otherwise well-off counties.
The main reason for persistent rural poverty was the low level of per capita
grain production.62 Grain was the key source of calories, providing around
85 percent in the late 1970s. Moreover, the lack of imports, the absence of
private commerce, and fewer spatial transfers by the state than in the 1950s,
meant that per capita production was a good guide to consumption.63 At the
top end, there were nine state-designated commodity grain bases which
supplied most of the state-procured grain. In these nine, per capita grain
output exceeded 800 jin in 1977. The Sanjiang plain base (northeastern
Heilongjiang), for example, produced 885 jin per head, the Lake Taihu base
(straddling the Jiangsu–Zhejiang border) produced 980 jin, and the highest
figure was the 1,010 jin recorded in the Pearl river delta base in Guangdong.64
Elsewhere, however, grain output per head was lower (Table 17.9). Across
great swathes of China, per capita production struggled to reach 700 jin per
person. Much of this was a consequence of adverse physical geography,
which militated against high grain yields in much of western China, on the
loess plateau, and in the southeast. Conversely, growing conditions were
more favourable on the (eastern) Yangzi river and in the “frontier” region of
Manchuria. The real surprise is the low level of production in the Huai, Hai,
and Huang river basins in eastern China, which encompass much of the north
China plain. This vast area, covering the bulk of Hebei, Henan, Shandong,
and Anhui provinces, was home to 23 percent of China’s rural population, yet
grain output per head in 1977 (306 kilograms) was well below the national
average of 355 kilograms. Much Maoist energy went into improving irrigation
in the region in order to expand multiple cropping, but per capita output
61
Liaoning nongyeju 辽宁农业局, 辽宁农业历史资料 (Historical Materials on
Liaoning’s Agriculture) (Shenyang, Liaoning nongyeju, 1982), pp. 174–6.
62
The suppression of sidelines in areas which had no comparative advantage in grain
production (mountainous areas or coastal areas with saline soil) did not help, but most
of the poor households in the 1970s were located in places where grain production did
make sense.
63
State grain procurements declined as a percentage of output between the 1950s and
1970s because excessive procurements were (rightly) blamed for exacerbating food
shortages during the Great Leap Forward. In the absence of large-scale spatial transfers,
the promotion of local grain self-sufficiency therefore became a key policy during the
1960s and 1970s.
64
Nongye quhua weiyuanhui 农业区划委员会, 全国综合农业区划报告
(Comprehensive Report on China’s Agricultural Regions) (Beijing, Nongyebu, 1980),
p. 157.
633

chris bramall
Table 17.9 Per capita grain output by agricultural region, 1977

Number of Share of national rural Per capita grain
Region counties population (%) output (kilograms)
Manchuria 179 7 518

Inner Mongolia and 129 3 342
along the Great Wall
Huang–Huai–Hai basin 383 23 306
Loess plateau 227 7 317
Middle and lower 527 29 396
Yangzi valley
Southwest 418 19 318
Huanan 193 10 317
Northwest 127 2 386
Tibetan plateau 147 1 575
NATIONAL 2330 100 710
Note. Grain output in unhusked kilograms. These unhusked data exaggerate the true
divide between the northern wheat region and the southern rice region because the
milling rate for rice (67 percent) was much less than for grains such as wheat (around
98 percent).
Source: Nongye quhua weiyuanhui 农业区划委员会, 全国综合农业区划报告
(Comprehensive Report on China’s Agricultural Regions) (Beijing, Nongyebu, 1980), p. 7
remained stubbornly low. The very persistence of low output testifies to the
inability even of the mass mobilization of labor made possible by the Maoist
model to overcome fundamental geographical constraints.
Urban Inequality
We can say less about inequality in the urban sector. Part of the problem is that
the urban surveys for the 1950s made little attempt to measure inequality. This is
especially true of the survey data collected by the SSB in the 1950s.65 Some
attempt to measure urban inequality was at least made in the 1960s. The 1964
survey, which covered 3,537 households, suggests a national urban Gini coeffi-
cient of around 0.2.66 The available data for a handful of provinces are broadly
65
For example, Guojia tongjiju 国家统计局 (National Statistical Office), 全国职工家庭
生活调查资料 1956, 1958, 1980 (National Survey Materials on Worker Living Standards
1956, 1958, 1980) (Beijing, Zhongguo tongji chubanshe, 1956, 1958, 1980). There is some
scattered information. For example, data on the size distribution for Shenyang suggest
a Gini of 0.12 (Liaoning tongjiju, 辽宁城市人民生活史料, p. 215). However, these
samples are too small and unrepresentative to tell us much.
66
Guojia tongjiju 国家统计局 (National Statistical Office), 中国统计年鉴 1985 (Chinese
Statistical Yearbook 1985) (Beijing, Zhongguo tongji chubanshe, 1985), p. 561.
634

consistent with this. For Liaoning province, for instance, we have data suggest-
ing that the urban Gini declined to 0.22 in 1962, and to just 0.18 in 1965.67 These
Liaoning figures are comparable with the Henan urban Gini of 0.23 for 1964.68
The urban income surveys ceased during the Cultural Revolution and did
not resume until 1977. For that year, Liaoning recorded an urban Gini
coefficient of 0.16 in 1977.69 We also have a figure for Hubei’s urban popula-
tion, which of course includes the major city of Wuhan, of 0.17. Both these
estimates fit well with the reported national figure of 0.16 in 1978.70 The same
national figure was reported by the World Bank’s mission to China for 1980.71
It is worth recognizing, however, that the surveys were small in scale and not
very representative of urban China. The national 1980 survey, for example,
covered 8,715 households but 1,200 of these were from Beijing alone; oddly,
Tianjin and Shanghai contributed only 500 households each, which was less
than Shenyang’s 600 households.72
These data suggest two conclusions. First, the data show that inequality
declined; the urban Gini had fallen from 0.2 in 1964 to 0.16 by 1978. This is
plausible given the impact of the Cultural Revolution on residual urban
income differentials. There were strict controls on private-sector commerce
and industry. There were limits to wage differentials within urban enter-
prises. And the virtual absence of rural-to-urban migration prevented the
emergence of slums and an urban underclass. Second, the overall degree of
urban inequality was low by world standards. Given controls on the private
sector, the virtual absence of urban unemployment, narrow wage differen-
tials, limits to rural–urban migration, and extensive income subsidies, it could
hardly be otherwise.
Urban inequality did nevertheless persist, and as much by design as by
accident. In part, this was because wage differentials were seen as an import-
ant way to motivate the urban workforce by rewarding those employed in
testing jobs in heavy industries such as coal mining and steel manufacture. In
67
Liaoning tongjiju, 辽宁城市人民生活史料, p. 81.
68
Henan sheng difangshi zhi weiyuanhui 河南省地方市志委员会, 河南省志—人民生
活志 (Henan Records – Living Standards) (Zhengzhou, Henan renmin chubanshe,
1995), p. 138.
69
This is an interesting survey because it covered 6,935 urban households drawn from six
Liaoning cities and seven counties (Liaoning tongjiju, 辽宁城市人民生活史料,
p, 854). In size terms, this was similar to the 1980 national survey.
70
Liaoning tongjiju, 辽宁城市人民生活史料, p. 81; Hubei tongjiju 湖北统计局, 湖北
省国民经济统计资料 1949–1978 (Statistical Materials on the Hubei Economy 1949–
1978) (Wuhan, Hubei tongji chubanshe, 1979), pp. 528–9.
71
World Bank, China: Socialist Economic Development, p. 89.
72
Liaoning tongjiju, 辽宁城市人民生活史料, p. 712.
635

chris bramall
1978, the average wage in state-owned coal mines was 810 yuan, compared
with 605 yuan in food processing. Regional wage differentials also existed
between the bigger cities. The average industrial wage in Shanghai was 881
yuan – it was higher still in “hardship” postings such as Tibet or Qinghai – but
only 691 yuan in Fujian.73 Within cities in the same provinces, there were also
significant differences. In Sichuan, wages in state enterprises (of all types)
were at their highest in Panzhihua at 711 yuan. They were lower in Chengdu
(645 yuan) and significantly lower in some of the smaller cities; the Wanxian
average was only 548 yuan.74 By the standards of market economies, these
were smaller differences, but they meant that significant urban income
inequalities persisted even in Maoist China.
The Urban–Rural Income Gap

Despite the low levels of inequality within both urban and rural sectors, the
overall degree of inequality in China in the late 1970s was still high by
international standards. One influential estimate derived a rural Gini for
1978 of 0.22 and an urban Gini of 0.17 – but an overall Gini of 0.32.75 This
reflected the significant gap in average per capita income between China’s
urban and rural sectors.
Most of the literature uses official data on the urban–rural gap, which are
themselves based upon the SSB’s estimates of personal expenditure per head
(Figure 17.3). In nominal terms, the gap rose from an average of 2.5:1 during
1952–1957 to 2.7:1 for 1974–1978. Adjusted for inflation, the gap averaged 2.6:1
during 1952–1957, rising to 3.1:1 during 1974–1978.76 There was some provincial
variation: in Liaoning, for example, the per capita income gap was about 1.7:1
in 1980.77 Nevertheless, a significant urban–rural per capita expenditure gap
was commonplace across China.
73
Guojia tongjiju 国家统计局 (National Statistical Office), 中国劳动工资统计资料
1949–1985 (Statistical Materials on China’s Labour Force and Wages, 1949–1985)
(Beijing, Zhongguo tongji chubanshe, 1987) p. 159; Guojia tongjiju 国家统计局
(National Statistical Office), 中国劳动工资统计资料 1978–1987 (Statistical Materials
on China’s Labour Force and Wages 1978–1987) (Beijing, Zhongguo tongji chubanshe,
1989).
74
Guojia tongjiju 国家统计局 (National Statistical Office), 1978 年城市国名经济基本
情况统计 (Statistics on the State of the Urban Economy in 1978) (Beijing, Zhongguo
tongji chubanshe, 1979), p. 143.
75
Adelman and Sunding, “Economic Policy and Income Distribution in China,” p. 453.
76
Guojia tongjiju 国家统计局 (National Statistical Office), 中国统计年鉴 1991 (Chinese
Statistical Yearbook 1991), p. 270.
77
Based on a 1980 sample of 490 peasant households and 960 urban households. Liaoning
tongjiju, 辽宁城市人民生活史料, pp. 840, 848.
636

350
300
250
200
yuan
150
100
50
1967
1974
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1968
1969
1970
1971
1972
1973
1975
1976
1977
1978
Peasant Non-peasant
Figure 17.3 The urban–rural gap

Note. Data are for personal expenditure in real terms. The calculation uses the SSB’s
deflator.
Source: Guojia tongjiju 国家统计局 (National Statistical Office), 中国统计年鉴 1991
(Chinese Statistical Yearbook 1991) (Beijing, Zhongguo tongji chubanshe, 1991), p. 270
At first glance, the widening of the gap is surprising given that a range of
explicitly pro-rural policies were pursued to prevent “urban bias.” These
included rural industrialization, the rapid expansion of rural education, and
the “sending-down” of skilled labour and educated urban youth to the
countryside. It appears, however, that these equalizing factors were offset
by centripetal forces. On the one hand, urban incomes were pushed up by the
growth of relatively highly paid urban jobs in higher-productivity heavy
industry or the party-state bureaucracy (which was mainly based in urban
centers). In addition, rises in urban participation rates offset relatively stag-
nant urban wages. On the other hand, rural incomes grew comparatively
slowly because of state extraction of the rural surplus (via the manipulation of
the internal terms of trade) to feed the urban state sector and to finance
defense industrialization.
Most of the literature on the urban–rural gap is content to leave the
discussion there. However, there are three important qualifications to
this story. The first centers on the definition of “rural” and “urban.” One
of the features of Maoism was the program of Third Front construction
initiated after 1964, whereby western China received high levels of
defense-related investment. This investment created cities out of green-
field sites; one famous example is the steel-producing city of Panzhihua
in southwestern Sichuan, which was classified as rural in the
637

chris bramall
1950s.78 From an analytical point of view, this presents a problem: should

Panzhihua be classified as urban or rural for the purpose of estimating the
change in the urban–rural gap? It was manifestly rural in the 1950s, yet
equally manifestly urban by 1978. This is important because treating
places like Panzhihua as urban makes it almost impossible by definition
for the urban–rural gap to close because any successful rural area ends up
being reclassified as urban because of its prosperity. The difficulty here
parallels the index number problem encountered in calculating growth
rates: should one use the prices (read locational classification) of the
first year (for which read rural) or the last year (urban)? In this particular
case, it makes more analytical sense to classify Panzhihua as rural
throughout the Maoist period because it is an example of successful
development in an initially rural setting; it is an illustration of the
countryside catching up.
A second difficulty in measuring the urban–rural gap is that the official gap
estimates ignore the hefty subsidies paid only to the urban population. These
subsidies covered foodstuffs, housing, heating, travel, and a range of welfare
benefits. For state employees, they amounted to 80 percent of the wage by
1978.79 Their impact was massive; they raise the urban–rural gap from around
3:1 to about 6:1 in the late 1970s.
The third problem relates to the prices used to measure non-marketed
peasant production. If production is valued at 1952 retail prices (instead of
state procurement prices), per capita peasant expenditure is raised by 26 per-
cent in 1952, and by 20 percent in 1957. According to one set of estimates, this
reduces the urban–rural gap from 2.4:1 to 1.9:1 in 1957.80 The same sort of
revision would be needed for the late 1970s. I do not attempt that here
because it raises intractable questions about how to compute shadow retail
prices in the context of the planned economy of the 1970s, but it is likely that
using shadow prices would markedly reduce the urban–rural gap.
The net effect of all this is difficult to determine because the three factors
discussed not only were significant but also operated in different directions:
classifying “new” urban areas as rural and using notional retail prices to value
non-marketed production would narrow the gap, whereas including urban
78
The site chosen for Dukou city, later renamed Panzhihua, straddled two counties.
79
Lardy, “Consumption and Living Standards in China,” pp. 163–5; T. Rawski, “The
Simple Arithmetic of Chinese Income Distribution,” Keizai Kenkyu 33.1 (1982), 17.
80
Guojia tongjiju 国家统计局 (National Statistical Office), 第一个五年计划全国国民
收入生产和积累资料 (Materials on China’s National Income, Production,
Consumption and Accumulation during the First Five-Year Plan) (Beijing, Zhongguo
tongji chubanshe, 1957), p. 16.
638

subsidies would widen it. It is probably safe to conclude that a significant

urban–rural gap did exist in 1978; it is hard to see that urban reclassification
and prices issues would eliminate a subsidy-adjusted income gap of around
6:1. However, we cannot go much further.
Conclusion
The revolutionary compact between the CCP and the Chinese peasantry in
the 1930s and 1940s certainly rested in part upon the nationalist commitment
to expel Japanese soldiers from Chinese territory and, more generally, for
China to “stand up” against Western imperialism. But it also rested on
a promise to deliver increases in living standards. This “promise of the
Revolution” was not broken. Daily food consumption rose by well over
10 percent between the 1950s and 1970s. In this regard, Maoist China fared
better than many other countries – Meiji Japan and Industrial Revolution
Britain come to mind – during the early stages of their industrialization.
China’s intra-urban and intra-rural distributions of income and consumption
were far more equal by the 1970s than during the 1930s. The proportion of the
population acquiring a primary education increased dramatically. And, per-
haps the greatest Maoist achievement, life expectancy at birth was twenty-
five years higher than in the early 1950s.
Nevertheless, little more than basic food security had been achieved.
Despite the rise in per capita real incomes and consumption, rural poverty
was widespread in the 1970s even on the north China plain. Diets were heavy
on grain and vegetables, but light on meat and vegetable oils. The great
divide between countryside and city remained a chasm; the gilded literati of
imperial China had been displaced by Party cadres and state employees, but
the peasantry remained an underclass. Maoist China fared much less well
than South Korea and Taiwan, despite starting from a rather similar point. In
many respects, the level of material living standards achieved in China by the
late 1970s was a meager return for the sacrifices that had been made.
Moreover, no weighing of the Maoist record on the scales of history can
ignore its appalling record on what Isaiah Berlin called “negative freedoms.”81
The Maoist regime certainly increased positive freedoms by expanding
educational opportunity and basic health care. However, the Chinese popu-
lation paid a terrible price. Mao himself admitted that the number of those
killed during the land reform of the early 1950s exceeded 500,000. The death
81
For a useful discussion of negative freedom, see Sen, “Food and Freedom,” pp. 770–1.
639

chris bramall
toll during each of the purges of the late 1950s, during the Cultural
Revolution, and in the early 1970s was equally high. China’s penal colonies
were full to overflowing. Its minority populations were repressed, and its
intellectuals were continuously persecuted. Basic freedoms, notably the
freedom to move locality and occupation and the right to speak free of
constraint, were ruthlessly suppressed. The Maoist regime was a ghastly
one in all these respects, and little has changed since 1976.
Whether any of this was an inevitable outcome of the Revolution is moot.
I have argued elsewhere that the concomitant to Mao’s commitment to
ensuring that China “stood up” – in the sense of providing military security –
was a squeeze on living standards; defense industrialization could not have
been financed in any other way. Moreover, the suppression of human rights
may have been necessary to extract the surplus to finance industrialization.
And unflattering comparisons between China on the one hand, and Taiwan
and South Korea on the other, are all very well, but China did not have access
to US aid during the 1950s, or to US markets during the 1960s and 1970s.
Nevertheless, China’s defense industrialization, and everything that went
with it, may not have been really necessary: the counterfactual proposition,
that China could have established better relations with the USA (and the
USSR) had Mao adopted a less hostile foreign policy before the rapproche-
ment with Nixon of the 1970s, remains one of the great historical imponder-
ables. But if one accepts the constraint imposed by the external threat, it may
be that China came close to doing as well as it reasonably could have done –
the disaster of the Great Famine aside – in terms of material living standards
between 1949 and the time of Mao’s death.
Further Reading
Babiarz, K.S., K. Eggleston, G. Miller, and Q. Zhang, “An Exploration of China’s Mortality
Decline under Mao: A Provincial Analysis, 1950–80,” Population Studies 69.91 (2015),
39–56.
Banister, J., China’s Changing Population (Stanford, Stanford University Press, 1987).
Bramall, C., In Praise of Maoist Economic Planning (Oxford, Clarendon Press, 1993).
Brandt, L., and B. Sands, “Land Concentration and Income Distribution in Republican
China,” in T. Rawski and L. Li (eds.), Chinese History in Economic Perspective (Berkeley,
University of California Press, 1993), pp. 179–206.
Gao, M., Gao Village (London, Hurst and Co., 1999).
Guojia tongjiju 国家统计局 (National Statistical Office), 各省自治区直辖市农民收入
消费调查研究资料汇编 (Compilation of Survey Materials on Rural Income and
Expenditure by Province) (Beijing, Zhongguo tongji chubanshe, 1985).
640

Guojia tongjiju 国家统计局 (National Statistical Office), 中国国内生产直核算历史

资料 1952–1995 (Historical Materials on Chinese GDP, 1952–1995) (Beijing, Dongbei
caijing daxue chubanshe, 1997).
Liu, T.-C., and K.-C. Yeh, The Economy of the Chinese Mainland (Princeton, NJ, Princeton
Matsuda, Y., “Survey Systems and Sampling Designs of Chinese Household Surveys, 1952–
1987,” Developing Economies 28.3 (1990), 329–52.
Nongyebu 农业部, 农业经济资料 1949–1983 (Materials on the Agricultural Economy,
1949–1983) (Beijing, Nongmuyuye bu, 1983).
Piazza, A., Food Consumption and Nutritional Status in the PRC (Boulder, CO, Westview
Press, 1986).
Potter, S., and J. Potter, China’s Peasants (Cambridge, Cambridge University Press, 1990).
Riskin, C., China’s Political Economy (Oxford, Oxford University Press, 1987).
Roll, C., The Distribution of Rural Incomes in China (London, Garland, 1980).
World Bank, China: Socialist Economic Development, vol. 1, The Economy, Statistical System
and Basic Data (Washington, DC, World Bank, 1983).
641

18
The Political Economy of China’s Great
Leap Famine
james kai-sing kung
Introduction
Indisputably, the Great Leap Famine of 1958–1961 stands out in Chinese
history as the most profound demographic catastrophe after six centuries of
nearly uninterrupted population growth.1 As some 30 million, or 5 percent of
a population of 660 million, were wiped out within a short period of three
years, it was by far the deadliest famine in human history.2 However, until
the demographic consequences of this catastrophe were fully revealed in the
1980s, the outside world assumed that China had solved its food problem
I would like to thank Cormac Ó Gráda for generously offering a number of insightful
suggestions on an earlier draft, and Bingjing Li for useful comments. My greatest gratitude
goes to Vicky Fangshu Jiang, whose enormously helpful assistance at an early stage greatly
facilitated my writing, and to Wenbing Wu for excellent research assistance. I am also
grateful for financial support provided by a GRF Grant (#17505519) and Sein and Isaac
Souede. All remaining errors are, of course, mine.
1
Unchecked population growth is the hallmark of a Malthusian regime. According to
Malthus, excessive population growth in the absence of technical change would typically
be corrected by wars, famines, and similar catastrophes. In the context of imperial China,
population growth since the late fourteenth century was curbed only twice – once
during the fall of Ming Dynasty in the seventeenth century, again during the Taiping
Rebellion in the late nineteenth century. Unlike the Great Leap Famine, however,
population loss during these episodes was caused by civil war. G. Clark, A Farewell to
Aims: A Brief Economic History of the World (Princeton, Princeton University Press, 2007);
T. Malthus, An Essay on the Principle of Population (London, Printed for J. Johnson, in
St. Paul’s Church-Yard, 1798).
2
Thirty million represents the mid-point between the lowest and highest estimates (more
below on this point). A death toll of this magnitude makes all other famines – including
the North China Famine of 1876–1879 which wiped out between 9.5 million and
13 million people – pale in comparison. Going beyond China, the well-known Irish
famine of 1845–1849 had a mortality of roughly a million people, while the 1973 Ethiopian
Famine had a mortality estimated to be between 40,000 and 200,000. B. Ashton, K. Hill,
A. Piazza, and R. Zeitz, “Famine in China, 1958–61,” Population and Development Review 10
(1984), 613–45; C. Ó Gráda (2000). Black ’47 and Beyond: The Great Irish Famine in History,
Economy, and Memory, vol. 8 (Princeton, Princeton University Press); P. Gill, Famine and
Foreigners: Ethiopia since Live Aid (New York, Oxford University Press, 2010).
642

The Political Economy of China’s Great Leap Famine
following the founding of the People’s Republic, when in fact the greatest
famine in human history was just beginning to unfold.3
Until recently, the Great Chinese Famine had remained the “least under-
stood event in human history.”4 The underlying causes of this human
tragedy were unavoidably complex, as it occurred simultaneously with
the Great Leap Forward (GLF) and all its attendant convoluted changes.
The GLF was an unorthodox development strategy conceived by Mao to
transform a predominantly agrarian China into a modern industrial state,
chiefly by mobilizing a disproportionate pool of surplus rural labor to
engage in economic activities outside agriculture, radically changing the
ways by which agricultural production was organized.5 Mao adopted
a plethora of untried strategies to industrialize agrarian China at an incred-
ibly rapid pace because he was constrained by a woeful lack of both capital
and the advanced technology required to achieve this ambitious goal.
These strategies resulted in his people bearing long-lasting, costly conse-
quences. First, based on an optimistic but unfounded belief that China
could increase its grain yields and output by leaps and bounds, Mao
demanded more grain from the provinces to finance industrialization.6 In
addition, recent research has found that part of this “surplus” was also
earmarked for export as China’s relationship with the Soviet Union turned
sour, obligating it to repay Soviet debt.7 Regardless, excessive grain pro-
curement has been found to have a discriminatory effect against the
peasantry and correlates significantly and positively with excess deaths.8
Were it not for the rigidity of central planning, the excessive consequences
of planned procurement could have been corrected upon discovering that
grain output actually fell short of the plan, which would have ameliorated
3
The statement was made by Lord Boyd Orr in May of 1959, cited in Ashton et al.,
“Famine in China.”
4
G.H. Chang and G.J. Wen, “Communal Dining and the Chinese Famine of 1958–1961,”
Economic Development and Cultural Change 46 (1997), 1.
5
The Great Leap Forward (dayuejin 大跃进) was part and parcel of the ideological slogan
of the “Three Red Banners,” which consisted also of “a general line for building
socialism” (shehui zhuyi jianshe zongluxian 社会主义建设总路线) and “the people’s
commune” (renmin gongshe 人民公社), all of which were promulgated at the second
plenary session of the Eighth National Congress on May 20, 1958.
6
T.P. Bernstein, “Stalinism, Famine, and Chinese Peasants,” Theory and Society 13.3 (1984),
339–77.
7
H. Kasahara and B. Li, “Grain Exports and the Causes of China’s Great Famine, 1959–
1961: County-Level Evidence,” Journal of Development Economics 146 (2020), 102513.
8
J.K.S. Kung and J.Y. Lin, “The Causes of China’s Great Leap Famine, 1959–1961,”
Economic Development and Cultural Change 52.1 (2003), 51–73; J.Y. Lin and D.T. Yang,
“Food Availability, Entitlements and the Chinese Famine of 1959–61,” Economic Journal
110.460 (2000), 136–58.
643

james kai-sing kung
the death toll.9 Others, however, interpret excessive grain procurement as

reflecting the extraordinarily strong incentives of provincial chiefs eager to
rise through the political ranks from being lower alternate members to
becoming full members of the Central Committee of the Chinese
Communist Party during these politically tumultuous times, by remitting
more grain to the center to signal their loyalty to Mao.10 Lower-level cadres
had no alternatives but to comply.11 Regardless of the cause, the excessive
procurement of grain was a key feature of the GLF and strongly associated
with the extra loss of life.
Second, to speed up industrial output, in particular of iron and steel, Mao
allocated a substantial proportion of the rural labor force – some 30 to
50 percent – to manufacturing these items in the countryside using primitive
“local methods” or tufai 土法, while confiscating farmers’ private belongings
such as cooking utensils and furniture as inputs to the “backyard furnaces”
sprinkled all over the countryside – a “strategy” with allegedly similar
catastrophic consequences.12 Similarly, to stabilize yields Mao also sent
many rural workers to work on irrigation projects during agricultural slack
seasons, allegedly to the extent of leaving the crops unharvested in some
places.13
Third, Mao also restructured the rural institutions as a cornerstone of his
bold development strategy. Specifically, to facilitate large-scale irrigation
works as well as to industrialize the countryside, Mao merged the agricultural
9
In addition, these authors find the corroborative evidence that famine severity was
highest in provinces with greatest grain output. An implicit corollary of this thesis is
that Mao would have been willing to cut back procurement should the circumstances
warrant it – a highly questionable assumption. See, e.g., T.P. Bernstein, “Mao Zedong
and the Famine of 1959–1960: A Study in Wilfulness,” China Quarterly 186 (2006), 421–45.
X. Meng, N. Qian, and P. Yared, “The Institutional Causes of China’s Great Famine,
1959–1961,” Review of Economic Studies 82.4 (2015), 1568–1611.
10
J.K.S. Kung and S. Chen, “The Tragedy of the Nomenklatura: Career Incentives and
Political Radicalism during China’s Great Leap Famine,” American Political Science
Review 105.1 (2011), 27–45; J.K.S. Kung, “The Emperor Strikes Back: Political Status,
Career Incentives and Grain Procurement during China’s Great Leap Famine,” Political
Science Research and Methods 2.2 (2014), 179–211.
11
According to Oi, lower-level village (brigade and team) cadres typically had enormous
incentives to negotiate with their leaders for a smaller grain quota. However, such
bargaining was largely ineffective, as their leaders did not want to risk punishment for
having delivered less grain than expected. J.C. Oi, State and Peasant in Contemporary
China: The Political Economy of Village Government (Berkeley, University of California
Press, 1989). Bernstein, “Stalinism, Famine, and Chinese Peasants.”
12
J. Becker, Hungry Ghosts: Mao’s Secret Famine (London, John Murray, 1996); W. Li and D.
T. Yang, “The Great Leap Forward: Anatomy of a Central Planning Disaster,” Journal of
Political Economy 113.4 (2005), 840–77.
13
Kung and Lin, “The Causes of China’s Great Leap Famine.”
644

collectives into gigantic people’s communes (renmin gongshe 人民公社).

Consisting of several thousands of farm households whose remuneration
was narrowly differentiated, Mao emphasized “perfect socialist conscious-
ness” while neglecting the importance of incentives for individual work in
boosting agricultural productivity. While the damaging effect of the com-
mune on work incentives is clearly recognized,14 what remains disputable is
whether the damage was because of the unwieldy size and narrow wage
differentials of the commune, or possibly because it prevented diligent
farmers from returning to individual household farming should they find
their coworkers not delivering the minimum effort required for efficient
team farming – a right they had had before they were forced into the
commune and to remain in it ever since.15 In any case, the consequences of
the impaired work incentives are also well rehearsed.
In addition to its dampening effect on work incentives, and operating on
the assumption that food would be provided free and unrestrictedly, the
commune also allegedly destroyed the incentives for people to economize on
food consumption by establishing a communal dining system, which argu-
ably “triggered” the great famine.16 The combination of a rigidly aggressive
procurement policy, the misallocation of a sizeable part of the rural labor
force, and the institution of a commune with negative repercussions on both
work and consumption incentives all render a critical analysis of the under-
lying cause(s) of the Great Leap Famine an exceedingly daunting exercise.
And, if the simultaneous occurrence of these vastly intricate events were
not complicated enough when unveiling the underlying causes of the great
Chinese famine, we must still reconcile the possible causes described above
with the Chinese government’s claim that the catastrophe in question was
largely an inadvertent consequence of “three consecutive years of bad wea-
ther” – a claim that, while in the minority, has nonetheless also received
scholarly support.17 Furthermore, presumably due to malnutrition resulting
from the famine, many have examined a variety of long-term consequences
14
D. Perkins and S. Yusuf, Rural Development in China (Baltimore, Johns Hopkins
15
J.Y. Lin, “Collectivization and China’s Agricultural Crisis in 1959–1961,” Journal of
Political Economy 98.6 (1990), 1228–52.
16
Yang thus likens the incentives underlying communal dining to a “tragedy of the
commons.” D. Yang, Calamity and Reform in China: State, Rural Society, and
Institutional Change (Stanford, Stanford University Press 1996). See also G. Hardin,
“The Tragedy of the Commons,” Science 162.3859 (1968), 1243–8. Chang and Wen,
“Communal Dining and the Chinese Famine.”
17
Y.Y. Kueh, Agricultural Instability in China, 1931–1990: Weather, Technology, and Institutions
(Oxford, Clarendon Press, 1995).
645

james kai-sing kung
of the Great Leap Famine in respect of health, education, and socioeconomic

status,18 while others have attempted to link endurance of this catastrophe to
the choice of farm institutions through the channel of “collective memory”
and to economic development more generally.19
While the underlying causes of excess deaths during the GLF is certainly an
issue of epic proportions, clearly it is not the only issue worthy of scholarly
attention. For instance, the reaction of peasants to the myriad monumental
changes imposed on them as a consequence of agricultural collectivization
and the GLF in general, and more specifically the extraordinary pressure
exerted upon them by officials to eke out the already meager amount of food
on which their survival critically depended, are clearly topics deserving
serious scholarly attention in the future.20 As another example, our chapter
is also reticent on the extent to which the famine in question was caused by
starvation specifically or by infectious diseases such as typhus, dysentery, and
malaria – a topic on which we have relatively scarce information.21 As will
become apparent, it is not easy to strike a balance in writing a review essay of
this nature. While every effort was made to include as broad a coverage of
everything related to the GLF as possible, it is inevitable that many important
18
Y. Chen and L.-A. Zhou, “The Long-Term Health and Economic Consequences of the
1959–1961 Famine in China,” Journal of Health Economics 26.4 (2007), 659–81; D. Almond,
L. Edlund, H. Li, and J. Zhang, “Long-Term Effects of Early-Life Development: Evidence
from the 1959 to 1961 China Famine,” in T. Ito and A. Rose (eds.), The Economic Consequences
of Demographic Change in East Asia (Chicago, The University of Chicago Press, 2010), pp. 321–
45; X. Meng and N. Qian, “The Long-Term Consequences of Famine on Survivors:
Evidence from a Unique Natural Experiment Using China’s Great Famine” (No. w14917)
(2009), National Bureau of Economic Research; T. Gørgens, X. Meng, and
R. Vaithianathan, “Stunting and Selection Effects of Famine: A Case Study of the Great
Chinese Famine,” Journal of Development Economics 97.1 (2012), 99–111.
19
Y. Bai and J.K.S. Kung, “The Shaping of an Institutional Choice: Weather Shocks, the
Great Leap Famine, and Agricultural Decollectivization in China,” Explorations in
Economic History 54 (2014), 1–26; E. Gooch, “Estimating the Long-Term Impact of the
Great Chinese Famine (1959–61) on Modern China,” World Development 89 (2017), 140–51.
20
Why, for instance, were Chinese peasants so docilely obedient even in the face of
excessive grain procurement, when it became clear that it had proceeded to a point
where it was a matter of life and death? Why did the resistance (or “counteraction,” as it
is called) that some peasants were observed to have put up fail to go beyond the
concealment of production and private distribution of crops, as the late Chinese
historian Gao Wangling discovered? What “survival strategies” other than, for
example, migration, were involved? W. Gao, “A Study of Chinese Peasant ‘Counter
Action’,” in K.E. Manning and F. Wemheuer (eds.), Eating Bitterness: New Perspectives on
China’s Great Leap Forward and Famine (Vancouver, UBC Press, 2011), pp. 272–94.
21
Apart from the frequent mention of edema in several accounts, outright starvation
appears to be an accepted cause of deaths in this great Chinese famine. C. Ó Gráda,
Eating People Is Wrong, and Other Essays on Famine, Its Past, and Its Future (Princeton,
Princeton University Press, 2015), pp. 145–7, contains a brief discussion of the subject
matter.
646

topics such as those mentioned above had to be left out because either little is
known about them or they go far beyond my limited expertise.
With this caveat in mind this chapter is thus intended to provide
a thorough review of the relevant literature around which my own expert-
ise revolves. We begin with a brief discussion of the various estimates of the
demographic consequences of the Great Leap Forward, as well as outlining
its salient features. We then provide a detailed review of the voluminous
literature examining the underlying causes of the famine, by classifying the
analyses into the categories of production, distribution, and consumption.
This will be followed by a summary of the growing literature on the
famine’s long-term impact, both on the well-being of individual survivors
and on institutional choice and economic development. A conclusion
follows.
Excess Deaths and Salient Features

Estimates of Excess Deaths
With the demise of Mao and the downfall of the “Gang of Four” (sirenban
四人幫) China entered a new era. Under the new leadership of Deng,
a pragmatic reformer, a population census was conducted in 1982 – an
exercise that compiled and released useful data on the country’s age
distribution, fertility trends, birth and death registrations, and other vital
statistics. Two striking features of the Great Leap Forward were revealed by
these statistical data – a sharp rise in the death rate and a drastic decline in
the birth rate. Taken together, the population declined by 13.5 million over
the two years of 1960–1961.22 A salient issue, in this context, is the number of
excess deaths caused by this tragedy, where demographers define excess
death rates as the difference between actual death rates and the rates
predicted by the linear trend calculated using population data both
before and after the Great Chinese Famine.23 Estimates made by these
22
A.J. Jowett, “The Demographic Responses to Famine: The Case of China 1958–61,”
GeoJournal 23.2 (1991), 135–46; Statistical Yearbook of China, 1983.
23
J.S. Aird, “Population Studies and Population Policy in China,” Population and
Development Review 8 (1982), 267–97; Ashton et al., “Famine in China”; J. Banister,
China’s Changing Population (Stanford, Stanford University Press, 1987); A.J. Coale,
“Population Trends, Population Policy, and Population Studies in China,” Population
and Development Review 7 (1981), 85–97; X. Peng, “Demographic Consequences of the
Great Leap Forward in China’s Provinces,” Population and Development Review 13.4
(1987), 639–70.
647

james kai-sing kung
demographers fall closely within the range of 23 to 30 million,24 the upper

end of which doubles the government’s claim of 15 million.25 Suspicious of
its validity, Cao Shuji made a herculean effort to reconstruct excess deaths
from county annals.26 By totaling the population difference between 1958
and 1961 on a county-by-county basis, and by taking into account natural
population growth and net migration, this demographic historian puts the
tally at 32.5 million, which is strikingly close to the upper-bound estimates
based on census data.27
Others have joined in the guessing game. For example, in a biography of
Mao, Chang and Halliday increase the estimate of excess deaths – “close to
38 million people died of starvation and overwork”28 – while Yang Jisheng,
a journalist, concludes that a number of 36 million “approaches the reality but
[it was] still too low.”29 More recently, historian Dikötter embraced the
estimate proposed by Chen Yizi, a member of an investigation team in the
1980s famous for his anti-Mao stance who claimed to have firsthand access to
“internal party documents,”30 based on which he contended that the famine
took 45 million lives “at a minimum.”31 While these revisions provide
additional reference points, the analytic rigor of their estimates has been
subjected to serious questioning.32
24
For example, Aird, “Population Studies and Population Policy in China,” and Coale,
“Population Trends,” put the death toll at 27 million; Ashton et al., “Famine in China”
at 30 million; C. Ó Gráda, “Great Leap, Great Famine: A Review Essay,” Population and
Development Review 39.2 (2013), 333–46 at below 30 million; and Banister, China’s Changing
Population, at 30 million. By extrapolating data based on fourteen provinces to the
remainder, Peng’s (“Demographic Consequences of the Great Leap Forward”) esti-
mates put the excess death toll at 23 million.
25
When estimating the death toll, most demographers make systematic adjustments for
the underregistration of deaths that probably resulted from the breakdown of the
registration system during the Leap, and from the ubiquitous omission of infant
mortality in the census survey. Banister, China’s Changing Population.
26
Cao Shuji 曹树基, 大饑荒: 1959–1961 年代中國人口 (The Great Famine and China’s
Population in 1959–1961) (Hong Kong, Shidai guoji chuban youxian gongsi, 2005).
27
A caveat is that Cao only managed to do so for seventeen provinces.
28
J. Chang and J. Halliday, Mao: The Unknown Story (London, Jonathan Cape, 2005), p. 456.
29
J. Yang, Tombstone: The Great Chinese Famine, 1958–1962 (trans. E. Friedman, Guo Jian,
and S. Mosher) (New York, Farrar, Strauss, and Giroux, 2012), p. 430.
30
Cited in Becker, Hungry Ghosts, pp. 271–2. According to Chen, the famine cost between
43 and 46 million lives. However, Bramall is skeptical of the objectivity of Chen’s
estimate. C. Bramall, “Agency and Famine in China’s Sichuan Province, 1958–1962,”
China Quarterly 208 (2011), 990–1008.
31
Dikötter essentially arrives at his number simply by applying a 50 percent upward
adjustment to Cao’s estimate. F. Dikötter, Mao’s Great Famine: The History of China’s
Most Devastating Catastrophe, 1958–1962 (New York, Walker & Co., 2010); Cao, 大饑荒.
32
The estimates by Yang and Dikötter are criticized as excessive as they relied heavily on
an “implausibly low pre-famine death rate” for 1957, for which Western demographers
such as Banister have proposed adjustments. However, Banister’s estimate of 30 million
648

Salient Features
The Great Leap Famine has several salient features. The first is that the
famine was disproportionately more severe in rural areas. While the death
rate in the cities was 13.7 per thousand in 1960, it was in excess of twenty-eight
per thousand in the countryside – an inequality that was allegedly rooted in
China’s grain procurement system having an inherent bias against the
peasantry.33 From this perspective, urban residents were guaranteed
a minimum grain ration, but rural people were only entitled to what was
left to them after grain was procured – the residuum. In normal times, when
adequate food was available in rural areas, the peasants had enough to eat.
However, during the famine, food ran short as a result of the exaggeration of
grain output figures and consequent overprocurement, resulting in the
difference in the availability of grain between urban and rural residents
widening.34 Crude evidence suggests that, while city dwellers were entitled
to 303 kilograms per capita per annum during the famine period, their rural
counterparts had only 223 kilograms, 26.4 percent less than urban consump-
tion and 19 percent short of the recommended benchmark of 275 kilograms.35
A second salient feature is that excess deaths were highly unevenly
distributed. In sharp contrast to the belief that mortality rates would be
lower in large agricultural provinces, ironically these provinces seem to
have suffered higher death tolls.36 Indeed, using provincial-level data,
Map 18.1a reveals that the famine was most severe in the southwestern
province of Sichuan, a large and topographically diverse agricultural prov-
ince; followed by Guizhou in the southwest and Anhui, another sizeable
agricultural province in central China; Henan and Shandong in the north; and
Gansu in the northwest.37 A strikingly related feature is that excess deaths also
may still be on the high side. Drawing upon a demographic study using “much higher-
quality data,” Zhao and Reimondos arrived at an estimate of life expectancy for the
famine population of 32.5 years, which is much higher than Banister’s 24.6 years.
Banister, China’s Changing Population; Z. Zhao and A. Reimondos, “The Demography
of China’s 1958–61 Famine,” Population 67.2 (2012), 281–308; Ó Gráda, “Great Leap, Great
Famine.”
33
J.K.S. Kung and J.Y. Lin, “The Causes of China’s Great Leap Famine, 1959–1961,”
Economic Development and Cultural Change 52.1 (2003), 51–73.
34
Lin and Yang, “Food Availability, Entitlements and the Chinese Famine.”
35
K.R. Walker, “Food and Mortality in China during the Great Leap Forward, 1958–61,” in
R.F. Ash (ed.), Agricultural Development in China, 1949–1989: The Collected Papers of Kenneth
R. Walker(New York, Oxford University Press, 1998), pp. 106–47.
36
Meng, Qian, and Yared, “The Institutional Causes of China’s Great Famine.”
37
Inspiringly, the demographic historian Cao Shuji found great disparity in the excess
deaths (17 percent) between two adjacent provinces in central China, Anhui and Jiangxi.
While Anhui was severely struck by the famine, Jiangxi came through unscathed. Chen
provides an interesting comparative analysis between the two cases. Cao, 大饑荒;
649

james kai-sing kung
vary substantially within the same province.38 The prefectural-level data

constructed by Cao allowed Kung and Zhou to compare variations in excess
deaths both between and within provinces, which they did by computing two
separate measures of standard deviation.39 They find that the two measures
are strikingly similar, suggesting that the severity of famine varied not merely
between the provinces but also within them (see Map. 18.1b).40 A more
nuanced finding of their analysis is that the standard deviation of the severity
of famine across prefectures in the famine-stricken provinces is much larger.
Using Sichuan as an example, the within-province standard deviation of 4.2 is
40 percent higher than the between-province standard deviation of 2.99,
suggesting that while for Sichuan as a whole the famine was indisputably
severe, some prefectures within it came through unscathed. Bramall finds the
same among the counties in Sichuan Province, in which the crude death rate
ranged from eight per thousand to a high of 109 per thousand against the
provincial average of thirty-nine and the national average of seventeen per
thousand.41 Map 18.1c, which is drawn using county-level data, reveals this
disparity vividly. The enormous variations that existed not just between
provinces but also within them require new explanations that may include
nuanced “local conditions” that may range from exogenous differences in
terrain and cropping patterns to endogenous variations in, for instance,
radicalism.42
Y. Chen, “Under the Same Maoist Sky: Accounting for Death Rate Discrepancies in
Anhui and Jiangxi,” in Manning and Wemheuer (eds,), Eating Bitterness, pp. 197–225.
38
In the light of the finding that the “three large(st) areas of severe famine” were “not
confined within provincial borders,” Garnaut makes the case that the Great Leap
Famine should be analyzed in terms of regions rather than provinces. However,
since a macro-region encompasses a spatial territory larger than that of a province, it
would render the econometric analysis of the causes of death even more imprecise. A.
Garnaut, “The Geography of the Great Leap Famine,” Modern China 40.3 (2014), 323,
Map 2.
39
Cao, 大饑荒. J.K.S. Kung and T. Zhou, “Political Elites and Hometown Favoritism in
Famine-Stricken China,” Journal of Comparative Economics, 2020, at www
.sciencedirect.com/science/article/pii/S0147596720300184?via%3Dihub.
40
The within-province standard deviation is 2.58, whereas the between-province standard
deviation is 2.99.
41
Bramall, “Agency and Famine in China’s Sichuan Province.”
42
As far as local radicalism is concerned, clearly some cadres (commune and above) were
more eager than others to toe the Maoist line. However, systematic evidence of this is
scarce. R. A. Thaxton Jr., Catastrophe and Contention in Rural China: Mao’s Great Leap
Forward Famine and the Origins of Righteous Resistance in Da Fo Village (New York,
Cambridge University Press, 2008), presents an exception. To counter this limitation,
Kung and Chen, “The Tragedy of the Nomenklatura,” used the frequency with which
high-yield “agricultural satellites” were reported by communes as a proxy for the
variation in local radicalism.
650

Death rate
(per 1,000)
7.5–10.0
10.1–20.0
20.1–30.0
30.1–40.0
40.1–43.5
No data

Map 18.1a Famine mortality by province, prefecture, and county, 1959–1961. (a) Average death rate (per thousand people) by province, 1959–1961. (b)
Total excess deaths in 1958–1961 as a percentage of 1958 population by prefecture. (c) Average death rate (per thousand people) by county, 1959–1961
Source: (a) data reported by the National Bureau of Statistics (NBS); (b) adapted from Cao Shuji 曹树基, 大饑荒: 1959–1961 年代中國人口 (The
Great Famine and China’s Population in 1959–1961) (Hong Kong, Shidai guoji chuban youxian gongsi, 2005); (c) adapted from H. Kasahara and B.
Li, “Grain Exports and the Causes of China’s Great Famine, 1959–1961: County-Level Evidence,” Journal of Development Economics 146 (2020), 102513
Excess deaths in
1958–1961 (%)
0.0–5.0
5.1–10.0
10.1–15.0

15.1–20.0
20.1–29.4
No data
Map 18.1b (cont.)

Death rate
(per 1000)
1.2–10.0
10.1–20.0
20.1–30.0
30.1–40.0
40.1–163.3
No data

Map 18.1c (cont.)
james kai-sing kung
Causes of Excess Deaths

Endeavors made by demographers to estimate the extent of population
loss have inspired economists, political scientists, and historians alike to
re-examine the underlying causes of the excess deaths experienced during
this tumultuous period. On the basis of the extant literature we can divide
the possible underlying causes into the three conceptual categories of (1)
production, (2) distribution, and (3) consumption. First of all, the “pro-
duction failure” hypothesis considers the famine to be the direct conse-
quence of a shortfall in agricultural output and accordingly a decline in
food availability – a failure primarily caused by exogenous weather
shocks.43 Conversely, analogous to Sen’s famous “entitlement
approach,”44 the “distributional failure” thesis attributes the famine to
a systemic failure to distribute the farm surplus effectively to those in
need.45 Finally, a third view attributes the Great Chinese Famine to
irrational consumption behavior caused by the rules associated with the
communal dining system – an “irrationality” that led to wasteful food
consumption and rapid food depletion. As we shall see, the historical
evidence and analytical rigor with which each of these perspectives is
advanced to account for the famine varies enormously; hence they are not
equally persuasive.
Before we analyze these perspectives in further detail, perhaps the reader
should be reminded of the fundamental fact that China was in fact still a poor
country by contemporary world standards in the mid-1950s, and as such it was
vulnerable to the risks of a rash and overly ambitious policy such as the Great
Leap Forward. Using the (admittedly crude) estimates of per capita GDP of
the countries selected in the Maddison Project, we illustrate in Figure 18.1
that during the 1950s China not only trailed far behind the world’s leading
economies such as the United States and United Kingdom, but was also
43
For example, see Kueh, Agricultural Instability in China.
44
A. Sen, Poverty and Famines: An Essay on Entitlement and Deprivation (New York, Oxford
45
For example, Lin and Yang, “Food Availability, Entitlements and the Chinese Famine,”
make the case that because the urban residents were guaranteed a minimum ration
while their rural counterparts were subject to procurement that failed to consider their
subsistence needs, the former were essentially protected in times of food shortages. To
some extent this theory finds support in the argument that, even at the peak of the
famine in 1960, the national grain output could provide up to 2,101 kilocalories per
person on a daily basis – almost three times as much as the required calorie intake for
survival, if the food was distributed equally among the people. Meng, Qian, and Yared,
“The Institutional Causes of China’s Great Famine.”
654

United States
Netherlands
United Kingdom
Germany
France
Italy
Japan
India
China 859
0 5000 10000 15000 20000

Figure 18.1 Average GDP per capita (in constant US$), 1950–1960
Source: J. Bolt, I. Robert, H. de Jong, and J.L. van Zanden, “Rebasing ‘Maddison’: New
Income Comparisons and the Shape of Long-Run Economic Development,” Maddison
Project Working Paper 10 (2018)
well behind rising Japan and comparable only to similarly poverty-stricken

India.46
Production Failure: Policy Blunder, Institutions, and Weather

Adversity
Ambitious but Untried Policy
For Mao, the overriding goal of China’s socialist economic development was
to close the gap with major Western powers such as the United Kingdom and
United States through the industrialization of a “capital-goods-producing”
sector – a goal that required a vast increase in both agricultural and industrial
output within a short period of time.47 In the absence of the advanced
technology required for industrialization and capital to finance importing
it – China’s comparative disadvantage – Mao looked to the peasants to foot
46
Indeed, while not as famous as the Great Leap Famine, a large-scale famine, one that
wiped out as much as 5 percent of China’s population, occurred during the late 1920s
and early 1930s. Bai and Kung, “The Shaping of an Institutional Choice.” Writing in the
1920s, W.H. Mallory, China: Land of Famine (New York, American Geographical
Society, 1926), aptly characterized China as a “land of famine.” Further, even during
the famine-free years of the 1950s, thousands were killed by hunger each year, with
many more fleeing regions hit by harvest failures. F. Wemheuer, Famine Politics in
Maoist China and the Soviet Union (New Haven, Yale University Press, 2014).
47
Industrialization in that context required the agricultural sector to provide a surplus not
only to generate export earnings but also to subsidize the “social wage” (cost of living)
of urban industrial workers in state-owned enterprises.
655

james kai-sing kung
the bill. Increasing the agricultural surplus sharply without injecting

resources into it thus became an item at the top of the chairman’s agenda.
Emboldened by the peasants’ display of “socialist enthusiasm” (shehui zhuyi
reqing 社會主義熱情) in responding to his massive call to undertake irriga-
tion projects in the slack winter months of 1957–1958, Mao decisively
amalgamated the advanced agricultural co-operatives (nongye gaojishe 農業
高級社) into people’s communes (renmin gongshe), in part to facilitate the
organization of large-scale public projects that required co-operation
between townships, and in part to industrialize the countryside.48 Under
the “banner” of the Great Leap Forward, Mao put together a package of
highly unorthodox development policies to achieve his unrealistic ambition
of overtaking the United Kingdom in steel and iron production within the
next twenty-five years and surpassing the United States within fifty to seventy
years.49 The core elements of this policy included the make-believe that it
was scientifically feasible to increase crop output sharply and hence tax it in
support of industrialization,50 the massive mobilization of the rural labor
force to engage in the production of steel and iron in the countryside, and not
least an erroneous but determined decision by the chairman to slash the
grain-sown acreage by close to 10 percent in order to grow oil-bearing and
other economic crops51 – all of which were conceived without considering
48
In addition, the commune was intended to provide a comprehensive array of welfare
including education (schools), health care (clinics), care services (daycare centers and
old people’s homes), weddings and funerals, and not the least the free supply of food in
communal mess halls.
49
Emboldened by the Soviet Union’s success of launching its first satellite (“sputnik”),
Mao ambitiously revised his targets of surpassing the United States in twenty to thirty
years. By May 1958, these targets were further revised to seven years for the UK and
eight to ten years for the US.
50
Examples abound in the name of what Becker terms “false science.” For instance,
thousands of new colleges and universities were created at the commune or township
level. Many scientists trained in the West were imprisoned or sent to perform manual
labor during the “Anti-Rightist Campaign” that preceded the Great Leap. Other
examples such as “close planting” and “deep ploughing,” which were seen as key to
increasing crop yields, instead respectively deprived plants of breathing space and
destroyed soil structure, whereas increased “fertilization” by adding broken glass and
bricks into the earth similarly stifled crop growth. See Becker, Hungry Ghosts.
51
As reflected in a conversation the chairman had with a county party secretary in Hebei
province in August of 1958. Mao’s assessment of the agricultural sector in producing
a huge surplus was clearly widely off the mark. The chairman asked, “How can you eat
your way through so much grain? What are you going to do with your surplus?” The
secretary replied, “We’ll exchange the surplus for machinery.” Mao then said, “And
what will happen if you’re not the only ones to have grain surplus and if every county
has them? What if no one will want it?” The secretary answered, “We are going to make
our sweet potatoes into alcohol.” But Mao continued, “How can you use so much?” K.
R. Walker, Food Grain Procurement and Consumption in China (Cambridge, Cambridge
656

the backwardness of Chinese agriculture in terms of mechanization, irriga-

tion, and fertilizers.52 However, evidence of the extent to which these false
scientific practices had led to a reduction in crop output remains largely
anecdotal, and hence unsubstantiated. We will therefore focus our review on
more solid historical evidence.
Misallocation of the Rural Labor Force and Commune

Organization
To increase iron and steel output beyond that produced by state-owned
enterprises, Mao dispatched 90 million strong, young farmers – representing
more than some 40 percent of the rural labor force – to undertake a variety of
irrigation projects such as the construction of dams, water channels, and
reservoirs, and to manufacture iron and steel at prospective iron ore deposits
using “local methods” (tufa). By one estimate, the diversion of labor resources
away from agriculture, which included a reduction in acreage sown with
grain by nearly 10 percent, accounted for 33 percent of the overall decline in
grain output between 1958 and 1961.53
Mao turned to the people’s commune to boost farm output in the absence
of technical change. After experimenting with gradualism for roughly half
a decade, Mao eventually lost his patience when Deng Zihui, the then
minister of agriculture, allowed 200,000 ill-managed elementary agricultural
co-operatives (chuji nongye she 初級農業社) to dissolve. After this, Mao
hastened the speed of agricultural collectivization by merging these smaller
co-operatives into “advanced” ones, and completely abolished the peasants’
private rights in land and other major farm implements.54 By 1957 nearly all
farm households belonged to an advanced agricultural co-operative. And in
one fell swoop, the Chinese peasants were pushed one step further and
became members of gigantic communes (c. 1958). It was following this
particular transition that individual work incentives became severely
52
In terms of mechanization, essential farm tasks were still performed by hand and with
primitive tools, farm managers or cadres were mostly illiterate and lacked management
skills, most land was yet to come under irrigation and was thus easily affected by natural
calamities, and fertilizers were primarily organic and hence unable to increase yields
sharply.
53
Strictly speaking, the sharp reduction only applies to 1959. Regardless, the implied
output elasticity is huge, a result that was probably confounded by other policy
blunders. Li and Yang, “The Great Leap Forward.”
54
For instance, the target was to increase the agricultural collectives from 650,000 to
1.3 million co-operatives.
657

james kai-sing kung
reduced, resulting in a low level of agricultural productivity for two decades,

until agricultural decollectivization came about to revive it.
On the communal farm, not only was the level of work organization and
income distribution raised from several hundred to thousands of households, but
to alleviate the monitoring burden of team production Mao replaced the com-
bined use of piece rates and time rates with an egalitarian wage system consisting
of exceedingly narrow differentials between various labor grades.55 Clearly, both
of these changes reduced the peasants’ incentive to put in their best effort56 –
a pitfall that Mao willfully neglected in his march towards Communism.
Lin, however, contends that scale alone cannot explain why China failed to
recover its total productivity factor in agriculture to the level of the early
1950s, as all the anomalies that occurred at the onset of the Great Leap had
already been corrected by the time the famine ended.57 While acknowledging
that the root cause of the supervisory problems lies in the sequential and non-
standardized nature of a labor-intensive agriculture,58 Lin appeals to a key
difference in the institutional arrangements before and after communaliza-
tion as key to what he refers to as the “productivity paradox” in Chinese
agriculture.59 Specifically, Lin argues that, given the costliness of monitoring,
55
The advanced agricultural co-operatives allegedly paid according to both piece rates on
the basis of discrete farm tasks assigned to workers, and the amount of time spent
each day working a myriad of tasks (the so-called “labor day” value), the latter
determined on the basis of observable characteristics (age, gender, etc.). The difficulties
that the agricultural collectives had experienced with these remuneration methods led
Mao to criticize them as “trivially complicated philosophy” (fansuo zhexue 繁瑣哲學),
and he dismissed them accordingly. Thus, on the communal farms, each worker was
assigned one of six to eight labor grades based on skills, strengths, and attitude – an
endeavor that completely eradicated the use of piece rates. While it was officially
recommended that the highest grade should earn more than four times that of the
lowest, the spread between them was so narrow that the difference between each
successive grade was a mere 0.5 yuan per month. A. Donnithorne, China’s Economic
System (New York, Praeger, 1967); L. Deng, H. Ma, and H. Wu, 當代中國的農業
(Contemporary Chinese Agriculture) (Beijing, Dangdai zhongguo chubanshe, 1992),
p. 156.
56
D. Perkins and S. Yusuf, Rural Development in China (Baltimore, Johns Hopkins
57
Lin, “Collectivization and China’s Agricultural Crisis.”
58
Typically, a cropping cycle spans several months, involving as it does the nonspecia-
lized farm worker shifting between myriad tasks dispersed over a large area at each
stage. Together, the sequential nature and spatial dispersion of labor-intensive agricul-
tural production renders monitoring an insurmountable task, leaving individual effort
imperfectly measured and rewarded on the margin. M.E. Bradley and M.G. Clark,
“Supervision and Efficiency in Socialized Agriculture,” Soviet Studies 23.3 (1972), 465–73;
P. Nolan, The Political Economy of Collective Farms: An Analysis of China’s Post-Mao Rural
Reforms (Cambridge, Polity Press, 1988).
59
It is a paradox in the sense that Chinese agriculture achieved total factor productivity
growth during both collectivization (bracketed 1952–1958) and de-collectivization
(bracketed 1979–1984).
658

the only way to resolve the incentive problem inherent in team farming is to
allow diligent farmers to “exit” team farming on discovering shirking by their
lazy partners, which they can easily infer by observing both the level of total
output and weather conditions – both of which are publicly available. But
those lazy workers who treasure the series of small but long-term gains from
co-operation might be willing to provide the minimum effort required to
meet the demands of their diligent teammates, thereby (and this is the only
means of) sustaining co-operative farming. Allegedly, this “exit right” was
permitted before the people’s commune was established in 1958, resembling
an institutional setup analogous to that of a “repeated game,” but vanished
altogether as communalization forcibly changed the structure of incentives
and payoffs to resemble those of a “one-shot” game, in the parlance of game
theory. This nuanced but fundamental change thus destroyed the self-
disciplining mechanism crucial to making collective farming feasible.
While novel, the hypothesis of the “exit right” has been criticized for its
tenuous grasp on historical facts.60 In particular, after becoming members of
the advanced co-operatives, peasants were not entirely free to leave them, as
local officials would be reluctant to let them go for fear of negative evaluation
of their performance.61 There is abundant evidence to show that, in order to
stop those who wanted to quit collective farming, local cadres would resort
to a combination of tactics ranging from moral persuasion to tortuous
actions, so much so that returning to private household farming was unlikely
to be a real option for many, despite what was stated in the Chinese
Constitution.
As an alternative explanation of the productivity paradox, Kung and
Putterman propose that the advanced agricultural collectives established in
1956–1957 might be nominal rather than real, in that the collectives, when
besieged with problems of labor supervision, wittingly decentralized the
organization of farm production all the way down to households via the
equivalence in organization of the elementary co-operatives, which
consisted of an average of some twenty-five households, while retaining
60
It has also been considered faulty on the grounds of game-theoretic reasoning. When
farmers are unable to leave a co-operative, as in the case of the commune, their
interactions arguably assume the properties of a repeated nature. In contrast, when
they are able to leave, they can choose to defect and leave after the first round of team
farming. See, among others, X.Y. Dong and G.K. Dow, “Does Free Exit Reduce
Shirking in Production Teams?”, Journal of Comparative Economics 17.2 (1993), 472–84.
61
J.K.S. Kung, “Transaction Costs and Peasants’ Choice of Institutions: Did the Right to
Exit Really Solve the Free Rider Problem in Chinese Collective Agriculture?”, Journal
of Comparative Economics 17.2 (1993), 485–503.
659

james kai-sing kung
the work point system and unified income distribution practices at the higher
(brigade) level – practices considered the cornerstone of collective
agriculture.62 Cleverly coined the “field responsibility system” (tianjian guanli
zeren zhi 田間管理責任制), this decentralized institutional practice circum-
vented the ideological stranglehold of essentially using contracting (bao 包) to
solve the incentive problem in collective farming. This alternative explan-
ation remains inconclusive, however, as the archival survey conducted by
Kung and Putterman covered only five (mostly) major grain-belt provinces.63
Further research is thus required to ascertain the extent to which such
nuanced flexibilities were indeed exercised in a highly politically charged
context.
Three Consecutive Years of Bad Weather

Interestingly, with the exception of the Chinese government – who attrib-
uted the precipitous fall in grain output to “three consecutive years of natural
disasters,” few scholars have examined in detail the role played by weather
conditions in 1959–1961 in causing the famine.64 Kueh represents a notable
exception.65 On account of the statistics collected on farm acreage “physically
struck” by natural disasters (shouzai mianji 受災面積) and on that “actually
affected” by natural disasters (chengzai mianji 成災面積), Kueh admirably
constructs a “weighted shouzai area index” as pffiffiffiffiffiffi
1 ffi
Lc Ln
ðAc Lc As Ln Þ, in which
he weighs chengzai area ðAc Þ and the balance of shouzai area ðAs Ac Þ by the
assumed yield losses (ðAs Ac Þ and Ln ). He then calculates the yearly
percentage deviations of the “weighted shouzai area” from the mean of
1952–1984 and uses the series obtained as a proxy for weather instability for
analyzing its impact on agricultural production.66 Specifically, the weather
instability index is constructed to predict the potential yield loss for each year,
against which the actual yield loss is compared to ascertain the effect of
62
J.K.S. Kung and L. Putterman, “China’s Collectivisation Puzzle: A New Resolution,”
Journal of Development Studies 33.6 (1997), 741–63.
63
Kung and Putterman, “China’s Collectivisation Puzzle.” The provinces were Hebei,
Henan, and Shandong in the north; Anhui in the center; and Guangdong in the south.
64
In that context, one cannot help but be reminded by Lin, “Collectivization and China’s
Agricultural Crisis,” that it is unlikely that a country the size and diversity of China will
be equally affected by bad weather for three years in a row, although Lin himself has
not examined the effect of weather on the famine catastrophe.
65
Kueh, Agricultural Instability in China.
66
Given that farm acreage “actually affected” by natural disasters (chengzai mianji), if
measured accurately, is a more direct measure of the effect of natural disasters, one
may simply use this (albeit more endogenous) measure instead of the more compli-
cated weighted index as Kueh does. Kueh, Agricultural Instability in China.
660

weather hazards on grain output.67 Based on this formulation, Kueh finds

that weather disturbance may have been the “single most important factor”
in the decline in yield during the Great Leap, explaining 75 and 95 percent of
the yield loss in 1960 and 1961 respectively.68 In a way, Kueh’s results may be
taken as lending support to the Chinese government’s claim that the famine
was indeed to a large extent a consequence of “three consecutive years of bad
weather.”69
A key issue associated with Kueh’s weather index is that shouzai and
chengzai areas were enumerated based on ex post crop loss information, and
are thus at best indirect measures of meteorological conditions.70 Moreover,
in the light of the various policy blunders alluded to earlier, these estimates
are likely to confound the losses caused by human error, resulting in obser-
vational error and overestimation of the effect attributable to the weather.
To go beyond Kueh,71 we turn to disaggregated meteorological data to
examine the role of weather in production failure, by modeling the produc-
tion variable, Ln , in province i and year t as
Yit ¼ αi þ βt þ γ1 Floodit þ γ2 Droughtit þ γ3 Freezeit þ Eit ;
where αi and βt denote province and year fixed effects, and F lood, Drought,
and F reeze stand for the three main specific weather hazards that presumably
would bear upon output. Specifically, we measure the degree of weather
hazards by the proportion to which output might be affected by flood,
drought, and freeze, based on the deviation of monthly precipitation and
67
Kueh’s methodology involves comparing three types of per hectare grain yields: (1) the
observed yield, (2) a trend yield variable derived from the application of fertilizer, and
(3) a weather-predicted yield calculated from regression equations for the period 1952–
1966 and 1970–1984, wherein the weather index is used as the key explanatory variable.
Going by this framework, the disparity between the observed and trend yield is taken as
the total actual yield loss, and the difference between weather-predicted yield and trend
yield is interpreted as the weather-explained yield loss. A comparison of the two gives
the proportion of yield loss caused by weather disturbances. Kueh, Agricultural
Instability in China.
68
Kueh, Agricultural Instability in China. Nevertheless, for 1959 Kueh considers the 9 per-
cent sown-area curtailment to be more damaging to the total output collapse than the
per hectare yield reduction caused by weather. Other estimates of the effect of the
weather on output reduction are considerably smaller. For example, Bramall finds that
drought explained 20, 11, and 9 percent of the output decline in each of the three famine
years in Sichuan Province, whereas Li and Yang, based on a “retrospective survey”
conducted in 1999, find that weather accounted for a 12.9 percent of the overall output
collapse in 1958–61. Of course, both estimates have estimation issues of their own.
Kueh, Agricultural Instability in China, pp. 216, 259; Bramall, “Agency and Famine in
China’s Sichuan Province”; Li and Yang, “The Great Leap Forward.”
69
Kueh, Agricultural Instability in China. 70 Kueh, Agricultural Instability in China.
71
661

james kai-sing kung
temperature from the norm for each province in each given year.72 We report
the two-way fixed effect estimates in Table 18.1.
To make our results comparable to those of Kueh,73 we use the deviation
of grain yield (interpreted as “yield loss”) as the dependent variable in column
1. Specifically, we calculated yields by normalizing total grain output with the
sown acreage, while yield loss is obtained by deducting the natural log of
yield for each given year over the 1953–1966 period from the natural log of the
mean yield for the entire pre-GLF period from 1953 to 1958.74 As shown,
adverse weather had a significant impact on yields: an additional
month per year affected by flood, drought, or freeze damage is associated
with an average reduction of yield of 0.029, 0.014, or 0.032 log points
(or roughly 2.9, 1.4, or 3.2 percent).75 Evaluated at the mean level of
weather adversity in 1959–1961, the combined weather-predicted yield loss
72
Data for monthly provincial temperatures for the 1954–1966 period are taken from C.J.
Willmott and K. Matsuura, Terrestrial Air Temperature: 1900–2010 Gridded Monthly Time
Series V 3.01, and Terrestrial Precipitation: 1900–2010 Gridded Monthly Time Series V 3.02.
These data sets provide monthly mean temperature and precipitation data at 0.5 × 0.5
degree resolution (the size of a 0.5-degree grid square at the central latitude of China (43
degrees north) is fifty-six kilometers high by forty-one kilometers wide.) Grid-level
values were interpolated from an average of twenty weather stations, corrected for
elevation. Weather data were then aggregated to the province–year–month level. The
same data set was used by Meng et al., “The Institutional Causes of China’s Great
Famine,” and Kasahara and Li, “Grain Exports and the Causes of China’s Great
Famine.” The computing formulas for the three weather shock variables are as follows:
1 X12 rainfall
floodit ¼ 1fZitm > 2g;
12 m¼1
1 X12 rainfall
droughtit ¼ 1fZitm < 0:5g;
12 m¼1
1 X12 temperature
freezeit ¼ 1fZitm < 2g:
12 m¼1
To use flood as an illustrative example, we consider a province p to be affected by flood

in month m of year t if the corresponding monthly rainfall has a Z-score above 2 (among
a sample of provincial rainfall between 1990 and 2010), and code the corresponding
province–year–month observation as 1, and 0 otherwise. Based on that, we add up the
number of months m in a year t affected by flood, and then divide the sum by 12, which
gives the proportion of year t affected by flood. We choose the rainfall Z-score
benchmark for drought to be –0.5 as a crude means to adjust for the skewness of the
data distribution.
73
74
Expressed in the equation, yield lossit ¼ logðaverage yield of 1953
58 in province iÞ logðyield it Þ; t = 1953, 1954, . . ., 1966.
75
These numbers were obtained by dividing the coefficients in column 1 of Table 18.1 by
12 (months).
662

Table 18.1 Weather shocks and grain output, 1953–1966

(1) (2)
Yield loss Output loss
Flood 0.342*** 0.365***

(0.101) (0.093)
Drought 0.170*** 0.109**
(0.042) (0.044)
Freeze 0.382** 0.048
(0.142) (0.147)
Year dummies Yes Yes
Province dummies Yes Yes
Observations 350 350
Number of provinces 25 25
R2 0.639 0.622
Note. The dependent variable in column 1 is the logarithm

value of grain yield loss per hectare. The dependent variable in
column 2 is the logarithm value of the total grain loss. Flood is
the proportion of a year affected by flood. Drought is the
proportion of a year affected by drought. Freeze is the
proportion of a year affected by freeze. Robust standard
errors clustered at the province level are in parentheses.
* significant at 10%; ** significant at 5%; *** significant at 1%.
is 0.071 log points (0.038 × 0.342( flood) + 0.316 × 0.170(drought) + 0.011 × 0.382
( freeze)).76 Measured against the average yield loss of 0.184 log points across
provinces in the Leap years, extreme weather conditions explain 38.4 percent
(0.071/0.184) of the loss of yield. Evaluated separately for each of the three
years, the weather-predicted yield loss is greater than the realized loss in
1959 – a finding that agrees with that of Kueh,77 whereas weather contrib-
utes to 28 percent and 26.1 percent of total yield loss for 1960 and 1961
respectively, substantially lower than the 75 percent and 95 percent esti-
mated by Kueh.78
In column 2, we estimate the effect of extreme weather on the deviation of
total grain output. As expected, bad weather, or rainfall anomaly more
specifically, lowers cropping prospects. Based on the average estimated effect,
weather explains 27.6 percent of the loss of output over the three years.
76
Respectively the average severity of flood/drought/cold across China in 1959–1961 is
0.038/0.316/0.011.
77
Kueh, Agricultural Instability in China. 78 Kueh, Agricultural Instability in China.
663

james kai-sing kung
Breaking down the average, again weather adversity more than accounts for
the output loss in 1959 and explains 20.8 percent and 19 percent of the output
loss in 1960 and 1961 respectively. Compared with the weather’s estimated
effect on grain yield per sown acreage, the magnitude in terms of output loss
is smaller, due perhaps to the reduction in sown grain acreage. Summing up,
the havoc wreaked by weather adversity with respect to both yield and output
was most severe in 1959, but tapered off in the subsequent two years.
As we do not control for policy and institutional factors in the regressions,
the effect attributed to weather is probably conflated and hence should be
interpreted as an estimate of the upper bound. However, taking the omitted
variable bias as given, our results based on provincial weather data refine the
result of previous analyses on an aggregate level, and lend support to the
claim made by Liu Shaoqi, China’s vice president, that weather adversity was
of secondary importance (accounting for roughly 30 percent) to policy and
institutional factors in causing the production failure (which accounted for
70 percent).
Excessive Grain Procurement as a Key Determinant of Excess

Death
There is now a growing consensus regarding the role played by (excessive)
grain procurement in causing excess deaths. Euphemistically termed the
Unified Purchase and Unified Sale System (tonggou tongxiao 統購統銷), the
practice of compulsory grain procurement was implemented in 1953 to
facilitate the transfer of surplus grain from the farm sector at below-market
prices to fuel industrialization. Before 1959, the gross procurement rate stood
at around 25 percent, but soared to 38 percent in 1959 and 1960, resulting in the
peasants suffering from acute food shortages. It was only in 1961, on the heels
of accumulating reports of starvation and related diseases, that the procure-
ment rate was reduced to 25 percent and further scaled back in subsequent
years.79
79
A puzzle remains, nonetheless. Before 1949, tenant farmers had to pay their landlords
a rent amounting to an average of 43 percent of the output value of their crops. While
the magnitude of rents varied across regions and by crops (wet rice versus dryland
farming) and depended on the “class status” of the farm households (upper, middle, and
lower), the difference was small. It is possible that the 38 percent of grain procurement
during the Great Leap represented only the average, and famine-stricken provinces or
subunits (such as prefectures and counties) probably suffered disproportionately higher
procurement ratios. The figure for provincial land rents in 1934 comes from Zhengmo
Chen 陳正謨, 中國各省的地租 (Land Rents in Various Chinese Provinces) (Shanghai,
Shangwu yinshuguan, 1936).
664

Against the sharp rise in grain procurement, Kung and Lin found, unsur-
prisingly, a positive correlation between grain procurement and excess
deaths, after controlling for a host of factors including per capita grain output,
ratio of rural population in a province’s population (a proxy for urban bias),
and several “political” variables such as communal dining – using mess hall
participation rates as proxy.80 While the level of grain procurement was
exogenously set by upper-level government, the actual procurement that
resulted was probably endogenously determined, depending on a gamut of
factors to be discussed below, ranging from the alleged rigidity of the
planning mechanism to the whims and ambitions of those who governed
the local jurisdictions responsible for the procurement exercise.
Bureaucratic Rigidity and New Procurement Policy

For Meng, Qian, and Yared, the root cause of the famine lay in the rigidity of
the central planning apparatus.81 Inspired by the surprising positive relation-
ship between agricultural productivity and the severity of famine, these
researchers found that the grain procurement targets were set by the plan-
ners in Beijing in proportion to the projected grain output based upon
historical production records (e.g. the past two to three years). These targets
then trickled down to the provinces, prefectures, and ultimately counties.82
Due to the enormous size of the country and the bureaucratic nature of the
planning apparatus (as the amount of output procured from each level
needed to be tallied and verified), it could take up to an entire year for
Beijing to ascertain the actual amount of grain produced and procured in the
previous year.83 While this lag posed no problem when projected output
growth did not deviate too much from actual output, it would impact
negatively on the peasants when actual output turned out to be significantly
80
Kung and Lin, “The Causes of China’s Great Leap Famine.” These other variables
include Party membership density, which is defined as the ratio of Party members per
10,000 people in a province, and time of liberation, which is ranked according to the
month and year in which a province was “liberated” by the Chinese Communist Party.
Yang, Calamity and Reform in China; Kung and Lin, “The Causes of China’s Great Leap
Famine.”
81
Meng, Qian, and Yared, “The Long-Term Consequences of Famine.”
82
Ashton et al., “Famine in China”; Oi, State and Peasant in Contemporary China; Walker,
“Food and Mortality in China during the Great Leap Forward.”
83
From a different perspective, political scientists tend to view the alleged “rigidity” as
a result of lower-level (commune and brigade) cadres succumbing to the extraordinary
political pressure generated by the process of grain procurement (Bernstein, “Stalinism,
Famine, and Chinese Peasants”; Oi, State and Peasant in Contemporary China), which is
clearly demonstrated by Da Fo village in Henan Province (Thaxton, Catastrophe and
Contention in Rural China).
665

james kai-sing kung
lower than expected against a procurement that had been ratcheted

sharply upwards. This may explain why Meng et al. found a positive
correlation between famine mortality and “production gaps,” which is
measured as the disparity between expected and realized grain production
during the famine.84 It was this bureaucratic rigidity in adjusting the
procurement rate in response to realized grain output that led Meng
et al. to conclude that the excess deaths that occurred during the Leap
were essentially a “planning failure” – an allegation supported by their
further proof that net procurement is positively correlated with historical
productivity.85
While the center may have responded sluggishly (if at all) to an unantici-
pated decline in grain output during the famine years, whether or not this
sluggishness was the product of the planning system remains debatable. For
example, the fact that the Central Committee held two sets of planning
accounts – an official “must-fulfill” account and a second “hoped-to-be-
fulfilled” account – the “two-books syndrome,” and used the latter as the
de facto benchmark against which to evaluate the performance of provincial
governments,86 contained an inherent bias toward higher levels of grain
procurement. When the unpublicized account became the effective target
as it percolated down the administrative hierarchies, grain output had
become grossly inflated by the time it reached the villages – a practice that
was in striking accordance with the prevailing political atmosphere known as
“winds of exaggeration” (fugua feng 浮誇風) – a tendency to boast of the
possibility of sharp output growth.
But the alleged “rigidity” may have been due less to the cumbersome
workings of the planning system than to the new procurement policy put in
place in April of 1958 – a policy that inadvertently gave rise to higher
procurement targets for the following reasons.87 Perhaps the most funda-
mental change was that procurement quotas were no longer set on the basis
of the average output of the past three years, but instead according to the
84
Meng et al., “The Institutional Causes of China’s Great Famine.” Based on the
documented planning mechanism, Meng et al. estimate “expected production” as
a two-year lag of the production figure multiplied by the three-year moving average
of the growth rates from two, three, and four years ago for each province. The
“production gap” is hence constructed as the difference between the realized and
expected output figures.
85
Meng et al., “The Institutional Causes of China’s Great Famine.” Both regressions do
not control for provincial fixed effects and contemporaneous production, however.
86
According to the “Sixty Articles on Work Methods” penned by Mao.
87
“國務院關於改進糧食管理體制的幾項規定” (Several Stipulations of the State
Council on How to Improve the Grain Management System), April 1958.
666

government’s predicted output over the subsequent year – a change that

allowed, or perhaps even encouraged, the exaggeration of grain output.88
The other change that may have given rise to inflexible grain procurement
was that, while provincial governments were allowed to revise procurement
quotas, they could only revise them upward. Of the two changes, the first
would probably give rise to an inflated grain output, especially when 1958,
a year of bumper harvest, was used as the base year for calculation,89 whereas
the latter tended to render the procurement policy “inflexible” to contem-
poraneous changes in output.
Excessive Grain Export as Another Culprit

More recently, by connecting grain procurement with exports,
Kasahara and Li shed new light on the complexity of excessive grain
procurement.90 Due to the souring of the Sino-Soviet relationship,
Beijing was forced to repay 1.5 billion yuan worth of Soviet loans. In
patent disregard for the economic costs entailed,91 Beijing exported up
to 4.1 and 2.7 million metric tons of grain in 1959 and 1960 respect-
ively – amounts that would have contributed to the caloric needs of
16.7 to 38.9 million peasants.92 The blow was not spread evenly among
the rural population, however, as the central government tended to
export those crops that fetched higher prices on the international
market – a strategy that subjected the regions specializing in relatively
“valuable” crops to heavier extraction.93 Using county-level data, and
by interacting historical cropping patterns with grain-specific inter-
national price movements as a measure of “export shock,” Kasahara
and Li found that regions with higher exposure to exports retained less
food and suffered from higher death rates.94 This “pegging” relation-
ship between procurement intensity and export exposure was
88
Kung and Zhou, “Political Elites and Hometown Favoritism.”
89
The higher procurement rate set for 1959 was indeed premised on the bumper harvest
of 1958.
90
Kasahara and Li, “Grain Exports and the Causes of China’s Great Famine.”
91
J.W. Garver, China’s Quest: The History of the Foreign Relations of the People’s Republic
(New York, Oxford University Press, 2015).
92
93
For example, given the price premium of rice and soybean, these two goods accounted
for 81 to 95 percent of total grain exports in 1955–1961. More importantly, on the heels of
their price hikes in 1957–1958, the exports of these two goods rose by 6.4 million tons and
2.0 million tons in 1958–1960, relative to the 0.06 million ton–0.99 million ton increase in
wheat, maize, and other grains.
94
667

james kai-sing kung
estimated to account for 15 percent of the excess deaths in 1960,95 with

the effect being greater in counties further from railroads, as the
logistic costs of disaster verification and relief transportation were
larger in areas not readily accessible by railroads.96 The geography of
mortality and its determinants – including the topography and trans-
port network – are clearly an important issue, especially in the context
of localized famine such as the case between adjacent counties and
prefectures.
Taken together, the coexistence of procurement rigidity and excessive
export inevitably raises the concern of how aware Mao was of the famine’s
severity, and his apparent lack of compassion and response to the grave
situation that was unfolding in the countryside. One may take the innocu-
ous view that the center was, perhaps, crippled by institutional barriers,
ignorant of the meltdown on the ground during 1959–1961, hence the lag in
adjustment to procurement. But this diagnosis is challenged by the radic-
alism rekindled in the wake of the Lushan Conference in July 1959,97 when,
after being challenged by Marshal Peng Dehuai, Mao single-handedly fired
all the cylinders of the Leap, raising the targets of grain and steel produc-
tion to realistically unattainable levels98 – an act that Bernstein interprets
95
This result may be overstated, however, given that their estimation fails to account for
the effect of the resale of grain to rural areas (the data for which are lacking for most
provinces), and the extent that resale of grain had a moderating effect on excess deaths,
as Kung and Zhou, “Political Elites and Hometown Favoritism,” found to be the case
for Henan Province.
96
Remoteness is a double-edged sword, however. If remoteness impedes the distribution
of relief because of higher transportation costs, it should also have deterred procure-
ment in the first place. Indeed, Garnaut, “The Geography of the Great Leap Famine,”
argues that the logistic difficulties of transporting large quantities of grain over vast
distances limited the reach of the extractive regime, and as a result saved many lives
from famine. Consistently, Gooch finds the same with terrain ruggedness and its
associated transaction costs. However, using the case of Sichuan as an example,
Bramall, “Agency and Famine in China’s Sichuan Province,” finds no apparent geog-
raphy of mortality in terms of either access to transportation or topography, despite the
fact that Sichuan exhibits enormous variations in the death rate across the counties.
E. Gooch, “Terrain Ruggedness and Limits of Political Repression: Evidence from
China’s Great Leap Forward and Famine (1959–61),” Journal of Comparative Economics
47.4 (2019), 827–52.
97
The Lushan Conference was originally designated to review and contain leftist extrem-
ism, but ended up seeing the purge of Marshal Peng Dehuai and the revival of the Leap
campaign. The twist occurred after Peng handed the famous “Letter of Opinion” to
Mao, in which he cautiously pointed out the Leap’s problems such as “winds of
exaggerations.” Mao took offence at the criticism and publicly charged Peng with
having “leaned towards the right for about 30 kilometers.” It then followed that the
rightist tendency was reinstated as the target of an ideological purge. Kung and Lin,
“The Causes of China’s Great Leap Famine.”
98
Kung and Lin, “The Causes of China’s Great Leap Famine.”
668

as willful neglect of the “expense on livelihood” in pursuance of political

goals, rather than unwitting ignorance.99
The Role of Political Elites: Central versus Provincial Officials

The disproportionate importance of procurement in causing excess deaths and its
highly endogenous origin inspired us to examine the possible exogenous deter-
minants of variations in grain procurement across provinces. By taking the
varying incentives of provincial leaders of different political ranks within
China’s centralized personnel system – the nomenklatura – as given, Kung and
Chen saw the Great Leap as providing an exceptionally rare opportunity for those
who were less senior in terms of political rank – the “alternative members” or
AMs – to climb the career ladder of what essentially was a “closed” market with
virtually no outside options, by procuring and remitting more grain to the center,
signaling their unswerving loyalty to, and support for, Mao’s fanatical campaign.
Given not only that their more senior counterparts – the “full members” or
FMs – held more positions, but also that the positions held were more
strategic and had distinctly more rights and privileges,100 there was
a powerful incentive for the AMs to get ahead. Further, unlike FMs, who
faced a prohibitive barrier to reaching the highest echelon as Politburo
members, AMs did not face similar barriers in their progress to becoming
FMs.101 Perhaps for this reason, AMs were found to have systematically
procured roughly 3 percent more grain in provinces they governed than
was procured in provinces governed by FMs.102 Using this as a benchmark,
99
Bernstein, “Mao Zedong and the Famine.” However, it would be a stretch to go as far
as to suggest that this willfulness was not accompanied by any personal struggle on
Mao’s part. Since the spring of 1960, a flush of reports of deaths and abuses came from
provinces including Guizhou, Guangdong, Shandong, and Liaoning, which were
treated by Mao with a flickering combination of lukewarm responses, doubt, “great
anger,” and self-deceptive negligence, reflecting his hesitancy and contradictory
motivations. Only in October 1960 did Mao come to “grasp the dimensions” upon
learning about the catastrophic situation in Henan, and supported abandoning the
Leap, which, however, failed to put an immediate end to the deep-rooted leftism.
100
For example, AMs were only allowed to express their opinions at plenary meetings;
they were ineligible to vote on resolutions. See Kung and Chen, “The Tragedy of the
Nomenklatura.”
101
At a notch above the FMs, Politburo members required the “prerevolutionary creden-
tial” of experience of the Long March, but a similar credential was not required of an
FM. Kung and Chen, “The Causes of China’s Great Leap Famine.”
102
However, Kung and Chen, “The Tragedy of the Nomenklatura,” also find that
the second in command in a province, the governor, did not exhibit the same career
incentives. This may possibly be due to the fact that, unlike the first party secretary, the
governor belonged to the government apparatus and was thus not predisposed to
engage in radical actions.
669

james kai-sing kung
political radicalism caused by “career concerns” can explain up to 16.83 per-

cent of the overall excess death rate – exceeding even that of excessive
exporting. On the basis of this finding, Kung employed a regression discon-
tinuity approach to carry out a further comparison of the incentives of those
AMs and FMs who were ranked near the cutoff threshold of ninety-seven (i.e.
there were ninety-seven FMs at the time), and to identify the causal effect of
membership status (i.e. AM versus FM) on excessive grain procurement
based on the exogenous ballot ranking of the Eighth National Congress.103
Doing so increases the magnitude from 3 percent for AMs as a group to
8 percent between those AMs on the verge of promotion and the lowest-
ranked FMs who only saw a distant prospect of further career progression.
A question arising from this analysis is whether political elites could only
do evil during the Great Leap. Inspired by the anecdote of Xi Zhongxun’s
appeal to arrange for “relief grain” to be shipped to his hometown of Fuping
county in Shaanxi province, Kung and Zhou examined the possible role of
members of the Central Committee (CC) in alleviating excess deaths during
the Great Leap Famine.104 In sharp contrast to the strong “career concerns” of
regional officials, political elites who worked in the central government –
specifically in the planning department and in charge of grain allocations –
had alleviated excess deaths by redistributing grain to their hometowns,
presumably on compassionate grounds. By regressing logged excess deaths
during 1959–1961 against the birthplace (hometown) of the Eighth National
Party Congress of 1958, Kung and Zhou found that having an additional
native CC member in a prefecture helped to reduce the excess deaths in that
prefecture by 46,500, accounting for 2.3 percentage points in the death rate
when evaluated at the mean.105 Moreover, while based on the evidence of
just one province (Henan), these authors found that the channel of “relief
grain” worked through the resale of grain or grain “sold back” to the
countryside rather than through a reduction in grain procurement.
Taking advantage of this insight into what might underpin excessive grain
procurement, perhaps we may reassess the joint significance of institutional
rigidity and political radicalism within the same analytical framework using
the following two-way fixed-effect model:
103
Kung, “The Emperor Strikes Back.”
104
Kung and Zhou, “Political Elites and Hometown Favoritism.” The father of Xi Jinping,
Xi Zhongxun, was one of the “Eight Elders” of the Chinese Communist Party and
a vice premier during 1959–1962.
105
670

Pit ¼ αi þ βt þ δ1 Qit þ δ2 Qit2 þ θRankit þ μYit þ Eit ;
where αi , βt , and εit are the same as previously defined, P stands for per capita
net/gross/resale/extra grain procurement depending on the choice of
dependent variable, and Q it and Q it2 stand respectively for the contempor-
ary per capita grain output and per capita grain output lagging by two years.
The panel covers the period from 1957 to 1965. Following Kung and Chen,
Rank is a dummy variable indicating the provincial leader’s political rank (i.e.
FM, AM, or nonmember NM) in the Central Committee, while Y includes
two control variables, per capita GDP and proportion of agricultural
income.106
The results are reported in Table 18.2. In all the regressions we control for
the fixed effects of province, year, and level of development (per capita GDP),
as well as for the relative importance of agriculture. Using per capita net grain
procurement as the dependent variable,107 we first examine the “institutional-
rigidity” hypothesis by assessing the relative importance of contemporaneous
versus historical grain output in determining the dependent variable. On its
own (columns 1 and 2), both variables exhibit a significant positive correlation
with net grain procurement, but the two-year lagged variable has a much
smaller estimated effect (0.154, column 2, versus 0.592, column 1). When put
in pairs (column 3), however, the lagged output becomes insignificant, but
contemporaneous output continues to be significant and has a similar mag-
nitude. This result is further confirmed when we interact these two variables
(separately) with the dummy variable of the Great Leap Forward (GLF)
(column 4). Together, this evidence does not bode well with the “planning-
rigidity” hypothesis, except that we must point out the caveat that this
analysis is performed at the province (instead of the county) level. In column
5, we include political rank in the regression as a “horserace” with planning
rigidity, and find that the variable AM is significant, reinforcing an earlier
finding that AMs were more aggressive as a group in squeezing their people,
even when controlling for current-year grain output. The two-year lag of
grain output remains insignificant in this regression. In column 6 we add in
the variable CC membership, which is defined as the number of Central
Committee (CC) members in a given province for every million people, the
purpose of which is to test whether political elites who served in the
106
Kung and Chen, “The Tragedy of the Nomenklatura.”
107
Net procurement is obtained by subtracting resale of grain from gross procurement,
whereas excessive net procurement is the difference between average net procure-
ment during 1955–1957 and the amount procured during 1958–1961.
671

Table 18.2. The effects of grain output, political rank, and CC membership on grain procurement, resale, and excess
procurement, 1957–1965
(1) (2) (3) (4) (5) (6) (7) (8) (9)
Current-year grain output 0.592*** 0.580*** 0.393*** 0.594*** 0.595*** 0.523*** -0.072*** 0.201***
(0.112) (0.108) (0.078) (0.111) (0.112) (0.130) (0.023) (0.063)
Two-year lag of grain output 0.154** 0.068 0.116 0.088 0.080 0.146 0.066* 0.005
(0.057) (0.045) (0.074) (0.070) (0.073) (0.095) (0.032) (0.064)
GLF # current-year grain output 0.286**
(0.120)
GLF # two-year lag of grain output −0.114
(0.095)

NM 0.042 0.074 −0.375 −0.450* 0.232
(0.914) (0.922) (1.143) (0.238) (0.484)
AM 0.828*** 0.821*** 0.727** −0.095 0.555***
(0.276) (0.279) (0.326) (0.066) (0.184)
CC membership −2.214** −2.619* −0.405 −2.779***
(1.062) (1.387) (0.563) (0.853)
Personal identity No No No No Yes Yes Yes Yes Yes
Controls Yes Yes Yes Yes Yes Yes Yes Yes Yes
Year dummies Yes Yes Yes Yes Yes Yes Yes Yes Yes
Province dummies Yes Yes Yes Yes Yes Yes Yes Yes Yes
Observations 225 225 225 225 225 225 225 225 225
Number of provinces 25 25 25 25 25 25 25 25 25
R2 0.368 0.268 0.370 0.381 0.507 0.508 0.483 0.421 0.525
Notes. The dependent variable in columns 1–6 is the per capita net procurement. The dependent variables in columns 7–9 are per capita
gross procurement, per capita resale, and per capita extra procurement. Current-year grain output is per capita grain output for the
current year. Two-year lag of grain output is per capita two-year moving average of grain output. GLF is a dummy variable, which equals
1 if the year is from 1958 to 1961, otherwise 0. NM is nonmember. AM is alternate member. CC membership ratio is the number of

Central Committee members per million population. Controls include lag of share of agricultural GDP and lag of GDP per capita. Robust
standard errors clustered at the province level are in parentheses. * significant at 10%; ** significant at 5%; *** significant at 1%.
james kai-sing kung
government might favor their “distant relatives” back in their “hometowns.”

Consistently with Kung and Zhou, the significantly negative coefficient
suggests that provinces with greater representation by these political elites
in the CC did suffer less in terms of net grain procurement.108
In the last three columns (7–9), we change our dependent variable to gross
procurement, resale or sale-back grain (fanxiao liang 返銷糧), and excessive
procurement respectively – all adjusted on a per capita basis, in order to
examine the channels through which procurement intensity was affected. In
column 7, we find that only contemporaneous but not historical grain output
has a significantly positive effect on per capita gross procurement, which
suggests flexibility. And while AM declines in significance (from 1 percent to
5 percent), it remains a significant determinant. We find the same with CC
membership, which, while similarly declining in the level of significance
(from 5 percent to 10 percent), remains significant.
In the case of resale grain (column 8), both contemporaneous and two-year
lagged output are significant but operate in opposite directions. The negative
effect of the current-year output on resale grain indicates, intuitively, that
a larger grain output is negatively associated with the amount of grain sold
back to the rural area. On the contrary, sale-back is positively (but only
marginally) correlated with the two-year lagged grain output and with
a comparable magnitude, suggesting that the resale amount in a given year
was probably financed by the provincial government’s stockpile from the
average of the previous two years, which from 1958 onwards was assigned to
deal with famines of an “average magnitude” (yiban zaihuang 一般災荒).109
As they were eager to remit more grain to the center, AMs were unlikely to
return more grain to the countryside, as revealed by the negative (albeit
imprecisely estimated) coefficient. Finally, in column 9, contemporaneous
production and AMs are jointly and positively significant in explaining the
“excessive portion” of procurement in comparison with the level in 1955–
1957 – a result that strikingly resonates with findings in the extant literature.110
108
109
Before 1958, intra-provincial grain transfer was controlled directly by the central
government, and provincial officials had to request permission for more grain sale-
back. However, the new policy that came into effect in 1958 delegated full control over
residual grain to provincial officials after delivery to the center was fulfilled. Z. Fan, W.
Xiong, and L.-A. Zhou, “Information Distortion in Hierarchical Organizations:
A Study of China’s Great Famine,” NBER working paper (2016); Kung and Zhou,
“Political Elites and Hometown Favoritism.”
110
Kung and Chen, “The Tragedy of the Nomenklatura”; Kung, “The Emperor Strikes
Back.”
674

Consumption: Communal Dining

A third view seeks the most proximate cause of the famine in “irrational”
consumption behavior as arguably manifested by the 2.65 million mess halls
established in the autumn of 1958 as an integral part of people’s communes.
Depending on the extent to which the rural population were herded into the
mess halls, which varied considerably from one province to another,
the egalitarian distribution and free supply of food – vividly described in
the slogan “Open up your stomach and eat as much as you can” – destroyed
participants’ incentive to economize on food, and hence resulted in wasteful
consumption.111 By early 1959, food shortages were reported to have occurred
in many places, including where there were allegedly bumper harvests.112
In establishing the causal importance of communal dining, Chang and
Wen developed a framework in which they eliminated many factors that
could have caused the famine’s onset;113 these factors ranged from outright
collapse of output and failure of distribution to demand shock and communal
mess dining. For instance, they argued that the famine was unlikely to have
been caused by a failure in food production, as 1958 was a year of bumper
harvest. Likewise, while they see procurement as having aggravated the
famine, it did not increase sharply until 1959, but the famine had already
started in late 1958 rather than in early 1959. By the same token, the lack of
a population explosion is also inconsistent with the famine being caused by
a sudden demand for food.114 By elimination, irrational consumption remains
the only likely “triggering” factor of the famine. As mentioned earlier,
participants in the mess halls were assumed to be driven by a mixture of
fear of deprivation and perhaps euphoria over the free (and unlimited) supply
of food, leading them to engage in self-destructive or what they call
111
Chang and Wen, “Communal Dining and the Chinese Famine.” As was documented
by Potter and Potter and cited in Chang and Wen, “in 20 days peasants finished almost
all the rice they had, rice which should have lasted six months.” S.H. Potter and J.
M. Potter, China’s Peasants: The Anthropology of a Revolution (Cambridge, Cambridge
University Press, 1990); Chang and Wen, “Communal Dining and the Chinese
Famine.”
112
Bo Yibo 薄一波, 若干重大決策與事件的回顧 (Several Important Policy Decisions
and Events in Retrospect) (Beijing, Zhonggong zhongyang dangxiao chubanshe, 1991).
113
Chang and Wen, “Communal Dining and the Chinese Famine.”
114
While there was no sudden increase in population, the fact that Mao sharply increased
the number of rural workers working in state-owned enterprises by 30 percent (from
99 million in 1957 to 130 million in 1961) to increase steel output put a greater burden on
grain procurement, as it required an additional 6 million tons of requisitioned grain to
be sent to Beijing, Tianjin, Shanghai, and Liaoning – cities and provinces where the
lion’s share of state-owned enterprises were located. To alleviate food shortages in
cities during the aftermath of the famine, the Chinese government sent 20 million of
these migrants back to the countryside and imported grain from abroad.
675

james kai-sing kung
“irrational” consumption behavior – a behavior that would rapidly deplete

the meagerly available food stocks.
One way of testing this hypothesis is to correlate the extent to which
a province engaged its rural populace in communal dining practices and
excess deaths, given the enormous variation that existed between provinces.
Prima facie evidence suggests that provinces with severe casualties, such as
Anhui, Guizhou, and Henan, all had participation rates of well over 90 per-
cent. By contrast, the less-afflicted regions, such as those in the remote
northeastern provinces, were far more moderate in implementing this
policy.115 In an attempt to expand on this claim, Liu, Wen, and Wei regressed
a province’s famine mortality against its mess hall participation rate at the
end of 1959 – the only year for which data on communal dining are available,
and found that communal dining can explain 8.7 to 16.8 percent of the
variations in excess death rates during 1958–1961.116 Their finding differs
from that of Kung and Lin,117 who found that the mess hall participation
rate was an insignificant determinant of excess deaths when competing
variables of political radicalism, such as Party membership density or time
of liberation, are controlled for.118 By contrast, grain procurement is consist-
ently significant in their regressions. Between the two ends – procurement
and consumption – the significance of (excessive) grain procurement implies
that the countryside would probably have been left with insufficient grain for
consumption, thereby rendering wasteful consumption a mere allegation.119
This would seem to be the case especially after the Lushan Conference;
Mao’s rekindled effort to get the Leap firing on all cylinders notwithstanding,
with the procurement of grain and steel escalated to unbelievably high levels,
the possible role played by communal dining in causing excess deaths should
not be exaggerated.120
115
Kung and Lin, “The Causes of China’s Great Leap Famine,” 60.
116
Y. Liu, J.G. Wen, and X. Wei, “Communal Dining System and the Puzzle of the Great
Leap Famine,” China Agricultural Economic Review 6.4 (2014), 698–716.
117
Kung and Lin “The Causes of China’s Great Leap Famine.”
118
While they have instrumented mess hall participation rates, their instruments with (1)
the ratio of the participation rate in advanced agricultural co-operatives in 1956 to the
participation rate of elementary agricultural co-operatives in 1954, and (2) population
density in the rural areas, are both endogenous.
119
In Da Fo village of Henan province, one of the famine-stricken provinces, substantial
grain rationing – which allegedly amounted to 60 percent of the villagers’ grain
output – had long put an end to so-called “overconsumption.” Thaxton, Catastrophe
and Contention in Rural China, p. 127. As the author also remarks, overconsumption, if it
ever existed, was “a practice of at least three harvests past” (p. 125).
120
In fact, official evidence suggests that, perhaps in response to early signs of food
shortages, by early 1959 communal dining had already lost its initial momentum and
rigor. Not only were seasonal operations allowed, but also, in many instances,
676

There are other reasons why communal dining is an unlikely culprit. In an

attempt to find out how communal dining was actually organized, I went
through the press archive on mess hall organization. Of a total of 120 pertinent
articles reporting on twenty-four provinces in the second half of 1958, I found
that over 75 percent of these reports made no explicit reference to unre-
stricted food consumption in terms of how the commune dining kitchens
were operated. Far from an “all-you-can-eat” scenario as conjured up by the
political slogan, food was effectively allocated on the basis of age, sex, and
labor grade, which is suggestive of a practice of differentiating food rationing
on the basis of differential work contributions (labor grade) and needs (age
and gender).121 In other words, while food might have been provided free, the
quantity was rationed, and hence the communal dining kitchen was not an
unmanaged common that brings ruin to all – at least not for the entire
country.
Long-Term Effects
In recent years, increasing attention has been paid to examining what may be
considered the “long shadow” cast by the Great Leap Famine over the well-
being of individual survivors, institutional choice, and comparative economic
development. What emerges from these studies is that the shock imposed by
the famine has exerted a persistent effect on these outcome variables of
interest through biological inheritance, human capital accumulation, and
even collective memory.
Individual-Level Outcomes
As individuals who were born or conceived during the famine (hereafter
“famine cohort”) enter midlife, it affords the researchers a unique opportun-
ity to examine how malnutrition occurring at the fetal and/or infant stages
might have led to adverse adulthood health outcomes,122 and, as
a consequence, disadvantages in a variety of educational and labor market
participation was reduced to a partial basis whereby only one meal a day was served
and in-home preparation was allowed. Above all, only 20 to 30 percent of the mess halls
remained nationally after this period of attrition. Bo, 若干重大決策與事件的回顧.
121
Kung, unpublished research notes. Thaxton, Catastrophe and Contention in Rural China,
similarly observes that unrestricted consumption in the mess hall of Da Fo village only
lasted for a few months; thereafter food was rationed by coupons.
122
The long-term effect of fetal and infant malnutrition on middle-age health is widely
known as the “Barker hypothesis” in the medical literature. See D.J. Barker, “The Fetal
and Infant Origins of Adult Disease,” British Medical Journal 301.6761 (1990), 1111.
677

james kai-sing kung
outcomes such as labor supply, educational attainment, and socioeconomic

status.123 In terms of health consequences, early-life exposure to the famine
has been associated with stunting, obesity, schizophrenia, diabetes, and
a higher likelihood of disability in adulthood. With its close link to nutrition,
disease, and the economic environment, attained height is frequently used as
an indicator of well-being. For example, in order to quantify the famine’s
long-term impact on human capital outcomes, Chen and Zhou drew upon
the variations in famine exposure across both regions and birth cohorts, and
found that, if not due to in utero nutritional deficiency, those who were born
in 1959 and 1960 would have grown 3.03 and 3.0 centimeters taller
respectively.124 In turn, the height deficiency is apparently associated with
a chain of unfavorable socioeconomic outcomes for the survivors. For the
1960 cohort, for instance, famine exposure as measured by an additional
excess death per thousand in the population is associated with a 2.1 percent
reduction in total labor supply and a 4.3 percent reduction in farm income.
A major issue associated with the difference-in-differences (DID) method
typically employed in studies of this nature is self-selection, in that infants
who survived the famine are likely to have (unobserved) stronger biological
survivability than those who perished, or these children may have been born
to physically and socioeconomically advantaged parents amid resource defi-
ciency; in either case, the resulting selection bias would have attenuated the
estimated effect of the famine. To overcome these challenges, Meng and
Qian improved the DID estimator using quantile analysis based on the
premise that attenuation would be less pronounced for individuals in the
upper quantiles of the outcome distribution.125 Adjusting the estimation this
way leads to a 2.8-centimeter height reduction because of in utero exposure to
famine. Yet another way to separate the selection effect from the stunting
effect is to make use of the idea of heritability, according to which only the
“tall genes” of the famine cohort (as opposed to environmentally induced
stunting) would be passed down to their offspring.126 On this assumption,
children of the famine cohort would on average be taller than those
unaffected by the famine, and their height could therefore be used to control
for selection within a model that maps the relationship between the heights
123
J. Currie and B.C. Madrian, “Health, Health Insurance and the Labor Market,”
Handbook of Labor Economics 3 (1999), 3309–3416, provide a comprehensive review of
studies linking poor health conditions to lower education and labor market attainment.
124
Chen and Zhou, “The Long-Term Health and Economic Consequences.”
125
Meng and Qian, “The Long-Term Consequences of Famine.”
126
Gørgens, Meng, and Vaithianathan, “Stunting and Selection Effects of Famine.”
678

of parents and of children. This approach thus allows identification of the

stunting effect separately from the selection effect. On the one hand, famine
has the negative – stunting – effect of reducing the height of its cohort by one
or two centimeters; on the other hand, this cohort is one or two centimeters
taller than the average because of their superior genes (which allow them to
survive) – the selection effect. The two effects cancel one another out.127
However, the assumption underlying Gørgens et al.’s methodology – that
children of the famine survivors were able to transcend their parents’ trau-
matic history and quickly rebound – is questionable in the light of studies
focusing on the intergenerational effects of the famine.128 For instance,
Almond, Edlund, and Zhang found that the Chinese famine had not only
altered the affected cohort’s sex ratio towards a higher proportion of females,
which was probably a result of higher male fetus susceptibility, but also
famine-born mothers were in turn 1.2 percent less likely to give birth to
boys, and 8 percent more likely to give birth to underweight babies.129
However, these findings by economists are disputed by medical researchers.
Using the Dutch Hunger Winter of 1944–1945 as context, Stein, Zybert, and
Lumey find no evidence of an excess of female births among deliveries of
human infants exposed to famine at any period of gestation.130 The same
disagreement applies to the claim of a higher risk of diabetes and its alleged
transmission to future generations. The allegation that the Great Leap
Famine was the origin of the prevalence of diabetes in China today is
similarly disputed by Li and Lumey,131 who argue that the observed preva-
lence of a wide range of chronic illnesses, including type 2 diabetes and
127
That is, just because the survivors are of normal height (the observed outcome), it
would be erroneous to suggest that the famine has done no harm to the survivors, as
the “harm” is compensated for by self-selection.
128
Gørgens, Meng, and Vaithianathan, “Stunting and Selection Effects of Famine.”
129
Almond et al., “Long-Term Effects of Early-Life Development.” This result finds
support from the Trivers–Willard hypothesis in evolutionary biology, which predicts
that, by comparison with parents in “good conditions,” those in “poor conditions”
would “favor” the female offspring, as the survivability of male infants is lower than
that of female infants; the chances are thus greater that daughters would survive
during famine. See, e.g., C. Ó Gráda, Famine: A Short History (Princeton, Princeton
130
A.D. Stein, P.A. Zybert, and L.H. Lumey, “Acute Undernutrition Is Not Associated
with Excess of Females at Birth in Humans: The Dutch Hunger Winter,” Royal Society
Biology Letters 271 (2004), S138–41.
131
P. Zimmet, Z. Shi, A. El-Osta, and L. Ji, “Epidemic T2DM, Early Development and
Epigenetics: Implications of the Chinese Famine,” Nature Reviews Endocrinology 14.12
(2018), 738–46. C. Li and L.H. Lumey, “Exposure to the Chinese Famine of 1959–61 in
Early Life and Long-Term Health Conditions: A Systematic Review and Meta-ana-
lysis,” International Journal of Epidemiology, 2017, 1157–70.
679

james kai-sing kung
schizophrenia, among adults born during the famine is really due to uncon-
trolled age differences between the treated group and those who were born
either before or after the famine. The health outcomes caused by stress in
utero are thus at best inconclusive.
Another long-term consequence for individuals who suffered from the
Great Leap Famine is educational attainment. However, the evidence is less
conclusive in this regard. For example, Meng and Qian estimated that while
prenatal famine exposure is responsible for 0.6 fewer years of schooling, the
same impact cannot be found from exposure in infancy – the latter defined as
individuals who were affected during their first year of life.132 In contrast, Shi
found that exposure to famine in infancy lowers the probability of completing
high school by 0.5 percent.133 In terms of marriage outcomes, Almond et al.
concluded that the male famine cohort is more likely to stay spouseless,
whereas Brandt, Siow, and Vogel found only negligible changes in the
marriage rates of survivors, possibly because their reduced marital attractive-
ness was mitigated by the smaller size of the cohort.134 Discrepancies between
research results are even more pronounced in the case of the famine’s impact
on the labor supply, with available estimates ranging from no effect to
a 13.9 percent reduction in total working hours.135
The lack of a consistent finding across studies is not unanticipated, in view
of the difficulty of separating the famine’s impact from that of later-life
events. This is especially pronounced for the non-health aspects of long-
term outcomes, as it remains unclear what and how subsequent environ-
mental factors would interact with the initial health shock. Perhaps
a particular topic that awaits illumination is the causal channels linking
exposure to famine, health capital, and other socioeconomic outcomes.
132
Meng and Qian, “The Long-Term Consequences of Famine.”
133
X. Shi, “Famine, Fertility, and Fortune in China,” China Economic Review 22.2 (2011),
244–59.
134
Almond et al., “Long-Term Effects of Early-Life Development.” L. Brandt, A. Siow,
and C. Vogel, “Large Demographic Shocks and Small Changes in the Marriage
Market,” Journal of the European Economic Association 14.6 (2016), 1437–68.
135
While Meng and Qian, “The Long-Term Consequences of Famine,” found that early-
childhood exposure to the famine reduced the labor supply by an average of 13.9 per-
cent, Shi, “Famine, Fertility, and Fortune in China,” found no significant effect of
infancy exposure. Between the sexes, Almond et al., “Long-Term Effects of Early-Life
Development,” found that exposure to famine increased the probability of not working
by 3 percent for women and 5.9 percent for men.
680

Institutional Change and Development

In addition to affecting individuals, the Great Leap Famine also provides
a rare opportunity to examine the possible role of history in shaping institu-
tional choice. For example, is the bad, collective memory experienced during
the Great Leap Famine associated with people’s communes, so that when the
opportunity to dismantle collective agriculture presented itself, as was the
case during 1979–1984, those who suffered the most from the Leap would be
the first to make the change?136 The relative autonomy and long period over
which provinces were allowed to switch to household farming presents
a good quasi-natural experiment for the possible bearing of history on
institutional choice.137 In this experiment, Bai and Kung interpreted the
variations in timing across the provinces as reflecting their perceived prefer-
ences for collective rather than family farm institutions, which in turn was
determined by their relative preference for more efficacious provision of
public goods such as irrigation or sharper incentives for individual work, on
the assumption that collective agriculture enjoys a comparative advantage in
the provision of public goods, but incentives for individual work are distinctly
greater on family farms.138 In short, this preference can be taken to represent
a trade-off between the two types of farm institution with respect to their
relative strengths and weaknesses.
Bai and Kung’s remaining analysis thus hinges on the question of what
determined the preference for one type of institution over the other across
the Chinese provinces. To do so, they constructed a panel data set with which
to examine the respective possible effects on institutional choice of the
strengths of collective farms (using irrigation as proxy) and their weaknesses
(using famine severity as proxy). For the dependent variable, the authors
assumed that the earlier a province chose to decollectivize, the more it must
have suffered from collective agriculture in general, and famine severity in
136
For instance, whereas Anhui province, a poverty-stricken province in central China,
jumped the gun and dismantled its collective farms before the Chinese Communist
Party officially sanctioned it, the northeastern provinces were recalcitrant and dragged
their feet even after the central government had explicitly recognized the virtues of
household farming and endorsed it. Bai and Kung, “The Shaping of an Institutional
Choice.”
137
In sharp contrast to the establishment of the people’s communes, which was largely
homogeneous and occurred over a narrow time window, it took nearly five years for
the reverse process to occur, which reflects the fundamental difference in approach to
implementing policy between Mao and Deng. Unlike Mao, Deng took his time to
demonstrate the productivity advantages of household farming in order to conciliate
conservative political opponents, many of whom still stood in the way of this institu-
tional change.
138
Bai and Kung, “The Shaping of an Institutional Choice.”
681

james kai-sing kung
particular. In sharp contrast to the conventional wisdom, which maintains

that weather adversity – specifically drought – was an exogenous shock that
predisposed villagers to switch over quickly to family farming,139 Bai and
Kung found that bad weather that occurred during decollectivization actually
slowed down the switchover to family farming, presumably because collect-
ive agriculture was still considered to be more efficient at providing
irrigation.140 However, in provinces that suffered heavily from the Leap’s
famine in terms of mortality, weather shocks occurring during decollectivi-
zation effectively served as a painful reminder of the weaknesses of collective
agriculture, thereby reversing the effect of bad weather on the choice of
institutions, by predisposing these provinces to make the switchover sooner
rather than later.141 To further ensure that this analysis also holds at the
village level, Bai and Kung employed the China General Social Survey
(CGSS) of 2005 (which contains a question asking the village cadre the year
in which they abandoned collective agriculture) to check robustness and
confirm it.142 Thus, through the channel of collective memory, the Great
Leap Famine offers social scientists a rare opportunity to examine the role of
history in institutional change.
Concerning its effect on development more generally, Gooch finds that an
increase of one standard deviation in excess deaths during the famine is
associated with a decrease in per capita GDP of 4,454 yuan half a century
later, although she is reticent on the mechanism causing this result – a topic of
future research.143
Conclusion
In this chapter we have reviewed several bodies of literature on China’s Great
Leap Famine – including its long-term consequences. Having killed an esti-
mated 30 million people (if not more), it remains the deadliest famine in
139
A production team by the name of Xiaogang in Anhui province became a household
name as they secretly divided their collective holdings among the farm households
after being struck by a severe drought, which made planting formidably difficult. The
story of these highly motivated peasant households spending several days working
extremely hard and eventually planting the wheat on the cracking dried earth speaks
volumes for the paramount importance of individual work incentives.
140
141
The same logic applies to irrigation. In provinces where irrigation had been built up
during the period of collective agriculture, peasants would be more confident in
dealing with weather risks even on their own, and this predisposed them to choose
household farming sooner.
142
143
Gooch, “Estimating the Long-Term Impact of the Great Chinese Famine.”
682

human history. A unique feature of this great famine is that it occurred in the
context of a bumper harvest, and under a regime whose leader was very
ambitious and sanguine about the future of his country and the people. But
a multitude of bold and untried policies, rooted in the fundamentally unreal-
istic goal of producing a substantially larger farm surplus to help fuel indus-
trialization, made this grand scheme an utter failure.
For more than two decades, social scientists have been fascinated by this
incident and tried to make sense of it. A variety of explanations have been
offered. While some emphasize that it was excessive grain procurement that was
primarily responsible for the observed excess deaths, other culprits included the
thorough destruction of work incentives under the commune system, while still
others put the blame on an irrational organization of communal kitchens.
However, as we have seen, some arguments are more persuasive than others
on the basis of logical reasoning, historical grounds, and not least analytical
evidence. To a large extent, there is perhaps a “consensus” that grain procure-
ment was a strong culprit. To this end, various attempts have thus been made to
understand possible exogenous variations – both economic (institutional rigidity
and excessive exporting, for instance) and political (career incentives) in the first
place. In furthering the analysis of the role played by the human factor in this
catastrophe, evidence has also been brought to bear on the suggestion that
political elites were not all sinister during these tumultuous times; those who
were not subject to the same strong incentives to climb the nomenklatura’s career
ladder as their regional counterparts were found to have exerted themselves to
ameliorate the food shortages faced by their “distant relatives.” As with many
other lesser known aspects of the Great Leap (and collective agriculture more
generally), more research is required to confirm these suggestive claims and to
reveal new findings.
As demonstrated by the works of Meng et al. and Kasahara and Li, research
on the Great Leap Famine has clearly benefited from disaggregated data at
the county level.144 We can thus expect more works of this nature in the
future. In the light of the still preliminary finding that the severity of famine
varied as much within a single province as between them (or, for that matter,
that there were variations involving subunits across provincial boundaries),
future work on the extent to which the severity of famine might have been
shaped by local variations in geography, history, economy, and politics
would certainly benefit from the availability of these fine-grained data.
144
Meng, Qian, and Yared, “The Institutional Causes of China’s Great Famine.” Kasahara
and Li, “Grain Exports and the Causes of China’s Great Famine.”
683

james kai-sing kung
Further Reading
Ashton, B., K. Hill, A. Piazza, and R. Zeitz, “Famine in China, 1958–61,” Population and
Development Review 10 (1984), 613–45.
Bai, Y., and J.K.S. Kung, “The Shaping of an Institutional Choice: Weather Shocks, the
Great Leap Famine, and Agricultural Decollectivization in China,” Explorations in
Economic History 54 (2014), 1–26.
Bernstein, T.P., “Mao Zedong and the Famine of 1959–1960: A Study in Wilfulness,” China
Quarterly 186 (2006), 421–45.
Bramall, C., “Agency and Famine in China’s Sichuan Province, 1958–1962,” China Quarterly
208 (2011), 990–1008.
Cao Shuji 曹树基, 1959–1961 年中國的人口死亡及其成因 (The Death Rate of China’s
Population and Its Contributing Factors from 1959–1961), 中国人口学 (Chinese
Population Science), 2005, 14–28.
Garnaut, A., “The Geography of the Great Leap Famine,” Modern China 40.3 (2014), 315–48.
Kasahara, H., and B. Li, “Grain Exports and the Causes of China’s Great Famine, 1959–
1961: County-Level Evidence,” Journal of Development Economics 146 (2020), 102513.
Kung, J.K.S., and S. Chen, “The Tragedy of the Nomenklatura: Career Incentives and
Political Radicalism during China’s Great Leap Famine,” American Political Science
Review 105.1 (2011), 27–45.
Kung, J.K.S., and J.Y. Lin, “The Causes of China’s Great Leap Famine, 1959–1961,” Economic
Development and Cultural Change 52.1 (2003), 51–73.
Kung, J.K.S., and T. Zhou, “Political Elites and Hometown Favoritism in Famine-Stricken
China,” Journal of Comparative Economics (2020), at www.sciencedirect.com/science/
article/pii/S0147596720300184?via%3Dihub.
Li, W., and D.T. Yang, “The Great Leap Forward: Anatomy of a Central Planning
Disaster,” Journal of Political Economy 113.4 (2005), 840–77.
Lin, J.Y., and D.T. Yang, “Food Availability, Entitlements and the Chinese Famine of
1959–61,” Economic Journal 110.460 (2000), 136–58.
Meng, X., N. Qian, and P. Yared, “The Institutional Causes of China’s Great Famine,
1959–1961,” Review of Economic Studies 82.4 (2015), 1568–1611.
Ó Gráda, C., “Great Leap, Great Famine: A Review Essay,” Population and Development
Review 39.2 (2013), 333–46.
Walker, K.R., “Food and Mortality in China during the Great Leap Forward, 1958–61,” in
R.F. Ash (ed.), Agricultural Development in China, 1949–1989: The Collected Papers of Kenneth
R. Walker (New York, Oxford University Press, 1998), pp. 106–47.
684

19
China’s External Economic Relations
during the Mao Era
amy king
When considering the question of China’s external economic relations dur-

ing the Mao era, the dominant narrative in the literature underscores the
following view: Mao’s China pursued a foreign economic policy that was
autarkic, isolated from the global economy, and locked into a Soviet-inspired
planned economy that provided limited incentives for economic inter-
dependence with the outside world. For some, China’s isolation from the
global economy was the result of its position in the Soviet bloc, which was
“heavily biased against foreign trade,” and from its adoption of a centrally
planned, socialist economic model that prohibited private interests from
pursuing foreign investment or trade.1 For others, Mao-era policies of autarky
were inspired by a form of xenophobia that stemmed from the country’s
experience of Western predations during the nineteenth century, resulting in
fear of economic dependence on foreign powers.2 Finally, others emphasize
the role of Mao’s revolutionary ideology in explaining China’s isolation from
the global economy; Mao’s tendency to view major international economic
institutions and norms as counterrevolutionary and “hostile” to the Chinese
state led him to disengage from international trade and other economic
opportunities.3
The volume of China’s external economic relations was undoubtedly
limited during the Mao era when compared to both the Republican (1912–
1949) and reform (1978–present) periods. During the Mao era, China did not
participate in international capital markets, did not permit private foreign
1
J.W. Garver, China’s Quest: The History of the Foreign Relations of the People’s Republic of
China (New York, Oxford University Press, 2016), pp. 367–8.
2
K. Reese, “Mainland China and Western Trade Credits,” Intereconomics 2 (1976), 45.
3
W. Hu, “China as a Listian Trading State: Interest, Power, and Economic Ideology,” in
G.J. Ikenberry, J. Wang, and F. Zhu (eds.), America, China, and the Struggle for World
Order: Ideas, Traditions, Historical Legacies, and Global Visions (New York, Palgrave, 2015),
p. 231.
685

amy king
direct investment in its economy, and pursued a government-led system of

foreign-exchange control that severely restricted the development of external
trade relations.4 Consequently, China under Mao was ranked thirtieth in the
list of the world’s top exporting countries, and its trade accounted for just 0.8
to 1 percent of world trade between 1952 and 1978. By comparison, during the
Republican era, Chinese trade accounted for approximately 2 percent of
world trade until the 1930s, when the effects of the Great Depression and
the War of Resistance against Japan took hold. Following Mao’s death in 1976
and the introduction of reform and opening under Deng Xiaoping in 1978,
China’s trade levels then increased exponentially (see Figure 19.1). By 1990,
China’s share of world trade had recovered to levels not seen since the late
1920s. By 2003, China accounted for 5.8 percent of world trade (Table 19.1).5
Yet China was not as closed off to the outside world during the Mao era as
these figures and dominant narratives suggest, particularly when viewed
within the context of other developing countries, and within the context of
its own tumultuous history of war. Within the first decade of the new
People’s Republic (1949–1959), Chinese foreign trade had grown faster than
trade growth experienced among other developing countries and all other
Asian countries. And while China’s membership of the Soviet bloc did shape
the country’s patterns of trade, particularly during the CCP’s first decade in
power, by the late 1960s more than two-thirds of China’s trade was taking
place with non-Communist countries, stretching across Asia, Africa, Europe,
the Americas, and Oceania.6 Characterizing China as “autarkic” or “isolated”
during the Mao era therefore obscures the ways in which the Chinese
economy was connected to the external world between 1949 and 1978.
China’s patterns of trade, its receipt of foreign investment and loans, and its
own distributions of foreign aid were shaped by a combination of five key
factors: its place in the Soviet bloc and its adoption of the Soviet economic
model, the Cold War geopolitical and economic order, the legacy of Japanese
empire and war with Japan, the impact of imperialism, and the CCP’s own
domestic political campaigns. Perhaps most importantly, though, the impact
of these factors was felt not only on the size and direction of China’s external
4
N.R. Lardy, China in the World Economy (Washington, DC, Institute for International
Economics, 1994), pp. 1–2.
5
W. Keller, B. Li, and C.H. Shiue, “China’s Foreign Trade: Perspectives from the Past 150
Years,” World Economy 34 (2011), 884; A. Maddison, Chinese Economic Performance in the
Long Run: 960–2030 A D (Paris, Development Centre of the Organisation for Economic
Co-operation and Development, 2007), pp. 55–60.
6
Y. Zhang, China in International Society since 1949: Alienation and Beyond (Basingstoke,
MacMillan Press in association with St. Antony’s College, Oxford, 1998), pp. 28–9.
686

China’s External Economic Relations during the Mao Era
Table 19.1 Chinese export levels, 1913–2003

Exports in Exports as % of GDP Exports in cur- Chinese exports as %
1990 prices in 1990 international rent prices of world exports in
US$ million US$ million US$ million current US$
1913 4,197 1.2 299 1.6

1929 6,262 2.3 660 2.0
1952 8,063 2.6 820 1.0
1978 15,639 1.7 9,750 0.8
1990 62,090 2.9 62,090 1.9
2003 453,734 7.1 438,230 5.8
Source: A. Maddison, Chinese Economic Performance in the Long Run: 960–2030 A D (Paris,
Development Centre of the Organisation for Economic Co-operation and
Development, 2007), p. 88
12,000
10,000
8,000
6,000
4,000
2,000
0
1950
1952
1954
1956
1958
1960
1962
1964
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
Exports Imports
Figure 19.1 Chinese total trade, 1950–2006 (in US$ million at current prices)
Source: State Statistical Bureau of the People’s Republic of China, Statistical Yearbook of
China (New York, Praeger, various years). All Statistical Yearbook of China current prices
used throughout this chapter are based on exchange rates issued by the People’s Bank of
China (1950–1978) and the Bank of China (1979–2006)
trade, loan, foreign aid, and investment ties, but also in CCP thinking
about China’s place in the global economy, and the role of foreign economic
policy in the country’s wider relations with the outside world. Chief among
CCP foreign economic policy ideas was the notion that economic
687

amy king
development – through rapid industrialization and access to advanced tech-

nology – was of critical import for “backward” countries who were attempt-
ing to catch up, re-establish economic and political sovereignty, secure
themselves in an era of industrialized military power, and obtain greater
power and influence on the world stage. Despite sweeping changes in the
nature of China’s engagement with the global economy since Mao’s death,
these core ideas have ongoing resonance in China today.
The Soviet Bloc and Adoption of the Planned

Economy
The Soviet Union was China’s most important economic partner during the
first decade of the new People’s Republic. Moreover, the Soviet economic
model’s influence shaped the Chinese economy long after the relationship
had been disrupted by the Sino-Soviet split. In the winter of 1949, senior CCP
leaders began discussing in earnest with their Soviet counterparts China’s
economic development goals. China’s economy had been ravaged by the
preceding years of war against Japan, by civil war between the Communists
and Nationalists, and by the Republican government’s adoption of Bretton
Woods trade and financial policies in the late 1940s. As a result, the country’s
industrial base had been badly eroded, agricultural production remained
lower than pre-1937 levels, and overly ambitious economic policies had
upended China’s trade balance with the United States and depleted the
country’s foreign-exchange reserves.7 Given these dire economic conditions,
Mao informed Anastas Mikoyan, the Soviet minister for foreign trade, in
February 1949 that “after the ending of the Civil War the main task of the
party will amount to economic construction.”8 However, Mao admitted that
the CCP had little experience in governing urban areas of China, and thus
would be heavily dependent on the Soviet Union for help in reconstructing
7
W.C. Kirby, “The Chinese War Economy,” in S.I. Levine and J.C. Hsiung (eds.), China’s
Bitter Victory: The War with Japan, 1937–1945 (New York, M.E. Sharpe, 1992), pp. 185–6; N.
R. Lardy, “Economic Recovery and the 1st Five-Year Plan,” in R. Macfarquhar and J.
K. Fairbank (eds.), Cambridge History of China, vol. 14, The People’s Republic, part 1, The
Emergence of Revolutionary China, 1949–1965 (Cambridge, Cambridge University Press,
1987), p. 144; for more on the impact of the KMT’s postwar trade and financial policies,
see the chapter by Grove and Kubo in this volume.
8
“Memorandum of Conversation between Anastas Mikoyan and Mao Zedong,”
February 5, 1949, History and Public Policy Program Digital Archive, APRF: F. 39, op.
1, D. 39, ll. 74–7. Reprinted in A. Ledovskii, R. Mirovitskaia, and V. Miasnikov, Sovetsko-
kitaiskie otnosheniia, vol. 5, book 2, 1946–fevral′ 1950 (Moscow, Pamiatniki Istoricheskoi
Mysli, 2005), pp. 78–80, trans. Sergey Radchenko (emphasis in original). See http://
digitalarchive.wilsoncenter.org/document/113323.
688

and developing large cities like Changchun, Anshan, Beijing, Tianjin, and
Shanghai.9 The Soviet Union, Mao reasoned, had long-standing economic
interests in northeast China, and in cities like Dalian “there is a whole array of
enterprises that we are in no position to exploit without Soviet assistance.”10
Even more fundamentally, the Soviet Union’s economic model – and its own
transformation from an economically “backward” country into a modern,
industrialized one – represented a powerful blueprint for the CCP. Soviet
economic ideology encapsulated a “scientific” approach to managing the
economy, and a central role for the state in both economic planning and
ownership of industry.11
Nevertheless, we should not overstate the influence of Soviet economic
ideology on Mao-era China; approaches to state-led industrial development
adopted by both the Guomindang and the Japanese in China during the 1930s
and 1940s laid the groundwork for the CCP to graft on Soviet plans for rapid
industrialization after 1949. As Morris Bian, Linda Grove, and Tōru Kubo
argue elsewhere in this volume, ideas about the necessity of state-led indus-
trial development took particular hold in China in the 1930s and 1940s as war
with Japan accelerated the need for Chinese military preparedness. In Free
China, the KMT government introduced a series of national plans for the
development of iron, steel, energy, chemicals, and heavy machinery to
underpin China’s military industries. In Japanese-controlled occupied China
and Manchuria, Japanese forces had similarly maintained strong state-led
approaches towards the development of heavy industry as a means of fueling
Japan’s war machine. The CCP was relatively late to understand the relation-
ship between heavy industry and national defense because the Party had
spent the years of war with Japan as a force fighting a guerrilla war based in
rural areas with little in the way of industry. But as the CCP took control of
China’s cities, and then confronted a new Cold War environment shaped
by war against the United States in Korea, Mao and his cadres quickly came
to understand the connection between economic – and particularly
industrial – development and the country’s ability to defend itself in
9
“Memorandum of Conversation between Anastas Mikoyan and Mao Zedong.”
10
“Record of Talks between I.V. Stalin and Chairman of the Central People’s
Government of the People’s Republic of China Mao Zedong,” January 22, 1950,
History and Public Policy Program Digital Archive, Archive of the President, Russian
Federation (APRF), f. 45, op. 1, d.329, ll. 29–38. Translated by D. Rozas. See http://
digitalarchive.wilsoncenter.org/document/111245.
11
W.C. Kirby, “China’s Internationalization in the Early People’s Republic: Dreams of
a Socialist World Economy,” China Quarterly 188 (2006), 876; for more on China’s
experience with the Soviet economic model, see Perkins’s chapter in this volume.
689

amy king
wartime.12 In his conversations with Mikoyan in early 1949, Mao reflected,

“In order to finally destroy the enemy, one should grow strong economic-
ally.” Moreover, acknowledging the legacy of Japanese and KMT-led eco-
nomic planning in the heavy industrial sector, Mao admitted that the
prospects for China’s industrial development were relatively favorable, par-
ticularly in northeast China, because nearly half of the industrial capital in
that region was already controlled by the state rather than by private
interests.13 Subsequently, throughout 1949 and early 1950, Mao, Zhou Enlai,
Liu Shaoqi, and other senior CCP officials held countless conversations with
Stalin and other Soviet counterparts to discuss the Soviet Union’s model of
industrial development, and the precise goods and technical assistance that
the CCP hoped to obtain from Moscow in order to roll out that model. By the
end of his otherwise difficult visit to Moscow in the winter of 1949–1950, Mao
had succeeded in negotiating an agreement for the Soviet Union to extend
a loan of US$300 million to China, which would help to finance China’s
import of industrial plants and equipment from the Soviet Union. The two
sides also reached agreement on the creation of a series of Sino-Soviet joint-
stock companies, which would manage railways, shipyards, mining, and
petroleum extraction across northeast China and Xinjiang, and agreed to
dispatch a large number of Soviet engineers and other technical experts to
assist with designing the construction of new industrial sites across China.14
Yet China’s signing of these trade and co-operation agreements with the
Soviet Union was never designed to preclude the development of Chinese
economic ties with countries outside the Soviet bloc. As part of their negoti-
ations with Moscow, CCP leaders explained their desire to take a relatively
relaxed approach towards foreign and private capital in China, particularly
during the Party’s first years of governing. Although the CCP had adopted
a policy at its Sixth Party Congress to confiscate foreign capital and national-
ize all private capital in China, Mao explained to Mikoyan that the CCP
needed to be “flexible” in implementing this policy. This was in part because
of the difficulty of disentangling foreign capital from Chinese capital, and in
part because the CCP was eager not to unnecessarily antagonize China’s
“bourgeoisie” immediately upon coming to power. Mao stressed to his
12
A. King, “‘Reconstructing China’: Japanese Technicians and Industrialization in the
Early People’s Republic of China,” Modern Asian Studies 50.1 (2016), 141–74.
13
“Memorandum of Conversation between Anastas Mikoyan and Mao Zedong,”
February 5, 1949.
14
A. Eckstein, Communist China’s Economic Growth and Foreign Trade (New York, McGraw-
Hill, 1966), pp. 138–9; Z. Shen and D. Li, After Leaning to One Side: China and Its Allies in
the Cold War (Washington, DC, Woodrow Wilson Center Press, 2011), p. 120.
690

Soviet counterparts that the Chinese economy would be socialist in charac-

ter, but that “we are not shouting about this so as not to scare someone
away.”15 This approach reflected the principles underpinning the CCP’s
postwar economic reconstruction policy of New Democracy, first adopted
in 1941 and maintained until the end of the Korean War in 1953. New
Democracy enabled the CCP to focus on rapidly rebuilding the Chinese
economy, and improving the country’s industrial base, while staving off
political opposition from rural landowners and other members of the “bour-
geoisie” by permitting a range of capitalist economic practices.16
At the same time, Moscow also counseled the CCP towards greater
openness to the non-Communist world. In January 1949, Mao explained to
the Soviet leadership that China’s foreign trade would take place primarily
with the Soviet Union and other members of the Soviet bloc, and that “only
those things, which [the USSR and the socialist bloc] do not need, will be
exported to capitalist countries.”17 However, over the spring and winter of
1949, Stalin and Mikoyan continually encouraged Mao to be willing to engage
in trade with capitalist countries under certain conditions. Stalin recognized
that such trade would likely be important for China’s economic develop-
ment, and could assist in reducing China’s economic dependence on the
Soviet Union.18 Just as importantly, such trade could also offer China a source
of political leverage in its dealings with the capitalist bloc, particularly on the
question of diplomatic recognition of the PRC. As Stalin explained to
a visiting CCP delegation in July 1949,
15
“Memorandum of Conversation between Anastas Mikoyan and Mao Zedong,” February 6,
1949, History and Public Policy Program Digital Archive, APRF: F. 39, op. 1, D. 39, ll. 78–88,
reprinted in Ledovskii, Mirovitskaia, and Miasnikov, Sovetsko-kitaiskie otnosheniia, vol. 5,
book 2, pp. 81–7. See http://digitalarchive.wilsoncenter.org/document/113352.
16
“Turning China into a Powerful, Modernized, Socialist, Industrialized Country,”
September 23, 1954, in Selected Works of Zhou Enlai, vol. 2 (Beijing, Foreign Languages
Press, 1989), p. 142.
17
“Memorandum of Conversation between Anastas Mikoyan and Mao Zedong,” January 31,
1949, History and Public Policy Program Digital Archive, APRF: F. 39, op. 1, D. 39, ll. 7–16,
reprinted in Ledovskii, Mirovitskaia, and Miasnikov, Sovetsko-kitaiskie otnosheniia, vol. 5,
book 2, pp. 37–43. See http://digitalarchive.wilsoncenter.org/document/112436.
18
“Cable, Filippov [Stalin] to Mao [via Kovalev],” April 19, 1949, History and Public Policy
Program Digital Archive, APRF: F. 45, op. 1, D. 331, ll. 24–5, reprinted in Ledovskii,
Mirovitskaia, and Miasnikov, Sovetsko-kitaiskie otnosheniia, vol. 5, book 2, pp. 120–1. See
http://digitalarchive.wilsoncenter.org/document/113356. “Cable, Stalin to Kovalev,”
April 26, 1949, History and Public Policy Program Digital Archive, APRF: F. 45, op. 1,
D. 3331 [sic, probably 331], L. 3, reprinted in Ledovskii, Mirovitskaia, and Miasnikov,
Sovetsko-kitaiskie otnosheniia, vol. 5, book 2, p. 126. See http://digitalarchive
.wilsoncenter.org/document/113357.
691

amy king
You have a really good magic weapon [fabao 法宝], which is that imperialism
wants to do business [maimai 买卖] with you. The imperialist countries’
economic crisis has already begun. I think the powers could quickly decide to
try and recognize you. You can first do good business with them and then
discuss the recognition issue.19
Stalin’s advice to Mao would significantly influence how the CCP
approached foreign trade during the 1950s and 1960s, although Stalin did
not foresee in 1949 that the onset of the Cold War would soon inhibit China’s
ability to trade with the “imperialist” world. Instead, over the course of the
1950s, the Soviet Union, and the Soviet bloc more generally, came to occupy
an increasingly important position in the Chinese economy. Unlike the rest of
the Soviet bloc economies, China never became a formal member of the
Council for Mutual Economic Assistance (COMECON).20 Rather, its trade
with the Soviet bloc was organized via a series of bilateral trade agreements,
modeled upon other trade agreements used by the COMECON members,
which it signed with the Soviet Union, Czechoslovakia, North Korea, East
Germany, Poland, Hungary, Bulgaria, Romania, Mongolia, Albania, and
North Vietnam between 1950 and 1954.21 The most important of these trade
partners was easily the Soviet Union, which became critically important in
aiding China’s postwar rebuilding and in giving China access to the industrial
goods it needed to make the leap from an agrarian to an industrialized
economy. China imported from the Soviet Union industrial plants and
equipment, machinery, petroleum and petroleum products, and metals and
metal ores. In return, China exported to the Soviet Union raw materials,
soybeans, foodstuffs, metals, textiles, and clothing. The latter two categories
of goods became increasingly important in China’s export profile as the
country rehabilitated its textile industry, and, by 1963, two-thirds of Chinese
exports to the Soviet Union comprised fabrics, clothing, and footwear. This
pattern of trade reflected China’s development strategy of rapid industrial-
ization, which underpinned the country’s First Five-Year Plan (1953–1957).
China was utterly dependent on access to Soviet capital goods to achieve this
rapid industrialization, particularly in the first decade of the Mao years. In
1950, machinery and other capital goods represented around 10 percent of
19
“Cable, Liu Shaoqi to Mao Zedong,” July 18, 1949, History and Public Policy Program
Digital Archive, 建国以来刘少奇文稿 (Liu Shaoqi’s Manuscripts since the Founding
of the PRC), vol. 1 (Beijing, Zhongyang wenxian chubanshe, 2005), pp. 30–7. Translated
for CWIHP by D. Wolff. See http://digitalarchive.wilsoncenter.org/document/113439.
20
Kirby, “China’s Internationalization in the Early People’s Republic,” 884.
21
Eckstein, Communist China’s Economic Growth and Foreign Trade, pp. 139–40.
692

China’s imports from the Soviet Union, but by the end of that decade capital
goods had swelled to more than 60 percent of Chinese imports from the
Soviet Union. Indeed, Alexander Eckstein notes that within the Communist
bloc, China absorbed more than half of the Soviet Union’s worldwide exports
of industrial goods and more than two-thirds of complete industrial plant.22
Pursuing the Soviet-inspired path of rapid industrialization required much
more than just access to Soviet imports, however. The Soviet Union’s
provision of industrial expertise to China, in the form of technicians, economic
advisers, engineers, and scientists, was an equally significant dimension of
China’s economic relationship with the Soviet bloc during the PRC’s first
decade. As Dwight Perkins argues elsewhere in this volume, China faced
a severe shortage of skilled engineers, scientists, and economists in its early
years. The country thus became highly dependent on the estimated 10,000 to
12,284 Soviet, and 1,500 Eastern European, technical experts who were sent to
China between 1949 and 1960.23 A further 10,000 Soviet military experts also
spent time in China in the 1950s, and many thousands of Chinese students
studied at Soviet universities and other training centers.24 Soviet bloc techni-
cians became a vital source of the specialist advice that China needed to
transform industrial goods into working plants, factories, and other industrial
sites. In May 1953, at the start of the First Five-Year Plan, for example, Li
Fuchun, then vice chairman of the PRC’s Central Economic and Financial
Commission, negotiated an agreement with Moscow to receive technology
and equipment to develop ninety-eight new Chinese industrial projects,
including coal mines, metal-processing factories, power plants, and oil
refineries.25 Yet without the simultaneous provision of Soviet technicians –
whose numbers peaked in 1956 – these industrial projects would never have
been realized. Together, Soviet industrial goods and expertise transformed
the Chinese economy and gave the Soviet Union a powerful means of
influencing China’s economic direction.26
Nevertheless, China’s economic relationship with the Soviet bloc posed
a range of challenges for CCP economic planners. The first was that Chinese
imports from the Soviet Union were relatively more important to China’s
22
23
Eckstein, Communist China’s Economic Growth and Foreign Trade, p. 169, Shen and Li, After
Leaning to One Side, p. 118.
24
Shen and Li, After Leaning to One Side, pp. 118, 127.
25
S.G. Zhang, Beijing’s Economic Statecraft during the Cold War, 1949–1991 (Washington, DC
and Baltimore, Woodrow Wilson Center Press and Johns Hopkins University Press,
2014), p. 61.
26
Shen and Li, After Leaning to One Side, pp. 117, 127.
693

amy king
economic planners than were the goods that China exported to the Soviet
Union. Between 1950 and 1955, as China was preparing for and carrying out its
First Five-Year Plan, Chinese imports from the Soviet Union increased more
rapidly than did the size of her exports to the Soviet Union (see Figure 19.2).
The structure of the Sino-Soviet trade agreement was such that it required
that trade be balanced on an annual basis, and China was simply unable to
export sufficient goods to the Soviet Union to meet its demands for Soviet
industrial imports. The result was a trade imbalance, which had to be
financed through a combination of increased Chinese exports and Soviet
loans.27 The trade imbalance was particularly acute during the Korean War
years (1950–1953), when China began importing both industrial goods and
military equipment from the Soviet Union. Paying for these imports required
China to expand dramatically its export of agricultural and raw-material
products, and to direct Soviet loans towards Korean War spending, which
occupied nearly 60 percent of China’s total budget.28 The Soviet Union was
generous in its provision of loans to China, which existed in the form of both
formal Sino-Soviet loan agreements – one of 1.2 billion yuan agreed to in 1950,
and another of 520 million rubles agreed to in 1954 – and a series of other ad
hoc loan arrangements. According to Feng-hwa Mah, Moscow provided
a total of 5.2 billion yuan in credits to China between 1950 and 1957.29
However, the generosity of these loans was somewhat undermined by the
fact that more than half of Moscow’s total 1950s loan package was diverted
towards financing China’s involvement in the Korean War.30 Moreover, the
terms of the Soviet loan agreements placed China in a highly precarious
position. The Soviet Union charged a low interest rate of just 1 percent on the
loans it extended to China, yet required that these loans be amortized over
a period of just ten years. This was a much shorter time frame than compar-
able international loans, and placed a very large debt-servicing burden on the
Chinese economy. Alexander Eckstein estimates that, over the course of the
1950s and 1960s, between 10 and 40 percent of China’s total export earnings in
27
Eckstein, Communist China’s Economic Growth and Foreign Trade, pp. 135, 154–5.
28
“From the Diary of N.V. Roshchin: Memorandum of Conversation with Chinese
Premier Zhou Enlai on 24 July 1951,” July 27, 1951, History and Public Policy Program
Digital Archive, AVPRF, f. 0100, op. 44, por. 13, pap. 322, ll. 44–57, translated by
D. Wolff. See http://digitalarchive.wilsoncenter.org/document/118735.
29
F.H. Mah, “Foreign Trade,” in A. Eckstein, W. Galenson, and T.C. Liu (eds.), Economic
Trends in Communist China (Chicago, Aldine, 1968), pp. 703, 728–9.
30
M. Nakajima, “Foreign Relations: From the Korean War to the Bandung Line,” in
Macfarquhar and Fairbank, Cambridge History of China, vol. 14, pp. 277–8.
694

1,200
1,000
800
600
400
200
0
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
Exports Imports
Figure 19.2 Chinese exports to and imports from the Soviet Union, 1950–1978 (in
US$ million at current prices)
Source: Statistical Yearbook of China (1981), p. 363
the Sino-Soviet trade relationship went towards its repayments of loans to the
Soviet Union.31
Second, as William Kirby has argued, the Soviet economic bloc displayed
inherent structural weaknesses that inevitably limited the development of trade
ties between Soviet bloc members. As planned economies, which did not use
freely floating prices or convertible currencies, Soviet bloc economies had to
engage in the painstaking process of continuous renegotiation of trade agree-
ments that met the requirements of both their foreign trade counterparts and
their own national economic plans. In addition, many of the countries within the
bloc were developing, war-torn countries, who sought to use the Soviet model
of rapid industrialization as a way to rebuild and transform their economies. The
result, however, was a considerable lack of diversity within the bloc; member
economies each focused on the development of heavy industry and thus found it
difficult to pursue trade on the principle of comparative advantage.32
Finally, though the Soviet economic model and Soviet industrial goods and
expertise provided significant economic benefits to China, membership of the
Soviet bloc separated China from economic partners that were more naturally
suited to China in terms of geography and historical familiarity. Located at some
distance from the Soviet Union and Eastern Europe, China’s exports and imports
to and from the Soviet bloc had to travel overland via the Trans-Siberian
Railway, which considerably added to the cost and duration of transport. One
31
Eckstein, Communist China’s Economic Growth and Foreign Trade, p. 161.
32
Kirby, “China’s Internationalization in the Early People’s Republic,” 882–7.
695

amy king
estimate suggests that geographical distance added an additional 30 percent to the

cost of China’s imports from the Soviet Union, when compared to the prices paid
by its Eastern European counterparts.33 At the same time, Soviet and Eastern
European goods and industrial advice were often deemed less suitable to Chinese
conditions than those from other historical economic partners, such as Japan. In
1949 and 1950, Chinese workers living in northeastern Chinese cities complained
that Soviet railway expertise paled in comparison to that possessed by the highly
proficient Japanese engineers who had been based in northeast China as part of
Japan’s colonial project since the 1930s.34 By comparison, Soviet experts were
provided with little or no understanding of or training in China’s history, culture,
or economic conditions, and typically had little in the way of Chinese language
skills. Rather, it seems that the primary factor shaping their selection to travel to
China was that they had the correct political background and displayed loyalty to
the Communist Party of the Soviet Union.35 Unsurprisingly, their resulting
experiences in China were rarely positive ones. The lack of language skills and
familiarity with the local Chinese context, coupled with China’s own lack of
technical expertise, led to poor communication between Soviet and Chinese
workers, and delays in applying the Soviet Union’s technical experience to the
Chinese context. As the 1950s progressed, Chinese officials began to complain
that Soviet and European imports were costly and that their economic planning
advice was not always well suited to Chinese conditions. In 1957, the CCP elected
to reduce its overall expenditure on technical advice from the Soviet Union, and
instead to draw on “fewer but better” Soviet experts.36 The following year, as the
number of Soviet bloc experts in China began to decline, Chinese foreign
minister Chen Yi began looking elsewhere for expertise, and in particular to
Japan. As Chen explained in a meeting with the Japanese president of the Japan–
China Import–Export Association,
China has a saying: “neighbours are dearer than distant relatives.” We have
built a relationship with Europe, but the road is far and it costs a lot in time
and money to reach, so it is not worthwhile. Japan is only separated by a strip
of water, and so it is relatively convenient . . . With respect to planning,
European experts’ plans are not compatible with, and do not suit, China’s
situation. Japanese experts’ plans would be relatively more suitable to us.37
33
34
King, “Reconstructing China,” 142–3. 35 Shen and Li, After Leaning to One Side, p. 123.
36
Shen and Li, After Leaning to One Side, pp. 129–31.
37
PRC Foreign Ministry Archive, File 105–00595-03, February 7, 1958, “陈毅副总理接见
日本中日输出入组合理事长南乡三郎谈话记录” (Records of the Meeting between
Foreign Minister Chen Yi and Japan’s China–Japan Export–Import Association
President, Nangō Saburō), pp. 2, 11–12.
696

It is difficult to overestimate the role that the Soviet Union played in aiding
China’s economic reconstruction, and facilitating China’s external economic
ties to the Soviet bloc, during the PRC’s first, crucial, decade in power.
Nevertheless, there were problems baked into the Sino-Soviet economic
relationship from the start, not the least of which was the significant devel-
opment gap that existed between China and her Soviet “elder brother.”38
This gap made the economic relationship a complementary one, but also an
unbalanced one, and by the end of the 1950s CCP leaders were increasingly
resentful of China’s structural dependence upon the Soviet Union. That
resentment would play an important role in catalyzing the Sino-Soviet split
and China’s eventual turn to the non-Communist world, as we shall see in
more detail in the sections that follow, although economic factors alone are
certainly not sufficient to explain the breakdown of the Sino-Soviet
relationship.39 Ultimately, the Soviet Union was the PRC’s first and most
important economic partner, and the positive and negative aspects of this
relationship would have a profound effect both on the pattern of China’s
external economic relations and on the ideas underpinning the CCP’s foreign
economic policy.
Cold War Geopolitical and Economic Blocs

The second critical factor shaping the pattern of China’s external economic
relations, particularly during the first decade of the PRC’s existence, was Cold
War competition between the United States and Soviet Union, and their
creation of rival economic blocs. If the Soviet Union was to define one half of
the geopolitical and economic blocs that shaped the global economy during
the Cold War, the United States was equally important in defining the other
half. However, the impact that the unfolding Cold War would have on
China’s patterns of trade was not immediately apparent. The United States
and key allies in Europe established a set of trade controls on Soviet bloc
countries in the winter of 1949–1950, which limited the export of goods that
could aid the development of their military and industrial capabilities.
Although China was initially included within this set of trade restrictions,
the United States remained China’s second most important trade partner,
behind the Soviet Union, until the end of 1950. As Rosemary Foot makes
clear, the US Truman administration elected not to cut completely its trade
38
R. Terrill, “China and the World: Self-Reliance or Interdependence?”, Foreign Affairs 55.2
(1977), 300–1.
39
697

amy king
with the new Communist nation because it knew that such a policy would
enjoy little support from among US allies, and because it recognized that
maintaining US trade with China might have the effect of limiting Soviet
influence on China.40
However, all this was to change in October 1950 when the CCP made the
decision to send forces across the Yalu river into Korea to aid North Korean
forces who had invaded the southern half of the Korean peninsula in July
that year. China’s involvement in the Korean War, and the failed efforts by
both the US and China to deter one another’s involvement in that war,
fundamentally transformed the political, military and economic relationship
between China, the United States, and US allies. The United States introduced
trade controls on China that were far stricter than those applied to the rest of the
Soviet bloc. The general Soviet bloc controls entailed a complete embargo on
a first category of goods that could contribute directly to a country’s military
capabilities, but permitted limited trade in a second category of industrial goods,
and free trade in a third category of goods, provided that those goods were
reported to a central organizing body known as COCOM.41 By comparison, the
export controls applied to China were far more restrictive. The United States not
only completely banned all American trade and shipping with China, but also
demanded that its allies block exports to China of goods in all three categories,
and created a further list of 200 goods that US allies should also embargo.
Ultimately, the US-led trade embargo on China was designed to limit access to
the goods and industrial technology that China needed to aid its Korean war
effort and its own economic development. As such, the US economic sanctions
became an integral part of the US containment strategy in Asia.42
The result of the trade embargo was an abrupt and distinct shift in the
pattern of China’s external economic relations. As depicted in Table 19.2, in 1950
the PRC’s leading trade partners included countries in the Western bloc such as
the United States and Great Britain, while others such as West Germany and
Canada enjoyed smaller but still significant trade ties with the PRC. However,
by 1952, the United States and its allies had experienced a major drop in their
trade with the PRC. The Cold War geopolitical order had a clear impact upon
the direction and composition of China’s trade.43 Throughout the 1950s, China’s
40
R. Foot, The Practice of Power: US Relations with China since 1949 (Oxford, Oxford
41
Foot, The Practice of Power, 52.
42
S.G. Zhang, Economic Cold War: America’s Embargo against China and the Sino-Soviet
Alliance, 1949–1963 (Washington, DC, Woodrow Wilson Center Press, 2001), p. 20.
43
698

trade with the Soviet bloc would occupy around two-thirds of China’s total
trade, while trade with the non-Communist world comprised the remaining
one-third (Figure 19.3). Moreover, because of strict controls on the export of
“strategic” military and industrial goods to China, the composition of goods
traveling between China and the non-Communist world was also quite differ-
ent to that between China and the Soviet bloc. Most non-Communist countries
could only export to China light manufactured consumer goods, foodstuffs,
paper, rubber, and chemical fertilizers, and what little machinery and equip-
ment was permitted under the US-led embargo. In exchange, China exported
agricultural goods, textiles, and iron and steel products to non-Communist
countries.44 Outside the Soviet bloc, Hong Kong represented the most import-
ant market for Chinese exports in the non-Communist world, particularly in
the early years of the PRC. China was a crucial source of food for Hong Kong,
while a large proportion of Chinese exports were rerouted from Hong Kong on
to third countries. The combination of Hong Kong’s extensive expertise in
banking, insurance, and shipping, and the fact that dozens of non-Communist
countries could more easily house trade officials and consulate staff in
Hong Kong than on the mainland, allowed Hong Kong to facilitate China’s
trade with the rest of the world.45
Table 19.2 China’s five largest trading partners, 1950–1952 (as % of China’s
total trade)
1950 1951 1952
Ranking Country % Country % Country %
1 USSR 30.0 USSR 41.3 USSR 54.9

2 United States 21.1 Hong Kong and 31.8 Hong Kong and 15.7
Macau Macau
3 Hong Kong 14.5 Czechoslovakia 3.6 German 5.5
and Macau Democratic
Republic
4 United 6.5 German 3.0 Czechoslovakia 5.0
Kingdom Democratic
Republic
5 Malaysia 5.4 Poland 2.6 Pakistan 4.2
Source: Statistical Yearbook of China (1981), pp. 359–71
44
45
699

amy king
80.00
70.00
60.00
50.00
40.00
30.00
20.00
10.00
0.00
1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965
Trade with non-Communist countries
Trade with Communist countries

Figure 19.3 China’s trade with Communist and non-Communist countries, 1950–1965
(percentage of China’s total trade)
Source: R.L. Price, “International Trade of Communist China, 1950–65,” in Joint Economic
Committee of the U.S. Congress (ed.), An Economic Profile of Mainland China (New York,
Frederick A. Praeger, 1968), p. 584
Beyond the makeup of China’s trading partners, the US-led embargo also
impacted CCP thinking about economic statecraft in two key ways. First, their
experience of economic sanctions helped the CCP to develop a greater under-
standing about the vulnerability of China’s economy to Cold War geopolitical
“shocks.”46 CCP leaders had not anticipated the US-led trade embargo and, when
it came, were surprised by how quickly the embargo disrupted the Chinese
economy. Indeed, the CCP had hoped that war in Korea might actually benefit
the Chinese economy by creating new opportunities to export strategic materials
such as tung oil, silk, and animal hides. The CCP was quickly disavowed of this
belief with the introduction of sanctions that made it almost impossible for US or
European firms and trading companies to ship to China, sign new trade deals
with China, or obtain letters of credit in US currency. Furthermore, without
access to new inflows of foreign currency, the CCP had to scramble to protect
the existing foreign currency reserves it possessed, and to resort to barter trade.47
The introduction of controls on the export of strategic goods also jeopardized
China’s ability to develop light and heavy industry. The CCP needed access to
machinery, construction supplies, and replacement parts to repair and rebuild
industrial sites around the country. Chinese officials had previously relied heavily
46 47
Zhang, Economic Cold War, pp. 83–93. Zhang, Economic Cold War, pp. 83–4, 90–1.
700

on the US, Japan, and Great Britain for access to these industrial goods.
However, given the strategic nature of most of China’s desired imports, the
US embargo prohibited American, Japanese, and British companies from signing
contracts for the export of these goods to the PRC, and pushed China even more
firmly in the direction of the Soviet camp to source strategic goods.
The second lesson Beijing learned through this experience was the instru-
mental value of economic statecraft. Beginning in the early 1950s, China
quickly responded to the US-led trade embargo by promoting trade with US
allies. In a way, China was mimicking the United States’ approach of using
economic policy as a tool to achieve its wider foreign policy and national
security objectives. However, unlike the US approach, which emphasized
denying trade with adversaries, China instead sought to use economic
inducements to achieve its foreign policy and security goals. In response to
the US-led embargo, the CCP deployed a strategy of economic inducement
and political persuasion to try to “drive a wedge” between the United States
and its allies, and ultimately disrupt the multilateral system of economic
sanctions against China.48 In September 1952, the Chinese Foreign Ministry
agreed to allow the newly established Chinese Committee for the Promotion
of International Trade (CCPIT) to provide Chinese industrial data to certain
imperialist countries, such as Britain and Japan, in order to boost bilateral
trade and obtain badly needed goods.49 Similarly, in April 1952, CCP officials –
under the leadership of Nan Hanchen, governor of the People’s Bank of
China – participated in an International Economic Conference in Moscow
that was designed to build trade between the socialist and capitalist blocs.
Nan lobbied delegations from non-Communist countries in Asia and Europe
to attend the conference, and in Moscow signed contracts with a host of
Western European companies to begin exporting goods to China.50
By the mid-1950s, China was carrying out a centralized campaign to
promote trade with US allies, many of whom were equally enthusiastic
about reinstating the lucrative China trade relationship. In 1954, during an
address to a delegation of British Labour Party members visiting China, Zhou
Enlai noted that countries such as Britain and France were desperate for
economic development and saw China’s population of 600 million as a vast,
48
Zhang, Economic Cold War, p. 46.
49
A. King, China–Japan Relations after World War Two: Empire, Industry and War, 1949–1971
(Cambridge, Cambridge University Press, 2016), pp. 84–5.
50
King, China–Japan Relations after World War Two; V. Zanier, “‘Energizing’ Relations:
Western European Industrialists and China’s Dream of Self-Reliance. The Case of Ente
Nazionale Idrocarburi (1956–1965),” Modern Asian Studies 51.1 (2017), 136–7.
701

amy king
untapped export market. For Zhou, trade ties were a vital “battle line” that China
could use to “improve its relations with Western countries.”51 Indeed, the issue
of economic sanctions became a source of great friction in the United States’
relationships with key allies such as Britain and Japan. The British Labour Party
and other British business groups lobbied hard for the relaxation of trade controls
against China which, they argued, were hurting British business interests.52
Similarly, in Japan, a coalition of business, the left-wing opposition, and politi-
cians from the ruling conservative Liberal Democratic Party joined forces to put
pressure on the Japanese government and the United States to allow Japan to
import from China low-cost raw materials and foodstuffs such as soybean, iron
ore, and coal.53 This pressure continued to mount throughout the mid-1950s, and
in 1957, unable to persuade its allies to maintain the tough line on China, the
United States watched as Britain, Japan, and a host of its other European allies
took the unilateral decision to relax partially their sanctions on China. Although
the United States would maintain strict sanctions against China until the 1970s,
the reintroduction of trade with US allies transformed China’s trade patterns. As
Figure 19.3 indicates, the proportion of China’s trade with the non-Communist
world increased steadily from 1959, and by 1963 the majority of China’s trade was
taking place with non-Communist countries. That year and thereafter, Western
European countries and Japan became regular exporters to China of industrial
equipment, machinery, and whole plants, and began eagerly hosting trade
exhibitions in China to encourage yet further trade.54
The Legacy of Japanese Empire and War with Japan

Thus far we have seen how China’s external economic relations during the
Mao era were shaped in large part by the CCP’s relationship with the Soviet
Union, and by the Cold War geopolitics that characterized China’s external
environment after 1949. Yet, throughout the Mao era, China pursued an
economic relationship with one country, Japan, that does not fit neatly the
ideological or geopolitical categories that we typically use to analyze the
CCP’s foreign economic policies. China’s economic relationship with Japan
between 1949 and 1978 was the legacy of a half-century of trade, investment,
51
“推进中英关系,争取和平合作” (Driving Forward Sino-British Relations, Striving for
Peaceful Co-operation), August 12, 1954, in 周恩来外交文选 (Selected Works on the
Diplomacy of Zhou Enlai) (Beijing, Zhongyangwenxian chubanshe, 1990), p. 81.
52
Zhang, Beijing’s Economic Statecraft, pp. 53–4. 53 King, China–Japan Relations, p. 102.
54
C. MacDougall, “Eight Plants for Peking,” Far Eastern Economic Review 23 (January 1964),
155–8; Zanier, “‘Energizing’ Relations,” pp. 165–8.
702

and people-to-people ties that had developed between the two countries on
the back of Japan’s colonial presence in China during the first half of the
twentieth century. That these economic ties not only endured but flourished
after 1949 – despite Japan’s brutal eight-year war with China, in the face of the
CCP’s deeply entrenched anti-imperialist ideology, and in the absence of
a diplomatic relationship between the two countries – is testament not only
to the deep economic complementarity between China and Japan, but more
importantly to the way in which Japan’s leap from agrarian to industrialized
nation shaped CCP thinking about its own development path.55
The large-scale, modern economic relationship between China and Japan
originated in the early 1900s, following Japan’s surprise victory over the Qing
Chinese government in the first Sino-Japanese War (1894–1895). Japan’s defeat of
China enabled the imperial Japanese government to compel China to sign the
Treaty of Shimonoseki, which granted most-favoured-nation status to Japanese
firms, citizens, vessels, and property in their dealings with China.56 Within
a decade of the war’s end, the trade relationship with Japan had grown to
15 percent of China’s total trade; Japan imported from China raw materials
such as coal and soybean, and exported to China textiles, metals, and
machinery.57 Not only did war between China and Japan help to facilitate this
flourishing trade relationship, but the war – and China’s shock defeat by its
smaller neighbour – actually encouraged reformers within the Qing government
to look to Japan as a source of ideas about how to modernize their country’s
military, government, education system, and industry. In the first decade of the
twentieth century, China sent some 25,000 students to study in Japan. Over the
next three decades, these students would bring home Japanese-inspired models
for trade exhibitions, industrial schools, and manufacturing centers.58 Yet a much
55
King, China–Japan Relations.
56
The granting of “most-favoured-nation (MFN) status” to foreign countries engaged in
trade with China first took place via the 1843 Treaty of Nanking, as a concession
obtained following Britain’s defeat of Qing China in the First Opium War (1839–
1842), and was later extended to the United States via the 1844 Treaty of Wangxia.
Japan had sought MFN status with China since 1870, but was not able to obtain such
rights until its victory over China in 1895. See S. Murase, “The Most-Favored-Nation
Treatment in Japan’s Treaty Practice during the Period 1854–1905,” American Journal of
International Law 70 (1976), 284–8.
57
Y.K. Cheng, Foreign Trade and Industrial Development of China: An Historical and Integrated
Analysis through 1948 (Washington, DC, Washington University Press, 1956), p. 18.
58
J. Lee, “Where Imperialism Could Not Reach: Chinese Industrial Policy and Japan,
1900–1940,” Enterprise and Society 15.4 (2014), 655–71; D. Reynolds, China, 1898–1912: The
Xinzheng Revolution and Japan (Cambridge, MA, Harvard University Press, 1993), p. 42;
M. Jansen, “Japan and the Chinese Revolution of 1911,” in J.K. Fairbank and K.C. Liu
703

amy king
darker strain of thinking also traveled from Japan to China during the first half of
the twentieth century. Inspired by the lessons of Germany’s defeat in World War
I, the imperial Japanese government grew increasingly concerned about their
country’s lack of natural resources and late push for industralization, and thus
sought to expand Japan’s colonial footprint in Northeast Asia. Manchuria, where
Japan had enjoyed rights to the Guandong peninsula since its victory in the
Russo-Japanese War (1904–1905), was now viewed by Japan as a rich source of
industrial and agricultural raw materials and as a future base for Japan’s heavy
industrial production. In 1932 the imperial Japanese government formally estab-
lished the Japanese-controlled state of “Manchukuo,” and over the next decade
would implement a strategy of state-controlled heavy industrialization and
agricultural development, and the creation of a market for Japanese exports.
Japan’s colonial presence in Manchuria bestowed important legacies for
China, which would influence CCP thinking and economic ties with Japan
throughout the Mao era.59 As discussed above, Manchuria served as an
important testing ground for Japan’s experiments with an economic planning
model that linked heavy industrialization with military modernization. The
heavy industries Japan left behind in Manchuria would become a valuable
foundation for the CCP as it pursued its own industrialization drive in
northeast China after 1949. Manchuria also became a critical meeting point
for Japanese and Chinese individuals who would play a role in helping to
build the China–Japan economic relationship during the Mao era.
Throughout the 1930s and early 1940s, Japanese technicians, engineers, bur-
eaucrats, and captains of industry helped to develop Manchuria’s economic
plans, industrial sites, and trade ties with Japan. After 1949, Japanese such as
Takasaki Tatsunosuke, who had served as vice president of Manchurian
Heavy Industries and the Anshan Iron and Steel Works, became an important
advocate for reviving Japan’s trade relationship with the PRC. In 1949–1950,
Takasaki provided the Chinese Foreign Ministry with the names of Japanese
industrial firms and price lists for goods that could be used to help reconstruct
northeast China, and in 1955 Takasaki – who had become the director of
Japan’s postwar Economic Planning Agency – served as an important voice in
the Japanese domestic debate pushing for the relaxation of Japan’s Cold War-
era economic sanctions on China.60 Similarly, Chinese such as Sun Pinghua,
who served in Manchuria’s Ministry of Economics during the 1940s, would go
on to play an influential role in developing the postwar economic
59
This paragraph draws on King, China–Japan Relations, Chapter 2.
60
C.W. Braddick, Japan and the Sino-Soviet Alliance, 1950–1964: In the Shadow of the Monolith
(Oxford, Palgrave Macmillan, 2004), p. 116.
704

relationship between the PRC and Japan. In 1955, Sun became the deputy
head of the PRC’s first trade delegation to Japan, which was instructed by the
Chinese Foreign Ministry and Foreign Trade Ministry to lobby Japanese
companies, politicians, and industrial experts to find ways to expand the
bilateral economic relationship. Finally, the CCP made good use of the
approximately 20,000 Japanese technicians and other skilled workers who
were left behind in northeast China at the end of World War I I. Between 1949
and 1953 these technicians were put to work across mines, research laborator-
ies, factories, and industrial sites in the northeast, and called on to train less-
skilled Chinese workers as the CCP endeavored to rebuild key industrial sites
in northeast cities such as Anshan and Dalian.61
Throughout the 1950s, the size of the China–Japan economic relation-
ship remained modest and was significantly outranked by China’s trade
with the Soviet bloc (Figure 19.4). This was due in large part to the Cold
War economic embargo, which severely restricted Japanese exports of
“strategic goods” to China. Nevertheless, between 1952 and 1962 the two
countries worked creatively to negotiate four rounds of “unofficial” trade
agreements, which were organized around different classes of goods that
Japan was permitted to export to China. The trade agreements were
“unofficial” because the two countries did not enjoy a diplomatic rela-
tionship – Japan recognized the Republic of China as the legitimate
government of China – and thus trade was negotiated between committees,
politicians, and business representatives on the Japanese side who could
plausibly describe themselves as one step removed from the ruling
government.62 No longer able to export heavy industrial goods or machinery,
Japan’s chief exports to China became agricultural machinery, textile machin-
ery, and light manufactured products such as radios and typewriters, and
increasingly chemical fertilizer, as the CCP pursued increasingly ambitious
agricultural targets in the lead-up to the Great Leap Forward. While the
bilateral trade relationship became a short-term casualty of the Great Leap
Forward, as we shall see below, by 1965 Japan had replaced the Soviet Union as
China’s largest trade partner.
Of course, the effects of Japanese colonialism and its aggressive war with
China also bestowed a powerful legacy on foreign economic policy thinking
61
King, “Reconstructing China.”
62
The major Japanese committees included the Japan Association for the Promotion of
International Trade and the China–Japan Trade Promotion Diet Members’ League. On
the Chinese side, the agreements were negotiated via the CCPIT or by representatives
from the Chinese Foreign Ministry.
705

6,000
5,000
4,000
3,000
2,000
1,000
0
1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978

USSR Japan Hong Kong and Macau United Kingdom Federal Republic of Germany
Figure 19.4 China’s bilateral trade with leading trade partners, 1950–1978 (in US$ million at current prices)
in Mao’s China. China’s defeat by a more industrially developed Japan in

the First and Second Sino-Japanese Wars underscored for successive
generations of Chinese leaders and reformers the importance of industri-
alization as a way to stave off colonial predation by more powerful
countries. As the Cold War unfolded and the PRC faced new wars in
Korea and Indochina, CCP leaders began reflecting on the history of
Japan’s industrialization process. Though the CCP was determined not
to follow Japan’s path in becoming a colonial power, Mao frequently
reflected on the “great benefits” that industrialization had brought to
Japan, and the valuable connections Japan had fostered between the
military and industrial halves of its economy.63 Moreover, as the only
Asian country to have made the leap from agrarian to industrialized
country, Japan served as a powerful exemplar for the CCP. The Soviet
Union’s economic model provided the blueprint for China’s centrally
planned economic system, but as Mao and other CCP officials so often
stated, Japan represented what it was possible for an “Eastern country”
(dongfang guojia 东方国家) to achieve. As Sino-Soviet relations soured in
the late 1950s and early 1960s, it was Japan’s development path, Japanese
industrial goods and technology, and Japanese industrial advisers that
increasingly provided the CCP with a useful basis for comparison.64
Finally, the CCP’s pursuit of an economic relationship with Japan, in which
Japan served as the more advanced, technologically developed partner, shaped
CCP thinking about questions of “self-reliance” (zili gengsheng 自力更生) in its
external economic relations. Over the course of the 1960s and early 1970s, Japan
became an increasingly important industrial partner for China, such that, by
1971, Japan supplied more than 60 percent of China’s total industrial plants and
advanced technologies.65 Nevertheless, this dependence on Japanese plants and
technology met the CCP’s definition of “self-reliance” for two reasons. First, as
Zhou explained in 1962, “China is emphasizing the approach of self-reliance,
but this certainly does not mean that we have to produce everything ourselves
to achieve self-reliance. In order to increase the speed [of development] we
63
“在关于日本经济政策和国防工业发展问题的一份材料上的批语” (Written
Comments on Materials Concerning Japan’s Economic Policies and Military
Industrial Development), July 16, 1961, in 建国以来毛泽东文稿 (Selected
Manuscripts of Mao Zedong since the Founding of the Republic), vol. 9
(Beijing, Central Documents Press, 1996), pp. 530–1.
64
King, China–Japan Relations, Chapter 6.
65
Y. Yokoi, “Plant and Technology Contracts and the Changing Pattern of Economic
Interdependence between China and Japan to 1989,” in C. Howe (ed.), China and Japan:
History, Trends, and Prospects (Oxford, Oxford University Press, 1996), p. 130.
707

amy king
ought to buy what we need.”66 Second, in pursuing the contracts for these
plants, the CCP cleverly negotiated deals that ensured that the technol-
ogy behind the plant equipment would be transferred to China, thereby
helping to pave the way for China’s eventual indigenous production of
industrial technology. So determined were Chinese negotiators in this
strategy, one observer argues, that “the level of technology and the
quality of the plant are usually of greater concern to the Chinese than
the final price.”67
The Legacy of Imperialism

The fourth factor to shape the pattern of China’s external economic relations
during the Mao era was China’s experience of semicolonialism during the late
nineteenth and early twentieth centuries. China’s experience of imperialism
impacted upon Mao-era thinking about economics, dependence, and national
security. CCP discourse on the exploitative nature of the international
economic order became particularly prominent in the mid-1950s, as China
began to associate much more closely with the experiences of other newly
independent countries, and influenced Beijing’s efforts to build trade and aid
relationships with them. In April 1955, the PRC attended the Asia–Africa
Conference, in Bandung, Indonesia, along with a group of twenty-nine
other newly independent countries from Asia and Africa. The “Bandung
conference,” as it came to be known, represented the first time that
a group of Asian and African countries had met in an international forum,
independent of the Western and colonial powers. At the Bandung conference
and in its aftermath, the CCP began to articulate new ideas about decolon-
ization as a key driver of world revolution, and that the imperialist inter-
national economic order had jeopardized the security of “backward” states
like China.68 In its place, CCP leaders called for a new international economic
order founded upon “equality and mutual benefit” (pingdenghuhui 平等互惠).
This concept had developed through conversations between Chinese prem-
ier Zhou Enlai, Indian prime minister Jawaharlal Nehru, and Burmese
president U Nu in 1954 and 1955, and became a key plank in the “Five
66
Quoted in Morizumi Kazuhiro 盛純和弘, 五十年の変遷―孙平化氏に聞く(Fifty
Years of Change: Interviews with Mr Sun Pinghua) (Beijing, Jinri zhongguo chubanshe,
1995), p. 157.
67
Reese, “Mainland China and Western Trade Credits,” 49.
68
On the former point, see J. Chen, “Bridging Revolution and Decolonization: The
‘Bandung Discourse’ in China’s Early Cold War Experience,” Chinese Historical
Review 15.2 (2008), 238.
708

Principles of Peaceful Co-existence” adopted at the Bandung conference. In

his speech to fellow countries at Bandung, Zhou Enlai, who led the Chinese
delegation, argued that “countries whether big or small, strong or weak,
should all enjoy equal rights in international relations,” including in the realm
of foreign trade.69 Yet Zhou continued with a warning: the “backward”
countries of Asia and Africa would not be able to enjoy equal or mutually
beneficial trade relations with the Western powers until they had escaped
from the “shackles of colonialism.”70
Zhou Enlai’s speeches at Bandung and his reports back to Mao Zedong
following the conference demonstrate that the CCP saw a clear link between
a country’s overall level of economic development and its ability to withstand
imperialist exploitation by more advanced foreign powers. But extolling the
virtues of economic development and mutually beneficial trading relations at
Bandung also represented the first step in a long-term strategy by the PRC to
“win over” countries in Asia and Africa, “dilute their relationships” with the
“American imperialists,” and obtain diplomatic recognition from those coun-
tries that did not yet recognize the PRC as the legitimate government of
China.71 In their preparations for Bandung, officials in the Ministry of Foreign
Affairs and Ministry of Foreign Trade argued that China could achieve these
goals by expanding “equal and co-operative” economic ties with Asian and
African countries, by warning them of US attempts to “manipulate
and monopolize” their markets and depress global prices for raw materials,
and by using the economic successes of China’s first five years of develop-
ment as a positive exemplar for other newly independent countries.72 Thus,
at Bandung, Zhou and his fellow delegates advocated a host of new measures
designed to help Asian and African countries move up the value chain and
become more than just agrarian exporters of raw materials.73 They called on
participating countries to diversify trade markets rather than relying on
former colonial masters to buy up raw material exports, break down old
colonial patterns of trade by sending representatives to each other’s coun-
tries, and create a permanent regional trade institution to help build and
69
“Zhou’s Speech to the Plenary Session of the Asian–African Conference,” April 19, 1955,
in China and the Asian–African Conference (Documents) (Beijing, Foreign Languages Press,
1955), p. 14.
70
“Zhou’s Speech to the Plenary Session of the Asian–African Conference,” p. 10.
71
“我参加亚非会议贸易活动方案(草案)” (Draft of China’s Plan for Its Participation in
the Trade Activities of the Asia–Africa Conference), March 8, 1955, PRC Foreign
Ministry Archive, File No. 207–00070-01.
72
“我参加亚非会议贸易活动方案(草案).”
73
“我参加亚非会议贸易活动方案(草案),” p. 17; “Final Communique of the Asian–
African Conference,” in China and the Asian–African Conference, pp. 68–71.
709

amy king
organize Asian–African trade ties.74 The Chinese delegation also endorsed the
idea of creating a long-term buy-and-sell agreement, or some other regional
price stabilization method, that would both stabilize and raise global prices
for raw materials.75
In advocating these measures, PRC officials made clear their view that
foreign trade and international economic ties were an essential characteristic
of the modern international system. It is therefore notable that their call for
“mutually beneficial” and “equal” trade relations between countries did not
equate to autarky or cutting oneself off from the global trading system.76
Instead, as Zhou argued in 1956, no country – not even the “large developed
countries” such as the United States and the Soviet Union – could be self-
reliant; instead, all were required to import certain “essential things.”77 Yet
despite their efforts to use the Bandung conference to catalyze
a transformation in regional trade practices, the PRC delegation had little
success in achieving many practical outcomes. A combination of countries’
ongoing economic dependence on their former colonial powers and new
Cold War geopolitical alignments made it very difficult to overcome existing
patterns of trade and foreign aid. Countries such as the Philippines, Thailand,
and Turkey, all of whom were US or NATO allies, believed that they “must
rely” on the United States for access to capital and technology, and were
hesitant to create a new permanent economic institution that might stand in
opposition to existing groupings such as the United Nations. Similarly, Zhou
believed that the nonaligned states such as India and Indonesia had opposed
China’s proposal to create an intra-regional bank or regional trade bloc
because they continued to feel “mostly reliant” on countries outside the
region for foreign aid, capital, and technology. Thus Zhou concluded that
even if fellow Asian and African nations had succeeded in achieving formal
political independence, they had yet to achieve real economic
independence.78
74
“周恩来关于经济合作问题致中共中央并毛泽东的报告” (Zhou Enlai’s Report to
Mao Zedong and the CCP Central Committee on the Problem of Economic Co-
operation), April 30, 1955, in 中华人民共和国外交档案选编(第二集)中国代表团
出席 1955 年亚非会议 (Selected Works from the Foreign Ministry Archive of the
People’s Republic of China (vol. 2): China’s Delegation to the 1955 Asia–Africa
Conference) (Beijing, Zhongguo tongji chubanshe, 2007), pp. 90–3.
75
“周恩来关于经济合作问题致中共中央并毛泽东的报告,” pp. 90–1.
76
“Zhou Enlai’s Report on the Asian–African Conference, Made at the Meeting of the
Standing Committee of the National People’s Congress,” May 13, 1955, in China and the
Asian–African Conference, pp. 44–5.
77
Selected Works of Zhou Enlai, p. 237.
78
“周恩来关于经济合作问题致中共中央并毛泽东的报告,” p. 92.
710

Despite these failures, the PRC took a number of unilateral steps at

Bandung and throughout the late 1950s and 1960s that were simultaneously
designed to meet its goals of transforming the international economic order
into a more equitable one for developing countries, and “winning over”
countries with ties to the United States and, later, the Soviet Union. At
Bandung, the Chinese delegation pledged that China would contribute
industrial equipment and technology to countries in Asia and Africa.79 Over
the next decade, despite its own low levels of economic development, China
provided nearly US$2 billion in grants, low-interest loans, technicians, and
industrial construction assistance to other developing countries. The major-
ity of China’s aid was dispatched to Communist countries also experiencing
political or military turmoil, such as North Korea – which was the recipient of
China’s first foreign aid grant in 1953 – North Vietnam, Hungary, and
Albania.80 Nevertheless, China was also a generous provider of foreign aid
to non-Communist and nonaligned countries in South and Southeast Asia.
Alexander Eckstein finds that between 1949 and 1964, China provided around
US$800 million in loans to non-Communist countries, almost half of which
was dispatched to neighboring countries such as Burma, Indonesia,
Cambodia, Laos, Sri Lanka, Nepal, and Pakistan.81 Moreover, China matched
its Bandung rhetoric by establishing trade missions in India, Nepal, Egypt,
Lebanon, and Syria, and demonstrated a willingness to trade with a host of
Asian and African countries on terms that were frequently unfavorable to
China.82 As the PRC’s relations with the Soviet Union began to decline in the
late 1950s and early 1960s, the PRC then used its competitive foreign aid
strategy to persuade countries like Albania, North Korea, and North Vietnam
to transfer their allegiance from Moscow to Beijing.83 China’s strategic use of
foreign aid also extended to Africa. In 1963, Zhou Enlai embarked on a two-
month, ten-country visit across Africa, where he outlined an economic
assistance strategy rooted in principles of self-reliance, independence, equal-
ity, and mutual benefit. Though China’s aid levels to Africa would not match
those provided by the Soviet Union, Zhou’s emphasis of these foreign-aid
principles during the Africa tour was designed to differentiate China’s
79
80
C. Garratt, “China as a Foreign Aid Donor,” Far Eastern Economic Review 31.3 (1961), 83.
81
Eckstein, Communist China’s Economic Growth and Foreign Trade, pp. 161–2, 214.
82
“我参加亚非会议贸易活动方案(草案)”; Zhang, Beijing’s Economic Statecraft, pp. 110–
15, 154–5; G.T. Hsiao, “Communist China’s Trade Treaties and Agreements,” Vanderbilt
Law Review 21.5 (1968), 637–43.
83
Eckstein, Communist China’s Economic Growth and Foreign Trade, pp. 167–8. Zhang,
Beijing’s Economic Statecraft, Chapter 5.
711

amy king
approach from that of Moscow, which, the CCP believed, had the effect of
cultivating dependency upon the Soviet Union.84
Finally, China’s experience of imperialism had a profound impact on how
China conceived of its own economic dependence upon the Soviet Union.
Throughout the late 1950s and the 1960s, the themes of “self-reliance” and
“independence” became increasingly important in Chinese foreign economic
policy thinking. At Bandung, Zhou Enlai exhorted his counterparts to resist
accepting foreign aid that came with “disadvantageous” political or economic
conditions.85 It is perhaps no coincidence that the same year as the Bandung
conference, China became increasingly unwilling to accept foreign aid from the
Soviet Union. In 1955, Soviet aid to China began to slow, and by 1957 it had
almost completely ceased.86 Part of the explanation for declining levels of
Soviet aid to China lies with Moscow’s own decision making and political
crises within the Soviet bloc, such as the 1956 Polish and Hungarian crises,
which drew Moscow’s attention and resources elsewhere.87 Nevertheless, Shu
Guang Zhang argues that throughout the late 1950s, Beijing became increas-
ingly wary of the fact that China’s economic dependence on Soviet foreign aid,
loans, and industrial goods and expertise made China susceptible to potential
Soviet efforts to influence China. As a result, CCP leaders adopted a range of
measures to try and limit China’s dependence upon the Soviet Union, despite
the reality that China was still very poor and in need of goods and expertise
from Moscow. For instance, the CCP worked hard to establish a “mutually
beneficial” or more “symmetric” trade relationship with Moscow by expanding
Chinese exports to the Soviet Union as a way to compensate for Soviet credits
to China, and by supplying the Soviet Union with agricultural products and
“strategic” raw materials, such as rubber, which Moscow was unable to import
from elsewhere. In addition, Beijing took great care to pay back Soviet loans on
time or even early, despite the cost this imposed on the Chinese economy.88
The Impact of Domestic Political Campaigns

The last factor to have shaped China’s external economic relations during the
Mao era was the domestic political campaigns and crises initiated by the CCP
84
R. Scalapino, “Sino-Soviet Competition in Africa,” Foreign Affairs 42.4 (1964), 641–2, 649.
See also Zhang, Beijing’s Economic Statecraft, pp. 116–17.
85
86
87
Zhang, Beijing’s Economic Statecraft, p. 70.
88
Zhang, Beijing’s Economic Statecraft, pp. 65–6.
712

throughout the 1950s and 1960s. These campaigns and crises were numerous,
but two interconnected events in particular – the Great Leap Forward and
the Sino-Soviet split – had the greatest impact upon China’s external economic
relations. To understand the events of the Great Leap Forward and the Sino-
Soviet split requires us to cast our eyes back to the mid-1950s, when China’s
economy, domestic political environment, and international relations were
rocked by Mao’s politicization of economic decision making, tensions with the
Soviet leadership, and growing awareness that the Soviet economic model was
not entirely well suited to China’s domestic conditions.89 In 1955, Mao ignored
the advice coming from Moscow and from his own Party’s more moderate
economic planners, such as Zhou Enlai and Chen Yun, and adopted
a particularly radical form of Stalinist economic planning designed to accelerate
rural collectivization and to achieve wildly ambitious targets in cotton, grain, and
steel production. This “Little Leap Forward,” as it came to be known, was
a precursor for the eventual Great Leap Forward launched in the late 1950s.
Despite their attractiveness in theory, the Little Leap and later Great Leap
Forward were impossible to achieve in practice; the policies adopted resulted
in disastrous shortages of industrial raw materials and agricultural products,
production bottlenecks, and widespread famine in China’s rural areas.90 In order
to achieve the demanded increases in agriculture and steel outputs, in 1958 and
1959 the CCP’s State Planning Commission stepped up China’s imports of rolled
steel and chemical fertilizer from the Soviet Union, Japan, and Western Europe
(as depicted in Figure 19.5).91 Indeed, China became a highly unpredictable trade
partner for these countries as its economic planners requested faster delivery
schedules and demanded higher and higher levels of trade.92
It was not long before the economic policies of the Great Leap Forward, and
the highly charged domestic political environment in China, began to have
a dramatic impact on China’s two most important economic relationships:
those with the Soviet Union and Japan. The Soviet Union was itself embarking
on a set of corrections to Stalin’s earlier economic policies under the new
leadership of Nikita Khrushchev, leading to discord between Mao’s own
89
For more on the Soviet model, see Perkins’s chapter in this volume.
90
A.L. Chan, Mao’s Crusade: Politics and Policy Implementation in China’s Great Leap Forward
(Oxford, Oxford University Press, 2001), pp. 17–24; F.C. Teiwes and W. Sun (eds.), The
Politics of Agricultural Cooperativization in China: Mao, Deng Zihui, and the “High Tide” of
1955 (Armonk, NY, M.E. Sharpe, 1993).
91
C.J. Mitcham, China’s Economic Relations with the West and Japan, 1949–1979: Grain, Trade
and Diplomacy (London, Routledge, 2005), pp. 19–20, 42–3; Zanier, “‘Energizing’
Relations,” pp. 159–63.
92
713

amy king
250.00
200.00
150.00
100.00
50.00
0.00
1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960
Rolled steel Chemical fertilizer
Figure 19.5 China’s imports of rolled steel and chemical fertilizer, 1950–1960 (in tens of
thousands of tons)
economic vision and that being pursued by China’s Soviet “elder brother.” Mao
now sought to achieve a socialist revolution in China that would be “fuller,
quicker, and more effective” than that of the Soviet Union.93 At the same time,
the Soviet leadership grew increasingly concerned about China’s domestic and
international behavior when Mao launched the second Taiwan straits crisis in
1958, in a bid to mobilize the domestic Chinese population around the eco-
nomic goals of the Leap.94 Sino-Soviet antagonism had a direct impact on the
Soviet technicians working in China, upon whom China remained highly
reliant. On July 16, 1960, Khrushchev announced that all remaining Soviet
technicians in China were to be immediately recalled back to the Soviet
Union. Within a month of this announcement, more than 1,300 Soviet techni-
cians had left China, taking with them both the blueprints and the expertise for
around 250 ongoing or planned construction projects, the effect of which was
made even more severe when Khrushchev announced the abrogation of
a dozen economic and technical aid agreements with China.95 In response to
Khrushchev’s move, the CCP tried to punish Moscow by suspending or
postponing unsigned trade contracts, canceling orders of plant equipment,
93
Shen and Li, After Leaning to One Side, pp. 156–7; Z. Shen and Y. Xia, “The Great Leap
Forward, the People’s Commune and the Sino-Soviet Split,” Journal of Contemporary
China 20.72 (2011), 861–8.
94
T.J. Christensen, Useful Adversaries: Grand Strategy, Domestic Mobilization, and Sino-
American Conflict, 1947–1958 (Princeton, NJ, Princeton University Press, 1996), pp. 204–5.
95
Eckstein, Communist China’s Economic Growth and Foreign Trade, p. 153; Zhang, “Sino-
Soviet Economic Cooperation.”
714

and refusing to supply basic consumer goods to the Soviet Union.96 Bilateral
trade between the two countries peaked in 1959, but declined sharply the
following year (Figure 19.1), and would never again reach the heights that the
two countries had enjoyed in the first decade of their alliance. The China–Japan
economic relationship faced a similar fallout from the politics of the Great Leap
Forward, but one that was far more temporary than that experienced between
China and the Soviet Union.97 As noted above, China and Japan had built small
but significant trade ties since the early 1950s, through three rounds of “unoffi-
cial” trade agreements. In 1958, the two sides were in the midst of negotiating
a fourth agreement that, if passed, would shift the relationship in a more official
direction by allowing the two countries to exchange trade representatives and
establish permanent trade offices in either country. However, when the
Republic of China (ROC) learned of these negotiations, Chiang Kai-shek
protested the agreement and threatened to launch an economic boycott of
Japan in retaliation. In its haste to respond to Chiang’s threats, Japan’s Kishi
government reiterated that it did not recognize Communist China, would not
pursue any kind of official relationship with the PRC, and would not allow the
PRC flag to be flown in Japan. In the midst of the back-and-forth between
Japan, the PRC, and the ROC, two Japanese youths tore down a PRC flag that
had been flying at a trade exhibition in the city of Nagasaki. The combination
of Kishi’s statement and the “Nagasaki flag incident,” as it came to be known,
inflamed an already tense domestic political environment in Beijing. In the
middle of 1958, the Chinese Foreign Ministry announced that it would not
permit trade with Japan unless Japan agreed to “three political principles”:
“That Japan should not (1) adopt policies hostile to China; (2) participate in any
plot aimed at creating two Chinas; and (3) obstruct the normalisation of
relations with China.”98 Over the next two years, trade between China and
Japan almost completely ceased, falling from US$80 million in 1958 to less than
US$200,000 in 1960, before beginning to slowly increase again in 1961 and 1962.
Yet even in the midst of the complete breakdown in Sino-Japanese trade
between 1958 and 1960, the CCP worked behind the scenes to find ways to
reopen the bilateral economic relationship. To do so, Zhou Enlai and other
senior CCP officials met not only with traditional “friends” of the CCP, such as
representatives from Japan’s Socialist and Communist parties, but also with
leading politicians and officials from Japan’s ruling conservative Liberal
Democratic Party. Through speeches, People’s Daily editorials and conversations
96
Zhang, Beijing’s Economic Statecraft, p. 88.
97
The following paragraph draws on King, China–Japan Relations, pp. 135–57.
98
R.K. Jain, China and Japan, 1949–1980, 2nd ed. (Oxford, Martin Robertson, 1981), p. 44.
715

amy king
with Japanese counterparts, CCP leaders from Mao downwards made it clear
that they were eager to work with Japan to find a way to reopen the economic
relationship once Prime Minister Kishi was no longer in power. By late 1959,
Mao, Zhou, and Liao Chengzhi – the Chinese Foreign Ministry’s leading Japan
hand – had each explained to visiting groups of Japanese their views about the
importance of the China–Japan trade relationship, not only because of its role
in restoring friendship and co-operation between the two countries, but also
because of the role that Japanese technology and industrial goods and expertise
could play in assisting China’s reconstruction.
Finally, the Great Leap Forward had unintended consequences on China’s
import and export of grain and other agricultural products. In 1958, China had
been required to expand dramatically its exports of grain and other agricultural
products in order to pay for the imports of steel and chemical fertilizer needed to
pursue the Great Leap Forward. This was a perverse outcome, which exacer-
bated growing domestic grain shortages within China.99 Yet within two years
the situation had reversed. In 1960 and 1961, the Chinese leadership acted to
alleviate domestic famine by stepping up imports of foreign grain. The two key
beneficiaries of this shift in policy were Canada and Australia, countries which
had developed limited trade ties with China in the mid-1950s, and which were
blessed with record grain surpluses at the very moment at which Beijing was
looking to expand its grain imports. In 1960 and 1961, Beijing signed multiple
agreements with the Canadian Wheat Board, the Australian Wheat Board, and
a host of other Canadian and Australian agricultural exporting firms for the
export of wheat, flour, barley, and oats to China. As Figure 19.6 indicates,
China’s imports from Australia increased ninefold between 1960 and 1961, and
twelvefold with Canada over the same period. Because China could not afford to
correspondingly increase its exports to these countries, the Canadian and
Australian governments – spurred on by domestic agricultural lobby groups
eager to expand the China trade relationship – agreed to fund this increased grain
trade through six-, nine-, and twelve-month credit agreements with China.100
Conclusion
China’s external economic relations underwent a dramatic shift after the Mao
era. The reform era saw the relaxation of central control over prices, foreign
exchange, and trade; the restoration of normal relations with the United
99
Mitcham, China’s Economic Relations with the West and Japan, p. 40.
100
Mitcham, China’s Economic Relations with the West and Japan, Chapter 3.
716

300
250
200
150
100
50
0
1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965
Chinese exports to Canada Chinese imports from Canada
Chinese exports to Australia Chinese imports from Australia
Figure 19.6 China’s trade with Canada and Australia, 1955–1965 (in US$ million at current
prices)
States; and China’s eventual entry into major international organizations

such as the International Monetary Fund, World Bank, and World Trade
Organization. Each of these moves helped to bring an end to the depressed
levels of trade that China had experienced under Mao, such that, by 1994,
China’s share of world trade had finally reached levels not seen since the
1920s.101 Yet the significance of the Mao era extends beyond the mere fact of
China’s relatively low levels of trade, its prohibition of foreign direct invest-
ment, and its absence from international capital markets and economic
organizations during this period. Rather, the significance of the Mao era
lies in how China’s economic interactions with the outside world during
these three decades shaped a range of CCP ideas about foreign economic
policy and China’s place in the global economy. First, the CCP viewed the
international economic order as one in which a state’s relative level of
development determined its vulnerability to exploitation at the hands of
more economically advanced, and thus powerful, states. This was an eco-
nomically deterministic view of the international system, in which the CCP
perceived the world as divided into camps of “backward” countries – agrar-
ian, nonindustrialized producers of raw materials – and “advanced” coun-
tries – those who had made the leap to industrialization and who
subsequently exploited “backward” countries for their raw materials and
101
Keller, Li, and Shiue, “China’s Foreign Trade,” 886; Maddison, Chinese Economic
Performance in the Long Run, pp. 85–6.
717

amy king
export markets. These ideas had their origins in Qing and Republican China
as the country encountered Japan’s highly coercive industrial development
path, and were reinforced in the Communist era as China confronted the
severe trade controls that formed a key plank in the United States’ Cold War
containment strategy. During the Mao years, the pursuit of rapid industrializa-
tion to achieve “catch-up” growth was the core organizing idea behind China’s
economic relationships with the Soviet Union and, later, Japan. It was also the
idea that motivated the CCP in its failed attempts to create a new international
economic order, premised on the notion of “mutual equality and benefit,” that
would allow developing countries in Asia and Africa to make the transition from
“backward” to “advanced” economies.
Second, China’s external economic relations during the Mao years were
shaped by ideas about “self-reliance,” which became an increasingly important –
though frequently misunderstood – slogan in Chinese political and economic
discourse in the late 1950s and early 1960s. Like the idea of “backwardness,” self-
reliance also had its origins in China’s historical experiences of war and semi-
colonization; Mao introduced the idea of “self-reliance” (zili gengsheng) during
China’s War of Resistance against Japan as a way to strengthen the CCP in the
face of food scarcity and economic isolation.102 Yet self-reliance was never simply
translated into policies of autarky or import substitution in Mao’s China. Rather,
the CCP viewed foreign economic assistance and access to overseas advanced
technology as entirely consistent with a policy of self-reliance.103 Indeed, one
author has suggested that self-reliance more accurately represented a spectrum
of thinking within the CCP. At times, such as during the politically charged
atmosphere of the Great Leap Forward or the Cultural Revolution, Mao and
others derided foreign influence in the Chinese economy, and China’s trade with
countries such as Japan and the Soviet Union suffered as a result. Frequently,
though, self-reliance was conceived in more expansive ways: while China should
avoid foreign control over its economy, CCP leaders acknowledged that in order
to become truly self-reliant, China would need to adopt technology from more
advanced countries and learn from the industrial development paths of others.104
102
L. Yang, “Self-Reliance,” in C. Sorace, I. Franceschini, and N. Loubere (eds.), Afterlives
of Chinese Communism: Political Concepts from Mao to Xi (Canberra, Australian National
103
Zhang, China in International Society since 1949, pp. 39–40.
104
Terrill, “China and the World,” 297–9. For a related argument that Chinese foreign
economic policy during the Mao era was animated by CCP debates over the correct
path to self-reliance, see L. Reardon, The Reluctant Dragon: Crisis Cycles in Chinese
Foreign Economic Policy (Seattle, University of Washington Press, 2014), pp. 5–8.
718

Third, China was both a target and a practitioner of economic statecraft

during the Mao era. From their very earliest months in power, CCP leaders
learned how economic instruments could be used as a tool of foreign and
security policy. In 1949, the Soviet leadership advised Zhou Enlai of the
potential for China to use its economic power as a form of positive induce-
ment in its relations with the West. Just as quickly, China was subjected to
US-led trade sanctions that not only transformed China’s trade partners and
restricted its access to a wide range of “strategic materials,” but also demon-
strated to the CCP how economic sanctions could be used as a weapon of
war. Absorbing these lessons over the next three decades, the CCP’s foreign
economic policy was frequently shaped by two, sometimes competing, goals.
The first goal was a desire to overturn what the CCP saw as an unjust
economic order used by imperialist countries to their advantage, and to
create instead an order that was more equitable for developing countries.
Yet paired with this ideal was a second, more tactical, understanding of
foreign economic policy as a form of statecraft. Throughout the Mao era,
the CCP sought to use methods such as access to its lucrative market, barter
trade, and unconditional foreign aid as a way to undercut the US-led trade
embargo, to persuade countries to extend China diplomatic recognition, to
make China more attractive to developing countries, and to draw an explicit
contrast between itself and the United States and the Soviet Union.
Ideas about industrialization as a way to overcome “backwardness,” about
self-reliance, and about economic statecraft have exerted an influence on
Chinese economic thinking beyond the Mao era. As a more powerful,
market-oriented, and economically open China embarks on new domestic
and foreign economic policy initiatives, traces of Mao-era ideas and experi-
ences resonate. The “Made in China 2025” plan, for example, is designed to
reduce China’s reliance on foreign technology by increasing the share of
indigenously produced materials and technology to 70 percent by 2025, and to
place Chinese manufacturing at the forefront of the world’s most advanced
industries, including aerospace, robotics, and shipping. Similarly, the “Belt
and Road initiative,” introduced by the Xi government in 2013, is an ambi-
tious plan to open up trade, finance, investment, and infrastructure links
between China and a network of countries stretching from North Africa to
the South Pacific. These initiatives draw on Mao-era ideals about using
advanced technology and industrialization to make China a more modern
and powerful country by 2049 – the centenary of the founding of the PRC –
and about reforming the global economy to increase the participation and
economic development – on “equitable” and “mutually beneficial terms” – of
719

amy king
developing countries.105 And yet, as in the Mao era, the ideals underpinning
these initiatives are also tempered by a darker vision of the global economic
order and the strategies needed to survive in it. This darker vision reflects the
views of a country whose experiences of colonialism, military–industrial
warfare, an unreliable Soviet ally, and Cold War economic statecraft have
taught its leaders that China can never truly escape dependency unless it
masters indigenous production of advanced technologies, and who believe
that China remains an unequal power in an international economic order led
by the United States and Europe. These twin commitments are at once
designed to reshape the global economy for the benefit of all, and to make
China a far more powerful state within it.
Further Reading
Cheng, Y.-K., Foreign Trade and Industrial Development of China: An Historical and Integrated
Analysis through 1948 (Washington, DC, Washington University Press, 1956).
Eckstein, A., Communist China’s Economic Growth and Foreign Trade (New York, McGraw-
Hill, 1966).
King, A., China–Japan Relations after World War Two: Empire, Industry and War, 1949–1971
(Cambridge, Cambridge University Press, 2016).
Kirby, W.C., “China’s Internationalization in the Early People’s Republic: Dreams of
a Socialist World Economy,” China Quarterly 188 (2006), 870–90.
Lardy, N.R., “Economic Recovery and the 1st Five-Year Plan,” in R. Macfarquhar and J.
K. Fairbank (eds.), Cambridge History of China, vol. 14, The People’s Republic, part 1, The
Emergence of Revolutionary China, 1949–1965 (Cambridge, Cambridge University Press,
1987), pp. 144–83.
Liu, T.-C., and K.-C. Yeh, The Economy of the Chinese Mainland: National Income and
Economic Development, 1933–1959, vols. 1, 2 (Santa Monica, CA, RAND Corporation, 1963).
Mah, F.-H., “Foreign Trade,” in A. Eckstein, W. Galenson, and T.-C. Liu (eds.), Economic
Trends in Communist China (Chicago, Aldine, 1968), pp. 671–738.
Mitcham, C.J., China’s Economic Relations with the West and Japan, 1949–1979: Grain, Trade
and Diplomacy (London, Routledge, 2005).
Price, R.L., “International Trade of Communist China, 1950–65,” in Joint Economic
Committee of the US Congress (ed.), An Economic Profile of Mainland China
(New York, Frederick A. Praeger, 1968), pp. 579–608.
Reardon, L, The Reluctant Dragon: Crisis Cycles in Chinese Foreign Economic Policy (Seattle,
University of Washington Press, 2014).
105
State Council of the People’s Republic of China, Action Plan on the Belt and Road
Initiative, March 30, 2015; State Council of the People’s Republic of China, Made in
China 2025, July 7, 2015.
720

Zanier, V., “‘Energizing’ Relations: Western European Industrialists and China’s Dream
of Self-Reliance. The Case of Ente Nazionale Idrocarburi (1956–1965),” Modern Asian
Studies 51.1 (2017), pp. 133–69.
Zhang, S.G., Beijing’s Economic Statecraft during the Cold War, 1949–1991 (Washington, DC
and Baltimore, Woodrow Wilson Center Press and Johns Hopkins University Press,
2014).
Zhang, S.G., Economic Cold War: America’s Embargo against China and the Sino-Soviet
Alliance, 1949–1963 (Washington, DC, Woodrow Wilson Center Press, 2001).
Zhang, Y., China in International Society since 1949: Alienation and Beyond (Basingstoke,
MacMillan Press in association with St. Antony’s College, Oxford, 1998).
721

20
The Chinese Economy in the Reform Era
barry naughton
Introduction
December 1978 was a political, economic, and social turning point for China.
As the balance of power within the top leadership shifted, a search for new
policies began that deepened into what came to be called “reform and
opening” and culminated decades later in a multistranded transition to
a market-based economy. This new policy orientation was accompanied by
a shift in development strategy that permitted China to take advantage of
its factor endowments and structural conditions and dramatically
accelerate economic growth. Thus 1978 marks not only the beginning of
“reform,” but also the start of the Chinese “economic miracle,” a remarkable
thirty-two-year period, through 2010, during which GDP grew at an annual
rate of 10 percent. Chinese economic structure and Chinese society were
utterly transformed. An extraordinary distance separates the vibrant upper-
middle-income, predominantly market-based, economy that is China today
from the troubled, isolated low-income country that was China at the end of
the Cultural Revolution. This chapter builds its narrative around the systemic
and structural changes that transformed China, especially in the thirty years
between 1978 and 2008.
It is important not to oversimplify the 1978 pivot. China was not a blank
slate, nor did policy makers flip a switch to turn China from a command into
a market economy. Extreme institutional distortions and poor policies held
back the Chinese economy, and although they were serious and deep, they
also provided the opportunity for a quick growth acceleration when they
were ameliorated. Economic policy is multidimensional, made up of many
“bundles” of policies, with each bundle having its own political and economic
logic. This chapter focuses selectively on a few key policy arenas: rural
reforms and economic opening in the early 1978–1982 period, and then,
after 1984, overall reform strategy and industrial and financial reforms.
722

Nearly every aspect of policy was contested, fought over through nearly
the entire transition period. Policy contention and uncertainty dogged the
Chinese leadership especially in the early years of reform, and it took many
years for policy initiatives to coalesce into a coherent approach to economic
reform. Gradually, though, successful reforms took hold, increasing the
growth and resilience of the economy. In turn, this created the conditions
that drove massive economic structural change and enabled the ensuing
“growth miracle.” There was always a great deal of contingency: the com-
position of political power was a key short-run determinant of policy at each
juncture, and policy approaches changed abruptly and dramatically. Policy
making was driven forward by crisis, but market reforms succeeded in pulling
the economy out of crisis by uncovering new sources of demand, improving
incentives and productivity, and fostering greater factor mobility and flexibil-
ity. In the early 1990s, an especially thorough and abrupt shift in political
conditions led to dramatic new departures in market reform policy.
This historical epoch come to an end during the first decade of the twenty-
first century. As the economy moved out of crisis, market-oriented reforms
lost momentum, and the underlying conditions for miracle growth were
naturally exhausted. This chapter seeks to give the transition process a degree
of coherence, while remaining true to its complexity. It is important to note
that the coherence sought here is not the economic coherence that emerged
ex post: economists have already shown persuasively how reform worked in
China.1 Rather, the intention is to show the coherence in the way Chinese
decision makers faced and reacted to challenges, perceived their choices and
options, and the way these choices shaped actual outcomes. It is a coherence,
in other words, that includes mistakes, retreats, and opportunities missed,
even in the context of a broad advance.
The Baseline
GDP and Well-Being
Where did China stand economically in 1978? By the seemingly straightforward
measure of GDP per capita (on which subsequent growth rates are based), the
picture was stark. China was a low-income country, among the poorest in the
world. Even after converting to purchasing-power parities (PPPs), in an effort to
1
Y. Qian, How Reform Worked in China: The Transition from Plan to Market (Cambridge,
MA, MIT Press, 2017). See also John McMillan and Barry Naughton, “How to Reform
a Planned Economy: Lessons from China,” Oxford Review of Economic Policy 8.1 (Spring
1992), 130–43.
723

barry naughton
Table 20.1 Development indicators, 1978

China Upper middle
Overachiever
Life expectancy at birth 66 65
Literacy 65.5% 68.8%
Energy consumption per capita 617 781
(kilograms of oil equivalent)
R & D share of GDP 1.5% N/A
GDP data
GDP per capita, constant 2010 US$ 307 2,674
GDP per capita, PPP 2011 US$ 1,536 c. 5,000
Underachiever
Urbanization 17.9% 34.8%
Caloric consumption 2,080 2,477*
* World average
correct for China’s highly distorted price system, China’s GDP per capita of
$1,500 was comparable to – but lower than – large sub-Saharan African countries,
such as Sudan or Congo. Yet, as the top panel (Overachiever) of Table 20.1
shows, many of China’s other indicators were far above what one would
expect from a low-income economy.2 Indeed, in life expectancy, literacy, and
perhaps most strikingly energy use per capita, China looked just like upper-
middle-income economies of the day. Life expectancy and literacy measure
human capital, while energy use is a good proxy for physical capital: China’s
endowment of both human and physical capital was much greater than we
would expect of a low-income economy. Moreover, by any measure of
institutional capacity, or social capital, China would again look completely
different from any other low-income economy. Rural society was organized
into a three-level institutional hierarchy consisting of communes, production
brigades, and production teams. The communes, corresponding to traditional
market centers, tied the countryside into a national economic network. Almost
every commune had a rural credit co-operative and a supply and marketing co-
operative, as well as schools and a rudimentary clinic. China was completely
different from any other low-income economy.
2
GDP per capita comparisons based on purchasing-power parity, 2011 price basis. This
and other indicators in Table 20.1 from World Bank, World Development Indicators,
accessed November 15, 2019. Life expectancy and literacy from China’s 1982 census; R &
D share from National Bureau of Statistics, 40 Years of China’s Science and Technology
(1990). Caloric consumption from Food and Agriculture Organization, http://faostat3
.fao.org/download/FB/FBS/.
724

One might look at these indicators and conclude that China was “really”
more developed than it seemed, and by extension that the Cultural
Revolution “wasn’t so bad.” Such arguments are mistaken. What the data
show most clearly is an economy grievously underperforming its potential.
Given the stock of physical, human, and social capital – accumulated at
enormous cost – the production of useful goods and services was appallingly
low. Other physical indicators help explain why (Table 20.1 lower panel,
“Underachiever”). China was sharply under-urbanized, well below middle-
income countries, and was even below the low-income average of 19.2
percent. Labor power was bottled up in the countryside, kept there by
China’s household registration system, which permitted almost no migration
to the city. Despite the compulsory retention of labor, China’s collective
farms weren’t producing enough to bring the population above subsistence
level. According to the Food and Agriculture Organization’s food balances,
China’s average caloric availability had hovered below 2,000 calories per
person for more than a decade. The 2,000-calorie threshold – which reflects
reasonably well the level at which population can avoid widespread malnu-
trition – was breached for the first time only in 1978.3 Finally, the calculations
of real income are based on the assumption that money income can buy
goods, whose scarcity is captured through price indices, but rationing was
pervasive in China, not only for grain and cotton cloth, but for mundane
goods such as soap and tofu.
Economic Challenges and Institutional Structure

Two existential challenges overshadowed everything else: agriculture and
population. First, could China feed itself? China had been importing modest
amounts of food grain annually since the collapse of the Great Leap Forward.
Could agricultural output be increased as population and consumption
demand increased? Nobody was sure. China’s extraordinary and distinctive
institutions reached into even small villages, and contributed to some aspects
of development, such as basic health and education. But these same institu-
tions tightly constrained many types of economic activity and clearly held
back food production and overall economic performance. The Cultural
Revolution variant of collective agriculture was especially restrictive.
3
Bramall (this volume) argues that FAO data understate actual caloric availability, due to
undercount of private plots and overcount of nonfood uses of grain, and estimates 2,566
kilocalories per day. Taking FAO as the lowest and Bramall as the highest estimates,
estimated consumption is still very close to subsistence levels, considering low diversifi-
cation of diet and prevalence of demanding physical labor.
725

barry naughton
Farmers were forbidden to sell grain in rural markets, and draconian limits
were put on private plots, household handicrafts, and even barnyard animals.
For fifteen years, Dazhai, a poor village in Shanxi province, had been
promoted as a Maoist model. Dazhai provided an indigenously developed
model of mobilizing labor to expand agriculture by hacking fields out of steep
slopes and building new irrigation systems. Dazhai abolished private plots
and local markets altogether, and also delinked farmer income from farmer
productivity by assigning work points in a public evaluation process of an
individual’s attitude and effort (rather than output). Dazhai also aggressively
pushed for larger-sized collectives, which further diluted the monitoring and
surveillance capacity of smaller teams. Post-Mao leaders had picked up and
promoted the Dazhai model to assert Maoist legitimacy, and well into 1978
they were insisting that one-third of all counties should become “Dazhai-
style” counties by 1980.4
Could this institutional system produce enough food to feed China?
Even those initially committed to the Dazhai model gradually conceded
that agriculture needed much more help, which meant reducing compul-
sory grain procurements, raising prices, and increasing access to credit.
This would allow farm collectives to purchase more modern inputs,
especially tractors, fertilizer, and improved seeds.5 But would that be
enough?
The second existential challenge was population. Between the census
years of 1964 and 1990, the population of working age (fifteen to sixty-four)
doubled, growing at 2.6 percent per year. In 1978, China was smack dab in the
middle of this population surge. Could all these people be fed? Could jobs be
found for them? In urban and rural areas alike, services and small-scale
manufacturing – huge sectors in most developing economies – were highly
restricted and woefully underdeveloped. Throwing open all these restricted
activities had the potential to unleash a quick output response and also rapid
growth of employment. All the restricted sectors – farming and farm sideline
activities, urban services and small-scale manufacturing – are among the most
labor-intensive sectors in a developing economy.
Today, when population growth rates have declined dramatically in China
and worldwide, it is easy to recognize that population growth brings
4
Fan Yinhuai 范银怀, 从大寨到大邱庄 (From Dazhai to Daqiuzhuang) (Beijing,
Zhongguo fazhan, 2013). Li Jingping 李静萍, 农业学大寨运动史 (History of the
Movement to study Dazhai) (Beijing, Zhongyang Wenxian, 2011).
5
F.C. Teiwes and W. Sun, Paradoxes of Post-Mao Rural Reform: Initial Steps toward a New
Chinese Countryside, 1976–1981 (Abingdon, Routledge, 2015), pp. 51–4.
726

economic opportunity as well as challenges. But at the time, China’s leaders

anxiously surveyed the population growth they saw around them, and
worried that they would be overwhelmed by the impending demographic
tidal wave. Birth control policies were tightened in 1977–1978, and the draco-
nian “one-child policy” was adopted in September 1980.6 The simple fact is
that China was following an inappropriately capital-intensive development
strategy: a country with abundant labor resources was giving priority to
development of the most capital-intensive industries, including steel, machin-
ery, and armaments.7 This inefficient development strategy held back
growth, of course, but more critically meant that China was unable to
respond to the profound challenge of feeding a rapidly growing population.
China was mired in a vicious cycle of shortage, poverty, and inefficiency.
Institutional fixes to short-run crises ended up creating long-term economic
distortions. Rural–urban migration had been halted in the 1960s after the
collapse of the Great Leap Forward, in order to guarantee grain to urban
residents. As a result, surplus rural workers were penned up in the country-
side. Even so, urban labor also seemed to be in surplus: low-productivity
make-work jobs were created to disguise unemployment, 17 million high
school graduates were sent to the countryside during the Cultural
Revolution, and urban labor markets disappeared as scarce jobs were
rationed. The “Soviet growth model” (see the chapter by Perkins in this
volume) was held in place by strict enforcement of twin state monopolies,
over grain purchases and over industry. Individuals who pursued their own
interests, selling household products, or moving to the city, faced severe
penalties. Distortion built on distortion. If the institutional obstacles could be
removed, there could be a quick economic rebound. However, policy makers
faced paralysis because of their ideological commitment to socialist institu-
tions, and because of the complex interdependencies among different institu-
tions. The potential benefits were enormous if China could escape from this
straitjacket.
6
Liang Zhongtang 梁中堂, 中国计划生育政策史论 (A History of China’s Planned-
Birth Policies) (Beijing, Zhonguo fashan, 2014). F. Wang, Y. Cai, and B. Gu, “Population,
Policy, and Politics: How Will History Judge China’s One-Child Policy?”, Population and
Development Review 38 (2012) (supplement), 115–29.
7
J.Y. Lin, F. Cai, and Z. Li, The China Miracle: Development Strategy and Economic Reform
(Hong Kong, Chinese University of Hong Kong Press, 2003).
727

barry naughton
Development Strategy and the 10-Year Plan

Chinese leaders, including Mao Zedong, had long dreamed of a strong,
industrialized China. During the Cultural Revolution, from 1966 to 1976,
industrialization was subordinated to political struggles and ideological pur-
ity, and long-run objectives for the economy lay dormant, but they were not
forgotten. In the mid-1960s, Zhou Enlai and Mao Zedong had laid out two
interrelated objectives for the year 1980. The first was to realize a “basically
self-sufficient national industrial system,”8 and the second was to “basically
realize agricultural mechanization.”9 Both objectives were strategic
responses to the massive setback of the Great Leap Forward (1958–1960):
with China now forced to import grain and relations with the Soviet Union
irreparably damaged, self-reliance was the only option. Zhou Enlai in 1964
embraced spending fifteen years building an independent industrial system,
until 1980, when it would launch a broader growth process with the “Four
Modernizations.” Mao Zedong took the same date as the target for agricul-
tural mechanization, which would replace the failed utopian experiment of
Great Leap-era communes.
These slogans and objectives had been in the background when Chinese
policy makers resumed coherent economic planning in 1972–1974, after the
most chaotic period of the Cultural Revolution. But planners also had a new
opportunity: China’s oil production had increased steadily through the 1960s
and 1970s, and after global oil prices spiked in 1973, China faced a lucrative
opportunity to export oil. A multiyear, US$4.3 billion technology import
program was put together, signaling that agricultural mechanization would
now include chemicalization, based on imported fertilizer factories (see the
chapter by King in this volume). Planners began work on a Ten-Year Plan to
cover 1975–1985, intentionally straddling the fast-approaching 1980
target year. Imported industrial plant, embodying the latest foreign technol-
ogy, would enable China to break the agricultural bottleneck. China’s devel-
opment strategy would consist of high investment, technological borrowing,
and massive provision of industrial inputs to agriculture. China would try to
outrun its problems of low systemic efficiency and an inappropriate
8
Zhou Enlai, “Report on the Work of Government,” December 21 and 22, 1964, partial
report in Peking Review, 1965.1. In January 1975 Zhou specified that the long-term target
had been 1980 (three five-year plans).
9
Zheng Yougui 郑有贵, “中国农业机械化改革的背景分析与理论反思” (Background
Analysis and Theoretical Reflection on the Reform of Chinese Agricultural
Mechanization),” January 6, 2015, at http://ww2.usc.cuhk.edu.hk/PaperCollection/Details
.aspx?id=1630.
728

development strategy, rather than resolve them, starting a new “big push” in
order to resume the industrial march to great-power status.
Political dysfunction prevented policy makers from following through on
this strategy. In 1975, Deng Xiaoping had been put in charge of policy and
seemingly designated Mao’s successor, but the very next year Mao fired Deng
and he was subjected to months of vitriolic criticism. These enormous political
swings meant that the Ten-Year Plan was never issued during Mao’s lifetime,
nor were the annual plans for 1975, 1976, or 1977 ever officially promulgated.10
The economy was on autopilot as the succession struggle intensified. This is
one of the paradoxes of the Chinese experience: it was a “planned economy”
with a development strategy, but without an actual plan.
Poor planning was evident throughout the economy. Chinese planners
had never achieved the detailed control of specific materials and enterprises
that Soviet planners routinely exercised. In 1977, a huge backlog of unfinished
construction, poorly located and poorly designed, sometimes outfitted with
indigenous technology solutions that didn’t work, presented painful choices:
throw good money after bad, or abandon projects on which billions had been
spent? These were the economic legacies of the Cultural Revolution: an
impressive endowment of human, physical, and institutional capital that
was being squandered; a crisis in agriculture and employment that was
about to become much worse; and a supposedly “planned” economy the
institutions of which had rotted from the inside out. When a new leadership
took over after the death of Mao Zedong in 1976, they had unprecedented
leeway to adjust policy, but also faced extraordinary challenges.
The Beginning of Economic Reform and the

Strategy of Economic Opening
Hua Guofeng emerged as China’s leader at the end of 1976, and Deng
Xiaoping rejoined the top leadership in the middle of 1977. Although there
was inevitably tension between Hua and Deng, they initially agreed to work
together. Co-operation was facilitated by the fact that Hua, Deng, and other
top leaders agreed on a simple economic strategy: rehabilitate the entire
economic mechanism, from top to bottom; return to the framework of the
aborted Ten-Year Plan; and achieve agricultural mechanization and prepare
for the “Four Modernizations” by 1980. Deng Xiaoping, who hated delays and
10
Liu Guoguang 刘国光 (ed.), 中国十个五年计划研究报告 (A Research Report on
China’s Ten Five-Year Plans) (Beijing, Renmin chubanshe, 2006), pp. 378–90, esp.
379–80.
729

barry naughton
consistently supported rapid growth, strongly supported this program.

However, Deng was not in charge of economic policy, which was instead
managed by three veteran vice premiers who all served continuously from (at
least) 1975 through 1982: Li Xiannian, Yu Qiuli, and Gu Mu.
As the economy recovered, Chinese policy makers had to address two basic
concrete issues. The first related to China’s biggest problem and the second to
China’s biggest opportunity. The biggest problem, as discussed earlier, was
agriculture: policy makers had already agreed to put more resources into
agriculture – in order to get more out – but they were uncertain and divided
about how far to go (as discussed below). Hit by bad weather and the adverse
impact of the Dazhai movement, agricultural output shrank in 1977. The biggest
opportunity was technology import. Largely cut off from most sources of global
technology for over a decade, Chinese leaders saw the enormous potential
benefit from expanded technology import. China’s industry recovered faster in
1977 than expected, growing 14 percent. With this uneven performance, policy
makers became entranced by a vision of bigger and faster import of industrial
technology, which, in the spirit of the Ten-Year Plan, would fuel faster growth
and eventually break the agricultural bottleneck once and for all.
Technology Import
During 1977 and 1978, all the top policy makers gave strong support to policies
intended to increase exports (of oil) and imports (of industrial machinery
embodying advanced technology). Working groups spread out from China
to learn basic things about their capitalist neighbors.11 In 1978, twenty top-
level missions visited fifty-one countries. What they saw astonished them:
Japan, in particular, after thirty years of sustained growth near 10 percent
per year, was prosperous, modern, and highly productive. Chinese visitors,
including Deng Xiaoping and Hua Guofeng, realized the extent to which
China had fallen behind Japan and the capitalist West. Moreover, the shib-
boleths of Marxist economics could be seen to be patently false: capitalist
workers were not impoverished, society was well ordered, and workers were
well educated. Vice premier Gu Mu led a team to five countries in Western
Europe and briefed top leaders in June 1978 for over seven hours, galvanizing
opinion to support reform and further opening.12 The capitalists were eager
11
Chen Jinhua 陈锦华, 国事忆述 (The Eventful Years) (Beijing, Zhonggong Dangshi,
2005); C.-J. Lee, China and Japan: New Economic Diplomacy (Stanford: Hoover Institution
Press, 1984), pp. 36–75.
12
Gu Mu, 谷牧回忆录 (Memoirs of Gu Mu) (Beijing, Zhongyang Wenxian, 2014), pp.
319–26.
730

to do business with China, Gu Mu reported, and were willing to provide

cheap credit. China’s most important partner was to be Japan, and Nippon
Steel’s Yoshihiro Inayama led the Japanese business elite in advocating
rapprochement with China. China and Japan signed a long-term trade agree-
ment in February 1978, providing a framework for the export of Chinese coal,
coking coal, and oil in return for advanced Japanese machinery. These visits
thus confirmed for top leaders like Deng Xiaoping the necessity of reform,
the possibility of rapid growth, and the virtue of rapidly expanding business
with the capitalist powers.13
Enthusiasm for the technology import program grew, and its scope doubled,
and then doubled again. A July 1977 draft had called for $6.5 billion worth of
industrial plant imports (through 1985), but by the summer of 1978 a two-
month-long State Council meeting approved a program to import a total of
$80 billion through 1985.14 In this giddy period, projected imports increased
more than tenfold in a single year. The Baoshan Steel Mill project emerged as
the flagship and proof of concept, located in Shanghai because Shanghai had
the highest skills and the best project management capabilities in the country.
An existing proposal for a large blast furnace in Shanghai was repurposed, and
morphed through several rounds of negotiation into a gigantic integrated steel
mill imported from Nippon Steel, Japan.15 It was a huge project, with initial
payments of 400 billion yen (about $2 billion), more than 1 percent of China’s
GDP, for the largest and most complex project China had ever built.
However, as the technology import program gathered momentum, it
spiraled out of control, and ultimately fell apart. Bureaucrats scrambled to
get their projects included, and projects were approved without serious
economic analysis. Site selection, technology choice, and careful analysis of
financing options were all neglected. Even Baoshan Steel, the best of the best,
ran into substantial problems with site preparation and co-ordination of the
many project components. In fact, there was no room for error, for China
had almost no foreign-exchange reserves. Reserve data were secret at the
time, but subsequent release revealed that China had only $167 million in
13
Xiao Donglian 萧冬连, “The Systematic Investigation and Use for Reference of Foreign
Experience in the Early Days of Reform,” 中共党史研究 (Research in Chinese
Communist Party History) 4 (2006), 22–32. E. Vogel, Deng Xiaoping and the
Transformation of China (Cambridge, MA, Harvard University Press, 2011), pp. 85–7,
118–19, 217–27, is especially good on the impact of foreign experience on Deng Xiaoping.
14
Li Zhenghua 李正华, “1978 年国务院务虚会研究” (A Study of the 1978 State Council
Theory-Oriented Meeting), Dangdai Zhongguoshi Yanjiu 17.2 (March 2010), 4–13.
15
C.-J. Lee, China and Japan: New Economic Diplomacy (Stanford, Hoover Institution Press,
1984), pp. 36–75. Chen, 国事忆述.
731

barry naughton
reserves at the end of 1978, enough for only five and a half days of import
coverage, while the standard “rule of thumb” for reserve adequacy is three
months’ worth of imports.16 Then oil output stopped increasing: after two
decades of steady growth, China drilled 7 million meters of wells in 1978
without discovering a single new field. Top-level commitment to the import
program began to waver, and support for the Baoshan project, which had
been solid through 1978, just barely held together for approval at the end of
December. This was the economic background for the Third Plenum turning
point.
The Party Work Conference and Third Plenum,

November–December 1978
Given its historical importance, it is remarkable that no thorough or reliable
account of the 1978 policy turning point exists. However, from existing
accounts, we can deduce a few significant economic policy events that
occurred during two consecutive meetings, a thirty-four-day Party Work
Conference, followed by the five-day Third Plenum, that extended through
most of November and December 1978. Such meetings are usually com-
pletely predictable. Work conferences permit discussion, but participants are
divided into closely monitored small groups, and the plenum is then con-
vened to rubber-stamp policies finalized behind closed doors. The year 1978
was completely different. The Work Conference was called to discuss agri-
cultural policy and the plans for 1979 and 1980. However, on the third day,
Party elder Chen Yun stood up in his small group and declared that economic
policy couldn’t be fixed until deep political issues from the Cultural
Revolution were addressed. Chen then proceeded to name names, putting
on the table literally scores of top leaders who had been purged by Mao and
imprisoned, and still not fully exonerated. The effect was electric. Support for
Chen’s speech spread rapidly to other groups, and boiled over into anger and
recriminations against Cultural Revolution beneficiaries, still in power, and
the demand that they be dismissed.17 In the wake of this unprecedented open
debate, the discussion of economic policy was completely derailed, but three
major outcomes were ultimately extremely consequential to the economy.
16
Revised data given in China Statistical Abstract, 2019, p. 164.
17
Yu Guangyuan 于光远, 1978:我亲历的内次历史大转折 (1978: The Historic Turning
Point That I Experienced), reprint (Hong Kong, Tiandi, 2008), pp. 56–67. Zhu Jiamu
朱佳木, 我所知道的十一届三中全会 (The Third Plenum That I Knew) (Beijing,
Dangdai Zhongguo, 2008), pp. 98–102.
732

First, Chen Yun returned to the top Party leadership, joining the Standing
Committee of the Politburo, which was thus dominated by Party elders.
Chen Yun is universally characterized as an economic policy “conservative,”
but his conservatism was rooted first and foremost in macroeconomic policy.
For Chen Yun, an out-of-control, heavy-industry-based “Leap” was the worst
nightmare in Chinese policy making, a thing he had repeatedly warned
against. If this “Leap” were to be funded by risky debt to the capitalists –
a group for whom Chen Yun never entirely lost his antipathy – it would be
even more dangerous. Thus, when, on December 10, Chen Yun gave a talk to
his small group about economic policy, it signaled the death knell for the
grandiose technology import program and, ultimately, the whole Ten-Year
Plan. Chen Yun was not initially given formal economic responsibilities, but
only three months later, in March 1979, he was put in charge of economic
policy, and immediately pushed through a three-year “Readjustment” (tiao-
zheng 调整). Readjustment meant reducing investment and growth, and
shifting resources to consumption.18 It implied that everything, including
reform, would be subordinated to the need to reduce stress on the economy.
It was to be the beginning of a multiyear program of macroeconomic
restraint.19
Second, Deng Xiaoping, who had skipped the first half of the Work
Conference in order to visit Singapore and Thailand, returned to find the
conference out of control. Deng immediately stepped in, assumed de facto
control, and, to close the conference, delivered one of the best speeches of his
life. Deng stepped up the reform rhetoric, stressing “thought liberation” and
“looking to the future.” Deng also insisted that the rancor against Cultural
Revolution beneficiaries be controlled, and that nobody be openly fired.
Deng thus succeeded in assuming leadership, moving the Party toward
reform, and maintaining the appearance of unity to the outside world.
Third, the Work Conference failed to resolve the questions about agricul-
tural policy. Debate about agricultural policy had been building for a year. A
key role was played by Wan Li, who in June 1977 had been appointed head of
Anhui, a major agricultural province with arguably more problems than any
other place. Wan Li had absolutely no experience with agricultural policy,
18
“Readjustment” is capitalized throughout this chapter to emphasize that it is a major
policy package, not a modest “adjustment” of policy. The earlier Readjustment of 1961–
1962, which the 1979 Readjustment echoed, was a complete reversal of policy in the
wake of the Great Leap Forward collapse.
19
Although Li Xiannian had arguably been the strongest proponent of the overambitious
technology import plan, Chen Yun was able to bring him along in supporting
Readjustment in 1979.
733

barry naughton
but had a reputation as an unusually bold, plain-spoken, and sometimes

abrasive leader, and was extremely close to Deng Xiaoping. Anhui had
suffered more than any other province in the Great Leap Forward famine,
with excess deaths accounting for a cumulative 10 percent of the population
over 1959–1961. Agricultural production and farmer confidence in the collect-
ive system had never really recovered. Wan Li quickly signaled his openness
to a series of bold experiments with “responsibility systems” in agriculture, in
which farm households or small groups took responsibility for production
on specific plots, turning over a procurement quota but keeping any add-
itional output. Wan Li publicized these actions and reported them to the
central government, which considered them legitimate, since they were
controlled local government “experiments.” However, they remained highly
controversial.
Wan Li had been scheduled to make a speech about Anhui policy at the
Work Conference, but it was canceled amidst the turmoil. In an acrimonious
exchange, Chen Yonggui, the chief proponent and beneficiary of the Dazhai
model, claimed that “everything [Wan Li] is doing is capitalism.” The draft
agricultural-policy documents were full of formulaic rhetoric and explicitly
prohibited distributing land to individual farm households. Amidst broad
dissatisfaction, the original drafts were discarded and party stalwart Hu
Qiaomu was put in charge of assembling a new document out of suggestions
put forward by the small groups. However, Hu’s draft pleased almost no one
and it maintained explicit prohibitions against giving households responsibil-
ity for land, tightly constraining experiments like those in Anhui. By one
account, Hu Qiaomu was asked to add a phrase allowing experimental
implementation in remote mountainous and poor areas, and replied, “No
way! No way! Put that in the document and it will spread around the
country . . . the dam will burst.”20 A provisional document was issued, with
the Dazhai model stripped out and the language toned down, but none of the
contradictions resolved. Despite the enthusiasm and sense of a breakthrough
fostered by Deng’s speech, little progress was actually made on core eco-
nomic issues.
20
Chen Guanren 陈冠任, 17 个省,自治区和直辖市改革启动纪实 (A Record of How
Reform Was Launched in 17 Provinces) (Beijing, Zhonggong Dangshi, 2009), p. 21. Wan Li
万里, “农村改革是怎么搞起来的?” (How Were Rural Reforms Kicked Off?), in Xu
Qingquan 徐庆全 (ed.), 中国经验: 改革开放30年高层决策回忆 (The Chinese
Experience: 30 Years of Reform and Opening, Recollection of High-Level Policy Makers)
(Jinan, Shandong renmin chubanshe, 2008), pp. 135–6. Li, Movement to study Dazhai, pp. 369–
75; Zhang Guangyou 张广友, 改革风云中的万里 (Wan Li amongst the Storms
of Reform) (Beijing, Renmin chubanshe, 1995).
734

What Changed after the Third Plenum?

The Third Plenum effectively terminated two mistaken policy initiatives –
the Dazhai model and the technology-import-based Ten-Year Plan – but
made no decisions on what would replace them. Politically, the landscape
had changed fundamentally, but the main economic issues were still unre-
solved. A broad commitment to undefined reforms – to which Deng
Xiaoping had given great impetus – opened up political space, permitting
proliferating experiments with decentralization, expanded autonomy, and
improved incentives. Economically, Chen Yun’s Readjustment opened up
space, by delaying and reducing big government commitments, and freeing
up resources for consumption and experimentation. New economic and
political options were on the table.
Readjustment allowed Chinese policy makers to sidestep one of the most
difficult problems that faces all reform programs. Reform policy makers relax
controls and decentralize decision making, only to be met with an explosion
of social demands. Citizens are emboldened to demand things they have long
wanted and finally have hope of achieving. Firms owned by an ill-defined
“public,” facing uncertain rules and budget constraints, jump at new oppor-
tunities. The resulting explosion of demand destabilizes the reform process,
leading to economic imbalances and inflation, and unnerving policy makers
who may not be able to satisfy surging expectations. In the wake of the 1979
Readjustment, this dynamic was damped down, because resources were
made available to meet emerging demands. For example, the 17 million
urban youth sent to the countryside during the Cultural Revolution,
emboldened, demanded the right to return home. In December 1978,
Deng Xiaoping asked them to be patient, promising that policy would
change sometime later. This position was untenable: three weeks after
the Third Plenum, the central government capitulated, and within months
10 million sent-down youth had returned home.21 Policy shifted to finding
work for these returned youth, by opening up new shops and workshops in
the cities. It took two years, but gradually the excess unemployment
created was worked down.
Readjustment made it possible for China to move away from a deeply
flawed development strategy. The technology import drive had collapsed
under the weight of its own short-term problems, but the long-term defects
21
Gu Hongzhang 顾洪章 (ed.), 中国知识青年上山下乡大事记 (Chronology of
China’s Educated Youth Sent to the Countryside) (Beijing, Zhongguo jiancha, 1997).
Chen, 17 个省, pp. 421–5. T. Gold, “Back to the City: The Return of Shanghai’s Educated
Youth,” China Quarterly 84 (December 1980), 755–70.
735

barry naughton
of the program’s strategic approach were even more important. The tech-
nology import program was premised on the idea that advanced technology
could provide a “quick fix” to the economy, without making more funda-
mental – and difficult – changes in the economic system and strategy. When
the import program was abandoned, it enabled China to move away from the
strategy it had been following for thirty years, of extracting resources from
the countryside and pumping them into capital-intensive heavy industry.
This had meant continuously suppressing domestic consumption, and, cru-
cially, keeping up the economic pressure on the countryside. When the
technology import program collapsed, at the end of 1978, the way was
open to shift development strategy comprehensively to consumption (on
the demand side) and labor-intensive sectors (on the supply side), bringing
the entire economy in line with its underlying factor endowment.22
Finally, Readjustment enabled economic problem solving and policy
experimentation that contributed to the long-run success of market transi-
tion. For example, as farmers were allowed into nonfarm activities,
a November 1979 document allowed “township and village enterprises” to
enter almost all sectors, abandoning previous policies to focus on “aid-to-
agriculture” enterprises, like fertilizer, agricultural equipment, and cement.
A radical new approach had been enabled that became a key component of
market transition. The top leaders, collectively speaking, were not yet
committed to radical reforms, but they had abandoned their commitment
to the old economic system.
Opening Up: One Step Ahead and One Step Behind

By pulling the plug on the overinflated technology import program, China
averted a short-run foreign-exchange crisis, but the fundamental economic
challenges were still severe. With oil output growth in doubt, China’s ability
to earn foreign exchange was even more limited than before, while her
appetite for foreign technology was greater than ever. The search for innova-
tive ways to earn foreign exchange, underway since 1977, now became more
urgent than ever. Inevitably, the search focused on Hong Kong.
22
D. Solinger, From Lathes to Looms: China’s Industrial Policy in Comparative Perspective,
1979–1982 (Stanford, Stanford University Press, 1991). B. Naughton, Growing Out of the
Plan: Chinese Economic Reform, 1978–1993 (New York, Cambridge University Press, 1995),
pp. 59–96.
736

The Hong Kong Connection

Hong Kong had been experiencing explosive growth since the late 1950s,
driven by manufactured exports. Tiny Hong Kong, with half a percent of
China’s population, exported twice as much as the entire Chinese mainland
in 1978. Both Chinese policy makers and Hong Kong entrepreneurs were
sensitive to the enormous opportunities presented by Hong Kong’s rise:
Hong Kong skills and trading networks together with cheap Chinese labor
and land could create a powerful economic combine that would earn abun-
dant profit and foreign exchange. Hong Kong was also an implicit rebuke to
Maoist economic policy. Local Chinese farmers had been slipping across the
supposedly sealed Hong Kong border for decades and now a new wave of
immigrants took advantage of China’s social relaxation. Almost 20 percent of
Shenzhen’s population lived in Hong Kong, and in Zhuhai population had
actually declined over twenty-five years due to emigration to adjacent
Macau. Deng Xiaoping identified the refuge flow as an international political
problem that could only be resolved by domestic economic policy changes.23
Already in 1977 an initiative began that was destined to transform China’s
export economy. A few “export-processing” contracts were signed, under which
Hong Kong businessmen sent raw materials to their home villages in the Pearl
river delta. The village would earn a processing fee for, e.g., sewing a zipper into a
pair of blue jeans, and return the product to the owner. Starting from nothing,
this “export processing trade” generated US$1 billion in exports by 1981.
Processing trade sidestepped China’s cumbersome customs procedures and
high tariffs, allowing exports to drive growth long before comprehensive reforms
were ready. Thirty years later, processing trade would grow from blue jeans to
laptop computers and earn an astonishing half a trillion US dollars per year.
Moreover, processing trade was not restricted to a zone, and Hong Kong’s
export-driven economy spilled out into the Pearl river delta and, eventually,
across the country.
“Export production bases,” mainly producing food for Hong Kong, already
existed, and in March 1978 Guangdong bureaucrats proposed diversification
into manufacturing and tourist centers. Shortly thereafter, a more innovative
initiative emerged from the China Merchants Steam Navigation (CMSN)
company, a state-controlled company which dates back to China’s 1872 “Self-
Strengthening” efforts. CMSN had been converted into a Hong Kong
23
Tao Yitao 陶一桃 and Lu Zhiguo 鲁志国 (eds.), 中国经济特区史论 (A Historical
Treatise on China’s Special Economic Zones) (Beijing, Shehui Kexue Wenxian, 2008), pp.
11, 15, 223–4.
737

barry naughton
corporation owned by the Chinese Ministry of Transportation after 1949.

Shenzhen native Yuan Geng, an aggressive, straight-talking, and visionary man-
ager, took over CMSN in mid-1978, and proposed moving CMSN’s existing ship-
breaking business from Hong Kong to Shenzhen.24 Crucially, Yuan proposed
that CMSN would own the land and host other land-constrained Hong Kong
businesses. In January 1979, after careful preparation, Yuan Geng visited Beijing
and proposed a two-square-kilometer development zone to Vice Premier Li
Xiannian, who was still in charge of economic policy. Li looked at the map Yuan
had thoughtfully provided and replied, “Land? Sure, you can have some; what if
I give you this peninsula?” circling the entire (fifty-square-kilometer) Shekou
peninsula.25 Yuan Geng prudently declined the generous offer, but broke ground
only six months later on his two-square-kilometer patch.
Special Economic Zones

Shekou hurtled forward in the second half of 1979, but there was profound
uncertainty about the definition and scope of special economic zones (SEZs).
The core idea – an export-processing zone – was easily accepted, since these
provisions were already incorporated in the proliferating export-processing
contracts. Provincial leaders in Guangdong had big ideas, and in mid-1979 they
received permission from the central government, including endorsement
from Deng Xiaoping, to move ahead with a bold interpretation of a special
economic zone. Crucially, the province was put in charge of drafting the
regulations. Deep disagreements ensured prolonged controversy and revision,
but the thirteenth draft was ultimately adopted as law in August 1980.
Four special economic zones were created, each oriented to a Chinese
business community outside the People’s Republic. Priority was given to
Shenzhen, adjacent to Hong Kong, which was given expansive borders, in
fact enveloping Shekou, which retained its autonomy. Provisions to allow
and attract foreign investment were multisided: foreign-invested firms would
have access to long-term land use rights at concessionary rates, income tax
holidays, and duty-free import of capital equipment. Beyond these concrete
provisions, the zones were granted a broad mandate that included “systems
and policies that are different from the rest of China, and which primarily
24
Yuan Geng 袁庚, “The Ten-Year Brilliance of Shekou,” in Xu, The Chinese Experience,
pp. 260–72. W. Huang, “The Tripartite Origins of Shenzhen,” in M. O’Donnell (ed.),
Learning from Shenzhen: China’s Post-Mao Experiment from Special Zone to Model City
(Chicago, The University of Chicago Press, 2017), pp. 65–85.
25
Tao and Lu, 中国经济特区史论, pp. 25–6. L. Li, Breaking Through: The Birth of China’s
Opening-Up Policy (Oxford, Oxford University Press, 2009), pp. 73–6.
738

employ market adjustment.”26 The zones would be allowed to grow outside

the scope of China’s command economy.
In essence, SEZ policy created a series of concentric circles. Shekou
became a curious zone within a zone. As a subsidiary of Hong Kong corpor-
ation CMSN, Shekou was a state-owned enterprise but exempt from the
normal bureaucratic hierarchy. It had the ability to borrow money and invest
in infrastructure, and it became a pioneer of a vast range of economic reform
experiments, including flexible land and labor policies that were impossible in
the rest of China at the time. Shekou was thus a uniquely dynamic actor
inside the zone. Shenzhen SEZ itself launched a much broader program of
urban development, commercial housing, and tourism, in addition to export-
oriented manufacturing. Outside the zone, the two provinces of Guangdong
and Fujian were given special powers, including almost complete budgetary
autonomy. Fujian’s inclusion demonstrated that Guangdong was not to be
a unique case, even though it took years for the Xiamen SEZ there (oriented
to Taiwan) to reach critical mass. Both provinces aggressively exploited the
broad grant of undefined policy autonomy to expand provisions for export
processing and favorable treatment for foreign investment over broad
regions.27 The SEZs thus became the symbolic center of a broad program
of economic opening. After a few years, Deng Xiaoping decided that the SEZs
had been his idea, and he became their advocate. This enhanced the symbol-
ism of the SEZs, and made them a signal of China’s commitment to openness
and acceptance of foreign businesses. Yet the idea remained controversial,
and the leadership commitment in fact subsequently wavered. Nevertheless,
the broad definition of SEZ policy, integrated into national legislation in
August 1980, endured.
The Shanghai Exception

Shanghai was a sharp contrast to Guangdong. The national government was
highly dependent on Shanghai for budgetary revenues, Shanghai’s industrial
skills made it a primary actor in industrial upgrading, and Shanghai’s high
incomes and tight organization were a point of political pride. Conservatives,
led by Chen Yun, wished to keep Shanghai under tight control. Chen Yun –
himself a Shanghai native – had a lifelong suspicion of the entrepreneurial
26
Communist Party Center, “Summary of the Meetings of Guangdong and Fujian
Provinces,” May 1980, cited in Tao and Lü, 中国经济特区史论, p. 30.
27
Tao and Lü, 中国经济特区史论, pp. 14–32, 218–46. Y.C. Jao and C.K. Leung (eds.),
China’s Special Economic Zones: Policies, Problems and Prospects (Hong Kong, Oxford
739

barry naughton
instincts of Shanghai businessmen.28 While experimentation could be accept-

able in peripheral provinces like Guangdong, it could not be countenanced in
the core of the planned economy. Private firms were springing up across
China, but Shanghai remained dominated by state-owned firms. When
foreign investment did begin trickling into Shanghai, it was primarily chan-
neled into joint ventures with state-owned enterprises, such as Shanghai
Auto, which dominated the Shanghai industrial and export economy.29
Chen Yun’s attitude toward Shanghai helps explain a surprising outcome.
When Chen Yun took control of economic policy in April 1979, he was
determined to cut back the overgrown technology import program, and
the Baoshan steel mill was obviously on the chopping block. Chen Yun
agonized, visited Baoshan, and convened four successive Beijing meetings.
Many were surprised when Chen gave Baoshan his blessing (though he
instructed managers to stretch out the timing and substitute domestic equip-
ment when possible). Chen felt that Baoshan had the right mix of govern-
ment control, local technological expertise, and proper sequencing, and that
it could serve as a model for future technology import.
The differential treatment of Shanghai and Guangdong led to an enduring
feature of the Chinese economy. Different provinces competed both economic-
ally and to secure favorable treatment from reform policy makers. Reforms
worked: while the share of Guangdong and Fujian in China’s exports soared
from 16 percent to 46 percent between 1978 and 1995, the lower Yangzi – with
Shanghai at its core – dropped back from 35 percent to 21 percent, losing its
traditional primacy. Not surprisingly, Shanghai and other regions began to
agitate for more autonomy and reform provisions. Domestically, almost every
province and central ministry set up subsidiary firms in Shenzhen to take
advantage of reform provisions, a total of 3,900 by the end of the 1980s.30
Competition for access to reform privileges became an important policy driver.
Rural Reform after the Third Plenum

Rural policy had been gradually relaxing from the Dazhai promotion of 1977.
Now, the state was putting more resources into agriculture, lowering
28
Chen Yun 陈云, “Several Important Orientations to Economic Construction,”
December 22, 1981, in 陈云文选 (Selected Works of Chen Yun), vol. 3 (Beijing,
Renmin chubanshe, 1995), pp. 306–7.
29
Y. Huang, Capitalism with Chinese Characteristics: Entrepreneurship and the State
(New York, Cambridge University Press, 2008), Chapter 4.
30
Tao and Lu, 中国经济特区史论, p. 32.
740

procurement quotas, paying higher prices, and providing more credit. Moreover,
a growing emphasis on autonomy for the collectives meant more freedom to
raise the crops and animals that farmers wanted as well. Despite these changes,
the core rural institutions and property rights remained largely unchanged
through 1979. As farmers gained more political and economic decision-making
authority, they began to demand greater institutional flexibility. Farmers began
to argue that, since government policy acknowledged their rights to decision-
making autonomy, they should also be allowed to change the way land use and
farmer compensation were determined.31
This demand set off a complex interactive process of bottom-up demands
and top-down policy change, as both farmers and policy makers probed to see
what they could achieve and what they should accept.32 Initially, the agricul-
tural collectives themselves were sacrosanct, a core tenet of PRC policy.
Rural land policy had been a critical component of Communist victory in the
civil war and thus of Communist Party legitimacy. Mao Zedong had person-
ally insisted on rapid collectivization (in 1955) and communes (in 1958), and
criticized proponents of “dividing up the land” and “going it alone” (in 1962).
To call for the dissolution of agricultural collectives was to directly repudiate
Mao Zedong. It is impossible to know what farmers thought, because for
twenty years government coercion had foreclosed their options. Farmers
who tried to leave collectives could easily find themselves in prison. In 1978,
farmers of Xiaogangcun in Anhui leased land to households, but signed – in
blood – a contract promising to raise each other’s children in case they were
arrested. Although these farmers were later celebrated as models of reform
(and courage), their fears at the time were not unreasonable.
As farmers in many regions tried out a huge variety of incentive arrange-
ments, different views among top leaders gradually coalesced into two camps.
The more reformist leaders, especially Wan Li in Anhui and Zhao Ziyang in
Sichuan, argued that farmers must be allowed to modify their own institutions,
without having to worry about constant official intervention (or punishment).
The more conservative leaders, such as Hua Guofeng, Li Xiannian, and Li’s
associate, Wang Renzhong, were not against institutional experimentation, but
for them this permission was premised on the continuing predominance of the
31
D. Kelliher, Peasant Power in China: The Era of Rural Reform, 1979–1989 (New Haven, Yale
University Press, 1992). K.X. Zhou, How the Farmers Changed China (Boulder, Westview
Press, 1996).
32
Lu Mai 卢迈, “The Policy Process of China’s Rural Reform,” 二十一世纪 (Twenty-
First Century) (Hong Kong) 50 (December 1998), 14–23. Du Runsheng 杜润生 (ed.),
中国农村改革决策纪实 (A Virtual Record of China’s Rural Reform Policy) (Beijing,
Zhongyang Wenxian, 1998), pp. 80–97. Wan, “农村改革是怎么搞起来的?”.
741

barry naughton
collectives, which they hoped would be strengthened by improved incentives and

higher incomes. The reformists had no desire to mount a frontal assault on
collectives, which would be self-defeating and create enormous uncertainty; the
conservatives had no desire to strangle experimentation, which would hobble
the rural economy. As a result, for about three years after the Third Plenum,
argument raged back and forth.
The institutional form at the center of this debate was “contracting output
to households” (baogandaohu 包干到户), which gave a household control
over a plot of land in return for the delivery of an agreed quantity of grain
after the harvest. Contracting output was the tipping point between two
competing conceptions of the collectives. Contracts initially for one season
could be (and were) easily extended for three years and longer, and the
households quickly gained control of land, “renting” from the collectives. For
conservatives, this was not only ideologically unacceptable, but also under-
mined their broader economic strategy of extracting surpluses from the
farmers to support industrialization. They wanted to be able to instruct
collectives to purchase tractors, machinery, and fertilizer coming from the
new factories they envisioned, and they wanted to go on mobilizing millions
of farmers every winter for construction of irrigation networks and roads.
Farmers were not enthusiastic about buying big tractors or being pressed into
annual labor gangs, reformers pointed out, but worked harder under institutional
arrangements they had a hand in creating.33 Thus, early on, reformers empha-
sized the importance of staying receptive to a range of “production responsibility
systems” (PRSs) and not tabooing household contracting. Whether reformers
understood from the beginning that household contracting would ultimately
lead to the dissolution of collectives and emerge as the dominant rural system is
unclear. In March 1979, a meeting of provincial agricultural cadres addressed the
question left open at the Third Plenum. All acknowledged that varieties of PRS
were spreading, but the more conservative central leaders, still in charge of
policy, came and stressed the importance of collective agriculture. Household
contracting could be permitted only in very special circumstances of isolated
households in the mountains.34 Still, no rollback was contemplated, and the door
to acceptance of household contracting was cracked open.
33
In economists’ terms, the strategy of equipping farmers with machinery was “labor-
augmenting,” which was not China’s greatest need. By contrast, both fertilizer and
incentive reforms were “land-augmenting.”
34
Liu Kan 刘堪, “回顾 1979 年7省农口干部座谈会” (Looking Back on the 1979
Conference of Agriculture Cadres from Seven Provinces), in Du, 中国农村改革决策
纪实, pp. 80–97.
742

In early 1980, Zhao Ziyang became acting premier and took over the
economics portfolio from Chen Yun, while Wan Li was promoted to head
agricultural policy in Beijing. These personnel changes must have sent
a strong message to local officials, who in many respects were caught in
the middle. The collective system made their job much easier, since it gave
them direct authority over rural resources, yet many saw that collectives held
back local growth. Local officials were charged with interpreting and imple-
menting changing central policies, so they were very sensitive to upper-level
signals. Local officials sometimes flip-flopped, as the example of Guizhou,
one of the poorest provinces, shows. The province was one of the few whose
backward, dispersed farmers were legitimately permitted to experiment with
household contracting, and various flavors of PRS spread in 1978–1979.
However, provincial officials worried that they were too far ahead of central
policy, and in December 1979 shifted gears and sent out teams to “rectify” the
PRS (and village officials). The result was a fiasco, with some farmers refusing
to plant if their contracted plots were not returned. After a couple of months,
provincial officials ended their “rectification,” and later swung to the other
extreme, adopting permissive policies province-wide. By the beginning of
1981, the province was sending work teams to the village again, but this time
to assist farmers signing written contracts with the collectives, clarifying
property rights for the land they had received.35
Not every place went to extremes like Guizhou, but everywhere local officials
made the most important practical decisions: counties typically adopted new
policies all at once.36 As farmers saw that the amount of coercion in the system
was being reduced, they were emboldened to push. The policy issues were
resolved through a gradual back-and-forth between top policy makers and
farmers. By mid-1980, Zhao Ziyang was in a position to demonstrate something
we might call “active compromise” – the ability to find a formulation for market
reform that soothed the fears of conservatives while opening up further policy
change. In his “Letter on Current Rural Policy,” Zhao expressed concern that
constant change in rural policy would disrupt the busy summer season under-
way. He suggested that household contracting should be permitted in poor and
backward areas. While household contracting was not appropriate in prosperous
areas with healthy collectives, if these areas were experimenting then they
35
Chi Biqing 池必卿 and Gao Chunsheng 高春生, “Circumstances of Guizhou
Implementing Household Contracting in the Whole Province,” in Du, 中国农村改
革决策纪实, pp. 268–98.
36
J. Unger, “The Decollectivization of the Chinese Countryside: A Survey of
Twenty-Eight Villages,” Pacific Affairs 58.4 (Winter 1985–1986), 585–606.
743

barry naughton
should continue without interruption.37 This created a presumption of permis-

siveness, which in fact supported the spread of PRSs. Ultimately, in January 1982,
the PRS, including household contracting, was officially affirmed. In 1983, in
most parts of the country, communes were abolished, with their governmental
functions transferred to the new townships, and income-earning activities spun
off as “township and village enterprises” (TVEs).
Despite its inconsistency, the interactive policy process through which the
state retreated provided a number of benefits. First, it prevented the process from
getting out of control, becoming a frantic division of assets, and it allowed policy
makers and the Communist Party to save face and protect the position of its
agents in the countryside. The change in rural society and property was enor-
mous, a genuine “second revolution” in the Chinese countryside, but it was
a controlled revolution. Second, it enabled an apparently fair and egalitarian
distribution of land. In part, this is because the initial distribution of land was seen
as provisional and was carried out locally, in a group process within the collect-
ive. Third, it allowed a negotiation about the bundle of private property rights
that farmers were achieving. Farmers “gradually enlarged the range of their
contracts, accumulated more and more private rights, and entered into
a network of market contractual relationships of a greater and greater
variety.”38 The end result was acceptable both to the farmers, who were allowed
to return to household farming, and to the policy makers, who until then had
seen collectives as an indispensable part of socialism. The collectives survived in
name as the ultimate owner of the land, but, given the opportunity to invest in
their own farms, farmers defected en masse from the collectives.
Achievement, Hesitation, and Renewed Reform

In 600 days, between the December 1978 Third Plenum and late summer 1980,
policy makers achieved astonishing results and locked in important economic
reforms. Despite this, toward the end of 1980, policy changed abruptly and
a much more conservative policy line emerged. Ironically, during this conserva-
tive period, policies already enacted unleashed a new economic dynamism that
propelled China into a new reform phase.
37
Zhao Ziyang 赵紫阳, “A Letter on Current Rural Policy,” June 19, 1980, in 赵紫阳文集
(Collected Works of Zhao Ziyang), vol. 1 (Hong Kong, Chinese University of
Hong Kong Press), pp. 44–5.
38
Q. Zhou, “How Deng’s Drama Unfolded (Part 1),” Caixin, August 27, 2019, 04:22 p.m.,
at www.caixinglobal.com/2019-08-27/how-dengs-drama-unfolded-part-1-101455513
.html.
744

Renewed Readjustment and Policy Vacillation

At the end of 1980, the emergence of inflation and renewed imbalances
triggered a new Readjustment.39 Officially, consumer price inflation in 1980
was 7.5 percent, and prices were certainly rising faster toward year’s end.
Chen Yun called for a do-over of the three-year Readjustment of 1979, which
he insisted had not been implemented thoroughly enough, and should be
restarted. The leadership fell in line behind Chen Yun: investment was cut
back sharply, effective in January 1981; many reforms were paused; and GDP
growth dropped to 5.1 percent, the lowest in the first decade of reform.
Macroeconomics drove policy change, but the shift was broad and long-
lasting, sustained through 1983. Austerity policies were adopted regardless of
their effect on market-oriented reforms: for example, cash balances of enter-
prises and local governments were frozen and borrowed by the central
government. Economic reforms came under sustained criticism and assault.
Until this time, despite disagreement about specific policies, the policy
environment had been buoyed by broad enthusiasm for vaguely defined
reform and the need to repudiate Cultural Revolution legacies. Now,
a systematic antireform view began to be articulated that linked reform to
corruption, crime, and a loss of national sovereignty. The SEZs, in particular,
came under sustained attack. Conservatives, including Chen Yun, criticized
the SEZs for breeding corruption and undermining national sovereignty. In
1982, central documents repeatedly declared that SEZs were premised on
maintaining the dominant position of public ownership and implementing
the same laws and basic economic system as the rest of China. This policy
principle threatened to cripple the SEZs’ role as a test bed for broader
reforms. Deng Xiaoping quietly acquiesced and personally signed off on
a 1983 crackdown on “spiritual pollution” that linked urban crime with
liberalization and foreign influence. Market reforms skidded to a halt.
A New Economic Dynamics

While politicians marked time, at the grass roots the economy was accelerat-
ing as the effects of the 1979–1980 reforms worked their way through the
system. After mid-1980, the agricultural PRS spread rapidly, and increasingly
took the more radical form of contracting land to households. From
39
This murky policy shift clearly had political as well as economic drivers. Chinese policy
makers were concerned about international issues such as the emergence of the Polish
Solidarity labor movement and the election of Ronald Reagan in the US. Domestically,
long-term personnel decisions after the removal of Hua Guofeng as Party head may
have touched off complex bargains among the most powerful elders, Deng and Chen.
745

barry naughton
14 percent at year’s end 1980, adoption jumped to 45 percent, then 80 percent

by the end of 1982, and it was finally nearly universal by 1983, at 98 percent of
all villages. With the spread of the production responsibility system from 1981
through 1984, agricultural growth accelerated. Economists have shown that
change in rural property rights accounted for the largest share of the prod-
uctivity increase.40 Agricultural output quickly responded. The vital grain
harvest surged from barely 300 million metric tons in 1978 to 387 million
metric tons in 1983, dispelling worries about China’s ability to feed itself.
Simultaneously, output of economic crops such as cotton and oilseed
doubled.
Equally significant, as farm households gained control of their cropping
and labor allocation decisions, they decided they could produce more crops
with fewer hours of (more intense) work. Farm households began releasing
family members for nonagricultural pursuits, notably to the TVEs which had
been thrown open during the course of 1979. Agriculture achieved the
remarkable result of rapidly expanding output while decreasing direct labor
inputs. Between 1978 and 1984, China created 80 million new jobs, almost
entirely by dramatically expanding labor-intensive subsectors. Roughly
a third were in newly diversified agriculture, a third in rural manufacturing
or construction, and a third in urban manufacturing or services. Urban
unemployment, which had surged in 1978–1979, now fell again, and availabil-
ity of consumer goods expanded dramatically. China could feed itself and
employ its massive population: rural reforms turned out to be a prerequisite
for China to sidestep the demographic tidal wave and reap the benefits of its
“demographic dividend.”
Growth of foreign trade was also rapid, although labor-intensive exports
took longer to begin their takeoff. Exports doubled between 1978 and 1982,
and while oil was still the largest item and mainstay of exports, labor-
intensive exports grew and slightly expanded their share. As these structural
changes fell into place, and the most drastic of the 1981 Readjustment policies
were relaxed, the economy accelerated. From 1983 through 1988, the econ-
omy roared ahead at over 10 percent per year, the first period of “miracle
growth” that China enjoyed.
40
J.Y. Lin, “Rural Reform and Agricultural Growth in China,” American Economic Review
82 (1992), 34–51, which also provides national PRS percentages. J. McMillan, J. Whalley,
and L. Zhu, “The Impact of China’s Economic Reforms on Agricultural Productivity
Growth,” Journal of Political Economy 97.4 (August 1989), 781–807. A. de Brauw, J. Huang,
and S. Rozelle, “The Sequencing of Reform Policies in China’s Agricultural Transition,”
Economics of Transition 12.3 (2004), 427–65.
746

China was enjoying quick economic returns from the abandonment of its
former, capital-intensive development strategy. Rural households pivoted
quickly to diversified household production and participating in nearby
township and village enterprises. In cities, young people were absorbed
into hastily created service and manufacturing jobs. New lower-cost combin-
ations of production factors that had previously been outlawed now experi-
enced explosive growth. In the TVE sector, growth was driven by
labor-intensive manufactures such as textiles and garments, furniture, and
plastic and metal products. These are the typical “early industrialization”
sectors that power the initial stages of economic development everywhere:
having skipped these sectors in its early command-economy growth, China
was now well placed to rapidly catch up. Vast labor-intensive sectors were
quickly created and opened up a powerful, sustainable growth path, which
used China’s underlying comparative advantage, and which China followed
for the next two decades and more.
These outcomes vindicated reformers and created a powerful argument
for further market-oriented reforms. Has there ever been a more unambigu-
ous demonstration of the power of incentives and markets than the adoption
of the PRS by Chinese farmers? Farmers worked harder and smarter with
control over their own land, and produced more with less. Chinese farmers
gave reform-oriented policy makers far more room for maneuver. In addition
to the demonstration effect, reforms unleashed new market forces that
created a demand for further reform. TVEs produced industrial goods and
services that competed with existing urban state-owned enterprises (SOEs).
TVEs were the opening wedge of an army of new entrants who broke down
the monopoly that SOEs had previously enjoyed over most industrial prod-
ucts. This created unprecedented pressure on SOEs and ultimately on the
government budget, which was dependent on profit remitted from SOEs.
Policy makers had to devise policies that would allow SOEs to respond to this
new competitive challenge. As economic conditions eased, the fact-based
argument in favor of reform strengthened, opening the way for a running
start to renewed market-oriented reforms.
Reform Renewal in 1984

In January 1984, Deng Xiaoping kicked off renewed reforms by visiting the
Shenzhen and Zhuhai SEZs. Having been a lightning rod for criticism, the
SEZs now served as a bellwether, symbolizing the fate of the entire economic
reform program. Emboldened by the success of rural reforms, Deng declared
the SEZs an unqualified success, and SEZ-like provisions were extended to
747

barry naughton
fourteen coastal cities in May. To drive the point home, the October 1
National Day parade in Beijing included a float from Shekou flouting its
most controversial slogan: “Time is money; efficiency is life.”
The job of translating enthusiasm into realistic policy making now fell to
the premier, Zhao Ziyang. Inevitably, his first step was to get buy-in from the
elders, especially Chen Yun. With pressure from Deng Xiaoping in the
background, and Zhao’s talent for active compromise very much in evidence,
all Standing Committee members signed off on a letter Zhao wrote on
September 9, 1984.41 At the subsequent Third Plenum in October, a broad
(but vague) new reform program was endorsed. China’s economy was
rechristened a “commodity economy” (a half-hearted version of a market
economy) and the scope of compulsory planning was to shrink, in favor of an
ill-defined “guidance planning” and increased market transactions. Enterprise
incentives and autonomy were to expand, and prices to be reformed.42 These
powerful ideas legitimated widespread reform experimentation, but didn’t
provide much concrete guidance: this was no reform blueprint. Instead, an
intense period of policy creativity and reform experimentation was launched
that lasted through the fall of 1988.
The 1980s
The five years from 1984 through 1988 were one of the most creative periods
of the entire reform era. Confronted with a treacherous political and eco-
nomic environment, Zhao Ziyang moved market-oriented reforms
decisively forward. Although the period was ended by the traumatic events
at Tiananmen Square in June 1989, the overall record was one of positive
achievement.
The Reform Challenge and the Political-Economy Matrix

Zhao Ziyang was in charge of day-to-day economic policy making, but above
him were the three powerful elders on the five-man Politburo Standing
Committee. Deng Xiaoping was generally an ally on economic reform issues,
but the support of Chen Yun and Li Xiannian was far from unconditional.
41
Zhao Ziyang 赵紫阳, “A Letter to the Standing Committee on Three Questions about
Economic System Reform,” September 9, 1984, in 赵紫阳文集 (Collected Works of
Zhao Ziyang), vol. 2, pp. 484–9, which includes the 批示 pishi (approval and commen-
tary) of the elders.
42
“Party Center Resolution on Economic System Reform,” October 20, 1984, in 赵紫阳
文集 (Collected Works of Zhao Ziyang), vol. 2, pp. 490–508.
748

Each of the elders had strong views on certain topics and substantial interest
groups arrayed behind them. Zhao had to achieve “buy-in” from many
powerful players, and many senior leaders could veto specific policies.
With many “veto-gates,” it was difficult to move policy decisively in any
given direction.43
At the same time, the economic policy choices had now become far more
complex than before. Reform could now “move into the city” from the
countryside, but the core of the planned economy was still intact. In this
system, state-owned factories dominated the industrial sector and responded
to plans and administrative commands from the government, and the gov-
ernment was still dependent on traditional institutions to steer the economy.
The state-owned factories made large profits, buying inputs at low state–state
prices and selling at high controlled prices. In turn, they turned the bulk of
those profits over to the government, which used them to fund its own
operations and invest in industrialization. The system was now fraying at the
edges, as firms had been given more autonomy and better incentives, but it
had not been fundamentally changed. Zhao Ziyang confronted the problem
of how to transform the core of the system.
It is customary to divide the elders into a reform camp under Deng
Xiaoping and a conservative camp under Chen Yun. There is much truth in
this, but it is important not to oversimplify the situation. Chen Yun was
conservative in important respects: he believed in cautious macroeconomic
policy, balanced budgets, low inflation, and government steerage of the
economy. Moreover, he was extremely sensitive to corruption and specula-
tion, and suspicious of capitalists and foreign businesses. Yet Chen was also
appreciative of the power of markets, and sensitive to the use of market prices
as signals of relative scarcities. He had long advocated for markets on a small
scale, for example for farm products. Chen famously likened the Chinese
economy to a “bird in a cage,” in which the job of planners like himself was to
make the cage as commodious as possible, but never to allow the bird to fly
away.
Chen Yun never systematically laid out his mature views on the socialist
economy, but we can extrapolate from his earlier views to sketch a version of
“conservative reform.” The breakthrough reforms of 1978–1982 had resolved
the immediate crises of food shortage, underemployment, and scarcity of
foreign exchange. The feeble institutional capacity and huge ideological
43
749

barry naughton
hurdles inherited from the Maoist era had been ameliorated, and the
small-scale markets Chen Yun had long advocated had been resuscitated. In
this environment, conservative reformers would have focused on improving
management in state firms and shoring up the state monopoly of select
critical sectors, while gradually stepping up planning. Such an approach
could have sustained rapid growth for a while, and left China with
a dualistic economy in which a modern state-dominated industrial sector
coexisted with a sprawling small-scale private rural, handicraft, and service
sector. Such a system is common in many developing countries, and would
have been a completely plausible outcome in China. However, such an
approach never really materialized in China. It would have required that
planning and state ownership be strengthened at a time when the driving
economic force was coming from dispersed, generally small-scale, labor-
intensive manufacturing and services that didn’t benefit at all from planning
or public ownership. It would have required somebody with remarkable
vision and determination to drive the economic system in this coherent, but
counterintuitive, direction. As Chen Yun aged and stepped back, and as
younger thinkers were increasingly taken with comprehensive marketization
and the promise of really dramatic productivity improvement, no such leader
emerged in China. As a result, the conservative camp became a brake on
reform, but never developed its own independent program.
The core industrial system confronted policy makers with an enormous
mass of entangled issues: how to reform industry (especially state-owned
industry), the fiscal and financial systems, and the price system. Policy makers
had to come up with a coherent package of policies, and they also had to
mobilize the political will to make difficult and disruptive changes. In the
USSR and Eastern Europe, especially Poland, an influential body of opinion
held that the only way to carry off such a difficult policy exercise was to do
everything at once, moving quickly so the economy could recover from the
shock and adapt to new economic and property relations. Typically, this “big-
bang” approach involved rapid price decontrol, with price liberalization and
macroeconomic stabilization combined, followed by rapid privatization.44
Rapid big-bang transition never had much appeal in China. The immense
pressure to find jobs for the rapidly growing labor force plus the intuitive
understanding that China had strong growth potential limited the economic
44
For example, O. Blanchard, R. Dornbusch, P. Krugman, R. Layard, and L. Summers,
Reform in Eastern Europe (Cambridge, MA, MIT Press, 1991); J. Sachs and W.T. Woo,
“Structural Factors in the Economic Reforms of China, Eastern Europe, and the
Former Soviet Union,” Economic Policy 4.1 (1991), 101–45.
750

appeal of the big bang. Politically, the desire to maintain Party control and
social stability meant that big bangs could not be countenanced. What was
Zhao Ziyang to do?
Looking for Room to Maneuver

Seeking to find a path between systemic stagnation and big-bang destabilization,
Zhao Ziyang was receptive to extraordinarily diverse influences. Zhao listened
to foreign economists, seeking institutional advice from the World Bank, and
from individuals as diverse as Milton Friedman and Alvin Toffler.45
Domestically, Zhao reached outside the economic bureaucracy to groups not
burdened by institutional interests, essentially empowering competing brain
trusts to suggest new policies. One group of senior economists, which evolved
into the Development Research Center, was informally led by Xue Muqiao, who
had long practical experience dating back to the 1940s. These veterans, given
their familiarity with socialist economic categories, suggested key transitional
concepts, such as the “commodity economy,” which legitimated initial steps
toward marketization even when conservatives vetoed the idea of a market-led
economy.46 Another group of young economists, predominantly in their thirties
and with no link to the traditional bureaucracy, provided many bold new ideas
that Zhao eagerly absorbed. Organized into the System Reform Institute and the
Rural Development Research Institute, these iconoclastic and activist economists
dreamed up new initiatives, and communicated intensively outside official
channels.47 Zhao presided over what sometimes seemed like a perpetual brain-
storming session. This reflected the excitement, idealism, and diversity that
characterized the 1980s more broadly in China, but also the extraordinarily
complex political and economic landscape through which Zhao had to navigate.
The Emergent Strategy of 1980s Economic Reform

Zhao Ziyang was able to cobble together a coherent approach to reform. At
its core, Zhao’s approach was to infuse the entire Chinese system with
incentives for economic growth. During the Cultural Revolution, virtually
45
A. Wood, “China: Long-Term Development Issues and Options, Past and Present,”
University of Oxford, TMCD working paper series No. 079 (January 2019). J. Gewirtz,
Unlikely Partners: Chinese Reformers, Western Economists, and the Making of Global China
46
B. Naughton, Wu Jinglian: Voice of Reform in China (Cambridge, MA, MIT Press, 2013).
47
I.M. Weber, How China Escaped Shock Therapy: The Market Reform Debate (New York,
Routledge, 2020). The September 1984 Moganshan meeting was a pivotal moment for
this group. Liu Hong 柳红 (ed.), 莫干山会议 (The Moganshan Meeting) (Beijing,
Dongfang, 2019), contains memoirs and documents.
751

barry naughton
all individual incentives had been sublimated to collective political objectives.

Early reforms had restored basic incentives to industry – things like hourly
and piece rates for workers and bonuses for managers – but now Zhao went
much further. Indeed, Zhao sometimes pushed it to extremes, supporting
a vast range of incentive mechanisms, tailored to specific situations, that
provided rewards for those who worked harder or smarter.
Industrial enterprise reform – pioneered in the late 1970s by Zhao when he
was still Party boss in Sichuan province – had at its core a system of profit
retention. In the mid-1980s, these reforms spread to almost all state firms,
providing firms with retained funds that they could use for bonuses, worker
housing or services, and investment.48 Within the firm, the “factory manager
responsibility system” gave managers top authority, displacing the Party secre-
tary. Within government, provinces signed up for a variety of budgetary deals
with the center that allowed them to retain revenues above a stipulated control
figure. Universities and research institutes were encouraged to set up subordin-
ate profit-making companies to supplement their budgets at a time when the
government couldn’t contribute much. Every sector had its own incentive
mechanisms.49 Incentives went beyond the purely financial: local government
officials began to have economic growth and fiscal revenue targets written into
their performance contracts. Promotion, it was made clear, would depend on
successful local economies.
Stronger incentives heightened tensions within the highly distorted bureau-
cratic economy. Prices were fixed and diverged widely from costs, and economic
opportunities were very unequally distributed. As incentives were strengthened
in general, individuals also had stronger incentives to chase rent-seeking oppor-
tunities, bend rules, and become corrupt. Two mechanisms were introduced to
contain these defects: “particularistic contracting” and the dual-track system.
Particularistic contracting occurred when enterprises (and organizations and
local governments) agreed to turn over to their superiors a specific amount of
profit or revenue, typically based on the previous year’s performance.50 This
equalized opportunity to some extent, and also allowed marginal retention rates
to be higher, further reinforcing incentives. The inspiration obviously came from
the rural reforms, and it was easy to implement. Profit contracting, however,
enmeshes firms in complex bargaining relations with their superiors – what will
48
Naughton, Growing Out of the Plan, pp. 97–135, 200–243.
49
Tian Jiyun 田纪云, “Ten Major Economic Reform Measures of the 1980s,” in Xu, The
Chinese Experience, pp. 56–78. Tian was vice premier during this period.
50
S. Shirk, The Political Logic of Economic Reform in China (Berkeley, University of
California Press, 1995).
752

next year’s contract be? – and it inevitably has a short half-life in a dynamic
economy.
Firms were allowed to operate on a “dual track.” They were required to meet
their legacy obligations – producing planned output with planned inputs and
selling at plan prices – but could also sell outside the plan at market prices
(conditional on fulfilling their plan). Thus every commodity had two tracks and
two prices (plan and market). Chinese planning had always been weak, so firms
generally had some surplus capacity to produce outside the plan. In the extreme
case, with a fixed plan, all of the action takes place on the market track, and the
firm faces market prices on the margin. With particularistic contracting and the
dual track, everyone is as well off as they were before (because of the contract),
but has been given a new opportunity to expand in a market direction. With
economic growth, the market track will expand and the economy will “grow
out of the plan.”51 The dual-track system also allowed state-owned enterprises to
trade at market prices with TVEs, which had never really been planned. Thus it
contributed to market integration, tearing down the barriers that protected the
monopoly industry system.
However, this system had two major defects. First, it made corruption much
easier, since simply reselling a low-price plan good on the high-price market could
be very lucrative and was hard to monitor. Second, particularistic contracting
made it harder to reform the economic parameters that govern the overall
economic system, such as the tax system. In essence, every enterprise had
a different tax rate. The dual track could drive prices to their market levels, but
it could never drive taxes, for example, to levels that were either uniform or
appropriate.
Facing these problems, Zhao Ziyang heard conflicting advice from propon-
ents of two very different approaches. The young economists around the System
Reform Institute proposed using the dual-track system as a full-blown transition
strategy. Recognizing its defects, they nonetheless argued that this would
incentivize managers to push for more autonomy and market-friendly strategies.
They argued that the economy was capable of rapid growth and could withstand
short-run disruptions and inflation, and urged quick movement through to
a market economy.52 A competing group of more established, usually older,
51
Naughton, Growing Out of the Plan. W. Byrd, “The Impact of the Two-Tier Plan and
Market System in Chinese Industry,” Journal of Comparative Economics 11.3 (September
1987), 295–308.
52
Both sides were reformers committed to a transition toward a market economy, and
neither side had sympathy for a “big-bang”-type reform. The young economists
preferred to lead with incentive and ownership reforms, delaying price reform; the
753

barry naughton
economists argued strongly that the priority was to first establish a “relaxed” low-
inflation macroeconomic environment. Once that was achieved, a two-step
price and tax reform should follow: an immediate, ambitious adjustment of
some of the most distorted prices – such as low energy prices – followed
promptly by full price liberalization. The debate was formulated in terms of
underlying principles and institutions, and often styled as “ownership reform”
versus “price reform.” In fact, differences over short-term macroeconomic policy
were often the focus of heated debate. As is so often the case, the policy
outcomes were not decided by the quality of the competing arguments but
rather by fluctuating macroeconomic conditions and by treacherous political
circumstances.
Economic Policy Cycles

Through the 1980s, there had been a clear pattern of macroeconomic
cycles, visible in Figure 20.1 below.53 Inflation peaked in 1980, 1985, and
1988, and later in 1995. Each of these inflation peaks also marks a phase of
a policy cycle, coming after a wave of reforms. Four clear combined policy
and macroeconomic cycles mark the first twenty years of reform. These
cyclical processes reflect the intrinsic reform problems discussed earlier.
Reforms lead to a surge in social and economic demands. This is particu-
larly true when firms have “soft budget constraints,” meaning that they
might avoid repaying their debts. Enterprise managers with soft budget
constraints will exploit their increased freedom of action to increase their
investments (which provide higher profit and status).54 Reform and decen-
tralization thus lead to rapid expansion of aggregate demand, outrunning
supply, and ultimately to inflation and/or shortages. Policy makers react to
these imbalances with top-down re-control, suppressing investment and
aggregate demand, and reversing or scaling back decentralization and
reform. After de-control succeeds, the stage is set for a new phase of
decentralization, reform, and expansion.
The distinctive feature of Chinese cycles is that phases of the cycles
correspond with regular shifts in the influence of top policy makers. Given
the dispersal of elder power between reformers (led by Deng) and conserva-
tives (led by Chen Yun), policy-making authority oscillated among these
older group preferred to lead with price reform (including parameters like taxes),
deferring incentive and ownership reforms.
53
National Bureau of Statistics, derived from monthly data at www.stats.gov.cn/tjsj.
Only annual data are available for 1978 through 1982.
54
J. Kornai, The Shortage Economy (Amsterdam, North-Holland, 1980).
754

policy makers along with the economic cycle. When conditions were favor-
able, reformers (Zhao Ziyang first among them) would seize the initiative,
often with Deng Xiaoping creating a protective umbrella of pro-reform
rhetoric. Reforms would generate an explosion of activity, optimism, and
accelerated growth, but would also lead to increased inflation and imbal-
ances. At this point, Chen Yun would find his voice, speaking out for
economic stability and against speculation. Conservatives would demand
macroeconomic adjustment, seek to restrict credit and investment, and
attack reforms which they had never fully accepted. It is striking that as
one side stepped forward, the other side would step back, avoiding an
irreconcilable split among the elders. Conservatives stepped in to damp
down the irrational exuberance of the high-reform phase, but without ever
really articulating an alternative vision of “conservative reform.”
The policy episodes described earlier fit easily into this cyclical framework,
which has long been noted by China analysts.55 In the first cycle, vigorous
reforms in 1980 led to imbalances and inflation that triggered a major inter-
vention by Chen Yun and a multiyear backlash against reforms. In the second
cycle, Deng’s visit to Shenzhen in 1984 presaged major reforms in 1984–1985,
followed by a surge of inflation and an unprecedented trade deficit. Under
pressure, Zhao Ziyang maintained day-to-day authority and carried out
partial re-control in 1985–1986. This second cycle was not quite complete
when Zhao contemplated adopting a “comprehensive reform package,”
tabled by the price reform group in late 1986. To have accepted the price
reform package, Zhao would have had to strictly carry through the macro-
economic control policies associated with the conservative group. Under
great political and economic pressure, Zhao declined to maintain tough
macro-control or adopt the comprehensive price reform program. His
default acceptance of profit contracting pleased many of his supporters, but
contributed to the acceleration of inflation that became evident within
months. Zhao was finding it difficult to thread a path between these conflict-
ing forces, and building pressures were soon to overwhelm him.
The Mid-term Crisis of Reform, 1989–1991

China had moved onto a high-growth track in the mid-1980s, which Zhao
Ziyang was understandably hesitant to disrupt. Inflation gradually acceler-
ated through 1987 and 1988 (Figure 20.1), and the gap between market and
55
J. Fewsmith, Dilemmas of Reform in China: Political Conflict and Economic Debate (Armonk,
NY: M.E. Sharpe, 1994).
755

barry naughton
30
Percentage increase over year previous periods

25
20
15
10
0
1979
1982
1984
1987
1990
1993
1996
1999
2002
2005
2008
2011
2014
2017
–5
Figure 20.1 Consumer inflation (1979–2019)
Note: quarterly data begin in 1982; annual data for 1979–1981 shown as gray squares.
plan prices also widened, creating unprecedented opportunities for corrup-

tion through arbitrage between the two prices. Public discontent increased,
and the more conservative elders, headed by Chen Yun, became disaffected.
This complex economic situation interacted with some sensitive political
events.
A cascade of political events began in December 1986 when Deng Xiaoping
deposed Hu Yaobang, the Party head and nominal top leader. Deng had
grown dissatisfied with Hu Yaobang’s performance in handling student
protests and other issues. Despite some differences over economic policy,
Hu Yaobang and Zhao Ziyang overall had formed a powerful team that
strongly favored reform. Zhao Ziyang was now “promoted” to take Hu’s
position as general secretary. At first, Zhao thought he could control the
situation, and the Thirteenth Party Congress in October 1987 announced an
unprecedented theoretical program for further reform. China was declared to
be in a “primary” stage of socialism, in which production forces were
underdeveloped and institutional diversity could be embraced. Led by
Deng Xiaoping and Chen Yun, all of the elders retired from the Standing
Committee, the apex of formal power.
However, Zhao’s promotion turned out to be a poisoned chalice. Zhao
gained in nominal authority, but the elders maintained their influence behind
the scenes and Zhao lost a key ally in Hu Yaobang. Moreover, when
756

promoted to Party general secretary, Zhao had to turn over direct control of
economic policy to his successor as premier, Li Peng. When surging inflation
led to a new re-control policy in September 1988, Zhao Ziyang was caught in
a power squeeze. Chen Yun was peering unhappily over his shoulder; Deng
Xiaoping was urging bold reforms; and the policy of re-control was actually in
the hands of two conservatives, Li Peng and Yao Yilin. Top-down controls
damped investment and growth, and quickly pushed the economy to its
maximum pain point, with growth slowing before inflation had begun to
come down.
It was in this fraught economic and political environment that, on April 8,
1989, Hu Yaobang had a heart attack in a Politburo meeting, and died in the
hospital a week later. Scattered student demonstrations exploded into mas-
sive nationwide memorials and protests, as demonstrators called for the
social and political liberalization advocated by Hu Yaobang, and an end to
corruption and inflation. The situation spiraled out of control: Zhao Ziyang’s
unwillingness to use military force against protesters in Tiananmen Square
on June 4, 1989, led to his deposition, the violent suppression of the demon-
strations, and the loss of power of an entire cohort of pro-reform policy
makers.
From June 4, 1989, through the end of 1991, conservatives dominated
economic policy. They blamed economic reform for a dazzling range of
negative outcomes: inflation, shortages of energy and key producer goods,
liberalism, and student unrest. Conservatives interpreted inflation and spor-
adic shortages as symptoms of deep structural imbalances and, to a surprising
extent, sought to roll back economic reforms. To control inflation, they
advocated stronger planning and increased supply of energy and key heavy
industrial products. Price controls were imposed on outside-plan markets,
and financial balances of both enterprises and local governments were frozen.
Conservatives were openly hostile to private business, and the small emer-
gent urban private sector shrank by half between 1988 and 1990 under
combined political and economic pressure. In rural areas, collective institu-
tions were re-emphasized and farmers pushed to join so-called “new co-
operatives.” The economy slowed to a standstill in the second half of 1989 and
the first half of 1990, as it worked through these drastic policy changes.
Conservatives were hobbled by their lack of a real program. The acute
issues they were most concerned with quickly disappeared: the inflation rate
dropped and even turned briefly negative in mid-1990, while shortages of
producer goods and electricity evaporated. Struggling to put together a plan,
a newly re-empowered Planning Commission drafted a Five-Year Plan for
757

barry naughton
1991–1995 that prioritized recentralization and a new program of investment

in “basic industries.” It estimated that GDP could grow at 6 percent annually,
and worker wages should grow only 2 percent annually.56 In fact, planners
grievously underestimated the potential of their own economy, which would
grow twice as fast in the next five years. It wasn’t the same economy as in
1978–1979 any more. It was three times as big, with a huge and vibrant small-
scale sector, and with much more resilience and flexibility, the direct result of
a decade of successful economic reform.
The Reform Breakthroughs of the 1990s

The years of conservative dominance after June 4, 1989, disappeared with
remarkable speed, and were then conveniently forgotten. The conventional
turning point is the “Southern Tour” (nanxun 南巡) that Deng Xiaoping took
in January–February 1992, in which he rebuked ideological opposition to
economic reforms and declared that “development is the only hard truth.”
Deng proclaimed that markets existed under both socialism and capitalism,
and therefore should not be subject to ideological litmus tests. These pro-
nouncements touched off a new round of reform. Clearly this was a catalytic
event, and Deng seemed to effortlessly overturn the entire policy orientation.
For an eighty-eight-year-old man, it was a remarkable display of political
theater.
But of course, the change ran much deeper. Through 1991, the ground had
been shifting underneath Chinese decision makers. The new Party leader,
Jiang Zemin, was trying to establish his own authority and gradually under-
stood that this would require a more pro-market position. Growth was
accelerating as conservative controls on the economy fell away. However,
Deng Xiaoping worried about his legacy as he saw that changes were being
blocked by conservative leaders who operated under Chen Yun’s patronage,
or at least believed that they did. Increasing frustration and anger on Deng’s
part led him to make his historic journey.57 Thereafter, everything changed
with breathtaking speed. The 1992 Party Congress forthrightly declared that
the reform objective was a “socialist market economy,” and in October 1993
56
Liu Guoguang, 中国十个五年计划研究报告, pp. 547–602, esp. 552, 557.
57
Zhou Ruijin 周瑞金, “改革开放 30 年的思考” (Reflection on 30 Years of Reform and
Opening), Reuters China, October 8, 2008, at www.reuters.com/article/idCNChina-2508
820081009; S. Zhao, “Deng Xiaoping’s Southern Tour: Elite Politics in Post-Tiananmen
China,” Asian Survey 33.8 (1993), pp. 739–56; Ma Licheng 马立诚, 交锋三十年 (Thirty Years
of Confrontation) (Nanjing, Jiangsu renmin chubanshe, 2008), esp. pp. 150–1.
758

a new Third Plenum adopted a comprehensive program to be implemented

by century’s end. Policy had swung 180 degrees in less than two years. Most
remarkable, almost all the proposed measures were actually implemented.
A Dizzying Shift in Political and Economic Conditions

What led to this remarkable change? First, the sense of crisis, already palpable
after the Tiananmen events, was intensified by internal and external events.
Internally, the erosion of state capacity continued unabated. To a certain extent,
the reform strategy followed under Zhao Ziyang was responsible for an inevit-
able erosion of budgetary revenues as a share of GDP and a weakening of the
instruments of the command economy. Reforms had relaxed the state monop-
oly over industry and allowed new entrants to compete away monopoly profits
on which government revenues depended. Conservatives had pledged to
reverse this trend but failed, and budgetary revenues were again declining as
a share of GDP (Figure 20.2).58 Externally, the gradual unraveling of the Soviet
system accelerated and culminated in the shocking dissolution of the Soviet
Union in December 1991. Socialist self-confidence was profoundly undermined.
Simultaneously, the ease with which the US military pushed Iraq out of Kuwait
(in February 1991) showed that the US had mastered a “revolution in military
affairs” that other countries, including China, hadn’t even begun. The conserva-
tive response to this internal and external turmoil was to hunker down and
reassert the purity of socialist ideals, but Deng Xiaoping and many others
understood that only accelerated reform could address the challenges.
Worldwide, it was obvious that market economies had outperformed command
economies, and the prestige of the US model was at an all-time high. Since the
conservatives had no viable economic program, their exposed position collapsed
when Deng pushed on it, and the policy spectrum narrowed substantially.
Second, within China, the makeup of power was changing. From the
Third Plenum through the end of the 1980s, a group of revolutionary elders
had held ultimate authority in a variety of fields, including the economic.
Now, the inevitable passage of time was bringing this dispersion of ultimate
authority among elders to an end. Several influential elders passed away in
1992–1993, including Li Xiannian, Hu Qiaomu, and Wang Zhen, and the
Central Advisory Commission, which had served as an instrument of elder
power, was abolished. Finally, Deng used this situation to push back at Chen
Yun in unmistakable fashion, using special economic zones as a symbolic
58
National Bureau of Statistics, 中国统计摘要 2019 (China Statistical Abstract 2019)
(Beijing, Zhongguo Tongji), pp. 68, 22.
759

barry naughton
35%
30%
Expenditures
25%
20%
15%
Revenues
10%
Fiscal
reform
5%
0%
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
Figure 20.2 Budgetary revenues and expenditures (share of GDP)
issue. Deng said, “when we set up the four SEZs . . . we didn’t think about
Shanghai’s superiority in human capital . . . if we had made Shanghai an SEZ
back then, it wouldn’t look [as backward] as it does today.”59 In fact, it was
Chen Yun who had insisted on the exclusion of Shanghai from the ranks of
SEZs. Now, Deng used the death of Li Xiannian to suggest to Chen Yun that
he make a gesture of support for the new policies. Chen Yun did so in his
“Memorializing Li Xiannian,” which was printed on the front page of the
People’s Daily. “We didn’t recognize the full importance of SEZs,” said
Chen.60 It signaled Chen’s capitulation and marked the end of an era of
elder dominance of policy.
59
“Key Points of Talks in Wuchang, Shenzhen, Zhuhai, and Shanghai,” January–February
1991, in 邓小平文选 (Selected Works of Deng Xiaoping), vol. 3 (Beijing, Renmin
chubanshe, 1994), p. 366.
60
Chen Yun, “悼念李先念同志” (Mourning Comrade Li Xiannian), People’s Daily,
July 23, 1992, 1. Reprinted in 陈云文选 (Selected Works of Chen Yun), vol. 3,
pp. 378–9. Chen’s personal secretary, Zhu Jiamu 朱佳木, tells of Deng’s suggestion
to Chen, “Chen Yun and Deng Xiaoping in the Early Reform Period,” Dangdai
Zhongguoshi Yanjiu 17.3 (May 2010), 14.
760

Finally, as the economy recovered from the post-Tiananmen crackdown, it

became clear that the structural conditions of the economy were changing.
The demographic tidal wave that had so worried policy makers had begun to
ebb, and the extraordinary growth of the labor force from 1964–1990 was
finally beginning to ratchet down. Jobs had been found in new labor-intensive
sectors, which made the economy stronger and more productive. The
urgency and difficulty of finding jobs for urban dwellers had declined.
Labor market reforms that might increase short-term unemployment
didn’t seem as inconceivable as they once had. Even more consequentially,
it might become possible to lower the barriers keeping rural workers from
migrating to the city (the hukou system). New economic opportunities were
opening up at the grass roots.
A New Turning Point

These rapid but profound shifts in political and economic conditions initially
outpaced the ability of policy makers to design appropriate responses. As the
policy environment liberalized, in the wake of Deng’s Southern Tour, there
was a nationwide scramble to take advantage of new opportunities. Local
government officials had already been incentivized to support economic
growth, and they had seen national GDP growth surge to over 10 percent
during the 1980s. Now, in 1992, growth was surging again – ultimately to
14 percent for the year – and there was a wild scramble to be part of this
growth. “Zone mania” spread, as local officials set up literally thousands of
“special zones” in every Chinese province, which gave them a flexible tool to
attract investment from outside, and generate lucrative development con-
tracts for friends and relatives.
Economic growth was accommodated with rapid credit growth, and
inevitably led to a new wave of inflation. As Figure 20.1 shows, inflation
began to creep up during 1992, and then took off, finally peaking in the fourth
quarter of 1994 at 27 percent. Inflation caused other problems: policy makers
had lowered the official nominal value of the renminbi (RMB) between 1989
and 1991, but inflation subsequently caused the real exchange rate to appreci-
ate, leading to a current-account deficit, and a modest drawdown of reserves.
Increased disequilibria led to a bigger role for the existing “swap market” for
foreign exchange, where transactions took place at a relatively uncontrolled
market rate. The swap market RMB plummeted to more than ten to the US
dollar for a period in 1993, opening up a huge difference with the official rate
and pulling almost all transactions into the swap market.
761

barry naughton
In some respects, economic control collapsed. It was during this period

that an image of “Wild West” Chinese capitalism was created, the sense that
anything was possible for those with money and influence. Early movers, in
a position to take advantage of unprecedented opportunities and lax supervi-
sion, established new private fortunes. In Shenzhen, a curious incident
occurred when officials pushed out ahead of national policy by establishing
a local stock market. Stock prices were soaring along with inflation, and there
was huge demand for shares. Local officials decided on a lottery, selling
10 million tickets for 100 RMB each, with each ticket giving a 10 percent
chance of buying a share. The gambling aspect only added to the appeal, and
a million people lined up the morning of August 10, 1992, but tickets sold out
within hours. For forty-eight hours, angry crowds roamed the city, denoun-
cing insider trading and corruption. The following day an additional tranche
of 10 million tickets was released, a corruption investigation was announced,
and the city gradually calmed down.61
Anxiety and uncertainty mounted as inflation rose to a peak in 1994.
Coincidentally, the inflation peak was almost identical to that in the fourth
quarter of 1988, but this time there were no surviving elders in a position to
demand a reform rollback. Zhu Rongji had been brought to Beijing from
Shanghai in 1991 as vice premier to manage the economy, and in July 1993 he
concurrently became head of the People’s Bank of China, the central bank, to
strengthen its authority and ability to stand up to local interest groups. Jiang
Zemin was able to consolidate his position at the top, with Zhu serving as top
economic policy maker, while Premier Li Peng stayed on as a (weakened)
counterweight. This created a much more decisive political system, a big
change from the dispersed, elder-dominated system of the 1980s, and Zhu
was able to take bold economic measures. Even so, inflation remained
a serious problem in the background for the next five years, finally dropping
below 5 percent in 1997. Steady restraint in credit and monetary policy was
indispensable, and relatively tight monetary policy became an essential
feature of the market reform measures of the late 1990s. Overall, then, the
political outcome of inflation was the opposite of that in 1989: it contributed
to the consolidation of a group of market-oriented leaders and to the policy-
making breakthrough at the 1993 Third Plenum.
61
Tao and Lu, 中国经济特区史论, pp. 77–8.
762

Another Third Plenum: A Reform Blueprint for Systematic

Reforms
At the Third Plenum of that Central Committee in the fall of 1993 policy
makers adopted a resolution that was a blueprint for the next stage of market-
oriented reforms.62 This contrasted sharply with earlier policy making, in
which measures had almost always been incremental, and grand reform plans
had repeatedly failed to gain traction. Remarkably, no authoritative agency
within the government was capable of producing the building blocks of this
further stage of reform. Instead, a loose grouping of outside economists
produced policy papers for the Party’s Economics and Finance Leadership
Small Group, which drafted the resolution.63
The reform policy debates that had previously divided the economic
community now receded: two inflationary episodes, combined with the
collapse of most planning instruments, made many previous contentious
issues obsolete. Economists now divided their policy recommendations into
several functional areas which could be tackled separately, sidestepping the
sequencing problems and need for prior macroeconomic stabilization that
had dogged the earlier comprehensive reform programs. The year 2000, only
seven years distant, was put forward as the target date for the creation of
a “nationwide integrated and open market system.” The resolution contained
fifty articles, the majority of which included a concrete, operational objective.
Remarkably, most of these operational goals were achieved. Table 20.2
shows twelve key provisions with an assessment of the extent to which
implementation was ultimately achieved. Key milestones are shown. In
most cases, implementation was carried out through specific measures that
are recognizably those envisioned by the drafting groups.
As Table 20.2 shows, the document contained both broad goals and
immediate actions. For the first time, implementation included a substantial
role for legislation, including foundational laws on corporations and banks.
While immediate legislation was needed to create a market regulatory
system, it was recognized that actually building a credible system was a long-
term and incremental process. The document was thus a realistic mix of
immediate, incremental, and long-term objectives. While earlier programs
62
Communist Party Central Committee, 关于建立社会主义市场经济体制若干问题
的决定 (Resolution on Several Issues of Establishing a Socialist Market Economic
System) (November 14, 1993) (Beijing, Renmin chubanshe, 1993).
63
In addition, the independent think tanks patronized by Zhao Ziyang were disbanded
after June 1989, while the official think tanks were paralyzed by ideological rectification.
Ironically, financial support from the Ford Foundation allowed one cluster of econo-
mists to remain active. Chen, 国事忆述; Naughton, Wu Jinglian, pp. 251–60.
763

Table 20.2 Twelve key provisions of the 1993 Resolution on Creating a Market System
Article Content Implementation Milestone
6 Corporatization of large firms, especially state-owned enterprises Yes, mostly 1994 Company Law
8 New ownership agency for state-owned enterprises Yes, late and 2003 State Asset Commission
partial
9 Diverse ownership economy; public ownership as “mainstay” Yes Incremental
11 Prices: market prices, handful of government-set prices Yes Immediate
12 Creation of wholesale, retail, and futures market institutions Yes Incremental
15 New regulatory agencies for fair market competition Yes, mostly Incremental
16 Government functions shift to indirect and regulatory control Yes, mostly 1998 government re-organization

18 Fiscal: increase revenues; unify and lower tax rates Yes 1994 fiscal reform
19 Accelerate financial system reform; separate central bank and Yes 1995 laws for central, commercial
commercial banks; new development banks and development banks
22 Macroeconomic control centralized Yes 1994 fiscal reform and
1998 financial reorganization
23–4 Wages: more flexible wage-setting; greater differentiation Yes Incremental
26–8 Create a multilayer system of social security No, mostly –
had been broad, vague, and sometimes overly ambitious, this one initiated
a multiyear program of market creation and market building.
The other fundamental difference between this reform program and 1980s
reform is that this program went forward under the presumption that the
central government would be strengthened. The pressure of inflation meant
that control over financial institutions should be tightened and centralized,
and credit and monetary policy exercised more authoritatively by an
empowered central bank. The fiscal power of the state needed to be strength-
ened since revenue had plummeted as a share of GDP (Figure 20.2). Thus,
while in the 1980s the most urgent need had been to begin dismantling an
omnipresent command-economy system, by the 1990s the command econ-
omy was largely gone, and the challenge to policy makers was to devise more
effective instruments, to strengthen the state without reverting to direct
controls. Fiscal reform was perhaps the most urgent. The proposed solution
was a mixture of tax reform and recentralization. Tax reform involved
shifting to lower and more uniform rates. The centerpiece was the value-
added tax (VAT), imposed in principle on all forms of economic activity.
Income tax rates for enterprises were unified across different ownership
types, and a personal income tax introduced.
The most contentious part of fiscal reform was the allocation of new taxes
to different levels of government, since this involved recentralizing taxation
authority (especially that over the VAT). In fact, the reform was designed so
that the central government would take a much larger share than its own
expenditure needs, and then redistribute the surplus to local governments.
This would obviously give the central government more control, and a much
stronger position vis-à-vis the local governments. Having just recently been
freed from the shackles imposed during the conservative ascendancy, local
officials were understandably reluctant to surrender fiscal resources. Zhu
Rongji spent over a month flying to most provincial capitals, jawboning
governors and Party secretaries, pressuring them to give up control over
their revenues in the national interest. Evoking national crisis, Zhu was
ultimately able to prevail.
To be sure, representing the national government, Zhu Rongji had abun-
dant resources with which to achieve the acquiescence of local government
officials in this recentralization. The center would guarantee a minimum
level of resources for each province, so they would not be worse off in
absolute terms. More fundamentally, Zhu was able to offer local officials
unprecedented freedom of action to operate more entrepreneurially, espe-
cially with respect to land development and investment. A new urban
765

barry naughton
economic system gradually grew from this, in which local governments

parlayed their control of urban land into a developmental resource. In the
long run, fiscal reform and recentralization achieved a dramatic turning point
in the trend of fiscal revenues to GDP (Figure 20.2). Moreover, by putting the
national budget on a sustainable foundation, Zhu was able to shift the battle
against inflation onto the monetary realm, enabling the gradual control of
inflation.
By the late 1990s, China had established the basic infrastructure necessary
to run a modern market economy. Some parts of the new system were
embryonic. Capital markets in general were underdeveloped, as the Shanghai
and Shenzhen stock markets made a modest and wobbly start, and the
issuance of corporate bonds was tightly restricted. Overall, though, the
pace of institutional construction was extraordinary. The panoply of new
institutions established a roughly level playing field, creating conditions
under which firms of different ownership forms could compete and
co-operate under common rules. China had emerged from the post-
Tiananmen chaos with a workable government commitment to market-
oriented economic reform, and followed through on this commitment.64
The High Tide of Reform

A final stage of heightened reform activity came in 1998–2000. Zhu Rongji
took over as premier in 1998 and increasingly put his personal stamp on
policy. Real GDP growth had been extraordinary during the 1990s, well over
10 percent annually from 1993 through 1996. However, inflation had also
remained high, and, with a sustained effort to control inflation, growth began
to slow in 1996 and 1997. Almost as soon as the banking system had been
restructured, an effort was made to tighten access to credit by commercial
banks, which in turn had no choice but to restrict the flow of credit to
enterprises. In essence, a hard budget constraint was being imposed on
banks, which passed on that hard budget constraint to the enterprises.
Market conditions shifted from chronic excess demand to a relative buyer’s
market, and firms were put under pressure to reorient their business strategy
and become more market-responsive (the shift in inflationary dynamics is
64
From this time, the English-language economic literature expands dramatically and
many good overviews are available. A. Kroeber, China’s Economy: What Everyone Needs
to Know, 2nd ed. (Oxford, Oxford University Press, 2020). B. Naughton, The Chinese
Economy: Adaptation and Growth, 2nd ed. (Cambridge, MA, MIT Press, 2018). L. Brandt
and T. Rawski (eds.), China’s Great Economic Transformation (Cambridge, Cambridge
University Press, 2008). J. Wu, Understanding and Interpreting Chinese Economic Reform,
2nd ed. (Singapore: Cengage Learning Asia, 2015).
766

clearly visible in Figure 20.1). Just as market conditions were softening, the
unexpected arrival of the Asian financial crisis in the second half of 1997
brought an additional shock. Growth slowed to below 8 percent in 1998 and
1999, and firms were under unprecedented market pressure.
In the event, Zhu decided to allow market forces to push many of the least
profitable state-owned enterprises out of business. China never embraced
a process of formal privatization, instead adopting a policy to “safeguard the
large firms, and let the small firms go” (zhuada fangxiao 抓大放小) (officially
in 1999). Under the conditions of that time, the most radical part of this policy
was certainly “let the small go,” since this implied that many small state
factories would be allowed to go bankrupt, and others would be quietly
acquired by their managers or by emerging entrepreneurs. In essence, small
publicly owned firms were left to fend for themselves, put on the same basis
as the now legitimated private businesses. “Safeguarding the large,” particu-
larly in retrospect, is equally significant: rather than tackling the difficult task
of large-firm privatization, policy makers decided to maintain but reorganize
the largest state firms, often setting up two or three large state firms in
a sector with controlled competition. Total employment in state-controlled
industry declined from 45 million in 1992 to 18 million by 2006, with the
largest reductions occurring in 1998 through 2003 (Figure 20.3). The pace of
“downsizing” was slower in nonindustrial enterprises, but even faster in
collective firms, which were predominantly industrial. Overall, urban public-
enterprise workers declined from 112.6 million to 52.3 million between 1992 and
2006. The workforce declined by 60.3 million workers, more than half. Industry
was transformed: state-owned enterprises (SOEs) accounted for 60 percent of
above-scale employment as late as 1998, but only 20 percent by the mid-2000s.65
This was a profound social shock and there was widespread dissatisfaction. In
the industrial northeast, layoffs were the largest, unemployment benefits ran
out, and there were large protests. Nationwide, large-scale unrest was avoided
and most laid-off workers received some transitional support.
Zhu Rongji also pushed for entry into the World Trade Organization
(WTO). The arduous negotiation was finally completed in 1999, and China
entered the WTO in December 2001, beginning a three-year phase-in period.
WTO membership was seen as being complementary to state-sector reforms.
Linking up with global norms and regulations (jiegui 接轨) would lock in the
65
National Bureau of Statistics, China Statistical Yearbook 2019, Tables 4-3, 13-3, 13-5,
combined with earlier years. Figure 20.3 shows employment in large (“above-scale”)
industrial enterprises. Data adjusted to include state-controlled corporations, as well as
traditional SOEs.
767

barry naughton
120
100
Total
80
Million Workers
60
40
State-owned and
state-controlled
enterprises
20
0
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
Figure 20.3 Industrial workers (total and state)
reform achievements, strengthen the government’s commitment to a new

regulatory regime, and, eventually, provide a new source of demand that
would speed the economic recovery. Thus, by the time Zhu Rongji reached
the end of his term as premier in spring of 2003, a fundamental transform-
ation had been completed. The institutional framework for a market econ-
omy had been created, and market forces had shifted the balance of the
economy decisively away from the state sector and towards a new, mixed,
corporate economy founded on substantial private ownership. The most
difficult years of the transformation were over, and China was poised for
a new growth acceleration, with the international market playing an
enhanced role.
The End of Reform and Post-transition China

As China entered the twenty-first century, demographic conditions began to
change rapidly, and it was clear that the end of miracle growth was on the
horizon. More surprisingly, after the Hu Jintao–Wen Jiabao leadership suc-
ceeded to power in 2002–2003, market-oriented reforms gradually lost
768

momentum. There was no dramatic policy turning point, and most outside
observers at the time believed that market-oriented reforms were moving
ahead. In retrospect, however, it is clear that the era of reform and opening
was winding down.
Broadening of the Policy Agenda

After 2003, the policy agenda steadily expanded. With the fiscal crisis over-
come (Figure 20.2), resources were available to finance a more activist
approach to government. To start with, that meant finally tackling overdue
social issues. Premier Wen Jiabao took a particular interest in rural areas,
where the collapse of the collectives had left institutional needs unfilled for
twenty years. Increased investment in basic rural education and health care
was overdue. However, farmers were already subject to direct taxes and
a burdensome system of improvised fees, so the resources could not be raised
locally. Wen addressed these problems with urgency: rural taxes and fees
were cut, increased fiscal redistribution from higher to local-level govern-
ments was introduced, and a program to support universal primary educa-
tion (through sixth grade) was adopted.66 Beginning in 2005, a system of
rudimentary rural health insurance was rolled out which provided publicly
supported access to health care for the overwhelming majority of rural
residents. China had exhausted its demographic dividend, and stood on the
threshold of an aging society, so efforts to patch up the national social
security system took on increasing urgency.
Policy makers now began a search for new drivers of economic growth in
the post-“growth miracle” era. High-technology industry moved to the top of
the policy agenda, and government-steered industrial policies returned, at
first modestly, and then expanded rapidly.67 When the global financial crisis
(GFC) hit, in 2008–2009, these policies were absorbed into a bold stimulus
program, which pumped the equivalent of 18 percent of GDP into the
economy through both fiscal and credit-based instruments. The perceived
success of this program led Premier Wen Jiabao to declare in the aftermath
that the “superiority of the socialist system consists in . . . the ability to
concentrate resources to accomplish big things.”68 This emphasis has marked
66
C. Wong, “Rebuilding Government for the 21st Century: Can China Incrementally
Reform the Public Sector?”, China Quarterly 200 (December 2009), 929–52.
67
L. Chen and B. Naughton, “An Institutionalized Policy-Making Mechanism: China’s
Return to Techno-industrial Policy,” Research Policy 45 (2016), 2138–52.
68
Wen Jiabao, “Report on the Work of the Government (2010),” Third Session of the 11th
National People’s Congress, March 5, 2010, at www.scio.gov.cn/xwfbh/xwbfbh/
wqfbh/2015/20150305/xgbd32605/Document/1395827/1395827.htm.
769

barry naughton
Chinese policy proclamations ever since. China is far from the fiscal crisis of
the early 1990s. In short, China shifted from a political economy in which
market-oriented reforms were the focus of policy makers and the prerequisite
for solving other problems, to a more “normal” political economy in which
competing demands jostled for the attention of policy makers.
Reform Stalls Out

The policies adopted by the Hu–Wen administration were potentially com-
plementary to continued market-oriented reforms, and were generally sup-
ported by reform-oriented economists and outside observers like the World
Bank. A repaired social safety net and support for domestic innovation would
clearly strengthen China’s market economy. Nevertheless, gradually the
impetus behind market-oriented reform weakened. The year 2005 saw the
last unambiguous case of centrally designed market-oriented reforms. In that
year, policy makers moved boldly to recapitalize and restructure state com-
mercial banks. Banks were relieved of 1.4 trillion RMB in bad loans (8 percent
of 2005 GDP), and then reorganized into corporations. Strategic investors
were brought in, and the banks were listed on international stock markets.
Designed under the previous administration, this was perhaps the last mani-
festation of the Zhu Rongji era. By contrast, other apparently ambitious
reforms were announced with some fanfare, but then faded away without
much effect. A “Program for Capital Market Development” and a “Charter of
Private Enterprise Rights” met this common fate.
Indeed, if anything, state firms gained status relative to private firms.
A number of prominent private entrepreneurs came under attack, and
some were jailed on corruption charges.69 Following a series of disastrous
mine accidents in 2005 and 2006, private coal mines in Shanxi province were
renationalized. To be sure, there was no massive rollback of privatization,
but the shrinkage of the state sector ground to a halt (Figure 20.3). Overall,
the economy continued to grow and diversify, so robust growth of the
private sector still took place. Despite widely acknowledged needs for
domestic capital market reforms, liberalization of the household registration
system, and further fiscal reforms designed to accommodate new social
expenditures, no major efforts were mounted. It was not until 2013 – well
after this time period – that a new “Third Plenum” made a significant effort to
renew reform momentum.
69
Ma, 交锋三十年.
770

Why Did Market-Oriented Reform Lose Momentum?

The reasons for the post-2005 slowdown of reform are, strictly speaking,
outside the scope of this chapter. Yet three observations may shed light on
the dynamics that were in play throughout the reform era. First, reforms
have up-front costs, while the benefits appear only after a lag. Reforms of the
1990s ultimately led to an acceleration of growth, as GDP growth rose above
10 percent from 2003 through 2010, notwithstanding the GFC in 2008–2009.
This was the final phase of China’s miracle growth era. Powering the
economy was the explosive post-WTO-entry growth of China’s exports,
which tripled between 2002 and 2006 to almost $1 trillion. Yet, in the lag
between policy and benefit, policy makers crafted a response in their own
interest, and that response was exactly the opposite of that of the early 1980s.
Successful rural reforms then emboldened reformers to push ahead, but
equally successful reforms simply induced complacency in the mid-2000s.
In the absence of crisis, Chinese policy makers found it difficult to keep the
reform process moving ahead. Leaders did not have sufficient motivation to
tackle a range of entrenched interests.
Second, a stronger and better-resourced state provided new policy levers
that allowed policy makers to achieve a diverse range of objectives as well as
more effective direct control. National leaders looked for instruments to
implement a range of social, economic, and strategic objectives that no
longer depended on market-oriented reform. In the state sector, the recovery
of profitability and growth of revenues made control of state firms an
attractive proposition, and gave policy makers incentives to block further
growth of competition in order to maintain rents. It also became much more
attractive to use financially stable state firms to contribute to industrial policy
objectives.
Finally, external events matter and the GFC of 2008–2009 had a profound
effect. Chinese policy makers interpreted the GFC as severely tarnishing the
free-market model exemplified by the US, while their massive stimulus
program was welcomed around the world, thus confirming the wisdom of
their own state-led policies. Although the reform slowdown clearly pre-dated
the GFC, the shock of the GFC thus sealed in the big change in policy
orientation and signaled the definitive end to the reform era. By coincidence,
the growth rate dropped below 10 percent after 2010, and has continued to
decline since. Thus both the reform era and the miracle growth era, born
together thirty-two years earlier, came to an end at about the same time.
771

barry naughton
Conclusion
By the year 2010, the Chinese economy was utterly different from the
economy on the eve of reform. It was, of course, much bigger and with
higher incomes, and it was also far more diverse and flexible, with much
higher skill levels and capabilities. Due to the extraordinary speed of growth,
the economy faced new structural challenges that were the inverse of those
of 1978. Then the largest challenge had been to quickly develop labor-
intensive sectors; by 2010, the need was to accommodate the looming
reduction in the size of the labor force, and move into higher-skill goods
and services. The one-child policy seemed urgent in 1978, but by 2005 it was
widely understood to be hobbling economic development, and it was finally
abolished in 2015. As growth inevitably slowed, an entirely new set of policy
choices emerged.
For thirty years, reform, opening, and market transition dominated the
concerns of policy makers. However, as this chapter has demonstrated, the
definition and objective of “reform” were constantly contested, and some
groups questioned the need for it altogether. Policy battles over specific
reforms were sustained and, at times, fierce. The narrative in which Deng
Xiaoping and pragmatic reformers came to power at the Third Plenum and
then instituted reforms is hopelessly oversimplified. Indeed, were it not for
the extraordinary success of rural reforms, the reform infant might well have
been smothered in its crib. Consensus was transitory and rare.
Rather, improvised solutions, remarkable policy entrepreneurs, and
hybrid organizations played a critical role throughout. Innovators like Yuan
Geng, who put together a peculiar corporate–governmental hybrid early on
in the Shekou zone, were essential in driving the process forward. The
township and village enterprises were primarily publicly owned, nominally
collective enterprises, yet they played a crucial role in introducing market
forces and competition. Even the PRS is best seen as an institutional innov-
ation, the outcome of a prolonged negotiation between the powerful state
and the farmers, with their long-standing demand for “land to the tillers.”
These processes of institutional innovation and policy entrepreneurship
required support from the top levels as well, leaders who had to provide
a protective umbrella and occasional decisive policy interventions. That top-
level support came most importantly from Deng Xiaoping, Zhao Ziyang, and
Zhu Rongji, three leaders whose role was indispensable.
Yet that fertile symbiosis was surprisingly delicate. It broke down twice in
the 1980s, at the end of 1980 and again after June 4, 1989, but was patched up
772

again both times. The precarious nature of the pro-reform ecosystem meant
that changes in the distribution of power among the leadership could drive
dramatic changes in policy. This was most obvious in 1992–1993, when
apparently modest shifts in power led to an unprecedented decisiveness
and focus on a broad complex of market-oriented reforms. Thus, while the
Chinese political system was broadly continuous during this period, modest
changes could drive really big changes in policy making. Most surprisingly,
the reform symbiosis fell apart altogether after 2005, when top policy makers
became distracted by competing objectives, and local policy entrepreneurs
found themselves with much less protection and much less space for experi-
mentation. Despite the astonishing, world-altering success of market-
oriented reform, the leadership nonetheless stepped back from further
reform policy making. Perhaps this reflects the truism that only in the face
of crisis will difficult reforms be enacted that create risk and threaten the
benefits of entrenched power holders. After 2005, as China’s exports acceler-
ated, the sense of crisis evaporated, and so did the urgency of reform.
Yet until this time, market-oriented reforms were sustained over multiple
decades, and, broadly speaking, reforms achieved their main objectives in the
end. It is not accidental that this period of sustained reform-oriented policy
coincided with China’s miracle growth phase. The dynamism and growth
potential inherent in the miracle growth phase acted like wind beneath the
wings of China’s economic reforms. Measures that were taken to liberalize
labor-intensive sectors in the early phase – household sidelines, small-scale
manufacturing, and urban services, to say nothing of agriculture itself –
unlocked productivity at the micro level. Just as important, they enabled
the economy to move onto a much more efficient and promising growth
path, bringing China’s prodigious human resources to bear. This broad truth
gave coherence to the inchoate process of reform policy making. Policy
makers got quick confirmation that reforms were succeeding, and they
were quickly reminded of the imminence of crisis if they did not get the
economy onto the right path.
The quick economic response to reforms helped dampen the inevitable
fluctuations and dilemmas in the reform process. Reforms face a deep-seated
dilemma: reform unleashes new demands immediately, while economic
gains are only manifest after many months, or even years. As we have seen
in the narrative of China’s reforms, these broad forces often played out in
a cyclical manner, as new demands disrupted economic equilibrium, causing
economic and social tension, and some reform backtracking. Yet with rapid
economic responsiveness, and throughout the miracle growth era, policy
773

barry naughton
makers managed to get the reform agenda back on track. This outcome can
be ascribed to the challenges that policy makers faced, the sense of crisis, and
the ultimate conclusion that only market-oriented reform provided viable
answers to the developmental challenges that China faced.
Further Reading
Brandt, L., and T. Rawski (eds.), China’s Great Economic Transformation (New York,
Byrd, W.A., China’s Rural Industry: Structure, Development, and Reform (Washington, DC,
World Bank, with Oxford University Press, 1990).
Chen Jinhua 陈锦华, 国事忆述 (The Eventful Years) (Beijing, Zhonggong Dangshi,
2005).
Du Runsheng 杜润生 (ed.), 中国农村改革决策纪实 (A Virtual Record of China’s Rural
Reform Policy) (Beijing, Zhongyang Wenxian, 1998), pp. 80–97.
Enright, M.J., Developing China: The Remarkable Impact of Foreign Direct Investment
(Abingdon, Routledge, 2016).
Li Langqing, Breaking Through: The Birth of China’s Opening-up Policy (Hong Kong, Oxford
Naughton, B., Growing Out of the Plan: Chinese Economic Reform, 1978–1993 (New York,
Naughton, B., The Chinese Economy: Adaptation and Growth, 2nd ed. (Cambridge, MA, MIT
Press, 2018).
Qian Y., How Reform Worked in China: The Transition from Plan to Market (Cambridge, MA,
MIT Press, 2017).
Tao Yitao 陶一桃 and Lü Zhiguo 鲁志国 (eds.), 中国经济特区史论 (A Historical
Treatise on China’s Special Economic Zones) (Beijing, Shehui Kexue Wenxian, 2008).
Vogel, E., Deng Xiaoping and the Transformation of China (Cambridge, MA, Harvard
Walter, C., and F.J.T. Howie. Red Capitalism: The fragile Financial Foundation of China’s
Extraordinary Rise (Singapore: Wiley & Sons (Asia), 2011).
Wu Jinglian, Wu Jinglian: Voice of Reform in China (ed. Barry Naughton) (Cambridge, MA,
MIT Press, 2013).
Yu Guangyuan 于光远, 1978: 我亲历的那次历史大转折 (1978: The Historic Turning
Point That I Experienced), reprint ed. (Hong Kong, Tiandi, 2008).
774

21
China’s Great Boom as a Historical
Process
Overview
Beginning in the late 1970s, China’s economy produced the largest growth
spurt in recorded history. This striking departure from the economic experi-
ence of the previous 200 years encourages onlookers to view recent eco-
nomic success as a “miracle” that requires neither economic nor historical
explanation. Such thinking ignores common elements that have shaped
China’s long-term economic trajectory: forces propelling spurts of innov-
ation and growth, restrictions that often impede these dynamic forces, and
enduring features of China’s polity that generate tensions between central-
ized authoritarian power and economic growth. Neglect of these historical
legacies invites misconceptions about the current boom’s origin and the
economy’s likely future path. History and economics figure prominently in
our analysis of both.
China has experienced repeated bursts of innovation and accelerated
growth. More than a century before China’s recent growth explosion, the
opening of coastal treaty ports, largely outside Qing jurisdiction, expanded
international and domestic commerce that served as conduits for new tech-
nology and ideas. Extension of foreign privilege to include the operation of
treaty port factories curtailed domestic opposition to modern manufacturing,
opening the door to long-term industrial expansion. During the early decades
of the twentieth century, these developments propelled structural change
and modern economic growth in two regions – the lower Yangzi area
adjacent to Shanghai and the northeast.
The authors, who are entirely responsible for what follows, gratefully acknowledge
advice from Debin Ma, Evelyn Rawski, Andrew Batson, Philipp Boeing, Chris Bramall,
Jeffrey Guarneri, Lyric Hale, Charles Hayford, Carsten Holz, Ruixue Jia, Wolfgang Keller,
Nicholas Lardy, Lillian Li, Stephen Morgan, Andrew Nathan, Kevin O’Rourke, Dorothy
Solinger, Jeffrey Williamson, Tim Wright, and Haihui Zhang.
775

Despite differences in timing, scale, and geographic scope, these episodes

share important commonalities. Innovation and growth arise primarily from
decentralized initiative rather than state direction. External opening – forced
or voluntary – and relaxation of domestic constraints encourage bottom-up
development. The opening of nineteenth-century treaty ports and late twen-
tieth-century special economic zones, the post-Mao shift from collective to
household farming, and the subsequent expansion of rural industry demon-
strate the potential of localized or sectoral innovation to unleash unexpected
and momentous consequences.
Episodes of growth acceleration coincide with interludes of state weakness
and retreat. The nineteenth-century opening of treaty ports, the imposition
of a free-trade regime, and, in 1895, ceding foreigners the right to establish
factories on Chinese soil all reflect Qing inability to resist foreign pressure.
Official weakness also facilitated financial innovation: “modern banking
gathered momentum, particularly through the 1920s, when central authority
was at low ebb.”1
China’s current boom began amid extreme state incapacity after the
Cultural Revolution had “effectively destroyed” China’s “apparatus of civil-
ian rule,” left “the legitimacy of the CCP [Chinese Communist Party] . . .
deeply shaken” and “severely damaged the national bureaucracy, leaving it
weak and divided” and rendering Beijing unable to “monitor compliance
with many kinds of orders.”2 Temporary withdrawal of central oversight
permitted local leaders and groups of households to defy official mandates by
reviving and extending short-lived rural reforms begun after the 1959–1960
famine.
Premier Zhu Rongji extended domestic-market liberalization by agreeing
to constraints on state economic actions as part of China’s 2001 accession to
the World Trade Organization (WTO).
The link between political frailty and economic dynamism is no accident.
The enduring features of Chinese political regimes – imperial, Republican,
and Communist – give rise to powerful tensions between authoritarian
control and the bottom-up institutional change, experimentation, and entre-
preneurship that foster productivity growth, the core component of long-
term economic advance. While Chinese states have become powerful
1
N. Horesh, Shanghai’s Bund and Beyond: British Banks, Banknote Issuance, and Monetary
Policy in China, 1842–1937 (New Haven and London, Yale University Press, 2009), p. 41.
2
A.G. Walder, “Bending the Arc of Chinese History: The Cultural Revolution’s
Paradoxical Legacy,” China Quarterly 227 (2016), 617–18; L.T. White I I I, Unstately Power:
Local Causes of China’s Economic Reforms (Armonk, NY, M.E. Sharpe, 1998).
776

China’s Great Boom as a Historical Process
champions of development, repeated episodes during the past 200 years

highlight consistent elite preference for systems that allow rulers to concen-
trate decision making and appropriate resources in ways that enhance their
control but ultimately limit economic advance.
Enduring Features of Chinese Political Regimes

What are these enduring features of the Chinese polity, which John Fairbank
described as resting on “ancient structures of social order and political values
that are too deep for rapid change?”3
The closely intertwined objectives of today’s Chinese rulers hardly differ
from the goals of Qing emperors. Both seek to maintain stability and regime
control, to harness domestic prosperity and advanced technology for military
and security purposes, and to match or overtake neighbors and potential
rivals.
State structure is equally consistent over time. Power resides in self-
perpetuating authoritarian hierarchies. Interlocking sets of economic, social,
and political ties that align interests within national elites and between rulers
and citizens enhance regime longevity. Checks and balances limiting state
action are notably absent. Custom and law promote order and harmony; they
legitimate and strengthen, rather than constrain, the state. There is little
tolerance for dissent. Official surveillance, nowadays reinforced by electronic
technology, identifies violators. Harsh penalties silence all but the most
determined critics.
The state promotes ideologies – Confucianism, Chinese variants of
Marxism, and, currently, elements of both – that portray the incumbent
polity and its leaders as founts of moral authority and bulwarks of stability.
Ideological commitment is an important criterion for official appointment.
Shared ideology offers a partial substitute for bureaucratic supervision,
allowing officials to enjoy wide discretion in governing as long as outcomes
satisfy their superiors’ expectations.
While severely limiting ordinary citizens’ voice in governance, Chinese
regimes strive, often successfully, to secure popular support. Meritocratic
systems of educational advancement and official recruitment – including the
former imperial examinations, the more recent civil service examinations
(guokao 国考), and the current system of competitive school and college
admissions – offer mobility paths that expand regime capability while
3
J.K. Fairbank, “The Unification of China,” in R. MacFarquhar and J.K. Fairbank (eds.),
The Cambridge History of China, vol. 14, The People’s Republic, part 1, The Emergence of
Revolutionary China, 1949–1965 (Cambridge, Cambridge University Press, 1987), p. 26.
777

legitimating elite privilege. Censorship, information control, and state mon-

opoly over educational curricula steer public opinion in directions that
benefit the incumbent regime.
This system invites widespread investment in scaling finely variegated
hierarchies of rank and distinction that bind individuals and groups to the
incumbent regime. The People’s Republic of China (PRC) has expanded the
traditional complex of individual recognition, which now embraces even
schoolchildren, and established new award ladders for firms and localities.
Today, as under the Qing, these distinctions, as well as promotion through
the state’s nomenklatura system, bring substantial accretions of wealth, pres-
tige, and security.
At every level, power and authority rest on personal patronage networks
in which long-term exchanges of money, favors, and loyalty build support for
leaders, while offering subordinates a combination of opportunity and pro-
tection. Leaders mobilize network supporters – local gentry and merchant
guilds under the Qing, multilevel coalitions of like-minded officials under the
PRC – to advance their policy agendas. Ambitious leaders need a constant
flow of resources to support adherents, enlist new clients, and compete with
rivals.
This financial imperative reinforces long-standing elite preference for
administrative structures that concentrate decisions in the hands of offi-
cials who enjoy wide discretion. Leaders seek personal control over
important decisions, in part to facilitate access to continuing resource
flows. Personal and network interests figure prominently in official and
private choices regarding appointments, promotions, contracts, and insti-
tutional arrangements.4
Despite episodic enforcement efforts, the culture of gift exchange that
permeates these personal networks infuses government systems and elite
culture with a comfortable tolerance for bribery. As prime beneficiaries of
irregular transactions, Communist elites, like their Qing and Republican
predecessors, ignore readily available disciplinary mechanisms: updating
land registers in the Qing era or publicizing officials’ personal and family
assets in today’s China. A popular ditty attributed to both Guomindang and
Communist leaders cynically portrays corruption as the lifeblood of Party
operations: “Fight corruption and destroy the Party, neglect corruption and
4
J. Osburg, Anxious Wealth: Money and Morality among China’s New Rich (Stanford,
Stanford University Press, 2013), offers a granular account of local networking.
778

destroy the country” (fanfu wangdang, bufan ze wangguo 反腐亡党, 不反则

亡国).
Injecting network interests into policy formation and public administra-
tion imposes costs that extend far beyond informal side payments. Networks
stifle competition. Reserving opportunities for insiders excludes interlopers.
When external competition does exist, insiders can leverage connections
(guanxi 关系) to sidestep inconvenient legal or regulatory requirements
that rivals cannot avoid. As a result, network involvement transforms appar-
ent market exchanges into landscapes pockmarked with efficiency-sapping
barriers and distortions. During the nineteenth century, Shannon Brown
finds, a “symbiotic coalition of Chinese merchants, organized in guilds, and
government officials – was quite effective in preventing innovation . . . [so
that] market forces alone could not overcome vested-interest opposition . . .
even in the transfer of a demonstrably superior technology.”5 Between 2004
and 2012, firms linked to members of China’s Politburo obtained land “for less
than half the price paid by their unconnected counterparts to obtain land of
comparable quality.”6
These foundations, which have survived the transition from empire to
People’s Republic, weave authoritarian hierarchy, individual ambition, and
personalist networking into a fabric that binds citizens to the state, motivates
widespread support for official priorities, and enhances security for both
rulers and subjects. Unfortunately, the same structures and mechanisms
impose costs that diminish economic performance. Openness to Western
technology and ideas, and rebuilding efforts following episodes of state
weakness, illustrate the structural tensions between authoritarian control
and economic growth.
The Promise and Danger of Openness

While recognizing the need to embrace technological advance as a vehicle for
building national strength, elite thinking harbors deep suspicion of the
institutional penumbra surrounding Western technology. Wei Yuan 魏源,
an early nineteenth-century reformer, supported “the adoption of Western
naval hardware and technology” while embracing “ideals, inspiration, and
5
S.R. Brown, “The Ewo Filature: A Study in the Transfer of Technology to China in the
19th Century,” Technology and Culture 30.3 (1979), 550–68; Brown, “Cakes and Oil:
Technology Transfer and Chinese Soybean Processing, 1860–1895,” Comparative Studies
in Society and History 23.3 (1981), 449–63.
6
T. Chen and J.K.S. Kung, “Busting the ‘Princelings’: The Campaign against Corruption
in China’s Primary Land Market,” Quarterly Journal of Economics 134.1 (2019), 223.
779

historical traditions [that were] wholly shaped by Yuan and Ming

precedents.”7
Several decades later, Zhang Zhidong 张之洞, a prominent official, popu-
larized this perspective in the epigram 中学为体, 西学为用, meaning that
China would utilize (yong 用) Western technology and devices while retain-
ing its own cultural essence (ti 体). This formulation reverberates across the
centuries, echoing earlier discussion surrounding the importation of
Buddhism and prefiguring the embrace of “self-reliance” by both Mao
Zedong and Xi Jinping.8
Twentieth-century nationalists viewed China’s treaty ports “not as spark
plugs or vital centers but as malignant tumors.”9 Such attitudes prompted
PRC planners to limit investment in coastal cities and sparked Cultural
Revolution attacks on individuals with overseas ties.
Xi Jinping has revived fears that the ideas, attitudes, and institutional
arrangements associated with Western technology and thinking threaten
the foundations of China’s polity. Limiting foreign travel by academics and
researchers, removing foreign textbooks from college curricula, and forbid-
ding classroom discussion of specific topics all follow this agenda. His signa-
ture “Made in China 2025” initiative, an inward-looking, Soviet-style plan to
pursue advanced technology with minimal international involvement,
reflects long-standing distrust of the “software” associated with imported
technology.
A contrary perspective welcomes the absorption of Western institutions
along with advanced technology. As early as 1859, Hong Ren’gan 洪仁玕,
a Taiping leader who studied and worked with Christian missionaries before
joining the rebels, produced a document that Stephen Platt describes as
offering “for the very first time in a Chinese context . . . a litany of proposals
that . . . would become catch-phrases for later Chinese reformers.”10 Hong’s
admiration for private business, democratic government, impartial news
reporting, rule of law, and open trade “entitle him to a place in the front
7
J.K. Leonard, Wei Yuan and China’s Rediscovery of the Maritime World (Cambridge, MA,
Harvard University Council on East Asian Studies, 1984), pp. 198–9.
8
W.T. DeBary, W.T. Chan, and C. Tan (eds.), Sources of Chinese Tradition, vol. 2
(New York, Columbia University Press, 1964), p. 82; The Encyclopedia of Buddhism,
at https://encyclopediaofbuddhism.org/wiki/Essence-Function.
9
R. Murphey, The Outsiders: The Western Experience in India and China (Ann Arbor,
University of Michigan Press, 1977), p. 228.
10
S.R. Platt, Autumn in the Heavenly Kingdom: China, the West, and the Epic Story of the
Taiping Civil War (New York, Knopf, 2012), pp. 59–61. F. Michael, The Taiping Rebellion:
History and Documents, vol. 3 (Seattle, University of Washington Press, 1966–1971), pp.
751 ff. provides a translation of Hong’s proposal.
780

rank of Chinese who tried . . . to commend Western ideas to the attention of

their countrymen.”11
The post-1978 reform era revived support for reduced state control, greater
market orientation and increased international openness in opposition to the
Mao-era tendency to repress dissent, suffocate private business, suppress
market allocation, and minimize global involvement. Strong resistance to
liberalizing initiatives obliged reformist Premier Zhao Ziyang to portray
policies that gave “full play” to market forces, embraced “the renewed
centrality” of foreign economic and technical exchange, and favored the
coast as steps toward “the initial stage of socialism” and the achievement of
“self-reliance.”12
In 2013, the CCP Central Committee, seemingly accepting recommenda-
tions from a team of Chinese and World Bank researchers,13 called for an
economy “centering on the decisive role of the market in allocating
resources . . . [and] greatly reducing the government’s role in the direct
allocation of resources.”14 At the same time, persistent concern over foreign
influence triggered fierce pushback.
A 2013 circular, widely cited as “Document 9,” cites a litany of “false
ideological trends,” including democracy, the rule of law, unfettered journal-
ism, and market opening. It excoriates proponents of these heresies for
aiming to “gouge an opening through which to infiltrate our ideology” and
even “denying the legitimacy of the CCP’s long-term political dominance.”
The authors conclude that allowing “any of these ideas to spread . . . will
disturb people’s existing consensus on important issues.”15
The Conflicting Objectives of State Rebuilding Efforts

Rebuilding efforts following episodes of governmental weakness reveal the
pull of traditional patterns of centralized authoritarian control. While fortify-
ing political power, these initiatives reinforce tensions between political and
11
K.W. So, E.P. Boardman, and C. P’ing, “Hung Jen-Kan, Taiping Prime Minister, 1859–
1864,” Harvard Journal of Asiatic Studies 20.1–2 (1957), 294.
12
J. Gewirtz, Unlikely Partners (Cambridge, MA, Harvard University Press, 2017), pp. 116,
191, 196.
13
World Bank and Development Research Center of the State Council, PRC, China 2030
(Washington, DC, World Bank, 2013).
14
“Decision of the Central Committee of the Communist Party of China on Some Major
Issues Concerning Comprehensively Deepening the Reform,” January 16, 2014, at china
.org.cn/china/third_plenary_session/2014-01/16/content_31212602.htm, accessed October
10, 2017.
15
“Communique on the Current State of the Ideological Sphere,” translation at chinafile
.com/document-9-chinafile-translation, accessed March 3, 2021.
781

economic goals inherent in the structure and operating mechanisms of

Chinese political systems. Consider the initial decade of Guomindang leader-
ship and then the PRC rebuilding efforts during the 1990s.
The Guomindang established its Nanjing government in 1927 following
a fifteen-year interregnum during which regional military leaders jousted
with a succession of weak administrations in Beijing. The Guomindang
focused on two objectives: control and development.
Control involved external and internal dimensions. While working to
regain tariff autonomy and to abolish foreign concessions, extraterritoriality,
and other trappings of the nineteenth-century treaty regime, Nanjing sought
to forge a military that could overcome domestic and foreign threats.
The Guomindang also set out to assemble a developmental state. While
shortages of time and money, along with military exigencies, hindered
implementation, detailed plans and initial achievements in multiple sectors
“became available to the Communists, and many planners and technicians
joined them . . . providing the nucleus for much that the Communists later
accomplished.”16
Conflict between political and economic objectives quickly emerged.
Guomindang leaders sought to weaken the bankers and industrialists
whose businesses had led the lower Yangzi region’s considerable growth
achievements. Top officials organized and invested in new companies, which
often “did little more than shift commerce from . . . private merchants” to
politically connected newcomers. The same officials steered official procure-
ment toward these new firms, which they endowed with “special privileges
or monopoly powers.” To compete, private operators sought partnerships
with officials or their relatives.17 In the mid-1930s, monetary changes and the
introduction of a fiat currency relaxed the discipline that China’s private
banking system had imposed on government spending and borrowing.
Sixty years later, the PRC launched its own rebuilding effort following the
near-anarchy of the Cultural Revolution and a decade of decentralized
development that further weakened the center. Although PRC leaders,
unlike their Guomindang predecessors, faced no external military threat,
the shadow of Soviet collapse hovered menacingly in the background.
During the spring of 1989, nationwide urban protests attracted support
among government and Party personnel. After the Tiananmen massacre,
16
A.N. Young, China’s Nation-Building Effort 1927–1937: The Financial and Economic Record
(Stanford, Hoover Institution Press, 1971), p. 388.
17
P.M. Coble Jr., The Shanghai Capitalists and the Nationalist Government, 1927–1937
(Cambridge, MA, Harvard Council on East Asian Studies, 1980), pp. 221, 232–5, 243–8.
782

China’s leaders struggled to solidify Party cohesion and central authority.

Success rested in part on resolving the frustrating imbalance between their
ambitious plans and the meager funds at their disposal following “a rapid and
dramatic erosion in the traditional tax base” that lowered both the GDP share
of government revenues and the center’s share of fiscal resources.18
To navigate this complex and risk-laden environment, state and Party
leaders advanced a policy agenda that combined recentralization and market
opening, features that appealed to multiple interest groups and therefore
promoted a broad policy consensus. Deng Xiaoping’s ringing endorsement of
growth electrified the nation. Fiscal and banking reforms reversed the decline
in resources available to the center. A succession of policies, including
relaxation of controls over labor mobility, state-sector restructuring, tariff
reduction, and exchange rate depreciation enlarged the scope of market
forces in both domestic and international transactions. Restoration of incen-
tives and expansion of market activity narrowed major gaps that had accu-
mulated within China’s sclerotic planned economy: domestic production
rose toward potential levels, while rising technology imports extended the
economy’s production frontier.
As high-speed growth continued, bold measures – privatizing urban housing
and many state enterprises and township and village enterprises (TVEs),
pushing whole sectors into market competition, dismissing millions of state-
sector workers, and ending material allocations – relieved the center of vast
fiscal burdens. Rapid growth of exports and of both domestic and foreign
investment further enlarged the array of resources subject to central influence.
The center, having enlarged its revenues and shed costly obligations, now
possessed ample financial resources to support both domestic and inter-
national objectives. The streamlined agenda for the domestic economy
focused on expanding infrastructure networks, strengthening a slimmed-
down state sector, nurturing “national champions” within the ranks of
centrally managed state firms, and absorbing strategic technologies.
Growing fiscal capacity, foreign-exchange earnings, and financial resources
enabled Beijing to expand overseas aid, outbound investment, Olympic
sponsorship, and other efforts to strengthen China’s international standing
and, by doing so, enhance the regime’s domestic legitimacy.
These measures delivered superlative results. Living standards rose. High-
speed growth vaulted China into global prominence as an industrial and trade
18
C.P.W. Wong and R. Bird, “China’s Fiscal System: A Work in Progress,” in L. Brandt
and T.G. Rawski (eds.), China’s Great Economic Transformation (Cambridge, Cambridge
783

powerhouse. Success bred confidence, encouraging the Party to relax its grip
on daily life.
Beneath the surface, however, these advances rest on structures and
mechanisms that recall Guomindang administration during the 1930s and
nineteenth-century Self-Strengthening efforts involving official–merchant
collaboration (guandu shangban 官督商办). Officials direct resources toward
firms they can influence – often at the expense of more dynamic alternatives.
Loyalty to leaders and responsiveness to official requests determine the
selection of managers. Individuals shuttle between corporate and govern-
ment positions.19 Webs of personal influence muddle the interests of leaders
and firms at every administrative level. Officials routinely commandeer
corporate resources to support personal or policy agendas. Their relatives
and cronies colonize important businesses. The need to shore up Party
structures battered by Cultural Revolution turmoil and frayed by the lure
of “plunging into the sea” (xiahai 下海) of private business dictated
a relaxation of financial discipline to satisfy the expectations of modestly
compensated officials and Party functionaries in a society that increasingly
measures status in monetary terms.20
Under the PRC, the vast reach of state economic influence magnifies the
impact of political intervention. After falling through the mid- to late 1990s,
the state’s share of GDP has remained remarkably constant at 45 percent,
with nonfinancial state-owned enterprises (SOEs) consistently accounting for
over 20 percent.21 A succession of SOE mergers has consolidated central
control within strategic industry and service sectors. Between 2003 and 2019,
central-level enterprise groups under the State-Owned Assets Supervisory
Commission (SASAC) fell from 186 to 97, while the number of subsidiaries
under these groups nearly doubled and their registered capital increased
more than fivefold.22 The state continues to dominate China’s financial
system, which has grown rapidly relative to GDP. Much of China’s rapidly
19
F. Liu and L.L. Zhang, “Executive Turnover in China’s State-Owned Enterprises:
Government-Oriented or Market-Oriented?”, China Journal of Accounting Research 11
(2018), 132–3, give examples of executives shifting between managerial and official posts
and note that “most SOE executives” hold administrative ranks that allow them to
occupy government positions.
20
Wage compilations show average 1993 salaries in government and Party organizations
lagging behind earnings of workers in high schools, physical education, hotels, ware-
houses, and construction. See 中国劳动统计年鉴 1994 (China Labor Statistics
Yearbook 1994) (Beijing, China Statistics Press, 1994), pp. 109–10.
21
A. Batson, “The State Never Retreats,” Gavekal Dragonomics, 1 October 2020, 6–7.
22
L. Brandt, R. Dai, and X. Zhang, “The Anatomy of China’s State-Owned Enterprises,”
unpublished MS, 2021.
784

growing overseas foreign direct investment comes from state or state-

connected firms.
Following two decades of transition from plan to market, liberalizing
reform slowed dramatically after China’s 2001 entry into the World Trade
Organization. While committing enormous resources to economic develop-
ment, the Hu Jintao and Xi Jinping administrations have retreated from the
market opening, global co-operation, and private initiative largely respon-
sible for China’s recent prosperity. Instead, they have promoted state enter-
prises, top-down decision making, self-reliance, and Party involvement in
business management, arrangements that past Chinese experience identifies
as potent sources of inefficiency. Pursuing breakthrough innovations rather
than more predictable efforts to narrow the gap separating domestic and
global production frontiers heightens the risk of disappointing outcomes.
Amid continuing expansion of China’s scientific, technological, and organ-
izational capabilities, multiple studies find a steep falloff in productivity
growth in the wake of the 2008 financial crisis. Deterioration in this core
component of China’s economic prospects underlines the continuing tension
between the demands of state building and the requirements of economic
growth, which we see as an inevitable consequence of the tradition of
authoritarian rule to which Chinese elites remain committed.
Nineteenth-Century Developments
Internal and external shocks diminished the power and authority of the
nineteenth-century Qing state. Domestic uprisings, most notably the mid-
century Taiping Rebellion, drained the imperial treasury and forced the
center to rely on provincial gentry to organize and finance regional armies.
At the same time, growing foreign pressure, initially from the European
powers and subsequently from Japan, undermined Qing sovereignty, result-
ing in the treaty port system described in James Kung’s Chapter 11 of this
volume.
Domestic rebellion in which incumbent Han elites supported imperial
Manchu rulers in defense of the status quo destroyed cities, turned fertile
agrarian regions into wastelands, and created waves of refugees. Foreign
incursions, by contrast, injected new technologies and breached trade restric-
tions, thus encouraging economic growth. Telegraphic communication and
steam transport lowered transaction costs and linked domestic and overseas
markets. Treaties eliminating trade barriers and limiting taxation of overseas
trade created new opportunities for Chinese farmers and consumers. Transit
785

passes intended to exempt foreign goods from internal taxes intensified

domestic competition by permitting Chinese merchants to avoid transit
taxes and other restrictions on internal trade.23 High domestic interest rates
encouraged foreign banks and mercantile houses to inject new funds into
China’s capital-scarce economy, lowering the cost of financing business
within the treaty ports and along major commercial routes linked to overseas
trade.24
The creation of semi-autonomous treaty ports unleashed a flood of innov-
ation, especially in Shanghai, which anticipated Shenzhen’s contemporary
role as a magnet for ambitious and entrepreneurial migrants, an entry port
for new ideas, and a hotbed of institutional innovation.25 The relative
obscurity of both locales – Shanghai as a county seat, Shenzhen as a sleepy
village – limited the capacity of conservative elites – degree-holding gentry in
nineteenth-century Shanghai, advocates of state-owned enterprise in late
twentieth-century Shenzhen – to obstruct innovation. In both instances,
local economic dynamism prompted competitive reactions elsewhere: self-
initiated open ports under the Qing,26 multiplication of special economic
zones in the PRC, and relaxation of restrictions on entry and competition in
both systems.
Despite their differing economic consequences, internal and external
challenges to Qing rule were mutually reinforcing. Domestic turbulence
limited the capacity of the Qing state to resist foreign incursions. Foreign-
controlled schools, newspapers, and publishers quickly transformed Shanghai
and other foreign-controlled locales into transmission belts for new ideas,
technologies, and institutional arrangements.27 The Taiping leadership, for
example, included men who had lived, worked, and studied in Hong Kong,
ceded to Great Britain in 1842.
This double-barreled assault on the Qing imperium opened new channels
of mobility entirely separate from the long-standing paths of academic
examination and mercantile degree purchase.28 The desperate struggle to
subdue the Taipings established military success as an alternate route to high
23
E. Motono, Conflict and Cooperation in Sino-British Business, 1860–1911: The Impact of the
Pro-British Commercial Network in Shanghai (New York, St. Martin’s Press, 2000).
24
Y.P. Hao, The Commercial Revolution in Nineteenth-Century China: The Rise of Sino-Western
Mercantile Capitalism (Berkeley, University of California Press, 1986), pp. 106–10, 345.
25
R.X. Jia, “The Legacies of Forced Freedom: China’s Treaty Ports,” Review of Economics
and Statistics 96.4 (2014), 596–608.
26
Kung, Chapter 11 in this volume. 27 Kung, Chapter 11 in this volume.
28
E. Kaske, “Fund-Raising Wars: Office Selling and Interprovincial Finance in
Nineteenth-Century China,” Harvard Journal of Asiatic Studies 71.1 (2011), 69–141, docu-
ments the growing sale of both degrees and offices.
786

office for men with little academic distinction.29 “Modern” schools in

Hong Kong and various treaty ports produced cosmopolitan graduates
whose technical knowledge, language skills, and business acumen marked
them as indispensable allies of the provincial magnates whose defeat of the
Taipings thrust them into national prominence.
These developments initiated a gradual rise in the economic payoff to
“modern” relative to Confucian education.30 As change spread beyond the
treaty ports to encompass new activities like railways, elite families began to
withdraw their sons from traditional schooling. The resulting erosion in a key
bulwark of the imperial system accelerated when China’s crushing defeat in
the 1894–1895 Sino-Japanese War, followed in 1900 by the rout of antiforeign
Boxer militias at the hands of a Western military expedition, forced trad-
itional elites to recognize the inevitability of sweeping change.
Notwithstanding the dynasty’s ignominious collapse following decades of
directionless economic fluctuation, the century’s closing decades substan-
tially enhanced China’s longer-term potential for economic advance.
Telegraphic communication, along with steam and rail transport, rested on
solid beachheads.31 Expanded access to modern education, along with the
multiplication of information flows, produced a considerable group of pros-
perous, cosmopolitan, often Western-educated elites.32 Domestic opposition
to Chinese-owned factories crumbled after the Treaty of Shimonoseki
allowed Japanese nationals and, thanks to most-favored-nation treaty provi-
sions, other foreigners to enter manufacturing. As with international trade
and domestic commerce, privileges won through foreign military pressure
encouraged domestic economic growth.
Beginning around 1900, a “wave of scientific translations [most] from
Japanese sources” broadcast new knowledge.33 Conservative resistance to
29
J.W. Esherick, Ancestral Leaves: A Family Journey through Chinese History (Berkeley,
University of California Press, 2011), pp. 67–8; D.R. Reynolds with C.T. Reynolds,
East Meets East: Chinese Discover the Modern World in Japan, 1854–1898 (Ann Arbor,
Association for Asian Studies, 2014), pp. 8, 229.
30
N. Yuchtman, “Teaching to the Tests: An Economic Analysis of Traditional and
Modern Education in Late Imperial and Republican China,” Explorations in Economic
History 63 (2017), 70–90.
31
R. Thompson, “The Wire: Progress, Paradox, and Disaster in the Strategic Networking
of China, 1881–1901,” Frontiers of History in China 10.3 (2015), 395–427.
32
Y.P. Hao, The Comprador in Nineteenth Century China: Bridge between East and West
(Cambridge, MA, Harvard University Press, 1970), p. 102, for example, places the
number of current and former compradors at 20,000 by 1900.
33
D. Wright, “Yan Fu and the Tasks of the Translator,” in M. Lackner, I. Amelung, and
J. Kurz (eds.), New Terms for New Ideas: Western Knowledge and Lexical Change in Late
Imperial China (Leiden: Brill, 2001), p. 235.
787

imported technologies, factory industry, and modern education diminished.

By 1911, China’s economy and society were far more open to competition and
change than in 1800 or 1850. The Guangxu Emperor’s 1893 edicts ordering
officials to halt the prior practice of seizing assets from returning overseas
migrants illustrates this growing openness.34 The farm sector, although far
from dynamic, comfortably supported growing urban and nonagricultural
populations in the lower Yangzi and Lingnan regions.
Despite these gains, substantial obstacles continued to restrict China’s
growth prospects. Modernizing advances remained local rather than regional
or national. The state, a key link in all latecomers to modernization,
remained weak and unfocused. In the late 1880s, “the Japanese government’s
published annual budget was a matter of amazement to many Chinese.”35
Writing in 1897, William Mayers described the operation of China’s central
government as “registering and checking the actions of various provincial
administrations [rather] than . . . assuming a direct initiative in the conduct of
affairs.”36 Even for the management of currency, “the Board of Revenue
couldn’t be the source of a coherent monetary policy. It had no power to
inspect the quality of provincial coins . . . [and] could comment on provincial
memorials [to the throne] only if they were referred to the Board.”37
The Republican Period

A tumultuous interregnum that began and ended with regime change,
China’s Republican era (1912–1949) witnessed extremes of political instability,
cultural ferment, and openness to international exchange, along with modest
economic growth, considerable expansion of state capability, and the emer-
gence of trends that foreshadowed future developments.
Following the Qing collapse, a succession of republicans, monarchists, and
military leaders failed to restore political unity. The Nanjing-based
Nationalist administration under Chiang Kai-shek (Jiang Jieshi) won inter-
national recognition following the successful Northern Expedition (1927). Its
sphere of actual control, however, was less than complete even before
34
M.R. Godley, The Mandarin-Capitalists from Nanyang: Overseas Chinese Enterprise in the
Modernisation of China 1893–1911 (Cambridge, Cambridge University Press, 1981), pp.
240–1.
35
Reynolds and Reynolds, East Meets East, p. 341.
36
Quoted in F.H.H. King, A Concise Economic History of Modern China (1840–1961)
(New York, Praeger, 1969), pp. 21–2.
37
King, A Concise Economic History of Modern China, p. 34.
788

Japanese armies forced the shift of its capital to Wuhan and later to
Chongqing.
Chinese elites, shaken by humiliating military setbacks and the Qing
collapse, plunged into an intense and disputatious search for cultural renewal.
Elite gentrymen who had formerly met modern innovations with visceral
hostility now invested in railways and joined newly established chambers of
commerce. Radical ideas, fostered in treaty port schools and championed by
students returning from overseas studies, leapt to the fore. As Chapter 14 in
this volume by Gao, Van Leeuwen, and Wang shows, new subjects, text-
books, and ideas spread far beyond coastal enclaves. Newspapers and radio
broadcasts amplified the circulation of novelty.38 In distant Shanxi, a school
principal reprimanded a traditionally educated teacher who encouraged
students to celebrate the lunar New Year.39 Hu Shi (1891–1962), a Cornell
University graduate and future Chinese ambassador to the United States,
cruelly mocked the ignorance of ordinary folk.40
Elite preference for authoritarian politics survived this intellectual turmoil.
Early English–Chinese dictionaries rendered “democracy” as “disorderly
administration by the many” and “abuse of power by the mean.”41 A 1903
visit to North America convinced the influential reformer Liang Qichao that
“resort to rule by . . . majority . . . would be the same as committing national
suicide . . . the Chinese people must for now accept authoritarian rule.”42
Nearly a century later, Andrew Nathan observes that most Chinese intellec-
tuals, including opponents of the Communist Party’s political monopoly,
continue to “fear the disorder they believe would flow from any weakening
of party control . . . [and] accept the party’s claim that political order . . .
requires leaders with strong authority.”43
The inflow of new ideas reflected a general climate of openness. China’s share
of global trade rose from 1.3 percent in 1913 to 2.1–2.3 percent during 1927–1929 and
3.7 percent in 1936; comparable PRC figures languished below 1 percent
38
W.H. Yeh, Shanghai Splendor: Economic Sentiments and the Making of Modern China, 1843–
1949 (Berkeley, University of California Press, 2007), p. 34.
39
H. Harrison, The Man Awakened from Dreams: One Man’s Life in a North China Village 1857–
1942 (Stanford, Stanford University Press, 2005), p. 97.
40
Chabuduo xiansheng 差不多先生 (Mr. Close-Enough), available at https://zh
.m.wikisource.org/zh-hans/%E5%B7%AE%E4%B8%8D%E5%A4%9A%E5%85%88%
E7%94%9F%E5%82%B3.
41
G.T. Jin and Q.F. Liu, “From ‘Republicanism’ to ‘Democracy’: China’s Selective
Adoption and Reconstruction of Modern Western Political Concepts (1840–1924),”
History of Political Thought 26.3 (2005), 479–80.
42
A.J. Nathan, Chinese Democracy (Berkeley, University of California Press, 1986), p. 60.
43
Nathan, Chinese Democracy, p. 231.
789

throughout 1968–1980, regaining the 1936 level only after the year 2000.44
Throughout the early twentieth century, China was also a major beneficiary of
foreign direct investment, much of it from advanced countries. By the 1930s,
China held more than 10 percent of the global stock of inbound foreign direct
investment and over 15 percent of the stock located in developing nations, with
the largest portion directed toward (mostly rail) transportation.45
Openness strengthened the economy, particularly in coastal regions where
modern education, returned overseas students and migrants, and frequent
interaction with foreign business stoked the transfer of technologies and the
spread of commercial knowledge among would-be Chinese entrepreneurs.
The history of numerous industries, among them mining, railways, banking,
department stores, textiles, and matches, reflects this beneficial mélange.46
While limited growth of fiscal revenue, much of it immediately needed for
the military, signaled the continuing restriction on governmental develop-
ment efforts,47 comparing the Nanjing decade (1927–1937) with circumstances
in 1880 or 1910 highlights major expansion of the state’s capacity to formulate
and implement effective development programs.
Unlike its imperial and Republican predecessors, the Nanjing-based
Guomindang administration pursued a well-defined economic agenda centered
on revenue expansion; extending control over banking, finance, and the monet-
ary system; developing military-linked production; deepening regional and
national economic integration; and building an officially directed education
system.
Public administration no longer resembled the Qing Board of Revenue,
which acted as a “transmission center of documents and repository for
ledgers . . . [that] rarely initiated policy.”48 Central government agencies,
ranging from the National Resources Commission and the Ministry of
Finance to the Cotton Control Commission, their staffs now bolstered by
highly trained professionals, many with advanced overseas degrees, designed
and began to implement a wide array of economic-policy endeavors.49
44
See the online appendix at www.cambridge.org/EconomicHistoryChina.
45
See the online appendix at www.cambridge.org/EconomicHistoryChina.
46
Among many others, see S. Cochran, Big Business in China: Sino-Foreign Rivalry in the
Cigarette Industry, 1890–1930 (Cambridge, MA, Harvard University Press, 1980); and E. Köll,
Railroads and the Transformation of China (Cambridge, MA, Harvard University Press, 2019).
47
1989), pp. 12–32.
48
M.B. Kwan, The Salt Merchants of Tianjin (Honolulu, University of Hawaii Press, 2001), p. 32.
49
W.C. Kirby, “Engineering China: Birth of the Developmental State, 1928–1937,” in W.
H. Yeh (ed.), Becoming Chinese: Passages to Modernity and Beyond (Berkeley, University of
790

Although the absence of political unification, rifts within the central

administration, budgetary weakness, and growing military pressure limited
progress, even critics chronicle advances such as the “successful work of the
National Economic Council . . . in improving the production of silk, cotton,
and tea.”50 Beyond Nanjing, provincial governments and educational institu-
tions initiated a variety of projects intended to distribute superior wheat
seeds, control silkworm egg disease, improve tea garden management,
upgrade equipment for handloom weavers, and so on.51
Political disunity did not preclude long-term policy co-ordination, in which
“different levels of government, regardless of . . . political fragmentation,
closely interacted” to advance shared objectives. Remarkably, by “1926,
prison reform across the country was impressive enough” to merit “a positive
assessment by a traveling committee of the [thirteen-country] Commission
on Extraterritoriality in China,” which advised that “extraterritoriality might
be abolished by foreign powers.”52
Political fragmentation and Japanese military pressure notwithstanding,
domestic and international openness, expansion of new skills and capabilities,
declining resistance to new technologies and ideas, and growing public sector
support contributed to modest but significant economic expansion and struc-
tural change during the decades preceding the outbreak of full-scale war in
1937. Two regions experienced the full array of developments associated with
modern economic growth. Chinese entrepreneurship powered growth in the
Shanghai-centered lower Yangzi area, with a population of 60 million, match-
ing Japan’s. In the northeastern region of Manchuria, populated by over
30 million, foreign investment, much of it from semiofficial Japanese compan-
ies, led a broad-based expansion. In both areas, growth of aggregate and per
capita output during the prewar decades approached or exceeded Japan’s.53
California Press, 2000), pp. 137–60; M. Zanasi, Saving the Nation: Economic Modernity in
Republican China (Chicago, The University of Chicago Press, 2006).
50
L.E. Eastman, The Abortive Revolution: China under Nationalist Rule, 1927–1937
(Cambridge, MA, Harvard University Press, 1974), p. 219.
51
T.H. Shen, “First Attempts to Transform Chinese Agriculture, 1927–1937,” in P.K.T. Sih
(ed.), The Strenuous Decade: China’s Nation-Building Efforts, 1927–1937 (New York,
St. John’s University Press, 1979), p. 220; L.M. Li, China’s Silk Trade: Traditional
Industry in the Modern World, 1842–1937 (Cambridge, MA, Harvard University Council
on East Asian Studies, 1981), pp. 188–96; R. Gardella, Harvesting Mountains: Fujian and the
China Tea Trade, 1757–1937 (Berkeley, University of California Press, 1994), pp. 146–69.
52
F. Dikötter, The Age of Openness: China before Mao (Hong Kong, Hong Kong University
Press, 2008), p. 15.
53
D.B. Ma, “Economic Growth in the Lower Yangzi Region of China, 1911–1937:
A Quantitative and Historical Analysis,” Journal of Economic History 68.2 (2008), 355–92;
K. Chao, The Economic Development of Manchuria: The Rise of a Frontier Economy (Ann
791

A small but dynamic modern sector led the way in both regions, with the
pace of industrial growth exceeding comparable figures for Japan, India, and
Russia/the USSR during the prewar decades.54 Although foreign firms bene-
fited from a head start, favorable treaty provisions, and superior access to
capital, Chinese-owned firms offered powerful competition: by 1933, they
contributed 73 percent of nationwide manufacturing output and 78 percent in
China proper.55
The expansion of manufacturing, with textiles and food processing in the
forefront, enlarged demand for cotton and wheat. Factory interests comple-
mented official efforts to improve rural storage facilities, promote standard-
ized crops, and expand rural credit.56 Transport improvements, along with
a monetary revolution that substituted paper notes issued by private banks
that were freely convertible to silver for unwieldy silver coins and bullion,
reduced transaction costs, magnifying the spread effects of urban-based
growth.57 Rising per capita incomes may have extended beyond the coastal
cities and their rural hinterlands to encompass the entire economy.58
While the quantitative dimensions of nationwide growth remain uncer-
tain, two decades of Guomindang rule introduced distinctive changes that
prefigured important elements of PRC economic structure, institutions, and
policy. State management displaced private control in banking and in import-
ant segments of manufacturing. Industrial expansion began to shift toward
military-linked producer industries even before 1937. Wartime pressures
intensified these trends and widened the geographic dispersion of industrial
activity.59
Arbor, Michigan Papers in Chinese Studies, 1983), pp. 14–15; R. Minami and F. Makino,
Asian Historical Statistics 3: China (Tokyo: Tōyō Keizai Shinpōsha, 2014), pp. 515–16.
54
L. Brandt, D.B. Ma, and T.G. Rawski, “Industrialization in China,” in K.H. O’Rourke
and J.G. Williamson (eds.), The Spread of Modern Industry to the Global Periphery since 1871
(Oxford, Oxford University Press, 2017), p. 199.
55
Brandt, Ma, and Rawski, “Industrialization in China,” p. 208; and Rawski, Economic
Growth, p. 74.
56
Zanasi, Saving the Nation, focuses on cotton improvement.
57
Rawski, Economic Growth, Chapters 3, 4; D.B. Ma, “Financial Revolution in Republican
China during 1900–37: A Survey and a New Interpretation,” Australian Economic History
Review 59.3 (2019), 242–62.
58
Rawski, Economic Growth, p. 342, concludes that nationwide per capita output rose by 22
to 24 percent between 1914–1918 and 1931–1936. This conclusion, however, rests on
estimates of agricultural output trends, which require considerable error margins.
59
Brandt, Ma, and Rawski, “Industrialization in China,” pp. 209–12. P. Schran, Guerrilla
Economy: The Development of the Shensi–Kansu–Ninghsia Border Region, 1937–1945 (Albany,
State University of New York Press, 1976), p. 153, cites contemporary accounts indicat-
ing that armaments production in the Communists’ Shaanxi base area represented
“crude work” that turned out limited quantities of “inferior arms.”
792

While government operations reflected the efforts of “the Guomindang

elite . . . to reform China’s administrative bureaucracy by adopting and
adapting American theories of public administration,”60 policy objectives
and industrial organization converged toward the preferences of the post-
1949 PRC administration. The organization and even the terminology
(danwei 单位) developed around state-owned industrial firms in wartime
China remain in daily use eighty years later.61 William Kirby describes the
Guomindang’s prewar efforts as the “birth of the developmental state,” and
notes that, following the emergence of the PRC, the Nanjing regime’s “main
industrial planning committee did not disband . . . [but] simply reported to
a new government.”62 Guomindang determination to subordinate banking
to the financial requirements of the ruling government and party and to limit
the scope of independent action on the part of leading enterprises, business
owners, and corporate managers foreshadows government–business rela-
tions in today’s China.63
The Guomindang years also witnessed a dramatic change in economic
ideology. Although many prominent officials and researchers – among them
T.V. Soong, H.H. Kung, Franklin Lien Ho, and H.D. Fong – boasted econom-
ics degrees from prominent US universities, expert opinion turned against
market outcomes. A 1941 account noted that “the urgent need for creating
a planned economic system has almost become a consensus both within and
outside the government.” A review of 574 essays published between 1938 and
1944 in “a leading economic journal” found “‘unanimous agreement’ on the
desirability of creating a planned economic system in China.”64
The Era of the Planned Economy

The establishment of the People’s Republic in 1949 ended a century marked
by multiple episodes of warfare, regime change, and monetary chaos that
severely limited economic growth. The new government installed a Soviet-
60
M.L. Bian, “Building State Structure: Guomindang Institutional Rationalization during
the Sino-Japanese War, 1937–1945,” Modern China 31.1 (2005), 38.
61
Bian, “Building State Structure,” 66.
62
Kirby, “Engineering China,” 137; W.C. Kirby, “Continuity and Change in Modern
China: Economic Planning on the Mainland and on Taiwan, 1943–1958,” Australian
Journal of Chinese Affairs 24 (1990), 135. After severing China’s northeast region and
establishing Manchukuo as a separate state, the Japanese authorities developed
a Soviet-style five-year plan for 1937–42; see Minami Manshū tetsudō kabushiki kaisha
chōsakai 南滿洲鐵道株式會社調查課 (ed.), 満州五カ年計画概要 (Summary of
Manchukuo’s Five-Year Plan) (Dairen, 1937).
63
Coble, Shanghai Capitalists. 64 Bian, “Building State Structure,” 60.
793

inspired plan system that governed China’s economy for three turbulent
decades.
Rapid Removal of Long-Standing Constraints on Growth

Firm nationwide political control, reinforced by universal presence of
Communist Party branches, provided the new government with an unpre-
cedented capacity to implement policy even at the village level with minimal
reliance on unofficial intermediaries. Sweeping and often violent campaigns
stifled potential resistance from landed and mercantile interests.
Fiscal expansion demonstrated the new regime’s control. The ratio of govern-
ment revenue to GDP, which had languished below 10 percent for centuries,
exceeded 20 percent throughout the planned-economy period.65 Growth initia-
tives benefited from political unity, the cessation of internal warfare, and the
return of monetary stability following destructive wartime hyperinflation.
Beginning in 1953, a succession of five-year plans pushed investment to new
heights. Focusing on upstream sectors linked to industrial expansion and
military hardware, new developments extended trends established during the
Guomindang regime’s final decade.66 Support from the Soviet bloc, which
provided the largest ever transfer of technology along with technical advice
and short-term loans, facilitated the emergence of new industries. Soviet
support clustered around 150 major projects, which absorbed nearly one-fifth
of overall investment spending under the First Five-Year Plan (1953–7).67
These plans combined the expansion and upgrading of production capabilities
with major investments in human resources. Local governments worked to
universalize primary-school enrollment. Literacy and vocational programs
improved adult skills. Publishing houses distributed cheap technical manuals.
Despite limited food supplies during and after the 1959–1961 Great Leap Famine,
improvements in sanitation, nationwide immunization programs, and cam-
paigns to improve maternal and infant health reduced mortality rates and
increased life expectancy.68
65
China Compendium of Statistics 1949–2008 (Beijing, China Statistics Press, 2010), pp. 9, 18.
66
Brandt, Ma, and Rawski, “Industrialization in China,” 199–200, 209–12.
67
Z.K. Dong 董志凯 and J. Wu 吴江, 新中国工业的奠基石 156项建设研究 (1950–
2000) (Foundations of New China’s Industry: A Study of 156 Projects (1950–2000))
(Guangzhou: Guangdong jingji chubanshe, 2004), p. 333; and Guojia tongjiju gudingzi-
chan touzi tongjisi 国家统计局固定资产投资统计司 (ed.), 1950–1985 中国固定资产
投资统计资料 (Statistical Materials on China’s Fixed Asset Investment, 1950–1985)
(Beijing, China Statistics Press, 1987), p. 50.
68
R. Hayhoe (ed.), Contemporary Chinese Education (Armonk, NY, M.E. Sharpe, 1984); D.
M. Lampton, Health, Conflict and the Chinese Political System (Ann Arbor, Michigan
Papers in Chinese Studies, 1974); K.S. Babiarz, K. Eggleston, G. Miller, and Q. Zhang,
794

Economic Outcomes: Growth, Incomes, and Productivity

Notwithstanding setbacks from the 1959–1961 famine and, on a lesser scale,
from the Cultural Revolution, GDP expanded briskly, with industry occupy-
ing a growing share of total output. China’s growth exceeded results in other
large, low-income nations, with real per capita output growing at an esti-
mated annual rate of 1.8–2.3 percent, which cumulates to a rise of 60 to
82 percent between 1952 and 1978.69
This growth, however, occurred primarily at the extensive margin, with
expansion powered by rising investment. Three decades of planning failed to
deliver productivity growth – the central ingredient in sustained economic
modernization. At the aggregate level, Perkins and Rawski find positive
annual growth of total factor productivity (TFP) during 1952–1957,70 after
which the trend turns negative, with an average annual decline of 0.5 percent
during 1957–1978.71 Sectoral studies show consistently poor productivity
results. For industry, authors whose work produces the most favorable
outcomes find the small increases during 1957–1978 “disappointing both in
comparative terms and in relation to the massive injections of technology and
human capital characteristic of Chinese industrial development.”72 Two
careful studies of plan-era agriculture arrive at similar outcomes: decline or
small gain during 1952–1957, long-term decline thereafter.73
In the absence of productivity growth, the rising share of investment in
overall expenditure restricted consumption opportunities, especially for the
“An Exploration of China’s Mortality Decline under Mao: A Provincial Analysis, 1950–
80,” Population Studies 69.1 (2015), 39–56; A.L. Piazza, Food Consumption and Nutritional
Status in the PRC (Boulder, Westview Press, 1986).
69
D. Morawetz, Twenty-Five Years of Economic Development, 1950 to 1975 (Baltimore, Johns
Hopkins University Press, 1977), p. 5; per capita income estimates, both in international
dollars, from Penn World Tables, v. 9.1, accessed June 23, 2020, and from A. Maddison,
Chinese Economic Performance in the Long Run (Paris, OECD, 1998), p. 40.
70
TFP is the quotient of separate indexes of output (usually GDP or value-added) and
a combined input measure. Rising (falling) TFP reflects increases (reductions) in
average output per unit of combined capital, labor, and materials.
71
D.H. Perkins and T.G. Rawski, “Forecasting China’s Economic Growth over the Next
Two Decades,” in Brandt and Rawski (eds.), China’s Great Economic Transformation, p. 839.
72
K. Chen, G.H. Jefferson, T.G. Rawski, H.C. Wang, and Y.X. Zheng, “Productivity
Change in Chinese Industry, 1953–1985,” Journal of Comparative Economics 12 (1988), 587;
see also R.M. Field, “Slow Growth of Labour Productivity in Chinese Industry, 1952–
81,” China Quarterly 96 (1983), 641–64.
73
S.G. Fan and X.B. Zhang, “Production and Productivity Growth in Chinese Agriculture:
New National and Regional Measures,” Economic Development and Cultural Change 50.4
(2002), 833; A.M. Tang, “Food and Agriculture in China: Trends and Projections, 1952–77
and 2000,” in A.M. Tang and B. Stone, Food Production in the People’s Republic of China
(Washington, DC, International Food Policy Research Institute, 1980), p. 28, using his
adjusted TFP measure.
795

80 to 85 percent living in the countryside. Nicholas Lardy finds, “Except for

a few years . . . average per capita food consumption [between 1949 and the
late 1970s] . . . does not appear to have reached the prewar level.”74 Urbanites,
most employed in the state sector, received benefits denied to villagers:
employment security, pensions, and subsidized food, health care, housing,
education, and transport. The historically modest gap between urban and
rural living standards – Charles Roll places per capita rural consumption at
“about 81–88” percent of the urban average during the 1930s and “approxi-
mately the same” in 1955 – subsequently widened dramatically.75 Yang and
Zhou cite a National Bureau of Statistics working paper showing that urban
per capita incomes in 1980 were more than triple the rural average.76 Mobility
restrictions and food rationing protected higher urban living standards by
limiting migration into the cities.
Explaining Productivity Stagnation

The PRC’s plan system ramped up investment outlays, but the new regime
created distortions and inefficiencies that completely offset anticipated prod-
uctivity benefits arising from national unity, monetary stability, strong gov-
ernment, growth-oriented policies, new technology, and improved human
capabilities. Why did three decades of economic planning fail to deliver the
anticipated material benefits?
The new system severely curtailed the engines of prewar growth: private
entrepreneurship, commercial competition, and market integration that
allowed growing circulation of commodities, information, capital, technol-
ogy, and individuals within and across China’s national boundaries. The
planned economy’s crude instruments – state-owned enterprises, inflexible
prices, and government-mandated production quotas, supply links, invest-
ment projects, and job assignments – sufficed for fulfillment of official targets,
but only at the cost of creating large pools of underutilized resources.
The planned economy’s corrosive effect on individual incentives was
particularly damaging to the rural economy. The collectivization of agricul-
ture frayed the connection between personal effort and reward for three-
quarters of China’s workforce. This encouraged widespread shirking, as
74
Nicholas Lardy, “Food Consumption in the People’s Republic of China,” in R. Barker
and R. Sinha (eds.), The Chinese Agricultural Economy (Boulder, Westview Press, 1982),
p. 159.
75
C.R. Roll Jr., The Distribution of Rural Incomes in China: A Comparison of the 1930s and 1950s
(New York, Garland, 1980), p. 124.
76
D.T. Yang and H. Zhou, “Rural–Urban Disparity and Sectoral Labour Allocation in
China,” Journal of Development Studies 35.3 (1999), 112.
796

individuals and households diverted resources toward private plots, which,

beginning in 1958, occupied less than 10 percent of cultivated acreage.
A Guangdong team leader explained, “People aren’t lazy all the time, just
when they do collective labor. When they work on their private plots, they
work hard,” adding that a task that formerly required six man-days of
household labor might consume sixteen man-days of collective effort.77
Incentive problems also limited industrial advance. Socialist planning,
discussed at length in Chapter 16 in this volume by Dwight Perkins, imposed
a framework of rigid prices, mandated production quotas, and state control
over the distribution of materials as well as intermediate and final products.
This system generates a panoply of dysfunctional responses observed in all
centrally planned economies. Neither firms nor individual workers benefit
from exceeding minimum requirements. Improvements in cost, product
quality, or customer service become uncompensated gifts to buyers or to
the state, which absorbs all profits. Factory managers prioritize physical
output targets at the expense of quality, cost, and customer service.
Unprecedented Gap between Actual and Potential Output

Divergence between rising capabilities and stagnant productivity signaled an
unprecedented gap between actual production and the level of output that
existing resources, technologies, and skills could deliver. The unexpected
growth explosion following the onset of reform in the late 1970s illuminates
the enormous scale of this latent potential. We focus on three areas: trade,
agriculture, and industry.
Latent Potential in International and Domestic Exchange

Except for the transfer of Soviet technology during the 1950s, China’s plan-era
economic strategy promoted self-reliance at the expense of participation in
domestic and international commerce. While a US-led boycott limited
China’s global trade options, all restrictions on domestic commerce and
much of China’s international isolation reflected the commitment of
China’s leaders to self-reliance and local self-sufficiency. Hostility to foreign
involvement terminated China’s prewar standing as a substantial recipient of
overseas investment. Curtailment of fruitful opportunities for domestic and
international exchange imposed major economic costs.
77
S.W. Mosher, Broken Earth: The Rural Chinese (New York, The Free Press, 1983), pp.
39–40.
797

During China’s absence from active engagement with global trade and
investment, which extended for nearly fifty years from 1937, rising post-
World War I I direct investment from advanced nations, steep reduction in
transaction costs, and major increases in trade flows, including exports of
labor-intensive manufactures from low-income countries, offered opportun-
ities that China ignored. China’s long withdrawal from international
exchange deprived the economy of benefits from imported technology and
from efficient utilization of available resources. Shifting to domestic suppliers
of capital equipment following the 1960 break with the Soviet Union had
a “catastrophic effect on the quality of equipment.”78 Clinging to self-reliance
also ignored a potential export bonanza in labor-intensive manufactures
arising from the availability of vast numbers of literate, underemployed
rural youths at wages far lower than in overseas rivals.79
Restricting domestic trade unraveled long-standing patterns of regional
specialization. Costs were particularly high in the farm sector, as limited
availability of outside grain supplies necessitated the conversion of fields best
suited to growing sugar, peanuts, rape, soybeans, and other commercial
crops to grain cultivation. These shifts reduced incomes for former producers
of cash crops and for their former customers, who mounted inefficient efforts
to replace cash crop purchases with local production.80
Latent Potential in Agriculture

Historically, Chinese agriculture operated close to the production frontier
determined by available land, labor, water, fertilizer, and technology. With
no “artificial barriers” to the diffusion of “new seeds, new crops, and better
cropping patterns . . . there was no great back-log of advanced but essentially
‘traditional’ technique . . . that could be exploited readily.”81 From the start of
the PRC, investment and new technology rather than land reform or collect-
ivization held the key to future agricultural growth. Collectivization initially
78
P. Zeitz, “Trade in Equipment and Technological Development: Evidence from the
Sino-Soviet Split” (unpublished, 2010).
79
Even though average Chinese industrial wages in 1991 reached 3.8 times the 1978 level,
a multinational comparison found 1991 hourly labor costs in China’s increasingly
export-oriented textile and garment sectors to be less than one-tenth of comparable
costs in Japan, Hong Kong, and Taiwan. See 中国统计年鉴 (China Statistics
Yearbook) (hereafter Yearbook) 1992, Table 4-33; and L. Moore, “The Competitive
Position of Asian Producers of Textiles and Clothing in the US Market,” World
Economy 18.5 (1995), 589.
80
N.R. Lardy, Agriculture in China’s Modern Economic Development (Cambridge, Cambridge
81
D.H. Perkins, Agricultural Development in China, 1368–1968 (Chicago, Aldine, 1969), p. 53.
798

sought to increase farm output and resource transfers out of agriculture

without diverting investment from industry to agriculture. But its adverse
side effects – erosion of incentives and “technological commandism” – delayed
effective implementation of major advances in new high-yielding seed varieties
and promoted uneconomic expansion of triple-cropping and agricultural mech-
anization prior to the revival of household farming in the late 1970s.82
The immediate post-reform surge in rural output and TFP beginning in
the late 1970s demonstrates the “gigantic waste of labor and resources”
resulting from plan-era rural policy.83 Extraction of resources from the
agricultural economy to support industrial production and investment occu-
pied the core of China’s plan-era growth mechanism. Sluggish farm perform-
ance tied the bulk of China’s workforce to the land, slowing the transfer of
labor to higher-productivity occupations. Slow growth of food output limited
the farm sector’s capacity to feed China’s cities, necessitating the diversion of
scarce foreign exchange to support grain imports. Undernutrition further
slowed the growth of farm output.
As China entered the 1970s, deteriorating agricultural conditions threat-
ened to upend the delicate balance among food production, grain procure-
ment, and rural nutrition. The procurement system, essential to feeding
China’s cities, showed increasing disarray. Sichuan, China’s most populous
province and a major victim of the 1959–1961 famine, lurched from grain
surplus to deficit amid the threat of renewed food shortages.84 Net procure-
ment, the grain available for transfer from rural to urban areas, declined in
most years, as did grain stockpiles, forcing a discomfiting choice between
higher grain imports and further reduction of reserves.85
Beyond its economic implications, the deteriorating extraction mechanism
reflected a severe erosion of central authority. Lax controls enabled rural officials
to divert grain to local advantage: Politburo member Li Xiannian 李先念
complained that collectives reported rising grain requirements for seed and
feed despite the absence of increases in cultivated acreage or meat production.
82
T.B. Wiens, “Technological Change,” in Barker and Sinha, Agricultural Economy, pp.
110–20; and Wiens, “The Limits to Agricultural Intensification: The Suzhou
Experience,” in US Congress, Joint Economic Committee, China under the Four
Modernizations (Washington, DC, US Government Printing Office, 1982), pp. 462–74.
83
W.J. Shan, Out of the Gobi: My Story of China and America (Hoboken, NJ, Wiley, 2019), p. 240.
84
F.S. Zhao 赵发生 et al. (eds.), 当代中国的粮食工作 (Grain Work in Contemporary
China) (Beijing, Zhongguo shehui kexue chubanshe, 1988), p. 145. Provincial Party
secretary Li Jinquan’s 李井泉 September 1975 submission to the State Council
demanded prompt attention to Sichuan’s request for procurement relief to avoid
“repeating the mistake of 1959.”
85
Zhao et al., 当代中国的粮食工作, pp. 166–7.
799

Latent Potential in Industry

In addition to the weak incentives mentioned earlier, the chief source of
latent industrial production potential stems from the plan system’s rigidity.
Even without considering planners’ limited access to timely and reliable
information, the primitive calculators available to Mao-era planners limited
the feasible number of product categories.86 Fine-tuning production quotas
to include, for example, assortment requirements for metal fasteners or
shoes, was impractical. The difficulty of modifying complex production
arrays meant that successive annual plans rarely incorporated major adjust-
ments. Frequent supply lapses encouraged firms to accumulate inventories.
In the late 1970s, “China . . . carried a much larger volume of inventories and
incomplete construction than . . . the Soviet Union,” where stockpiles were
far greater than in market economies.87
Both during the plan era and today, widely varying capabilities across firms
in specific industries amplify inefficiencies arising from weak exit mechanisms
for poor performers, a problem that persists today.
Industrial policies generated additional sources of latent capacity. During
1953–1978, “heavy” industry absorbed 43 percent of overall state-sector basic
construction expenditure and 90 percent of outlays for industry.88 This
approach lavished resources on capital-intensive operations that often
churned out low-quality products. Although coastal producers generally
delivered superior performance in terms of quality, cost, and productivity,
planners directed the bulk of investment toward interior regions. This
reached a peak under the “Third Front” program, which channeled over
40 percent of national investment during 1963–1975 to a massive and largely
uneconomic heavy industry complex in China’s central and western regions
intended to guard against possible invasion.89 Emphasis on local self-
sufficiency encouraged the proliferation of inefficient local production.90
86
China’s material allocation system, which included fewer than 600 items, was “much
less extensive than the Soviet” system, which spanned “as many as 65,000” items. C.P.
W. Wong, “Ownership and Control in Chinese Industry: The Maoist Legacy and
Prospects for the 1980s,” in U.S. Congress, Joint Economic Committee, China’s
Economy Looks toward the Year 2000, vol. 1, pp. 577, 603.
87
B. Naughton, Growing Out of the Plan: Chinese Economic Reform 1978–1993 (Cambridge,
88
Guojia tongjiju gudingzichan touzi tongjisi, 1950–1985 中国固定资产投资统计资料
(Statistical Materials on Chinese Fixed Capital Invesment in 1950–1985), pp. 43, 44, 97.
89
B. Naughton, “The Third Front: Defence Industrialization in the Chinese Interior,”
China Quarterly 115 (1988), 351–86.
90
A. Donnithorne, “China’s Cellular Economy: Some Economic Trends since the
Cultural Revolution,” China Quarterly 52 (1972), 605–19.
800

Even as food supply issues stalled China’s economic growth, the widening
gap between actual and potential output both within and beyond the farm
sector offered the possibility that suitable reforms could rapidly generate
large increases in output. In addition to directly raising agricultural produc-
tion, rural reform promised to promote economy-wide growth by accelerat-
ing the reallocation of labor into nonagricultural activities in which returns
were even higher. This is exactly what happened.
The Reform Era

China’s economy entered the reform era in difficult straits. Three decades of
socialist planning had expanded the scale and scope of industry and upgraded
its technical capabilities; the new system also delivered notable advances in
education, public health, and life expectancy. Despite these gains, massive
inefficiency kept the economy far below its potential. Lagging food produc-
tion left hundreds of millions underfed and threatened to destabilize key
flows underpinning the economy’s advance.
In sharp contrast, four decades of reform have brought a remarkable
transformation. Some metrics now identify China’s economy as the world’s
largest. Rapid structural change has steeply reduced the importance of
agriculture, with the primary sector’s share of aggregate output falling
from 27.7 percent in 1978 to less than 10 percent beginning in 2009. Official
estimates show that primary-sector employment has fallen even faster, from
83.5 percent in 1978 to half or less beginning in 1997 and 26.1 percent in 2018.
Industry and services have moved to the forefront, with services gradually
taking the lead, surpassing industry’s share of employment in 1994 and output
in 2012. Massive population shifts have raised the urban share of China’s
population to 60 percent.91 China has emerged as a great trading nation,
a global science and innovation powerhouse,92 and both a leading recipient
and a major source of overseas investment.
Our analysis emphasizes the twin processes of economic transition –
the shift from plan to market in the allocation of resources, and structural
transformation, most notably the movement of people and resources out
of agriculture and into industry and services. Along with productivity
91
Yearbook 2019, Tables 2-7, 3-2, 4-2. Official sources overestimate employment in the
primary sector, which includes forestry and fisheries as well as agriculture.
92
R.B. Freeman and W. Huang, “China’s ‘Great Leap Forward’ in Science and
Engineering,” in A. Geuna (ed.), Global Mobility of Research Scientists: The Economics of
Who Goes Where and Why (London, Academic Press, 2015), pp. 155–75.
801

improvements within individual sectors, the transfer of resources along

productivity-enhancing paths toward nonagricultural activity, non-state
enterprises, and coastal locations delivered more than three-fourths of the
increase in per capita incomes during the first three decades of the reform
era, with the rest coming from capital deepening and rising education
levels.93
The central role of productivity growth and resource reallocation
during China’s long boom conceals deep-rooted tensions between eco-
nomic advance and the state’s noneconomic objectives. Rapid productiv-
ity growth in non-state industry and services has elevated returns to
investment and thus sustained the incentives for high rates of capital
formation. By the start of the global financial crisis in 2008, the non-
state sector’s share of investment had increased from slightly more than
10 percent in 1978 to nearly half.94 The rest went to the state sector, where
returns to capital were often negative and productivity growth was only
a third or a quarter of what the non-state sector delivered. Despite these
dismal economic returns, China’s leaders continue to promote state-
sector investment, which over the last decade has averaged roughly
20 percent of GDP, to advance multiple noneconomic objectives – patron-
age and network building, national security, and demonstrations of
national might. In the wake of the 2008 crisis, policies that steer resources
toward the state sector and extend official intervention in private-sector
management threaten to curtail China’s economic growth.
We divide the reform era into three phases: reform from below, extending
into the early 1990s; the following decade and a half of more organized,
centrally directed reform initiatives; and the current period, beginning with
the global crash, dominated by top-down innovation plans.
Stage 1: Reform from Below – Decentralized Initiative and

Central Reaction
Reform commenced in the villages. While scholars dispute the relative
importance of spontaneous grassroots action and local government decisions
in the rapid shift from collective to household cultivation, the impotence of
central leadership is indisputable. Major documents issued by central CCP
bodies in 1979 and 1980 bristle with calls for restoring rural workers’
93
X.D. Zhu, “Understanding China’s Growth: Past, Present, and Future,” Journal of
Economic Perspectives 26.4 (2012), 108.
94
L. Brandt and X.D. Zhu, “Accounting for China’s Growth,” University of Toronto
Department of Economics Working Paper 394 (2010), Figure 2.
802

production enthusiasm (shengchan jijixing 生产积极性), while prohibiting

household cultivation, lauding collectives as the “unshakable foundation” of
agrarian progress and denying that household activity could support “the
establishment of modern agriculture.”95
Subsequent developments highlight the center’s irrelevance. Noting that
contracting to households had aroused “great enthusiasm among the
masses,” the summary of a 1981 agricultural reform conference notes that
“since reality has already outrun the [1980] directive . . . delegates suggested
that the Center promptly formulate new documents reflecting the new
circumstances.”96
Restoration of household farming, along with partial decontrol of rural
marketing and individual entrepreneurship, propelled swift increases in both
agricultural output and productivity,97 even as millions abandoned farming for
newly emerging opportunities in industry and services. Sichuan and Anhui,
provinces that had suffered the most during the Great Leap Famine, led these
rural reforms.98 The suddenness of the ensuing shift from near-stagnation to
rapid growth, which generated nationwide improvements in rural incomes and
food availability, reveals the centrality of institutional changes that simultan-
eously restored incentives, encouraged greater work effort, and allowed agricul-
ture to exploit the untapped potential of new seeds, chemical fertilizer, and
expanded irrigation accumulated under the collective regime.99
Alongside these rural developments, growing awareness that prolonged
isolation had stranded Chinese industry and technology far behind its East
Asian neighbors, as well as North America and Western Europe, inspired
plans for a big push to upgrade domestic technology and equipment.100 The
95
中国农业年鉴 1980 (Beijing, Nongye chubanshe, 1981), pp. 57–8; 中国农业年鉴 1981
(Beijing, Nongye chubanshe, 1982), pp. 409–10.
96
“全国农业经济问题讨论会纪要” (Summary of the National Symposium on
Agricultural Issues), 农业经济问题 (Agricultural Economic Issues ) 10 (1981), 2. Also
A. Watson, “Agriculture Looks for ‘Shoes That Fit’: The Production Responsibility
System and Its Implications,” World Development 11.8 (1983), 713.
97
J.Y. Lin, “Rural Reforms and Agricultural Growth in China,” American Economic Review
82.1 (1982), 46, attributes 48.69 percent of the output growth during 1978–19884 to
decollectivization. Fan and Zhang, “Production and Productivity Growth,” Table 5,
find that, with 1952 = 100, TFP in agriculture (based on constant 1980 prices) jumped
from 67 in 1978 to 82 in 1982 and 129 in 1992.
98
D.L. Yang, Calamity and Reform in China: State, Rural Society, and Institutional Change
since the Great Leap Forward (Stanford, Stanford University Press, 1996).
99
J.K. Huang and S. Rozelle, “Technological Change: Rediscovering the Engine of
Productivity Growth in China’s Rural Economy,” Journal of Development Economics
49.2 (1996), 337–69.
100
D.H. Perkins, “Reforming China’s Economic System,” Journal of Economic Literature
26.2 (1988), 618.
803

collapse of this effort, which quickly outran China’s puny export earnings,
prompted hesitant urban reforms aimed at “enlivening” operations within
the plan system by modestly extending state enterprise managers’ decision-
making authority and expanding opportunities to buy and sell industrial
materials and products.
The dual-track system, which preserved administered prices for plan-
related distributions while allowing market sales of above-plan output,
broadened market opportunities and sharpened incentives within the state
sector.101 It also encouraged the growth of more efficient producers, particu-
larly benefiting TVEs clustered in coastal provinces. Dual pricing created
market-based price signals in nearly every sector, a critical step in expanding
market-oriented reform, and modestly sharpened incentives within the state
sector. At the same time, the arrangement preserved rents accruing to plan
participants. This reduced opposition to market reform, but created lucrative
opportunities to resell underpriced goods acquired through plan allocations
at higher market prices.
Expansion of overseas trade and investment, led by the creation of special
economic zones, added an international dimension to China’s boom. China’s
opening coincided with efforts by Taiwan and Hong Kong entrepreneurs,
responding to rising wages in their home markets, to find low-cost venues for
labor-intensive export production. The combination of local land and labor
along China’s coast with the market knowledge, manufacturing experience,
and financial resources of these operators shifted growing numbers of rural
workers into manufacturing jobs and brought rapid growth of factory
exports.
Although the initial reforms affected the entire economy, the largest
impact occurred outside the cities and beyond the state sector. Unlike rural
reform, which often involved little more than lifting restrictions that had
suppressed long-standing patterns of production and marketing, urban
reform required the construction of new and unfamiliar institutions, to
which state enterprises, managers, and workers, many with no experience
of market discipline,102 would have to adapt. Such changes inevitably encoun-
tered opposition from entrenched interests.
Not surprisingly, individuals and firms on the fringes of the plan system
took the lead. Rural incomes jumped upward, narrowing the gap with city
101
W. Li, “The Impact of Economic Reform on the Performance of Chinese State
Enterprises, 1980–1989,” Journal of Political Economy 105 (1997), 1080–1106.
102
State-owned industrial firms numbered 15,190 in 1955 and 83,400 in 1980; see N.R. Chen,
Chinese Economic Statistics (Chicago, Aldine, 1967), p. 182; and Yearbook 1981, 204.
804

folk.103 Rural firms soon penetrated urban markets, slashing the profits of
state-owned rivals.104 Collective and privately owned firms gained a foothold
in the new export sector. Relaxation of mobility restrictions sparked the
initial phase of what later developed into a tidal wave of migration into
China’s cities; the late 1970s and early 1980s saw the return of many urbanites
“sent down” to rural villages, while villagers sought opportunities to fill gaps
created by the plan system’s repression of retail and service businesses.105
With increases in output, productivity, profits, and revenues clustered in
rural areas and in non-state enterprises under the supervision of local gov-
ernments, the center found itself scrambling to fund its priorities. Both the
ratio of government revenue to GDP and the center’s share of overall
revenue, much of it derived from SOE profits, declined.106 The center’s
unwillingness to reduce urban real incomes by imposing higher grain prices
saddled the state budget with growing outlays to bridge the gap between
rising grain costs and lower fixed retail prices. A further obstacle arose when
state-owned commercial banks, responding to reform-enhanced profit
motives, steered resources to emerging non-bank financial institutions
(NBFIs) that extended credit to fast-growing collectives and private firms.
The state now lacked sufficient budgetary and banking support to imple-
ment plans for expanding employment, wages, and investment in the lagging
state sector. Urban SOE employment increased more than 50 percent during
1978–1994. The center turned to the People’s Bank of China (PBOC), China’s
central bank, to extend lending to the commercial banks, which used these
additional resources to implement the credit plan’s provisions for “state
sector working-capital and fixed investment needs.”107 This short-term
response proved costly, as PBOC intervention caused increases in money
supply and prices, rekindling memories of wartime hyperinflation – a key
103
D.Y. Yang and F. Cai, “The Political Economy of China’s Rural–Urban Divide,”
Stanford Center for International Development, Working Paper No. 62, 2000, p. 32,
find that in real terms the urban–rural ratio for consumption (not income) dropped
from 2.9 in 1978 to 1.9 in 1985, then rebounded to 2.5 in 1992.
104
B. Naughton, “Implications of the State Monopoly over Industry and Its Relaxation,”
Modern China 18.1 (1992), 14–41.
105
D.J. Solinger, Chinese Business under Socialism: The Politics of Domestic Commerce, 1949–
1980 (Berkeley, University of California Press, 1984), p. 325, notes that the number of
shops, restaurants, and commercial centers “under commercial departments, in urban
and industrial and mining areas,” dropped from 1 million to 180,000 between 1957 and
1978.
106
Wong and Bird, “China’s Fiscal System,” 433.
107
L. Brandt and X.D. Zhu, “China’s Banking Sector and Economic Growth,” in C.
W. Calomiris (ed.), China’s Financial Transition at a Crossroads (New York, Columbia
805

ingredient in the CCP’s victory over the Guomindang.108 Official interven-

tion to limit monetary growth by constricting the supply of credit to the
dynamic non-state sector restrained inflation, but also lowered the overall
growth rate. The result was a series of stop–go cycles in which periods of
accelerated growth led by non-state firms alternated with intervals of reduced
credit and output growth.109
Despite these tensions, which helped to spark the unrest that culminated in
top-level purges and violent suppression of mass protests in 1989, this initial
stage of reform delivered an astonishing turnaround that accelerated the
growth of overall output. In stark contrast to the plan era, the initial reforms
increased personal incomes and released several hundred million villagers
from the scourge of absolute poverty.110
For the first time, China experienced widespread productivity growth,
reflecting the joint impact of transition and development. Transition partially
restored market exchange, market prices,111 personal mobility, and openness
to entry and competition from both domestic and overseas firms and prod-
ucts. This enabled China’s first ever large-scale shift out of agriculture, as
non-primary employment more than doubled, adding over 150 million work-
ers between 1978 and 1992.112
Deng Xiaoping’s endorsement of growth and rejection of long-standing
egalitarian emphasis highlighted an unprecedented alignment of incentives,113
as a widely shared preference for growth now united villagers seeking to
escape collective control, workers hungry for bonuses, managers and bankers
pursuing profits, and officials whose career prospects and informal incomes
now rested increasingly on raising output.114
108
K.N. Chang, The Inflationary Spiral: The Experience in China, 1939–1950 (Cambridge, MA,
MIT Press, 1958).
109
L. Brandt and X.D. Zhu, “Redistribution in a Decentralized Economy: Growth and
Inflation in China under Reform,” Journal of Political Economy 108.2 (2000), 422–39.
110
M. Ravallion and S.H. Chen, “China’s (Uneven) Progress against Poverty,” Journal of
Development Economics 82.1 (2006), 1–42.
111
By 1990, market prices governed just over half of retail transactions and exchange of
agricultural products; for production materials, the share of market pricing was 36.4 percent.
H. Dinh, T.G. Rawski, A. Zafar, L.H. Wang, and E. Mavroeidi, with X. Tong and P.F. Li,
Tales from the Development Frontier: How China and Other Countries Harness Light
Manufacturing to Create Jobs and Prosperity (Washington, DC, World Bank, 2013), p. 77.
112
Yearbook 2019, Table 4–2.
113
E.F. Vogel, Deng Xiaoping and the Transformation of China (Cambridge, MA, Harvard
University Press, 2011), 242, dates this from 1978, when “allowing some regions and
enterprises to get rich first” was a major theme of Deng’s December 13 speech to the
Central Party Work Conference.
114
H.B. Li and L.A. Zhou, “Political Turnover and Economic Performance: The Incentive
Role of Personnel Control in China,” Journal of Public Economics 89 (2005), 1743–62.
806

Along with remarkable economic advance, China’s initial reforms exposed

a fundamental duality between the economy’s dynamic segments, which
clustered outside the cities and beyond the state sector,115 and the lagging,
resource-hungry state sector. A stark performance gap separated the two:
between 1980 and 1992, growth of output, labor productivity, and TFP in
state-owned industries was only a half to a third of that in collective and
private firms.116 Even so, Beijing continued to see the state sector as central to
its pursuit of multiple objectives, many extending beyond narrowly eco-
nomic outcomes, and as a portfolio of resources available to supplement
state appropriations and to reinforce loyalty within the ruling coalition.
Heavy reliance on the state sector explains why, despite its evident
economic weakness, the annual “flow of resources through the financial
system making its way to the state sector” during 1978–1994 amounted to 15
to 20 percent of GDP.117 As China gradually recovered from the tempestuous
events of 1989, further reform seemed essential to resolve a fundamental
conflict between the desire for continued rapid growth and the drain from
large-scale transfers to underperforming segments of the economy.
Stage 2: Major Reform Initiatives Extend Market Forces and

Restore Central Control
Suppression of the June 1989 Beijing protests left China’s central leadership
badly shaken. Ousting CCP general secretary and former premier Zhao
Ziyang and his allies while mobilizing the army to terminate public protests
fractured the top echelons of power and blurred lines of control over routine
economic administration.
The economy stumbled: employment growth during 1988–1989 dropped
to less than one-third of the average over the preceding decade, while
nominal investment outlays declined for the first time since 1980–1981.118
The GDP share of government revenue and expenditure, which had stabil-
ized at the end of the 1980s following a decade of decline, resumed its
downward march.
115
Y.S. Huang, “How Did China Take Off?”, Journal of Economic Perspectives 26.4 (2012).
116
G.H. Jefferson and T.G. Rawski, “Enterprise Reform in Chinese Industry,” Journal of
Economic Perspectives 8.2 (1994), 48, 56.
117
Brandt and Zhu, “China’s Banking Sector and Economic Growth,” 96–9. Over this
period, more than 60 percent of capital formation, and two-thirds of all new banking
loans, went to the state sector.
118
Employment data from Yearbook 1991, Table 4-8; investment data from Yearbook 1991,
Tables 5-20, 5-35; and from 1950–1985 中国固定资产投资统计资料, pp. 49, 216.
807

Despite this unlikely start, Deng Xiaoping’s 1992 “Southern Tour” ignited an
avalanche of growth that outstripped the impressive early reform achievements.
This renewed growth rested, in turn, on constructive actions that swept aside
multiple constraints and further expanded the influence of market forces, while
restoring the power and authority of the CCP and the central state. Major
reforms affected public finance, banking, state enterprises, and market opening.
Fiscal Restructuring
Tax reform implemented in 1994 reversed the long decline in the GDP share
of fiscal revenue, increased the central government’s claim on overall rev-
enue, and, perhaps most important for re-establishing central authority,
ensured that province-level units, “including Shanghai and Beijing,” became
“dependent on central transfers to finance expenditures.”119
Bank Reform
During the 1980s, the main source of investment funding shifted from
budgetary grants to bank loans. State enterprises, the main recipients,
“turned increasingly to bank credit without much concern about their future
ability to repay.”120 This led to an epidemic of payment arrears: estimates
show that, by 1998, half or more of bank loans were “non-performing.”121
During the late 1990s, the central government took major steps to rectify
this dangerous situation. Newly created asset management companies pur-
chased vast tranches of bad loans, thereby recapitalizing the floundering
state-owned commercial banks. The center increased its control over the
financial system: shuttering weak financial firms, closing down most NBFIs,
reorganizing the central bank’s subnational branches to reduce the influence
of provincial and local leaders, and increasing the influence of high-level
officials in the appointment and promotion of bank executives.
The removal of bad loans, coupled with the establishment of policy banks to
shoulder the burden of noncommercial finance, greatly strengthened the lending
capacity of China’s four giant commercial banks. Although politically directed
lending continued, the commercial element in bank operations deepened.122
119
Wong and Bird, “China’s Fiscal System,” 437.
120
B. Naughton, The Chinese Economy: Transitions and Growth (Cambridge, MIT Press,
2007), p. 306.
121
Brandt and Zhu, “China’s Banking Sector and Economic Growth,” 128.
122
J. Stent, China’s Banking Transformation (New York, Oxford University Press, 2017).
A textile executive commented, “Banks are not the same as before. Now if you have no
money and can’t repay, they won’t lend to you” (May 1996 interview).
808

State Enterprises
The focus of reform shifted from flows (of new workers, new investments,
and above-plan output) toward more complex realignments affecting embed-
ded resource stocks, including workers and entire firms. Beijing’s vision of
the state sector’s role narrowed, with textiles, food processing, and other
industries now classed as “competitive,” implying that preservation of state-
sector dominance, and even the survival of individual firms, were no longer
essential.
Privatization, often via management buyouts, multiplied, as did bankrupt-
cies and closures. The overall number of state-owned enterprises plunged
from 262,000 to 112,000 between 1997 and 2007; for industry, the total declined
from 103,300 in 1992 to 20,680 in 2007.123 Severe culling eliminated over one-
third of state-sector personnel, formerly endowed with (often heritable)
lifetime tenure; between 1996 and 2000 alone, the state-sector headcount
plunged from 113 million to 67 million.124
Bottom-up initiatives originating with provincial and local authorities,
which had gained control over large segments of state-owned industry
following decentralization programs in 1957 and 1970,125 dominated these
downsizing efforts. Subnational governments welcomed opportunities to
shed the burden of maintaining weak enterprises, including TVEs and
other collective enterprises as well as state-owned firms, that could not
withstand intensifying market pressures.
The center, by contrast, acted to strengthen enterprises under its direct
control. Following the 2003 creation of the State-Owned Assets Supervisory
Commission (SASAC), policy effort focused on the complex and rapidly
expanding operations of roughly 100 giant state-owned enterprise groups in
key commodity (petroleum, grain), manufacturing (steel, aluminum, air-
craft), infrastructure (railroads, electricity, telecoms), and financial (banking,
insurance) sectors. These efforts helped to maintain the state’s share in GDP
while increasing the share of the state sector under central government
control.
123
K.J. Lin, X.Y. Lu, J.S. Zhang, and Y. Zheng, “State-Owned Enterprises in China:
A Review of 40 Years of Research and Practice,” China Journal of Accounting Research
13 (2020), 34; Yearbook 1995, Table 12-1; Yearbook 2008, Table 13-8 (including state-
controlled industrial units).
124
Yearbook 2005, Table 5-4. 125 Wong, “Ownership and Control.”
809

Market Opening
The scope of market-based transactions continued to expand. Rapid growth
of highway and water transport, much of it in the hands of unregulated
private operators, contributed to the erosion of local protectionism and
interprovincial trade barriers.126 Analysis based on monthly data for ninety-
three products in thirty-six major cities found that “prices did converge”
during 1990–2003, and that “the patterns of convergence . . . were highly
comparable” to observations from “the United States, Canada, and European
countries” – all indicating the powerful influence of market forces.127
Employment became increasingly market-based. The former system of job
assignments faded, as graduating students and employers sought mutually
advantageous matches. Market expansion unleashed a torrent of internal migra-
tion – a familiar phenomenon in China’s modern history.128 In 2001, Premier Zhu
Rongji bluntly advised “laid-off workers . . . to find jobs on the private labor
market.”129
SOE reform and sweeping privatization of collective enterprises, together
with modest improvements in the legal protections surrounding private
ownership and Jiang Zemin’s 2001 decision to admit entrepreneurs to
Communist Party membership, improved the position of private business.
These changes, along with widespread privatization of collective firms,
spurred explosive growth in the private sector’s share of output and espe-
cially employment. Between 1992 and 2007, urban private employment rose
from 10.6 million to 78.9 million; in the countryside, 2007 private-enterprise
employment surpassed 110 million.130 These trends benefited from
126
Yearbook 2010, Tables 16-4, 16-8, 16-24, shows that between 1990 and 2007, China’s truck
fleet increased from 3.7 to 10.5 million vehicles; during the same period, the length of
highways as well as the annual volume of freight carriage along inland waterways
more than tripled.
127
C.S. Fan and X.D. Wei, “The Law of One Price: Evidence from the Transitional
Economy of China,” Review of Economics and Statistics 88.4 (2006), 694.
128
In addition to overseas migrations, major domestic population movements include
Qing-era migration into Sichuan, the resettlement of lands devastated by the Taiping
wars, and large-scale population movement into Manchuria during the late nineteenth
and early twentieth centuries. See M. Bastid-Bruguiere, “Currents of Social Change,”
in J.K. Fairbank and K.C. Liu (eds.), The Cambridge History of China, vol. 11, Late Ch’ing,
1800–1911, part 2 (Cambridge, Cambridge University Press, 1980), pp. 582–6; T.
R. Gottschang and D. Lary, Swallows and Settlers: The Great Migration from North
China to Manchuria (Ann Arbor, University of Michigan Center for Chinese Studies,
2000).
129
Q.W. Zhu, “Domestic Market Fuels Growth,” China Daily, August 6, 2001, 4.
130
Yearbook 2011, Table 4-2; 中国乡镇企业及农产品加工业年鉴 2008 (electronic edi-
tion, no page or table numbers, accessed June 29, 2020). Both urban and rural
employment include individual proprietorships.
810

“extremely rapid growth of credit to private and individual businesses”

following the 1994 implementation of China’s Company Law.131
In the late 1990s, sweeping privatization of urban housing created
a property market that hugely increased the wealth of urban households,
creating opportunities for new owners to finance private businesses and
overseas education for their children.132
Along with domestic opening, China moved to rejoin the global economy.
Hesitant initial steps, notably the opening of tiny special economic zones,
developed into a powerful push to regain and then surpass China’s prewar
footprint in global trade and investment. Tariff reductions and other meas-
ures implemented ahead of China’s 2001 accession to the World Trade
Organization created “one of the developing world’s most open trade
and FDI regimes,” highlighting China’s growing involvement in cross-
border flows of commodities, investment, technology, information, and
individuals.133
Rapid expansion of international trade and investment added momen-
tum to domestic growth. China’s share of global merchandise trade grew
from 0.9 to 2.2 percent between 1980 and 1992 – neither exceeding the
prewar figures noted above – to 2.7, 3.6, and 7.7 percent in 1995, 2000, and
2007. China’s trade share overtook Japan’s in 2004.134 Rising foreign direct
investment (FDI), much of it from Taiwan and Hong Kong, and often
directed toward export-oriented manufacturing, along with authorization
of growing numbers of domestic firms to conduct international trade,135
brought considerations of cost and profit to the fore, shifting trade patterns
toward the underlying structure of comparative advantage. Chinese firms
began to join international supply chains, accelerating the spread of man-
agement skills.
Beginning in the 1990s, large FDI inflows enabled China to recover its
prewar standing as a major destination for overseas investment. China’s
share of the global FDI stock housed in developing nations, which exceeded
131
Lardy, Markets over Mao, p. 102.
132
H.M. Fan, G.L. Gu, W. Xiong, and L.A. Zhou, “Demystifying the Chinese Housing
Boom,” in M. Eichenbaum and J.A. Parker (eds.), NBER Macroeconomics Annual 2015
(Chicago, University of Chicago Press, 2016), pp. 105–66.
133
L. Branstetter and N.R. Lardy, “China’s Embrace of Globalization,” in Brandt and
Rawski, China’s Great Economic Transformation, p. 676.
134
Post-1949 figures from https://data.wto.org, accessed July 14, 2020.
135
Branstetter and Lardy, “China’s Embrace of Globalization,” p. 635, note the number of
companies authorized to conduct international trade: twelve in 1978, 800 in 1985, and
35,000 in 2001.
811

15 percent during the 1930s, achieved similar levels again by the late 1990s.136
While China has consistently been among the top three recipients of FDI
since the early 1990s, its share of the worldwide FDI stock in 2019 remains
below half of the 1930s figure of 11 percent.137
Outcomes
Market opening encouraged accelerated structural change that moved
resources toward more productive uses. The primary sector’s GDP share
dropped from one-fifth to one-tenth between 1992 and 2007, while the tertiary
(service) sector’s share jumped from 36 to 43 percent. The official measure of
China’s primary-sector labor force peaked in 1991; by 2007, it had declined by
83.7 million. Employment growth clustered in the service sector, which
added 113 million workers during the same years.138
The growing influence of market forces pulled resources into coastal
regions, which increased their weight in overall production and investment
while dominating export production and absorption of incoming foreign
investment.139 The share of China’s eastern region in overall fixed investment
jumped from about one-third prior to 1975 to over 60 percent during the mid-
1990s.140 A 2008 survey clearly demarcated the geographic locus of economic
dynamism: of 140 million internal migrants who had left their home counties,
70 percent originated in China’s central or western regions, and 62 percent
had moved to eastern provinces, which housed 43 percent of the national
population.141
Growing internationalization intensified the impact of domestic-market
opening on competition, cost reduction, and quality improvement. Tariff
reductions and other liberalization measures implemented ahead of China’s
136
The Asian financial crisis temporarily lowered China’s FDI inflows and its share of the
global FDI stock.
137
Calculated from UNCTAD, World Investment Report 2020, Annex Table 1; these data
exclude FDI flows into Hong Kong.
138
Yearbook 2019, Tables 3-2, 4-2.
139
X.J. Jiang, FDI in China: Contributions to Growth, Restructuring and Competitiveness
(New York, Nova Science Publishers, 2004), p. 82, notes that, as of late 2001, 86 percent
of FDI had located in China’s eastern region.
140
NBS, “固定资产投资水平不断提升对发展的关键性作用持续发挥” (The
Ongoing Rise in the Level of Fixed Asset Investment Continues to Play a Key Role
in Development), at 70prc.cn/2019-09/19/c_138404706.htm, posted September 19,
2019, accessed June 29, 2020.
141
Yearbook 2009, Table 3-4; 2008 年末全国农民工总量为 22542 万人 (At the End of
2008, the Total Number of Migrant Workers Nationwide Was 225.42 Million), at stats
.gov.cn/ztjc/ztfx/fxbg/200903/t20090325_16116.html, accessed July 13, 2020. The data
on regional origins and destinations are limited to migrants with fixed employment.
812

WTO accession represented “a watershed” that forced widespread cost

reductions.142 Growing competition from imports and from an expanding
array of domestic producers created pressures that increased productivity and
reduced both the level and the dispersion of sales markups.143
Foreign-invested firms occupied a “vital role . . . [in] transfers of technol-
ogy, production and organizational skills, managerial know-how, and mar-
keting expertise” that powered “robust progress” in China’s “capacity to
manufacture a growing array of internationally competitive products.”144
Overseas firms, eager to capitalize on low Chinese costs, promoted domestic
supply chains to feed their Chinese assembly plants. Along with the arrival of
overseas component manufacturers, these supply networks absorbed thou-
sands of local firms: by the year 2000, “of Motorola’s 700-odd suppliers in
China . . . more than 400 are domestic.”145
These changes generated striking economic results. Following a brief
slowdown in the wake of the 1989 disturbances, rapid growth resumed:
measured at international prices, per capita income rose at an annual rate
of 6.4 percent during 1992–2007.146 As in the initial reform phase, productivity
growth, dormant prior to 1978, continued as the primary driver of expansion
for the entire economy and for industry, the largest sector.147
The period between 1992 and the 2008 global financial crisis represents an
interlude of relative political calm in which contentious debate about the
long-term objective of economic policy continued even as major reforms
delivered large and tangible benefits to advocates of both market transform-
ation and state-led development.
Liberalizing reformers rejoiced as openness, entry and competition swept
across large swathes of China’s economic landscape. Jiang Zemin’s dual 2001
initiatives, first opening the CCP to private entrepreneurs, and then propos-
ing a “socialist market economy with Chinese characteristics,” fanned expect-
ations of gradual convergence to market outcomes. Beyond economics,
the broad liberalizing agenda of disgraced former CCP general secretary
Zhao Ziyang “happened by evolution,” with growing “separation of
142
Branstetter and Lardy, “China’s Embrace of Globalization,” p. 656.
143
L. Brandt, J. van Biesebroeck, L.H. Wang, and Y.F. Zhang, “WTO Accession and
Performance of Chinese Manufacturing Firms,” American Economic Review 107.9 (2017),
2784–820; Y. Lu and L.H. Yu, “Trade Liberalization and Markup Dispersion: Evidence
from China’s WTO Accession,” American Economic Journal: Applied Economics 7.2 (2015),
221–53.
144
L. Brandt, T.G. Rawski, and J. Sutton, “China’s Industrial Development,” in Brandt
and Rawski, China’s Great Economic Transformation, pp. 622–3.
145
Jiang, FDI in China, 29. 146 Calculated from Penn World Tables v. 9.1.
147
Perkins and Rawski, “Forecasting,” 839; Brandt et al., “WTO Accession.”
813

responsibilities and spheres of authority,” leaders chosen “for their policy-

relevant expertise . . . economic policy-makers at all levels suffer less and less
frequently from intervention by the ideology-and-mobilization specialists,”
while “neither the top leader nor the central Party organs interfere as much in
the work of other agencies” as in the past, and “ideological considerations
have only marginal, if any, influence on most policy decisions.”148
Developments between 1992 and 2007 equally reinforced the position and
prospects for state-led development. The collapse of the Soviet Union
alarmed Chinese elites. Fears that China might experience similar centrifugal
pressures reinforced CCP claims that it alone could ensure national unity and
guide China to a position of global prominence. Patriotic education cam-
paigns promoted “national greatness,” echoing early twentieth-century pol-
itical discourse. A string of diplomatic triumphs – the 1997 return of
Hong Kong, 2001 entry into the World Trade Organization, and the selection
of Beijing to host the 2008 summer Olympics – highlighted the CCP regime’s
capacity to deliver benefits extending far beyond economic growth.
In tandem with growing market influence, developments between 1992
and 2007 multiplied the power of the central state. Beijing maintained strong
control over large segments of the economy, including major upstream
industries (petroleum, electricity), railroads, and large segments of the service
sector (finance, telecoms). Fiscal and banking reforms massively enlarged the
central state’s command over resources, while state-sector downsizing,
urban housing privatization, and the termination of urban food subsidies
eliminated large fiscal burdens. Economic success created vast pools of
discretionary funds: between 1992–1993 and 2007, central government rev-
enue, state enterprise assets and profits, nationwide financial deposits, and
foreign-exchange reserves each rose far more rapidly than China’s GDP.149
Giant centrally supervised enterprise groups, some with thousands of subsid-
iaries, amassed 2007 profits equivalent to 4 percent of GDP.150 Their opaque
corporate structures, along with booming infrastructure spending, multiplied
opportunities to distribute rents, a key link in maintaining elite support, on
a grand scale. One account describes state-directed investment as “the prime
enabler of corruption.”151
148
A.J. Nathan, “China’s Changing of the Guard: Authoritarian Resilience,” Journal of
Democracy 14.1 (2003), 11–13.
149
All measured at current prices. See the online appendix referenced in note 44 above.
150
B. Naughton, “SASAC and Rising Corporate Power in China,” China Leadership Monitor
24 (2008), 2.
151
J. Du, Y. Lu, and Z.G. Tao, “Government Expropriation and Chinese-Style Firm
Diversification,” Journal of Comparative Economics 43 (2015), esp. 166–8; J. Osburg,
814

Deep resource pools enabled the implementation of large, top-down

development projects, notably a major initiative to develop China’s west-
ern region, begun in the year 2000, and the initial phase of building
national networks of expressways and high-speed rail lines. Beyond
these specific programs, the incoming leadership group headed by Hu
Jintao and Wen Jiabao abandoned former premier Zhu Rongji’s downsiz-
ing of central government scale and functions in favor of a more activist
approach. Beginning in 2003, the new leaders shifted technology upgrad-
ing “expenditure . . . towards domestic research and development . . . and
away from technology import,” raised “direct government expenditure
on techno-industrial projects,” and instituted a steep rise in “the number
of industrial policies” that supported “specific sectors, firms, or
technologies.”152
The fifteen years prior to the 2008 financial crisis witnessed rapid
evolution of China’s economy. Growth flourished, largely driven by rising
productivity. Domestic and international opening enlarged the influence
of market signals and pressures. Reforms also expanded the state’s com-
mand over resources, encouraging a turn toward governmental activism.
With movement toward marketization “stalled out” following the 2003–
2004 turn toward governmental activism, the overall weight of market
elements in China’s economy began to recede in advance of the 2008
global crash.153
Stage 3: Toward State Capitalism

The 2008 global financial crisis enhanced state influence in China, as in all
major economies. Beijing responded to the steep downturn with a blizzard of
new credit, most channeled through state-controlled entities and directed
toward urban infrastructure. Following a rapid recovery, growth continued,
although at considerably reduced rates that some analysts view as
exaggerated.154
“Global Capitalisms in Asia: Beyond State and Market in China,” Journal of Asian Studies
72.4 (2013), 824.
152
L. Chen and B. Naughton, “An Institutionalized Policy-Making Mechanism: China’s
Return to Techno-industrial Policy,” Research Policy 45 (2016), 2141.
153
B. Naughton, “The Return of Planning in China: Comment on Heilmann–Melton and
Hu Angang,” Modern China 39.6 (2013), 651.
154
Y.Y. Hu and J.X. Yao, “Illuminating Economic Growth,” IMF Working Paper 19/77
(2019); W. Chen, X.L. Chen, C.T. Hsieh, and Z.M. Song, “A Forensic Examination of
China’s National Accounts,” Brookings Papers on Economic Activity 1 (2019), 77–141.
815

State Control to the Fore

Economic policy redoubled the emphasis on state leadership and adopted
a new trajectory in which cutting-edge innovation supplants technological
catch-up as the key driver of expansion. President Xi’s “China Dream” sees
domestic prosperity and technical advance as twin springboards for
a nationalist agenda targeting regional and global leadership across multiple
arenas: innovation, trade, investment, diplomacy, science, and the military.
Two signature policies, “Made in China 2025” and “One Belt, One Road”
illuminate current economic priorities. Both contrast sharply with the rec-
ommendation of greater openness, entry, competition, and market allocation
in China 2030, a major 2013 study by the Development Research Center under
China’s State Council and the World Bank.
Made in China 2025, a long-term program developed by the Chinese
Academy of Engineering, a bastion of top-down planning, establishes time-
tables for attaining an array of advanced manufacturing milestones, often
including specific figures for output volume and domestic or even global
market share.155 With its focus on quantitative targets and neglect of compe-
tition, prices, and costs, this program, while dealing with a new set of
industries and technologies, embodies a top-down, nonmarket strategy that
echoes China’s plans of the 1950s. Its nonmarket approach resembles subse-
quent initiatives, especially the 2006 “National Medium- to Long-Term Plan
for the Development of Science and Technology” and the 2010 “Decision of
the State Council on Accelerating the Fostering and Development of
Strategic Emerging Industries.”
The Belt and Road program proposes a vast network of energy and
infrastructure facilities spanning the entire Eurasian land mass, with exten-
sions to Africa and Latin America. This initiative, which combines aid,
lending, trade, and diplomacy, seeks to deepen China’s ties with low- and
middle-income nations, in part to offset weakening demand growth for
Chinese products in advanced markets.156 This agenda showcases Chinese
capabilities in design, finance, management, construction, and hardware
manufacture linked to an array of upstream industries, many awash in excess
production capacity. While China continues as a leading global destination
155
J. Wübbeke, M. Meissner, M.J. Zenglein, J. Ives, and B. Conrad, “Made in China 2025:
The Making of a High-Tech Superpower and Consequences for Industrial Countries,”
MERICS Papers on China No. 2, 2016.
156
The share of China’s exports to advanced nations declined from 54.6 to 47.7 percent
between 2007 and 2018. Yearbook 2008, Table 17-8; Yearbook 2019, Table 11-5.
816

for foreign investment, Belt and Road projects spearhead its emergence as
a major source of outbound international investment.
These huge programs represent the leading edge of official economic
intervention, which has achieved a scale without historical precedent.
China’s government spending exceeds its US counterpart.157 Beijing’s control
over financial resources extends far beyond official budgets. China’s state-
dominated financial system remains responsive to official directives, as do
managers of China’s world-leading foreign-exchange reserves and the leaders
of nonfinancial state enterprises, whose combined assets eclipse those of the
500 largest US companies.158
This multiplex arsenal supports outlays of astonishing breadth and scale.
Some 90 percent of companies with A-shares listed on the Shanghai exchange
received government subsidies in 2016. The China Integrated Circuit Industry
Investment Fund, established in 2014, “invested in more than 70 projects and
companies” following initial fund-raising. Subsequent contributions lifted
funding to US$51 billion. China’s shipbuilding industry, which reported
2005 output of RMB 125.7 billion, received “policy support” valued at RMB
550 billion between 2006 and 2013.159
Government intervention extends beyond China’s national borders.
UNCTAD data show that China’s stock of outbound FDI, much of it in the
hands of state enterprises, now exceeds the stock of inward FDI. Overseas
lending, partly in support of Belt and Road projects, represents a further
extension of official activity: year-end 2018 debts of “73 of the world’s poorest
countries” held by the Chinese state and state-owned financial institutions
amounted to US$104 billion, matching the total ($106 billion) owed to the
World Bank.160
Chinese advances in multiple segments of technology-intensive activity –
Internet software, supercomputers, electric vehicles, high-speed rail, green
energy, high-voltage power transmission, artificial intelligence, and genetics,
157
See the online appendix referred to in note 44.
158
See the online appendix referred to in note 44.
159
D.H. Xu 徐东华 (ed.), 中国装备制造业发展报告 2017 (Report on the Development of
Equipment Manufacturing Industry in China 2017) (Beijing, Shehui kexue
wenxian chubanshe, 2017), p. 87; B. van Hezewijk, “Big Fund = Big Impact?
‘Winning the Future’ of the Semiconductor Industry,” August 24, 2019, at
www.linkedin.com/pulse/big-fund-impact-winning-future-semiconductor-industry-van
-hezewijk; TX Investment Consulting Co., Ltd., “全求船舶制造业特续景气,国内造
船企业加速整合” (Accelerate the Consolidation of Domestic Shipbuilding for the
Continued Prosperity of the Shipbuilding Industry) (February 28, 2007), 7; P.J. Barwick,
M. Kalouptsidi, and N.B. Zahur, “China’s Industrial Policy: An Empirical Evaluation,”
NBER Working Paper 26075, 2019, 2.
160
“The Debt Toll,” The Economist, July 4, 2020, 63.
817

among others – demonstrate the new strategy’s capacity to promote innov-

ation. At the same time, multiple constraints limit the effectiveness of the vast
resources deployed in pursuit of innovation.
Constraints: Ongoing, New, and Resurrected

China’s economic system channels vast resource flows into unproductive
activities. Top-down selection of priorities steers investment in directions
that often clash with domestic capabilities and with China’s international
comparative advantage. Politics pervades the allocation process, delivering
resources and opportunities into the wrong hands, while bypassing worth-
while industries, projects, and proprietors.
SOE priority status has survived decades of underperformance. From 1978
to 2007, the state sector “contributed essentially zero to aggregate growth in
total factor productivity.”161 Additional evidence confirms the deleterious
impact of state ownership on growth, profitability, and structural change.
Entry barriers and subsidies allow plodding, overstaffed state firms to remain
profitable;162 at the same time, soft budget constraints exempt long-time
money losers from financial discipline, dragging returns downward.163 The
growing complexity of SOE structures conceals payoffs to allies, wealth
extraction, and waste. Negative consequences of state ownership extend
beyond the SOEs themselves to encompass the sectors and regions they
inhabit: “in almost every dimension – the rate of start-up of new firms, size of
firms, TFP, and wages . . . new firms are weaker where the SOEs are more
dominant.”164
Announcement of official priorities sparks rampaging investment as offi-
cials, agencies, companies, and organizations pursue the anticipated cornu-
copia of financial and reputational bounty. In 2016, a “robot craze” prompted
local governments to announce 2020 output targets that amounted to
a considerable multiple of overall demand projections.165 Inflated R & D
161
Zhu, “Understanding China’s Growth,” 119.
162
Insiders at one of China’s largest energy firms regard two-thirds of the company’s
workforce as superfluous (personal communication).
163
N.R. Lardy, The State Strikes Back: The End of Economic Reform in China? (Washington,
DC, Peterson Institute for International Economics, 2019), pp. 52, 55, 89, shows
declining return on assets for state firms after 2007, with the share of loss makers
regularly exceeding 40 percent.
164
L. Brandt, G. Kambourov, and K. Storesletten, “Barriers to Entry and Regional
Economic Growth in China,” University of Toronto, Department of Economics,
Working Paper 652, January 5, 2020.
165
Wübbeke et al., “Made in China,” 25.
818

spending,166 low-quality patents,167 phantom companies,168 unaudited ven-

ture funds,169 and dubious projects burden Chinese industrial policy with
long tails of excess.
The ubiquity of procedures that allow “particularistic bargains” rather
than “universal rules” enables officials to distort seemingly market-based
transactions to benefit favored participants.170 Officials can readily manipu-
late government-managed auctions and supplier certification processes to
steer business opportunities toward preferred clients.171 In return for access to
urban real estate at discounted prices, companies associated with relatives of
top leaders accelerate the promotion of provincial officials.172 Similarly priv-
ileged “princelings” orchestrate lesser rivulets of efficiency-sapping resource
diversion in every locality and sector.
Xi Jinping’s emphasis on top-down strategizing and enthusiasm for
the “dominance” (zhuti diwei 主体地位) and “leading role” (zhudao diwei
主导地位) of public ownership and state-controlled enterprises enlarges
these costs. Casting state-owned enterprises as lead actors in national eco-
nomic strategy diminishes prospects for favorable outcomes. The growing
sway of official mandates over financial resources, investment opportunities,
and approval mechanisms stifles decentralized experimentation and limits
private-sector options.173 New constraints, beginning with the installation of
frontier innovation as the centerpiece of China’s policy agenda, expand the
burden of system costs.
166
“中国科研经费水分大:‘节省’经费发‘福利’ 经济参考报,” June 3, 2007, at techweb
.com.cn/news/2007-03-06/162748.shtml; Y.T. Sun and C. Cao, “China’s Research Is
Work in Progress,” China Daily, May 11, 2015.
167
A.G.Z. Hu, P. Zhang, and L.J. Zhao, “China as Number One? Evidence from China’s
Most Recent Patenting Surge,” Journal of Development Economics 124 (2017), 107–19;
P. Boeing and E. Mueller, “Measuring Patent Quality: Development and Validation
of ISR Indices,” China Economic Review 57 (2019), available at https://browzine.com/
articles/332678339.
168
R.C. Dai, X.Y. Liu, and X.B. Zhang, “Detecting Shell Companies in China,” presenta-
tion at ASSA annual meeting, January 4, 2020.
169
N. Xiang, “Rise of Trillion-RMB Government Funds Reshapes China’s Investment
Landscape,” January 13, 2017, at chinamoneynetwork.com/2017/01/13/rise-of-trillion-
rmb-government-funds-reshapes-chinas-investment-landscape, accessed September 11,
2017.
170
S.L. Shirk, The Political Logic of Economic Reform in China (Berkeley, University of
California Press, 1993), p. 336.
171
H.B. Cai, J.V. Henderson, and Q.H. Zhang, “China’s Land Market Auctions: Evidence
of Corruption?”, RAND Journal of Economics 44.3 (2013), 488–521.
172
Chen and Kung, “Busting the ‘Princelings’.” The authors note that recent anticorrup-
tion efforts appear to have reduced these discounts by 40–50 percent.
173
S. Heilmann, Red Swan: How Unorthodox Policy Making Facilitated China’s Rise
819

Current policy replacing market-propelled catch-up with officially man-

dated innovation targets adds both cost and risk. Investing in activities that
enjoy a comparative cost advantage is widely seen as a key contributor to
China’s recent boom. This has meant that Chinese firms, often working
within the anonymity of global supply chains, have pursued incremental
advances rather than “‘moonshot innovations’ – not for them ‘iPhone
envy’.”174 With “Made in China 2025” in the forefront, current policy stands
this approach on its head, focusing precisely on “moonshot innovations”
spanning a vast spectrum from large-scale passenger aircraft and space
exploration to genetics and nanotechnology.
Attempting frontier innovation in a middle-income economy with
a limited command of the human, industrial, and organizational resources
that underpin innovation systems in advanced nations multiplies the risks
associated with any such effort. Surveys of China’s engineering industries
highlight weaknesses in precision, durability, quality control, software devel-
opment, and commercialization of research results – all critical to innovative
success.175 Growing hostility to foreign involvement, especially in strategic
and advanced sectors, invites premature import substitution, further com-
pounding the dangers surrounding the main thrust of China’s current eco-
nomic agenda.
Structural change has added constraints in two areas: services and urban-
ization. The tertiary or service sector, now the largest contributor to both
output and employment, includes retail, hospitality, and other low-skill,
labor-intensive industries. The technology-intensive service segment
includes entrepreneurial and innovative operators such as Baidu, DRI, and
Huawei, along with state-owned financial and telecom giants whose main
asset is the official umbrella that protects them from competition.
Despite the achievements of a few globally competitive firms, weak
performance predominates. Exclusion of private operators limits competi-
tion and raises costs in air and rail transport, finance, insurance, and telecom-
munications, among others. The protectionist nature of China’s innovation
policy is evident in digital services, where China ranks as the global leader in
restricting cross-border trade.176
174
G.S. Yip and B. McKern, China’s Next Strategic Advantage: From Imitation to Innovation
(Cambridge, MA, MIT Press, 2016), pp. 82–3.
175
Annual issues of D.H. Xu 徐东华 (ed.), 中国装备制造业发展报告, address these
issues in considerable detail.
176
OECD Services Trade Restrictiveness Index: Policy Trends up to 2020 (Paris, OECD, 2020),
pp. 12–13.
820

Massive internal migration reflects both the attraction of vibrant urban

economies and the distortions associated with decades of policy discrimin-
ation against rural areas. National policy often appears to conflate cause and
effect, anticipating that enlarging city boundaries, reassigning farmland to
nonagricultural pursuits, and relocating villagers into high-density housing
clusters will somehow elevate productivity. Municipal governments, reflect-
ing concern over the cost of providing health and education benefits as well
as urban contempt for migrants’ low cultural level, hesitate to absorb these
newcomers, and sometimes seek to drive them away.
Revival of pre-reform obstacles to growth completes the roster of con-
straints that limit China’s growth prospects.
China’s current leader has resurrected the pre-reform personality cult. As
under Mao, many actions must once again await the leader’s personal
decision. Deng Xiaoping’s pragmatism fades as specialized bureaucracies
give way to party loyalists. China’s constitution now decrees that “east,
west, south, north, the party leads on everything.”177
These changes add fresh burdens to the economy. Party review of business
decisions in state and even private firms will complicate already labyrinthine
decision mechanisms. Growing pressure on private firms “to set up party
committees with an increasing say over strategy” steers activities in direc-
tions that deliver political rather than commercial returns. Not surprisingly,
available data show declining profitability for non-state industrial and service
firms.178 Educational quality must suffer as teachers shelter behind rote
learning and academics give way to “Xi Jinping thought.” As in the past,
increased emphasis on orthodoxy and suppression of dissent, the bedfellows
of politics in command, will attenuate the critical thinking essential to
innovation.
Strident emphasis on “autonomous” (zizhu 自主) innovation built upon
“independent Chinese intellectual property” illustrates how growing nation-
alist preoccupation has curtailed involvement with foreign firms, technolo-
gies, and components. Enhanced focus on security and on civil–military
177
N. Grünberg and K. Drinhausen, “The Party Leads on Everything,” Merics China
Monitor, September 24, 2019, 10.
178
“The New State Capitalism: Xi Jinping Is Trying to Remake the Chinese Economy,”
The Economist, August 15, 2020. NBS data show the return on assets for above-scale
private industry falling from 12 to 14 percent during 2010–2012 to just over 7 percent in
2018–2019. For services, see L. Brandt, “Policy Perspectives from the Bottom Up: What
Do Firm-Level Data Tell Us China Needs to Do?”, in R. Glick and M.M. Spiegel (eds.),
Policy Challenges in a Diverging Global Economy (San Francisco, Federal Reserve Bank of
San Francisco, 2015), p. 297.
821

integration sharpens this nationalist policy edge. With foreign businesses

complaining that “strong-arm tactics . . . marked difficulty in getting licenses”
and deportation of foreign managers make them “feel unwelcome in China,”
it is hardly surprising that the number of foreign-invested enterprises and
their share in both output and exports began to decline well in advance of the
abrupt deterioration of US–China relations in 2020.179 Rising barriers led the
European Commission to identify China as “the EU’s most restrictive trading
partner.”180
Trade disruptions involving rare earths, cars, beef, barley, medical sup-
plies, sports, and tourism, among others, have become a routine instrument
of China’s foreign policy, encouraging foreign partners to diversify away
from China. Domestic activities suffer as well: even in scientific fields,
researchers face restrictions on participation in international projects and
conferences. Foreign textbooks now arouse suspicion: in an apparent excep-
tion, business schools are “mostly spared from curbs on the use of imported
textbooks.”181
Strong conflict between the vast resources mobilized to support China’s
innovation ambitions and the daunting obstacles hindering China’s economic
progress invites a review of recent productivity trends, which combine
multiple factors into a single measure of economic advance.
Productivity
Ongoing decline in the size of the labor force and in the share of GDP going
to investment dictates the dependence of future growth on increases in TFP,
which measures the level of output per unit of combined inputs. Socialist
planning raised output amidst stagnant productivity. Reform abruptly
reversed this failure. Multiple studies track China’s transition to “intensive”
growth – with the majority of output expansion attributable to higher
productivity rather than increased quantities of labor and capital inputs –
for three decades from 1978.
179
R. Legaspi, “More U.S., Foreign Businesses Feel Unwelcome in China,” China Topix,
January 9, 2015 at chinatopix.com/articles/31659/20150109/more-us-foreign-
businesses-feel-unwelcome-in-china.htm, accessed July 25, 2020. Yearbook 2019,
Tables 13-3, 13-9, show sharp reduction in foreign-invested industrial firms along with
employment and share of overall industrial output after 2007. L. Brandt and K. Lim,
“Accounting for Export Growth in China,” MS, 2020, use China’s trade transactions
Customs data to show a decline in the share of exports by foreign firms.
180
“Report from the Commission to the Parliament and the Council on Trade and
Investment Barriers 1 January 2018–31 December 2018,” Brussels, n.d., 28.
181
“MBAs with Chinese Characteristics,” The Economist, February 15, 2020, 57.
822

Beginning in 2008, however, we see a return to “extensive” growth

powered by larger inputs. A succession of studies using national, provincial,
and enterprise-level data point to a marked decline in productivity growth
since the eve of the global financial crisis.182 The size of the private sector and
the scale of productivity deterioration suggest that declining performance
encompasses both private and state enterprise, with areas of stagnant or
declining productivity dwarfing pockets of dynamism.
China enters the reform era’s fifth decade with its economy far larger and
more sophisticated, its people more prosperous and better educated, its
command of modern technology far greater, and the expertise of its policy
makers far deeper than in 1978. Despite these astonishing advances, the
revival of plan-era policy approaches and political strategies now confronts
China’s economy with the same challenge it faced in the 1970s: how to
overcome self-imposed obstacles that prevent improvements in knowledge
and capabilities from generating intensive growth that outruns the accumu-
lation of resources.
Conclusion
China’s boom, a major event in global economic history, has transformed
a poor, backward, isolated economy into a prosperous and dynamic global
giant. This stunning departure is no miracle, but rather the consequence of
readily understandable changes in core elements of China’s economy. The
restoration of economic incentives, reflecting Deng Xiaoping’s call to “let
some people get rich first,” invited every individual, enterprise, and official to
pursue income-enhancing opportunities. Gradual opening of domestic and
international markets, along with partial relaxation of long-standing restric-
tions on entry, competition, and mobility, expanded the universe of available
choices.
Modest institutional opening prompted a rush to exploit the untapped
potential accumulated under socialist planning. Initial opportunities clustered
in the countryside, where thousands of enterprises and millions of villagers,
freed from the shackles of collective farming and enforced self-sufficiency,
182
D. Dollar, “China’s New Macroeconomic Normal,” unpublished, 2016; C.E. Bai and
Q. Zhang, “Is the People’s Republic of China’s Current Slowdown a Cyclical
Downturn or a Long-Term Trend? A Productivity-Based Analysis,” Manila, Asian
Development Bank Institute Working Paper No. 635, 2017; S.J. Wei, Z. Xie, and X.
B. Zhang, “From ‘Made in China’ to ‘Innovated in China’: Necessity, Prospect, and
Challenges,” Journal of Economic Perspectives 31.1 (2017), 549–70; Brandt and Lim,
“Accounting for Export Growth in China.”
823

streamed into long-forbidden markets and occupations. Decentralized move-

ment of labor, materials, and capital toward financially rewarding activities
brought massive change: hundreds of millions left farming, millions of new
firms emerged, and vast resources poured into China’s coastal provinces.
Long before the recent boom, Qing-era Chinese society harbored elements
favorable to economic growth. Wide dispersion of entrepreneurship, commer-
cial acumen and sophistication, universal regard for education, informal contract
enforcement mechanisms, and competent local administration all contributed to
the initial reform response and its subsequent extension. These growth-
enhancing features supported Qing-era prosperity and commercialization, but,
enmeshed in tightly interlinked economic, political, and social institutions, lacked
the capacity to generate an economy-wide response to the appearance of new
markets and new technologies in the nineteenth and early twentieth centuries.
During the twentieth century, growing state strength and the gradual
buildup of physical and human capital eroded long-standing obstacles to
growth. Wartime disruption and then the deficiencies of the PRC plan system
delayed the realization of these gains. Beginning in the late 1970s, the combin-
ation of reforms that broke both old and new barriers to growth and Deng
Xiaoping’s effort to harmonize the incentives of government and citizens
unleashed a boom that revealed the full power of China’s economic system.
Remarkable economic gains have not eliminated the tension between the
demands of political stability and economic development that pervades
China’s governance arrangements. Systematic misallocation via networking
cements elite loyalty and promotes critical support for regime survival, but
the long-term economic cost is staggering. The rent seeking that honey-
combs policy implementation propels high levels of income inequality and
causes massive waste183 – as when large shares of funds awarded for con-
structing public projects vanish into private pockets before work
commences.184
183
Analyses of contemporary inequality find that the top 1 percent of households receive
roughly 15 percent of overall income. See https://wid.world/country/china, focused
on 2005–2015; and T. Piketty, L. Yang, and G. Zucman, “Income Inequality Is Growing
Fast in China and Making It Look More Like the US,” at https://blogs.lse.ac.uk/
businessreview/2019/04/01/income-inequality-is-growing-fast-in-china-and-making-it
-look-more-like-the-us. These estimates resemble those for the late Qing: C.L. Chang,
The Income of the Chinese Gentry (Seattle, University of Washington Press, 1962), pp. 327–
8, finds that gentry families comprised 2 percent of China’s population and received
24 percent of overall income during the 1880s.
184
Participants indicate that skimming may absorb 30 percent of costs for airports or
stadiums and mention higher figures for road building (personal communication).
824

Long before China’s post-1978 growth explosion, Qing territorial expan-

sion, suppression of mid-nineteenth-century rebellions, and the PRC’s recov-
ery from both self-inflicted and external shocks demonstrated the durability
and resilience of Chinese authoritarian systems. The most dynamic episodes
of change and growth, however, cluster around interludes of state weakness,
when ruptures in the carapace of restrictions surrounding elite interests
enable China’s populace to deploy its remarkable commercial talents.
Shanghai’s pre-1937 development into Asia’s premier financial complex, as
well as a commercial hub and manufacturing center, illustrates this potential.
Several decades later, post-Cultural Revolution erosion of central authority
enabled nationwide rural reforms. The astonishing boom that followed
demonstrated the capacity of unheralded “peasants” to lift China’s vast
countryside onto an elevated growth trajectory that liberated hundreds of
millions from absolute poverty even as crumbling commune finances
reduced funding for social welfare. The subsequent surge in private entre-
preneurship extended “development from below” into the urban economy,
where private firms garnered large shares of output and employment wher-
ever they managed to gain a foothold.
The 1990s spawned a unique concatenation of expanded market opening
with massive growth and centralization of state-controlled fiscal and financial
resources. SOE reforms decanted tens of thousands of enterprises and tens of
millions of workers into the grip of market discipline, while sweeping reduc-
tions in barriers to international trade and investment intensified domestic
competition, elevated quality standards and forced widespread reductions in
profit margins. While the multiplication of state-controlled resources stabil-
ized a regime shaken by Tiananmen, the economic benefits of market
opening extended robust productivity growth until the 2008 global financial
crash.
Long-standing tension between market- and state-led economic strategies
resurfaced following China’s 2001 WTO entry. Unlike the 1990s, there is little
sign of mutually acceptable initiatives. The market economy vision, most
clearly articulated in the 2013 document China 2030, anticipates a retreat of the
state, and especially of state-owned enterprises, from the “commanding
heights” of an open economy led by private business – changes that would
sharply reduce the resources available to state and Party leaders.
Aside from a brief flurry in 2013, when a Central Committee decision
endorsed the notion of building an economy in which “market forces
dominate,” the rival vision of state economic leadership has captured the
imagination of China’s ruling elites. Support for state direction over market
825

dominance came from many sources. The economic success of Japan,

Taiwan, and South Korea has built a global constituency promoting
government entrepreneurship as the wellspring of technological develop-
ment. Many Chinese viewed the absence of globally prominent Chinese
firms, brands, and technologies as signaling the failure of openness to end
China’s economic subordination to former colonial powers. Concern about
China’s need to develop its own military technology bolstered nationalist
objections to economic opening. Unavoidable reliance on state interven-
tion to alleviate the 2008 financial crisis reinforced this view of market
frailty and deepened support for increased government management of the
economy.
The administration of Xi Jinping has moved decisively toward state
control. Core elements extend practices familiar from seventy years of
Chinese economic planning. Policy directives, notably Made in China
2025, set overall strategy and lay out investment priorities. State-owned
enterprises take the lead in implementing top-down initiatives. The cur-
rent policy constellation incorporates new dimensions and revives former
practices.
Reflecting China’s recent economic advance, the current array of strategic
industries and technologies includes many new entrants. Recent plans for
both well-established and novel sectors revolve around bold plans to reach
and then extend global technological frontiers.
China’s effort to redirect development from widespread, decentralized
incremental efforts that add value through improvements in cost, quality,
and design to more concentrated pursuit of targeted innovations in a narrow
range of products and technologies faces formidable challenges. Extending
technological frontiers is always a high-risk proposition. Launching
a “breakthrough” strategy from a middle-income platform beset by weak-
nesses in key domestic supply chains and limited downstream demand adds
fresh layers of risk.
Assigning vast resources to a talented and highly motivated corps of
domestic researchers will surely deliver successes – already visible in State
Grid’s technical advances in high-voltage electricity transmission and in the
commercial achievements of firms like Alibaba, Pinduoduo, and Tencent.185
When measured against the enormity of the world’s largest economy,
however, even considerable numbers of isolated breakthroughs may fail to
185
Y.C. Xu, “The Search for High Power in China: State Grid Corporation of China,” in
L. Brandt and T.G. Rawski (eds.), Policy, Regulation and Innovation in China’s Electricity
and Telecom Industries (Cambridge, Cambridge University Press, 2019), pp. 221–61.
826

deliver economy-wide productivity increases, leading to a Soviet-style out-

come in which the occasional sputnik illuminates galaxies of mediocrity.
Looking beyond efforts to scale the heights of advanced technology, the
absence of major reforms during the two decades following China’s 2001
entry into the WTO has burdened the economy with an immense backlog of
costs. Excess capacity in steel, electricity, and many other industries; state-
sector firms often bulging with surplus employees; and zombie companies
held together with patchworks of subsidies, loans, and tax concessions
exemplify the distortions that permeate every corner of China’s vast eco-
nomic landscape. Past outcomes invite expectations that strengthening Party
control and promoting self-reliance will accelerate the pace of cost accretion.
The decade following the global financial crisis has seen a return to the
plan-era pattern in which growth arises almost entirely from the accumula-
tion of labor and capital. Mounting signs of a steep fall-off in productivity
growth warn that the current state-led economic strategy may prematurely
terminate China’s remarkable growth explosion.
Some will see this skepticism as “misleadingly wrong” and “encouraging
a complacent and dangerous underestimate of China’s potential
trajectory.”186 China’s growth potential is indeed large. With its remarkable
human resources, competent public administration, and per capita income
roughly one-fourth the US level, China faces an unmistakable opportunity to
navigate a lengthy runway of intensive growth.
For the moment, however, China’s leaders have turned away from
openness and competition, the conventional tools for traversing the path
from middling to high levels of productivity and income. China’s current
policy constellation ignores abundant evidence, much of it from China
itself, highlighting the benefit of shifting from plan to market, redistribut-
ing resources from state to private firms, and allowing increased access to
foreign firms, imported products, and external technologies. Unless
China’s leaders once again demonstrate that they are “imaginative and
flexible” and can “shift policy decisively, comprehensively, and without
regard to procedural or legal niceties,”187 disappointment seems more
likely than triumph.
Whatever the outcome and whatever its future course, China will con-
tinue to grapple with dilemmas that have bedeviled two centuries of mod-
ernization efforts. How can China embed a creative, freewheeling culture of
186
“The New State Capitalism.”
187
T. Orlik, China: The Bubble That Never Pops (New York, Oxford University Press, 2020),
pp. 198–9.
827

economic and technical innovation within an authoritarian system whose

leaders feel threatened by unorthodox thinking? How can China resolve the
concern arising from fears that indiscriminate opening to Western technol-
ogy and ideas endangers the edifice that Confucian and Communist thinkers
have long seen as the foundation of authoritarian rule and social stability?
Further Reading
Breznitz, D., and M. Murphree, Run of the Red Queen: Government, Innovation, Globalization
and Economic Growth in China (New Haven, Yale University Press, 2011).
Elvin, M., and G.W. Skinner (eds.), The Chinese City between Two Worlds (Stanford,
Stanford University Press, 1974).
Huang, Y.S., Capitalism with Chinese Characteristics (Cambridge, Cambridge University
Press, 2008).
Mülhahn, K., Making China Modern: From the Great Qing to Xi Jinping (Cambridge, MA,
Rozelle, S., and N. Hell, Invisible China: How the Urban–Rural Divide Threatens China’s Rise
(Chicago, The University of Chicago Press, 2020).
Schell, O., and J. Delury, Wealth and Power: China’s Long March to the Twenty-First Century
(New York, Random House, 2013).
Skinner, G.W. (ed.), The City in Late Imperial China (Stanford, Stanford University Press,
1977).
So, B.K.L., and R.H. Myers (eds.), The Treaty Port Economy in Modern China: Empirical
Studies of Institutional Change and Economic Performance (Berkeley, University of
California Institute of East Asian Studies, 2011).
Walder, A.G., China under Mao: A Revolution Derailed (Cambridge, MA, Harvard University
Press, 2015).
Walter, C., and F. Howie, Red Capitalism: The Fragile Financial Foundation of China’s
Extraordinary Rise (Hoboken, John Wiley, 2011).
Whiting, S.H., Power and Wealth in Rural China (Cambridge, Cambridge University Press,
2009).
Zelin, M., J.K. Ocko, and R. Gardella (eds.), Contract and Property in Early Modern China
828

Index
Page numbers in bold refer to content in tables; page numbers in italics refer to content in
figures and maps.
absolutism, 45 authoritarian control, 775, 776, 779, 781

accounting systems, 202 authoritarian hierarchy, 777, 779
Africa, 711 authoritarian politics, 49, 274, 789
Aglen, Francis, 263 authoritarian rule, 43, 228, 264, 785, 789, 828
agriculture, 87–8. See also Great Famine authoritarian system, 44, 532, 825, 828.
(1958–1961) See also regionally decentralized
collectivization, 566, 567, 574–7, 583–4, authoritarianism (RDA)
586–8, 681–2, 798–9 aviation, 487–8
farm size and productivities, 103
freight transportation, 475–6 Bai, Ying, 393, 519, 681–2
land institutions, distribution, and credit, Balazs, Étienne, 5
115–20, 118, 121 Bandung Conference (1955), 708–11
long-term output trends, 88–99, 89, 91, 92, Bank of China (BoC), 220–2, 223, 226–8, 272,
93, 94 292, 297–8, 312
models of, 99–107, 105, 106 Bank of Communications (BoCom), 215,
new elements (Qing era to 1949), 107–15, 220–1, 222, 223, 226–8, 297–8, 312
110, 116 Bank of the Great Qing, 215
pre-1979 living standards, 599–601, 600, banks, 281, 282, 283, 283, 284–6, 285. See also
633–4, 634 native banks (cash shops)
Qing-era land reclamation, 51–4 apprenticeships, 504
Reform Era, 725–6, 740–4, 745–6, 803 banker networks, 317–19, 318
rural finance (1930s–1940s), 298–307, 301, commerce and industry finance, 311–12
302, 303, 305, 306 competition with the nationalist state,
sideline production, 104–7, 106 226–8, 296–8
Almond, Douglas, 679 co-operation with the nationalist state,
American Silver Purchase Act (1934), 238, 298 228–32
Anhua, Hunan, 69, 70 depositing money, 235, 320, 568
Anhui province, 629, 631, 803 evolution of the money system, 216–20,
annual budget appropriations, 190 217, 218
anti-Japanese base areas, 159, 163 geography of, 287–9, 289, 290, 292–4,
Anti-Rightist campaign, 550–1, 580 358, 378
apprenticeship, 503–4, 513 government bonds, 277
arsenal industry, 127, 171–3 issuing of banknotes, 44, 210, 215, 219,
Asian Financial Crisis (1997), 767 220–4, 221, 226–7
Australia, 716, 717 Reform Era, 766, 770, 805, 808
829

Index
banks (cont.) Bolshevism, 541–3, 546–7

relation with university development, bondholders, 41, 230
396, 397 bonds, 44, 229–32, 255, 263, 264, 272
rise of Chinese institutions, 40–1 stock markets, 313–16, 314, 315
social status of, 317–19, 318 wartime, 277
state’s takeover of, 222–4, 568 Boserupian framework, 100–3, 101
statistical data, 21–2, 22, 24 bottom-up development, 46, 240, 741, 776, 809
treaty ports system, 373–8, 374, 375, 377, Boxer Rebellion (1899–1901), 130, 244, 255,
379, 380 259–60
Bao Shichen 包世臣, 59, 75, 84 Boyd Company, 128
Baodi, Hebei, 145 Brandt, Loren, 119, 680
Baoshan Steel, 731, 740 Bretton Woods Agreement (1944), 160, 688
barefoot doctors, 626 Britain
Bastid, Marianne, 251 business organization, 345
Baxian, Sichuan, 336 consumption, 622
Beijing, 41, 43, 47, 85, 86, 88, 91, 96, 110, 122, 166, industrialization, 16, 354
182, 183, 207, 224, 226, 227, 232, 233, 234, infrastructure, 467
239, 242, 243, 248, 251, 252, 253, 254, 255, investment, 439, 440
256, 258, 260, 261, 262, 263, 264, 265, 266, loans, 275
270, 271, 272, 273, 275, 277, 280, 281, 284, trade, 20, 72–3, 422–5, 435, 701
286, 292, 294, 295, 297, 298, 299, 307, 309, treaty ports system, 360–1, 367
312, 313, 316, 317, 320, 322, 323, 328, 333, 336, British American Tobacco Company, 112,
341, 356, 361, 371, 401, 426, 460, 468, 469, 131, 477
471, 489, 490, 492, 495, 500, 501, 504, 506, British East India Company, 423
509, 513, 517, 524, 526, 528, 537, 544, 557, Broadberry, Stephen, 622
564, 580, 581, 585, 611, 614, 617, 620, 624, Brown, Shannon, 779
628, 629, 630, 630, 631, 632, 634, 636, 637, Buck, John, 88, 102, 120, 615
638, 641, 658, 665, 667, 675, 689, 691, 692, Burgess, John, 514
693, 701, 702, 707, 708, 709, 710, 711, 712, bus transportation, 486
715, 716, 721, 726, 727, 729, 730, 732, 734, business organization, 324–5
735, 737, 738, 740, 741, 743, 751, 759, 760, early twentieth-century developments,
762, 763, 774, 776, 783, 784, 794, 799, 800, 348–51
803, 807, 808, 809, 814, 815, 817 family firms, 325–32
banks, 292–4 private ordering, 332–42
currency use, 225, 226–7 Western influence, 342–8
education, 514, 518
family firms, 327 Cai Yuanpei 蔡元培, 525
infrastructure, 180, 476, 485, 487 California School, 27, 28, 87, 100
living standards, 64, 95, 96, 614 Campbell, Cameron, 65
Olympic Games (2008), 814 Canada, 716, 717
Tangshan earthquake (1976), 595 Canton System, 422, 427
warlord era finance, 261–6, 262 Cao Shuji 曹树基, 648, 650
Beijing Stock Exchange, 313 Cao Yu 曹禺, 317
Beijing–Hankou railroad, 474, 476, 481, 483 capital flow, 235, 236, 439, 440
Beiyang Xuetang, 130 capital markets, 448, 450
Belt and Road initiative, 719, 816, 817 capitalism, 7, 27–8
Bergère, Marie-Claire, 144 censorship, 778
Berlin, Isaiah, 639 central account offices, 350–1
Bernstein, Thomas P., 668 Central Air Transport Corporation
bimetallism, 208–9 (CATC), 487
birth control, 65, 727 Central Bank (PBoC), 220, 222, 223–4, 231, 272,
birth rates, 63, 65 297, 568, 569, 578, 580, 762, 805
body shares, 337–8 Central China Cocoon and Silk Company, 158
830

Index
Central China Promotion Company, 158 China National Aviation Corporation

central control, 32, 258, 259, 274, 581, 784, 807–15 (CNAC), 487
centralization of state-controlled fiscal and Chinese Economic Yearbook, 306
financial resources, 825 Chinese Monetary History, 8
centralized authoritarian control, 781 Chinese Revolutionary Alliance, 406
central-level enterprise groups, 784 Chinese Textile Construction Company, 163
centrally supervised enterprise groups, 814 Chong, Ja Ian, 255, 266
centrally planned command system. See Chongqing, 159, 189, 487
planned (command) economy Christianity, 391
ceramics, 132 Chuan, Han-sheng 全汉升, 67, 68
Ch’en Huan-chang 陈焕章, 1 Chuxing Company, 351
Ch’ien, Tuan-Sheng, 514 civil service exams, 28, 405, 500, 501, 502, 519, 777
Chairman Mao. See Mao Zedong 毛泽东 Cixi, Empress Dowager of Qing, 38, 171
Chan, Wellington, 344, 348 class struggle, 585–6, 592
Chang Chung-li 张仲礼, 499, 507 climate change, 50
Chang, Gene Hsin, 675 Cohen, Myron, 333
Chang Jung, 648 Cohen, Paul, 26–7, 34
Chang Yu-fa 张玉法, 357, 367–9, 385 Cold War, 692, 697–702
Changchun, Jilin, 294 collectivization, 657–60, 725, 741–4, 798
Changshu, Jiangsu, 146 Comintern, 541, 543, 544
Changzhou, Jiangsu, 59 commerce. See also business organization;
Chao Kang 赵冈, 92, 100, 101, 119 international trade
Chartered Bank, 40 apprenticeships, 504
Chefoo Convention (1876), 258 domestic trading, 66–9, 370–3, 372
Chen Duxiu 陈独秀, 544 economic effects from new technology,
Chen Hansheng, 5 444–7, 445, 446
Chen Hongmou 陈宏谋, 52, 54 financial capital, 307–12
Chen Jitang 陈济棠, 193 Maoist era, 568, 797–8
Chen Qiyuan 陈启源, 134, 149, 341 money market integration, 232–4
Chen Shuo, 669 commercial law, 37
Chen Yi 陈毅, 696 Common Program (1949), 548
Chen Yizi, 648 communal dining, 675–7
Chen Yonggui 陈永贵, 734 communes, 553–4, 583, 645, 657–9, 675–7, 724
Chen Yun 陈云, 565, 571–2, 574, 581, 586, 732–3, communication infrastructure, 457–64
739–40, 745, 748, 749–50, 755, 756, 759 Communist institutional origins, 531–3
Chen Yuyu, 678 analyzing institutional evolution, 533–4
Cheng Linsun, 286 creation of a totalitarian regime, 548–51
Chenggong, Yunnan, 521 empire comparisons, 534–40, 535, 538, 539
Cheong Weng Eang, 341 origins of totalitarianism, 541–3
Cheung Sui Wai, 68 regionally decentralized authoritarianism
Chiang Ching-kuo 蒋经国, 545 (RDA), 556–63
Chiang Kai-shek 蒋介石, 188, 196, 267, 272, regionally decentralized totalitarianism
274, 463, 466, 485, 545, 715, 788. See also (RDT), 551–6
Jiang Jieshi totalitarianism’s introduction to China,
China Dream, 816 543–8
China Economic Statistics Institute, 203 company founders, 148–52
China Health and Nutrition Survey company law, 181, 182, 200, 345, 347–8, 469, 811
(CHNS), 621 competition, 29, 740, 779, 786, 788, 812, 816,
China Integrated Circuit Industry Investment 820, 825, 827
Fund (CICIF), 817 conflict between political and economic
China Merchants Steam Navigation objectives, 782
Company (CMSNC), 169, 174–6, 307, 308, Confucian Classics, 28, 42, 43, 170, 359, 499,
344, 465, 492, 737 501, 510
831

Index
Confucianism, 2, 29, 34, 42, 169, 390, decentralization, 39, 537, 555, 581, 584, 588, 754.
496, 510 See also regionally decentralized
consolidated tax, 274 authoritarianism (RDA)
Constitutionalists, 405–11, 407, 408 decentralization programs (1957 and 1970), 809
constraints, 776, 794, 818–22 decentralized development, 782
consumption decolonization, 708
communal dining, 645, 675–7 defense industries, 127–8, 154, 155, 171–3, 186–7
Maoist era trends, 599–601, 601, 604, 604, deflation, 73
615–24, 617, 621, 623, 631–2, 632 demographics. See population changes
Qing era, 62, 64–5, 94–6 Deng Xiaoping 邓小平, 548, 562, 586, 587, 595,
Republican era, 109 596, 647, 729, 731, 733, 735, 737, 739, 745,
contracts, business, 334–6 747, 748, 755, 756, 758–60, 783, 808
copper, 74–5, 208, 211–13 Deng Zihui 邓子恢, 657
debasement of, 224–6 depositing money, 234–5, 320, 341, 568
corruption, 778 Dernberger, Robert, 355–6
cotton developmental state
modern industries, 23, 23, 111, 131–2, 141–6, central state enterprises, 188–91
143, 155 Guangdong enterprises, 192–6
postwar era, 163 Guizhou enterprises, 198–202
Qing-era industry, 70–1, 73, 104, 113, 426 ideology and policy of, 184–8
wartime production, 159 public versus private enterprise, 202–6
Council for Mutual Economic Assistance Shanxi enterprises, 196–8
(COMECON), 692 Dexin, 483
counterfeiting, 212, 214, 225 Dikötter, Frank, 628, 648
Crop Reports, 615 Dimitrov, Georgi, 545
crop yields. See agriculture Ding Richang 丁日昌, 171
cross investment, 340 Ding-wu Disaster, 98
Cuba, 488 Discovering History in China, 26
Cultural Revolution (CR), 532, 554–6, 557, Doctrine of Nationalism, 187
592–3, 594–5, 728, 732, 751 Doctrine of People’s Livelihood, 187
currency, 208–11 domestic public debt, 41, 44
Daoguang depression, 74, 74–7 Dongting lake, 52
government–private sector relationship, double-cropping, 60, 102
224–32 Dream of the Red Chamber, 316
major currencies overview, 211–16 droughts, 51, 98, 623
Maoist era, 566 Du Xuncheng 杜询诚, 21, 292, 373, 374
money market integration, 232–5, 238 dual-track system, 753–4, 804
Nanjing era, 44 Duan Qirui 段祺瑞, 265, 266
postwar era, 162 Dundas, Henry, 423
pre-1950 period, 24, 25 Dutch East India Company, 423
role of banks, 219–20, 221
role of the state, 220–4 East India Company, 72, 423
tax collection, 250 Eastern Bloc nations, 558–9, 560,
692, 695–6
Da Long, 130 Eckstein, Alexander, 693, 711
Da Sheng, 149, 150 ecological decay, 50–7
Dai Yi 戴逸, 48 Economic Growth in Prewar China, 164
Dalian, Liaoning, 234 Economic Principles of Confucius and His School,
Dane, Richard Morris, 262 The (1911), 1
Daoguang depression, 73–7 education, 496–7, 777. See also overseas study
Daoguang, emperor of Qing 道光皇帝, 49, Cultural Revolution era, 594–5
51, 54, 56, 73, 85, 97, 426 modern universities, 394–9, 395, 396, 397,
Dazhai, Shanxi, 726, 734 400, 404
832

Index
new system, post-1905, 509–18, 510, 511, 516, First Five-Year Plan (1953–1957), 489, 490, 567,
517, 518 579, 692, 694, 794
role of missionaries, 392, 392–4, 393 First National Finance Conference
socioeconomic changes, 519–26, 520, 523 (1928), 270
traditional system, pre-1905, 359, 497–508, First World War (1914–1918), 144, 311
499, 501, 506, 508 Five Small Industries (FSI) campaign, 555
egg products, 138 floods, 50–1, 52, 54, 98, 623
Eighth Party Congress (1956–1958), 571, 580, Foochow Naval Shipyard, 172
581, 584, 670 Food and Agriculture Organization (FAO),
Eighth Route Army, 570 615, 616, 725
electricity supply, 358, 382, 385, 386, 489 Food and Money Semi-Monthly (Shihuo
Elvin, Mark, 9, 87, 100 banyuekan 食貨半月刊), 4
employment levels, 588–9, 590, 601, 601, 746, Foot, Rosemary, 697
767, 801 foreign-aid strategy, 711–12
energy industry, 191 foreign debt (FD), 356, 365–70, 367, 368
Eng, Robert, 339 foreign direct investment (FDI), 22, 356–7, 370,
engineers, 396–9, 397, 400, 520, 570 414–17, 439, 439–42, 440, 442, 790
England. See Britain impact of treaty ports, 365–70, 367, 368, 369, 371
entrepreneurs, 127, 132, 148–52, 350–1, 810 railroad lines, 472
Essay on the Principle of Population, 58 Reform Era, 738, 740, 811–12, 817
Europe foreign firms, 373–6, 374, 375, 379, 440–2
living standards comparison, 62–3 railroad construction, 466–7
market prices, 67 Reform Era, 813
trade, 70, 425–6 relation with university development,
European Union (EU), 822 396–9, 397
exchange rates, 74–5, 160–1, 250, 352, 761 foreign influence, 442–8, 445, 446, 449, 450
exchange reserves, 814 foreign trade. See international trade
exploitation, 119–20 Four Cardinal Principles, 562
export-oriented industrialization, 132–40 Fourth Five-Year Plan (1971–1975), 579, 591,
594, 596
Fairbank, John, 26, 419, 777 France, 701
family firms, 325–32, 348–9 freight transportation, 473–80, 491, 491–2
family management, 150 Friedman, Milton, 751
famine, 51, 485, 554, 587. See also Great Famine frontier innovation, 819–20
(1958–1961) Fu Yiling 傅衣凌, 7
Fan Xudong, 342, 350 Fujian province, 631, 739, 740
Fang Xing 方行, 62–3, 64 Fukang (native bank), 310
farm size, 88, 102–3, 103 Furong, Sichuan, 83
Farmers Bank of China, 299 Fuzhou, Fujian, 172
Faure, David, 181, 343
Fei Xiaotong 费孝通, 5, 136 Gamble, Sidney, 514
feudalism, 3, 6, 7 Gang of Four, 596
Feuerwerker, Albert, 8, 9, 461 Gansu corridor, 287
fiat money system, 44, 238 Gansu province, 289
financial institutions, 280–7, 282, 283, 285. See Gao Gang 高岗, 548
also banks; native financial institutions Gao, Mobo, 617
business and commerce, 307–12, 308, 310 Gao Wangling 高王凌, 49, 61
geography of, 287–94, 288, 289, 290, 291, 293 Gaojia yan dike, Qingjiang, 56
rural finance, 298–307, 301, 302, 303, 305, 306 Gaoyang, Hebei, 145, 151
social status of, 316–21, 318 Gaozong, emperor of Qing, 78. See also
state’s role, 294–8 Qianlong Emperor
stock markets, 312–16, 314, 315 gap between actual and potential output,
financial trade associations, 319 797–801
833

Index
gap between urban and rural living Halsey, Stephen, 254, 492
standards, 796 Han dynasty, 417
Ge Zhaoguang 葛兆光, 30, 38 Han river, 52, 53
gender inequality, 521–3, 523 handicrafts, 124–7, 126
gentry degrees, 82 and agriculture, 104–7
Germany, 187, 467, 524 Communist era, 568
Gesellschaft (impersonal social relations), 6 new export-oriented industries, 138–40
Global Financial Crisis (2008), 561, 769, 771, 815 socialization of, 163
Gooch, Elizabeth, 682 textiles industry, 131, 133, 138, 139, 140, 144–6
grain procurement, excessive, 664–74, 672 Hangzhou Electric Light Company, 346–7
grain tribute, 55–6 Hangzhou, Zhejiang, 483
Grand Canal, 55–6, 246, 248 Hankou, Hubei, 69, 70, 217, 218, 309, 444
Great Depression, 111, 136, 238, 267, 271 Hanyang Iron Works, 179
Great Divergence debate, 10, 87 Hanyeping Company, 178–80, 307
Great Famine (1958–1961), 642–7 Hao Yen-ping, 343
communal dining’s impact, 675–7 Hart, Robert, 252, 254, 259, 459
excess death estimates and salient features, He Changling 贺长龄, 56
647–50, 651 He Liping 贺力平, 76
excessive grain procurement, 664–74, 672 He Ziquan 何兹全, 4
long-term effects, 677–82 health care, 626, 769
production failures, 655–64, 663 Henan province, 485, 583, 614
Great Leap Forward (GLF), 532, 552–4, 580–5, Hengjialou dike, Henan, 55
585, 643–7, 656, 712–16, 728 Hengzhou, Hunan, 65
Great Northern Telegraph Company, 460 Heshen 和珅 (Qing official), 49, 50
gross domestic product (GDP), 18–20, 19, 23, high-level equilibrium trap, 100, 106
238, 239, 245, 446, 531, 723–9, 724 highlands settlement, 52–4
Maoist era, 596, 597, 597–600, 598, 607–8, 655 Hino Kaisaburō, 4
Reform Era, 559, 561, 760, 766, 807, 812 Ho, Chun-Yu, 228
treaty ports, 358, 388, 389 Ho, Hon-wai, 251
Gu Mu 谷牧, 730 Ho, Ping-ti 何炳棣, 9, 58, 97
Gu Yanwu 顾炎武, 30 hoarding, 234
guandu shangban enterprises, 344–5 Hong Kong, 286, 294, 345, 422, 434, 582, 699,
Guangdong Enterprise Corporation (GDEC), 737–8, 811, 814
193–6 Hong Kong and Shanghai Banking
Guangdong province, 38, 72, 136, 138, 192–6, Corporation (HSBC), 40, 296
212, 614, 628, 738, 740 Hong Liangji 洪亮吉, 58, 84
Guangxi province, 320 Hong Ren’gan 洪仁玕, 780
Guangxu, emperor of Qing, 38, 171, 364, 788 honorary titles, 295
Guangzhou (Canton), 57, 70, 192, 316, 360, 422, Hou Yangfang 候杨方, 99
514, 523 household contracting, 743
guilds, 66, 342 household division. See family firms
Guiyang, Guizhou, 202 household subsidiary activities, 104–7
Guizhou Enterprise Corporation (GZEC), Housheng Cotton Mill, 311
198–202 Hu Jintao 胡锦涛, 785
Guizhou province, 198–202, 631, 743 Hu Qiaomu 胡乔木, 734
Guo Family Shop, 326–7 Hu Shi 胡适, 789
Guo Le, 149 Hu Yaobang 胡耀邦, 562, 756–7
Guo Moruo 郭沫若, 3 Hu Yuanhe lineage trust, 331
Guo Songtao 郭嵩焘, 41, 173 Hua Guofeng 华国锋, 596, 729, 741
Guo Songyi 郭松义, 90 Huaibei district, 81
Huainan district, 81
Hai river basin, 50–1 Huang Jingbin 黄敬斌, 63, 64
Halliday, Jon, 648 Huang, Philip, 87, 100, 102, 105
834

Index
Huang Zongxi 黄宗羲, 30 steamships and railroads, 464–71, 468

Huang Zongxian, 134 innovation, 101, 439, 772, 776, 786, 818, 819–20,
Huang Zunxian 黄遵宪, 38 821, 826
Hubei Coal Mining Company, 176 insurance companies, 292, 293
Hubei province, 52, 71, 80, 629, 631 intellectual elites, 523–6
Hubei Textile Bureau, 351 interest rates, 119–20, 121, 210, 236–8, 237, 448
Huguang Loan (1911), 261, 264 rural finance, 299, 302, 305
Huizhou, Guangdong, 93, 98 International Development of China, 185
hukou system, 588 international relations (Maoist era), 685–8
Hunan province, 38 Cold War geopolitical and economic blocs,
Hundred Days Reform (1898), 38 697–702, 699, 700
Hundred Flowers movement, 580 impact of domestic campaigns, 712–16
Hupei Cotton Cloth Mill, 178, 179 legacy of imperialism, 708–12
hyperinflation, 45, 566 legacy of Japanese empire, 702–8
Soviet Union and the planned economy,
ideological change (1850–1950), 15–18 688–97
age of culture, 41–3 International Settlement, Shanghai, 32
age of institutions, 38–41 international trade, 414–17, 434, 687, 687
age of machines, 34–7 commodity-level trade, 448–55, 451, 454
economic statistics, 18–26, 19, 21, 22, 23, 25 country composition, 433–5, 435, 436
end of the financial revolution, 44–5 export-oriented production, 132–40
Qing political regime, 28–31 integrated statistics, 442–4, 443
Western impact and Chinese response Maoist era, 596, 686, 691–6, 695, 697–702,
framework, 26–8, 32–4 699, 700, 704–5, 706, 709–10, 713, 714,
Imperial Academy, 500 715–16, 717
imperial examination system, 536, 777 Maritime Customs records, 129, 429–33, 431
imperial land system, 535 opium, 75–7
Imperial Post Office, 459 postwar era, 160–2
Imperial Telegraph Administration, 176, 461–3 pre-1839 era, 417–28
import-substitution industries, 140–8, 141, 143 pre-1950 period, 20–1, 21, 35–7
incentive problems, 116, 117, 558, 583 Qing maritime trade, 71–3
incentive-compatible (IC) transformation, Qing trade networks, 57
534, 546–7 Reform Era, 730–2, 737, 811
incentives, 231, 234, 533–4, 555, 560, 752–3, 796–7 steamships, 464–5
income inequality, 627–31, 628, 630, 634–9, 637 tea export, 69–70
income trends (Communist era), 602–4, volume of, 436–8, 438
608–14, 611, 612, 614 interregional trade. See commerce
India, 439, 517, 624 involution, 99–107
industrialization. See handicrafts; modern Irish potatoes, 109
industries iron industry, 582, 656
inefficiency, 598, 727, 785, 796, 800, 801 Iwai Shigeki, 279
inflation, 61, 63, 73, 162, 277, 590, 591, 754, 755, Iyigun, Murat, 31
756, 757, 761–2, 766
infrastructure, 152, 182, 357, 457–8 Jamieson, George, 248, 340
and agriculture, 112, 475–6 Japan
freight transportation, 473–80, 491, 491–2 consumption, 622
interruptions to railroad development, cotton industry, 142–3
471–3 education, 515, 517
motor roads and aviation, 484–8 gross domestic product (GDP), 19, 19
passenger transportation, 480–4, 481 industrialization, 126, 131–2, 146–7,
postal and telegraph communication, 707–8, 730
459–64 industry in occupied areas, 156–9, 704
postwar rail expansion, 488–92, 491 infrastructure, 463, 464, 471, 479–80, 493
835

Index
Japan (cont.) Kaiping Coal Mines, 176–7

investment, 439 Kaixian Gong Village, 106
legacy of empire, 702–8 Kang Youwei 康有为, 38, 42
Meiji influence on China, 38–9 Kangxi, emperor of Qing 康熙皇帝, 51, 69, 71,
overseas study, 401, 404–11, 408, 524 73, 391, 420
scholarship, 2, 4, 6 Kasahara, Hiroyuki, 667
silk production, 114, 116, 136–8 Kaske, Elisabeth, 82, 470
technical expertise, 696, 705 Katō Shigeshi, 4
trade, 72, 434, 702, 704–5, 715–16 Keller, Wolfgang, 448
treaty ports system, 367 Keynesian economics, 6
Western impact on, 31 Khrushchev, Nikita, 713–15
Jardine, Matheson & Co., 134, 142, 426, Kiakhta, Russia, 69
441, 466 Kincheng (Jincheng) Banking Corporation, 311
Jian family, 349 Kirby, William, 695, 793
Jian Yujie, 149 Kitamura Hironao, 67
Jian Zhaonan, 149 Koito Seisakujo, 157
Jiang Jieshi 蒋介石, 43. See also Chiang Kai- Köll, Elisabeth, 180
shek Kong Xiangxi 孔祥熙, 267
Jiang Qin, 503 Korean War (1950–1953), 694, 698
Jiang Qing 江青, 557 Kraus, Richard, 67, 68
Jiang Rongzhuang, 338 Kueh, Yak Yeow, 660–1
Jiang Zemin 江泽民, 758, 762, 810, 813 Kuhn, Philip A., 49
Jianghan Plain, 54 Kung, James Kai-sing, 393, 503, 650, 659, 665,
Jiangnan Arsenal, 127, 172 669–70, 676, 681–2
Jiangnan region, 57, 59 Kuran, Timur, 31
commerce and industry, 68, 136, 144 Kuznets, Simon, 20, 390
consumption, 62
living standards, 64 labor allocation, 573–4
population change, 63 laissez-faire principle, 170
Jiangsu province, 56, 272, 317 land acreage, 97–8
infrastructure, 468 land distribution, 119–20, 566, 602
population change, 59 land institutions, 115–19, 118
taxation, 264 land tax, 246, 248, 261, 267–70, 273
textiles industry, 115, 135, 144 Lardy, Nicholas, 796
Jiaqing, emperor of Qing 嘉庆皇帝, 49–50, Late Qing Reform (New Policies, 1903–1911),
54, 55, 78, 79, 82 112, 130, 244, 256
Jiaxing, Zhejiang, 60 latent capacity, 800
Jicheng bao, 346 latent potential
Jilin City, Jilin, 57 in agriculture, 798–9
Jilin province, 631 in industry, 800–1
Jin Guantao 金观涛, 29 in international and domestic exchange,
Jinan Puyi sugar factory, 114 797–8
Jingdezhen, Jiangxi, 66 League of Nations, 448
Jingzhou, Hubei, 54 Lee, James, 60–1, 65
Jin-Pu Railway, 180 Legalism, 170
Jiuda Salt Refinery, 350 Lenin, 541, 543, 544, 546
Joffe, Adolf, 544 Leninist party, 540, 541, 558
joint household economies. See family firms; Levine, Ross, 232
private ordering Lewis, W. Arthur, 355
joint-stock companies, 175, 344, 345, 690 Li Bingjing, 667
Joint-Stock Companies Act (1856), 345 Li Bozhong 李伯重, 59–60, 68
Ju Qingyuan 鞠清远, 4 Li Chihua, 679
junxian system, 535, 537, 538 Li, Dan, 228
836

Index
Li Dazhao 李大钊, 544 warlord era (1912–1927), 261–3, 266, 472

Li Fusun, 230, 232 wartime, 275
Li Hongzhang 李鸿章, 35, 38, 141, 171, 174–5, Lü Xuehai, 138
180, 307, 343, 466 Lumey, L.H., 679
Li, Lillian, 51 Lushan Conference (1959), 668, 676
Li lineage, 333 luxury goods, 57, 73, 426
Li Peng 李鹏, 757, 762
Li Quanshi 李权时, 268 Ma, Debin 马德斌, 64, 137
Li Siyou lineage trust, 332 Ma Jianzhong 马建忠, 173
Li Xiannian 李先念, 738, 741, 748, 799 Ma Jinhua, 266
Li Yingzhou, 338 Ma Junya, 144
Li Yuanhong 黎元洪, 351 Macartney, George, 1st Earl Macartney, 15,
Liang, Ernest, 473 423–4
Liang Fangzhong 梁方仲, 5 Macau, 422, 737
Liang Qichao 梁启超, 33–4, 38, 41–2, 43, machinery, adoption of, 385–8, 388
350, 789 Mackay Treaty (1902), 260, 264
Liang-Huai salt district, 81, 84 macroeconomics. See money markets
Liao dynasty, 51 Maddison, Angus, 607
Liao Zhongkai 廖仲恺, 545 Made in China 2025, 719, 780, 816, 820
Liaoning province, 65, 614, 635 Mah, Feng-hwa, 694
life expectancy. See mortality rates maize, 108
lijin duties, 248–51, 251, 252, 258–60, 264, 265, Makino Fumio, 125
267, 270, 478 Malthus, Thomas, 58, 60
Lin, Justin Yifu, 658, 665, 676 Malthusian framework, 100–3, 101
Lin Man-houng, 77 management styles, 148–52
Lin Zexu 林则徐, 427 Manchuria, 23, 127, 156–7, 160, 471, 479–80, 488,
lineage trusts, 328–32, 349 615, 704–5, 791
Lingnan region, 50 Manifesto of the First Nationalist Party
Linxiang, Hunan, 69 Congress, 185
literacy rates, 505 Mann Jones, Susan, 49
Liu family, 333 Mao Zedong 毛泽东, 27, 45, 545, 550
Liu Hongsheng 刘鸿生, 149, 351 agricultural decisions, 566, 577, 728
Liu Mingchuan 刘铭传, 180 Communist institutional origins, 551–5
Liu Qingfeng 刘少奇, 29 death of, 595
Liu Shaoqi, 545, 586, 587, 664 Great Famine (1958–1961), 668
Liu Ta-Chung, 608, 615 Great Leap Forward, 643–5, 713–14
Liu Yan 刘晏, 168 Soviet relations, 688–92
Liu Yuan, 676 transition to the Great Leap Forward,
living standards (Maoist era), 606–7 580–4
agricultural output, 633–4, 634 Maritime Customs Service (MCS), 33, 40,
food consumption trends, 615–24, 617, 621, 252–4, 253, 271, 273
623, 632 government bonds, 227, 229, 231, 263, 264
income inequality, 627–31, 628, 630, postal system, 459–60
634–9, 637 tax collection, 36, 71–3, 248, 258–9
income trends, 608–14, 611, 612, 614 trade records, 20, 128, 415, 429–33, 431, 442–4,
mortality record, 624–6, 625 451–2, 454
per capita GDP trends, 607–8, 723–5 maritime trade, 71–3, 85
rural poverty, 631–4 market
living standards (Qing era), 62–6 forces, 569, 572–3, 578, 580, 747, 767, 783,
loans, 237. See also rural credit 807–15, 825
Maoist era, 690, 694 integration, 232–40, 236, 237
Nanjing Decade, 272 opening, 810–12, 825
Qing era, 259, 261, 296 outcomes, 812–15
837

Index
market’s role in allocating resources, 781, money market, 208–11

801, 812 centrally planned command system, 577–80
Marks, Robert, 50 government–private sector relationship,
marriage, 66 224–32
Marxism, 3, 6–7, 27, 45, 544 integration of, 232–40, 236, 237, 239
Marxism–Leninism, 541, 544, 552 major currencies overview, 211–16
match production, 147 role of financial institutions, 216–20, 217,
Mawei Shipbuilding Yard, 128 218, 221
May Fourth Movement (1919), 266 role of the state, 220–4, 294–8
Mayers, William, 788 moonshot innovations, 820
Meiji era, Japan, 16, 17, 31, 38–9, 42, 130, 167, Morck, Randal, 338
405, 493 Mori Tokihiko, 143
Meng family, 327, 348 mortality rates, 63, 624–6, 625
Meng Zhang, 57, 337 Great Famine (1958–1961), 647–50, 651
Meng Xin, 665, 678 Motono, Eiichi, 36–7
Mentougou coal mines, 336 motor vehicles, 486
merchants Mukden Incident (1931), 471
as business managers, 344–5 munitions industry, 154, 379, 382, 383
business organization, 334, 337–8, 340, 342–3 Murphey, Rhodes, 25
currency use, 214 Muslim Rebellion (1862–1873), 98
industrial entrepreneurship, 148–9 Myers, Ramon, 9, 82
pre-1839 trade, 420, 422
state–enterprises relationship, 168, 170, Naitō Konan, 2, 4
173–81, 784 Nakai Hideki, 142
steamship companies, 465 Nakamura Jihei, 67
metalworks, 178, 582, 656 Nan Hanchen, 701
Mexico, 215 Nanhai, Guangdong, 134, 341
migration, 108, 573, 589, 821 Nanjing University, 112, 120
Mikoyan, Anastas, 688, 690 Nanjing, Jiangsu, 43, 66, 114, 371
military institutions, 248 Nanyang Brothers Tobacco Company, 113,
Minami Ryoshin, 125 131, 149, 349
Ming dynasty, 7, 69, 208, 244, 419, 421 Nanyang Gongxue, 130
Minimum List of Commodities (MLC), 448 Nathan, Andrew, 789
mining industry, 23, 311, 336–7, 635, 770 National Agricultural Research Bureau
Minong, Taiwan, 333 (NARB), 615
minting bureaus, 211 National Capitalism and the Old Chinese
missionaries, 359, 390–4, 391, 392, 393, 402, 514 Government, 373
Miyazaki Ichisada, 4 National Commercial Bank, 312
modern industries, 124–7, 126 National Defense Planning Commission, 203
company founders and management, National Federation of Education, 521
148–52 National Resources Commission, 153, 154,
education’s role, 519–21, 520 189–91, 203, 205
export-oriented production, 132–40, 135, 139 native banks (cash shops), 214, 235, 281, 282,
financial capital, 307–12, 308, 310 283, 284–6, 285, 291, 291, 295, 297, 298
import substitution, 140–8, 141, 143 commerce and industry finance, 307–12,
industrialization from abroad, 127–8, 129 308, 310
Maoist era, 568–80, 571, 584, 800–1 rural finance, 299, 300
postwar reorganization, 160–4, 161, social status of, 317
165 Native Customs, 248, 253, 259, 273
role of treaty ports, 357–8 native financial institutions, 213, 216, 217. See
technology transfer, 128–32, 146 also native banks (cash shops)
wartime, 152–60, 154, 155, 165 native-place affiliations, 340–1
Mokyr, Joel, 31, 390 natural disasters, 98–9, 99, 576, 595, 660–4, 663
838

Index
navigation transport, 174–6 People’s Bank of China. See Central Bank

Needham, Joseph, 390 (PBoC)
network interests, 778–9, 802, 824 People’s Commune, 553–4, 583, 645, 657–9,
networks, of bankers, 317–19, 318 675–7, 724
New Culture Movement, 525 Perkins, Dwight, 9, 59, 68, 90, 96, 232, 505, 795
New Democracy, 691 personal networks, 778–9
New World crops, 53, 93, 108–9 petroleum industry, 191
newspapers, 344, 346 Piazza, Alan, 607, 616
Ni Yuping, 85 Ping County coal mine, 307
Niida Noboru, 6 planned (command) economy, 45, 567–80, 749
Ningbo, Zhejiang, 294, 317, 364, 446, 522 productivity stagnation, 796–7
Ningjin, Hebei, 75 Platt, Stephen, 780
Nippon Railway, 493 Plekhanov, Georgi, 543
Nippon Steel, 731 Pomeranz, Kenneth, 16, 62–3, 70, 83
Nishijima Sadao, 7 population changes
North China Development Company, 157 diffusion of New World crops, 108
Northern Expedition (1927), 788 Maoist era, 588–9, 590, 726–7
Northwestern Industrial Corporation (NIC), Qing era, 57–66, 98–9, 99
196–8 Reform Era, 801
porcelain, 66
oil production, 596, 728 postal system, 459–60
on-the-job training (OJT), 131 postwar industrial reorganization, 160–4, 161
one-child policy, 727 Potter, Sulamith, 617
openness, 17, 455, 691, 739, 779–81, 788, 789, 813 Pottinger, Henry, 428
opium, 73, 75–7, 111, 249, 250, 259, 423, 426–7 poverty, 631–4
Opium War (1839–1842), 15, 31, 35, 48, 73, 124, price increases. See inflation
213, 247, 342, 360, 391, 414, 427 primary school enrollment, 794
Opium War (1856–1860), 254, 361, 428 primary sector. See agriculture
Oriental culture, 2, 3 Prince Gong (Yixin), 171
overseas study, 400–11, 402, 403, prison reform, 791
408, 524–5 private ordering, 332–42
private plots, 567, 574, 584, 587, 617–18, 619,
Panzhihua, Sichuan, 636, 638 726, 797
paper money, 210, 215–16, 219, 226–7 privately owned firms, 155, 373, 493. See also
role of the state, 220–4 foreign firms
Paris Peace Conference (1919), 265 privatization, 559, 562, 603, 750, 767, 809,
partnerships, 150, 151. See also family firms 810–11
private ordering, 332–42 producer co-operatives, 566, 567, 574–6, 583,
Party Work Conference (1978), 732–4 657, 724
passive investors, 340 productivity growth, 233, 598–9, 801–2, 806,
patronage, 36, 59, 458, 778 813, 815, 822–3, 825
pawnshops, 280, 281, 282, 283, 284, 285, 295 fall in, 658, 785, 795, 796–7, 823, 827
rural finance, 299, 300, 306 property rights, 83, 115, 117, 535, 554, 559–60
social status of, 316, 320 provincial finances, 245, 251–5
Pearl river delta, 737 Qing centralization, 255–61, 256, 257
Peking University, 525 public finance, 244–6
Peng Dehuai 彭德怀, 584 Nanjing Decade, 267–72, 269
Peng Nansheng 彭南生, 146 Qing centralization, 255–61, 256, 257
Peng Shihong, 197 Qing fiscal governance, 77–82, 80, 246–55,
Peng Xinwei 彭信威, 8 251, 253
Peng Yuxin 彭雨新, 69 warlord era (1912–1927), 261–6, 262
Peng Zeyi 彭泽益, 66–7, 76 wartime, 272–7, 276
pengmin (shed people), 53 public-health campaigns, 626
839

Index
Puqi, Hubei, 69 remittance houses, 281, 282, 283, 283–6, 285,

purchase shares, 300 287, 288, 295–6, 297
Putterman, Louis, 659 Reorganization Loan (1913), 261, 264, 266
Report on Conditions of Chinese Industry, 203–4
Qi Xin Cement Co., 181 responsibility system, 586, 660, 744, 745
Qian Changzhao, 187, 189, 191 Restoration Society, 406
Qian, Nancy, 665, 678 revenue figures, 262, 760
Qianlong, emperor of Qing 乾隆皇帝, 15, Nanjing Decade, 269
48–9, 77, 79, 82, 421, 423–4. See also Qing era, 78–82, 80, 256, 257
Gaozong, emperor of Qing wartime, 276
Qianmen, Beijing, 327 Revive China Society, 406
Qimen, Anhui, 333 revolutionists, 406–11, 407, 408
Qin Shihuang 秦始皇, emperor of Qin, 45, 55 rice prices, 63, 67
Qing political regime, overview, 28–31, 48–50 rice trade, 52, 67–9, 72
Qingdao–Ji’nan railroad, 474 Rishengchang, 338
Qiyang, Hunan, 67 road development, 484–7
qiying baotai formula, 48 Roll, Charles, 796
Qu Fulu, 338 Rong family, 149, 311, 321, 349
Quan Hansheng 全汉升, 4, 5, 178 rotating credit societies, 280, 281, 282, 283, 285
Rubin, Jared, 31
railroad army corps, 489 Ruifuxiang, 327, 348
Railroad Bureau, 472–3 rural construction, 583, 586
railroad development, 180, 260, 466–71, 468 rural co-operative treasuries, 287, 292, 293,
freight transportation, 473–80 299–300, 301, 302, 303, 306
interruptions to, 471–3 rural credit, 119–20, 121, 298–307, 301, 302, 303,
passenger transportation, 480–4, 481 305, 306
postwar expansion, 488–92 rural reforms, 740–4, 746, 771
Rawski, Thomas, 20, 22, 164, 233, 465, 473, 475, Russell & Co., 134
499, 505, 795 Russia, 69, 439, 471. See also Soviet Union
Readjustment Program (1979–1981), 733, (USSR)
735–6, 745 origins of totalitarianism, 541–3
Red Flag, 553 totalitarian influence on China, 544–6
Red Guards, 592
Reform Era, 722–3, 801–2 S.C. Farnham and Co., 128
baseline, 723–9, 724 salt industry, 80–1, 337
beginning of economic reform, 729–36 family firms, 328, 329–30, 331–2
breakthroughs of the 1990s, 758–68, 764, Sichuan province, 66, 83–4
768, 807–15 taxation, 247, 249, 250, 260, 262–3, 264,
creativity of the 1980s, 748–58, 756, 802–7 271, 273
culmination and post-transition, 768–71 Sands, Barbara, 119
opening up foreign exchange, 736–40 Santiaoshi district, Tianjin, 130
post-Global Financial Crisis (2008), 815–23 seagoing junk business, 170
renewed reform and dynamics, 744–8 seasonal labor, 104–5, 105
rural reform, 740–4 Second National Finance Conference
regional competitions, 559–61 (1934), 270
regional currencies, 209, 213 self-initiated ports (SIPs), 357, 358, 362, 363–5,
regional merchant groupings, 340 365, 366, 371
regionally decentralized authoritarianism long-term effects of, 388, 389
(RDA), 532, 556–63 modernization, 376–8, 388, 388
regionally decentralized totalitarianism overseas study, 403
(RDT), 532–3, 551–6 universities, 394–9, 396, 397, 400
Reinhardt, Anne, 465, 492 self-reliance, 707, 711, 718, 781, 797
Remer, C.F., 439 Self-Strengthening Movement, 35, 38, 167
840

Index
arsenal industry, 171–3 Sichuan province, 470, 803

business organization, 345 ecological changes, 52
modern industries, 343–4, 357, 379 financial institutions, 288, 300
officials and business enterprise, 173–81 Great Famine (1958–1961), 650
origins of, 577–8 living standards, 631, 636
semi-industrial firms, 151–2, 163, 164 modern industries, 153, 593
sericulture, 114–15, 116, 135, 136 salt industry, 66, 250
sex ratios, 65 silk industry, 135, 158
Shaanxi province, 52, 75 silk industry, 111, 131–2
Shandong province, 468, 485 business organization, 339
Shang Yue 尚鉞, 7 export-oriented, 133–8, 135, 139
Shanghai freight transportation, 475
banks, 40, 217–19, 218, 223, 227, 229, 237, 272, sericultural improvements, 114–15, 116
292–4, 296, 311 wartime production, 158–9
cotton industry, 23, 142, 143, 145, 159 Silk Road trade, 417
education system, 518 silver, 208–10, 213–15
infrastructure, 461, 466, 483, 487 Daoguang depression, 74, 74–7
Reform Era, 739–40, 760 money market integration, 232–5, 238
silk industry, 133–4, 136 role of banks, 219–20
trade, 70, 436–8, 438, 442–4, 443, 446, 731 role of the state, 220
treaty ports system, 371, 786 Sinking Fund Commission, 44, 229–32
Western impact and investment, 27, 32, Sino-American Trade Commercial Treaty
363, 441 (1946), 162
Shanghai Commercial and Savings Bank, 217, Sino-Foreign Salt Inspectorate, 262, 271, 273
311, 320 Sino-Japanese War (1894–1895), 31, 33, 37, 38,
Shanghai–Hangzhou railroad, 484 130, 179, 244, 255, 259, 401, 463, 467, 703
Shanghai Merchant Steamship Company, 150 Sino-Japanese War (1937–1945), 185, 189, 197,
Shanghai–Nanjing railroad, 475, 481 198, 205
Shanghai Native Bankers’ Association, 40 modern industries, 152, 154, 155
Shanghai Stock Exchange, 313, 315, 817 public finance, 272–7, 276
Shanghai xinbao, 346 Skinner, G. William, 9, 71
Shangyu, Zhejiang, 326 small-scale firms, 151–2, 164, 571, 767
Shanxi province, 196–8, 284, 288, 339, 521 Smithian growth, 102
Shaoxing, Zhejiang, 317 smuggling, 80, 430, 478
sharecropping, 116 soap production, 147–8
shareholder partnerships, 332–42 Social History Debate, 3
shares market, 346–7 Soda Saburo, 135
shed people (pengmin), 53 SOEs. See state-owned enterprises (SOEs)
Shekou, Guangdong, 739 Sogabe Shizuo, 4
Shen Baozhen 沈葆桢, 172 soil fertility, 53
Shen Xin company, 149, 150 Sommer, Matthew, 66
Shenbao, 346, 347 Song dynasty, 215, 328, 342
Sheng Xuanhuai 盛宣怀, 130, 174, 177, 179–80, Song family, 43
296, 461 Song Ziwen 宋子文, 44, 267, 271, 272
Shenyang, Liaoning, 234 Songjiang–Taicang region, 145
Shenyi tang, 331 South China Sea, 418, 420
Shenzhen, Guangdong, 737, 738, 747, 762, 786 South Manchurian Railway, 156, 287, 471, 479–80
Shi Zhihong 史志宏, 90, 93, 97 Southeast Asian trade, 72
Shiba Yoshinobu, 9 Southern Huai salt industry, 250
shipbuilding, 127–8, 493, 817 Southern Tour (1992), 758, 808
Shiue, Carol H., 448 Soviet Growth Model (1949–1978)
Shunde, Guangdong, 135 centrally planned command economy,
Siam, 72 567–9
841

Index
Soviet Growth Model (1949–1978) (cont.) state-sector downsizing, 809, 814

command system in 1956–1957 period, Stauffer, Milton, 391
569–80, 571 steam engines, 385–8, 387, 388
early years, 565–7 steamships, 464–6
economic recovery 1961–1964, 585–91, steel industry, 156, 582, 656
590, 591 stock markets, 312–16, 314, 315, 762
evaluation of economic perfomance, 597, Study of China’s Ancient Society (1930), 3
597–605, 598, 600, 601, 604 Sudō Yoshiyuki, 6
organization and policies 1965–1978, 591–7 sugar, 62, 114
transition to the Great Leap Forward, Sugihara, Kaoru, 132
580–5, 585 Suiping, Henan, 553
Soviet Union (USSR), 187, 479, 593, 688–97, Sun Lanzhi, 76
713–15 Sun Pinghua, 704
Cold War geopolitical and economic blocs, Sun Yat-sen 孙逸仙, 185, 188, 261, 263, 266,
697–702 406, 493, 543
loans and aid, 275, 712 Sun Yefang 孙冶方, 586
railway development, 488 Sun Yingde, 341
technical expertise, 490 Sun–Joffe Manifesto, 544
trade, 695 Sunrise (Cao Yu), 317
Spanish pesos, 209, 214 surveillance, 777
special economic zones (SEZs), 738–9, 745, Suzhou, Jiangsu, 466
747, 759, 804 Suzuki Chusei, 53
Spence, Jonathan, 390 sweet potatoes, 108, 618
sprouts of capitalism, 7, 27 symbolic capital, 170
Stalin, 545, 549, 691–2 System Reform Institute, 753
Standard International Trade Classification
(SITC), 448 Taiping Rebellion (1851–1864), 33, 35, 60, 63, 85,
state capitalism, 153, 815–23 98, 124, 127, 171, 224, 244, 343, 429, 466, 786
state incapacity, 759, 776 Taiwan, 339, 811
State Planning Commission, 577–8, 582, 713 Taiyuan, Shanxi, 197
State Statistical Bureau (SSB), 608, 609, 618, Takasaki Tatsunosuke, 704
619, 634 Tanaka Issei, 63
state trading companies, 578 Tanaka Masatoshi, 7
state weakness, 776, 779, 825 Tangshan earthquake (1976), 176, 595
state’s non-economic objectives, 802 Tang–Song transition, 2, 4, 6, 417
state–enterprises relationship, 167, 344. See also Tao Xisheng 陶希圣, 4
planned (command) economy Tao Zhu, 56, 81
arsenal industry, 171–3 Tawney, R.H., 25
developing central state enterprises, 188–91 tax collectors, 250, 254
Guangdong enterprises, 192–6 taxation. See also lijin duties
Guizhou enterprises, 198–202 Nanjing Decade, 267–72
ideology of the developmental state, 184–8 Qing era, 79–80, 244, 247, 248–51, 251, 252–5,
officials and business enterprise, 173–81, 784 253, 258–60, 429–30
origin of self-strengthening, 577–8 Reform Era, 765, 808, 812
public versus private enterprise, 202–6 remittance houses, 295–6
Reform Era, 809, 818 silver payments, 213, 214
Shanxi enterprises, 196–8 trade, 36–7, 248–51, 251, 429–30
state-led development, 152, 813 warlord era (1912–1927), 261, 262–6
State-owned Assets Supervisory Commission wartime, 272–7
(SASAC), 562, 784 tea, 62, 69–70, 72, 110, 426
state-owned enterprises (SOEs), 66, 127, 152, teacher-training schools, 518, 518, 522
184–6, 189, 555–6, 740, 747, 753, 767, 784, technical education, 128, 130–1, 132, 173, 359
809, 826 technological changes
842

Index
economic effects from, 444–7, 445, 446 postwar rail expansion, 488–92, 491
pre-1950 period, 34–7 steamships and railroads, 464–71, 468
technology transfer, 128–32, 146, 730–2 Treaty of Nanking (1842), 21, 252, 360, 371, 379,
telegraph communication, 460–4 391, 427–8, 464
Telford, Ted, 65 Treaty of Shimonoseki (1895), 37, 142, 259, 362,
Temple Destruction Movement, 512 364, 376, 441, 703
tenancy system, 116–17, 118 Treaty of Tientsin (1858), 391, 428, 464
Teng, Ssu-yu, 26, 419 treaty ports system, 15, 40, 44, 354, 356, 358–9
textiles. See cotton industry; silk industry early industrialization, 124, 128
TFP (total factor productivity), 598, 795, 799, foreign investment and debt, 365–70, 367,
807, 822 368, 369, 370, 371, 439, 439–42, 440, 442
Thaxton, Ralph A., 619 geography of, 365, 369
Third Five-Year Plan (1966–1970), 591, 594, 596 hinterland economic zones, 427–8
Third Front, 593–4, 637, 800 human capital development, 390–9, 391, 392,
Third National Finance Conference (1941), 273 393, 395, 396, 397, 400
Third Plenum (1978), 732, 735–6 long-term effects of, 388, 389
Third Plenum (1993), 763–6, 764 Maritime Customs records, 429–33, 431
Three Doctrines of the People, 185, 186, market integration, 370–3, 372
192, 511 modernization, 373–88, 374, 375, 377, 378, 379,
Three Gorges, 98 380, 382, 383, 385, 387, 388
Tiananmen Square Protests (1989), 562, 757 origins of, 360–3, 361, 362
Tianjin, 130, 147, 217, 218, 294, 317, 444, 461, overseas study, 402–4, 403
467, 476 spillover benefits, 447–8
Tianjin Arsenal, 128 taxation, 36, 252
Tianjin Soap Company, 147 trade, 414, 427
Tianjin Technology and Business Triads, 543
University, 520 tributary system, 29, 246, 418–20
Tianjin–Pukou railroad, 467–70, 474, 478, 483 Tsinghua University, 524
timber industry, 337 Twitchett, Denis, 9
Ting Zhili, 515
tobacco, 112–13, 274, 477 underutilized resources, 796
Toffler, Alvin, 751 United States of America (USA)
Tokugawa era, Japan, 31, 42, 72, 505 Cold War trade embargo, 697–702
Tong Kingsing 唐景星, 174, 177 gross domestic product (GDP), 239, 246
Tongcheng, Anhui, 65 infrastructure, 488
Tongmenghui (Alliance Society), 526 loans, 275
Tongrentang, 327, 335 overseas study, 524
Tongzhi Restoration (1861–1875), 35 trade, 136, 161–2, 435
Tongzhi, emperor of Qing, 171 Vietnam War, 593
total factor productivity. See TFP (total factor universities, 394–9, 395, 396, 397, 400, 404, 524, 594
productivity)
Township and Village Enterprises (TVEs), van de Ven, Hans, 252, 259
746, 747, 753, 783, 804 Vietnam, 593
Toyota company, 157 Viraphol, Sarasin, 72
transfer of technology. See technology Vogel, Hans Ulrich, 75
transfer Voitinsky, Grigori, 544
translated works, 38 von Glahn, Richard, 61, 68, 74, 77, 82
transportation infrastructure, 457–8
freight transportation, 473–80, 491, 491–2 wage levels
interruptions to railroad development, Maoist era, 573–4, 590–1, 591, 600, 602,
471–3 614, 636
motor roads and aviation, 484–8 Qing era, 61, 64, 95, 96, 105
passenger transportation, 480–4, 481 Wan Li 万里, 733–4, 741, 743
843

Index
Wancheng dike, Hubei, 54 World Trade Organization (WTO), 767, 776,

Wang Fansen 王汎森, 17, 30 785, 811, 814
Wang Jiaxiang 王稼祥, 545 Wright, Mary, 35
Wang lineage trust, 329–30 Wright, Tim, 144
Wang Luman, 296 Wu Chengming 吴承明, 10, 69, 73, 109, 275
Wang Qisheng 王奇生, 524 Wu Dingchang, 198–9, 200
Wang Renzhong 王任重, 741 Wu Hui 吴彗, 90, 92, 94
Wang Yangming 王阳明, 43 Wu Songdi, 233
Wang Yaosheng 王耀生, 502 Wu Xiaozhen, 65
Wang Yeh-chien 王业健, 63, 73, 75, 79, 82, 245 Wuchang Uprising (1911), 406
Wang, Yejian 王业健, 237 Wuchang, Hubei, 63
Wang Zhiyi, 54 Wuhan, Hubei, 614, 635
Wangquantang, 327 Wuxi, Jiangsu, 114, 115, 135, 137
Warnings to a Prosperous Age, 130
Washington Naval Conference (1921), 265 Xi Jinping 习近平, 780, 785, 816, 819
water control, 54–6 Xi Zhongxun 习仲勋, 548, 670
Weber, Max, 3, 393 Xianfeng, emperor of Qing 咸丰皇帝, 224
Wei Xian, Shandong, 145 Xiang river, 52
Wei Yuan 魏源, 56, 79, 170, 779 Xiangyin county, Hunan, 54
Wei Xiahai, 676 Xianning, Hubei, 71
Weifang, Shandong, 152 Xiaogangcun, Anhui, 741
well-field system, 169 Xiaoshan, Zhejiang, 63
Wen Jiabao 温家宝, 769 Xicun, Guangdong, 192
Wen lineage, 331 Xie Nanming, 149
Wen Xiang 文祥, 171 Xin Deyong, 98
Wen, Guanzhong James, 675, 676 Xinhai Revolution (1911), 359, 400, 401, 406
Weng Wenhao 翁文灏, 186, 187, 188, 189, 191 Xinjiang region, 49, 77, 289
Wen-hsin Yeh, 317 Xiuning, Anhui, 333
Western culture, 2, 41–3 Xu Dixin 许涤新, 109
Western impact, 354–60, 779–81 Xu Rongting, 351
business organization, 342–8 Xu Run, 307
firms and banks data, 21–2, 22 Xu Xinwu 徐新吾, 145
foreign investment and debt, 365–70, 367, Xue Muqiao 薛暮桥, 751
368, 369, 370, 371
human capital development, 390–9, 391, 392, Yan Xishan 阎锡山, 196–7, 288
393, 395, 396, 397, 400 Yan Yutang, 149
impact–response framework, 26–8, 32–4, 40 Yan Hongzhong, 236
long-term effects of treaty ports, 388, 389 Yan’an Rectification Campaign (1942), 545, 550
overseas study and political change, 400–11, Yan’an, Shaanxi, 565
402, 403, 408 Yang, Dennis Tao, 796
self-initiated ports (SIPs), 363–5, 365, 366 Yang Fan, 338
treaty ports and market integration, Yang Jisheng 杨继绳, 648
370–3, 372 Yang Lien-sheng, 8
treaty ports and modernization, 373–88, 374, Yang Nianqun 杨念群, 34
375, 377, 378, 379, 380, 382, 383, 385, 387, 388 Yang Xifu, 52
treaty ports system, 36–7, 360–3, 361, 362 Yanglinzhai, Hunan, 54
Western scholarship, 8–11 Yangzhou, Jiangsu, 81, 84
White Lotus Rebellion (1796–1804), 54, Yangzi river, 52, 362, 464–5, 467
79, 246 Yao Yilin 姚依林, 757
Will, Pierre-Étienne, 52 Yared, Pierre, 665
Wittfogel, Karl, 3 Yeh, Kung-chia, 608, 615
working capital, 338–40 Yellow River, 55, 246, 247
World Bank, 609, 624, 635, 751, 770, 781 Yinghe (Qing official), 56
844

Index
Yong An company, 150 Zhao Erxun, 114

Yongtai Company, 114, 137 Zhao Ziyang 赵紫阳, 562, 741, 743, 748–9,
Yongzheng, emperor of Qing 雍正皇帝, 51, 82 751–2, 755–7, 781
Yongzhou, Hunan, 67 zhaoshang system, 168, 170
Yoon, Wook, 462 Zhejiang province, 60, 82, 115, 272, 317
Yu Jieqiong, 74 Zheng empire, 419
Yu Zhiqing, 149 Zheng Guanying 郑观应, 130, 141, 149, 173, 461
Yuan Geng 袁庚, 738 Zhili Gongyiju, 131, 148
Yuan Shikai 袁世凯, 130, 173, 181, 226, Zhili province, 50, 130, 174, 468, 485, 502
261–4, 266 Zhou Enlai 周恩来, 552, 595, 707, 709–12,
Yuchtman, Noam, 520 715, 728
Yue Fengyi, 327 Zhou Heng Shun factory, 130
Yue Xianyang, 327 Zhou Xuexi 周学熙, 149, 181, 185
Yuezhou, Hunan, 364 Zhou Yumin, 252
Yugan, Jiangxi, 521 Zhou Ziqi, 264
Yunnan province, 153, 629 Zhou Hao, 796
Zhou Li-An, 678
Zeng Guofan 曾国藩, 35, 38, 43, 171 Zhou, Titi, 650, 670
Zhang Fangzuo, 131 Zhu Jiaye, 521
Zhang Jian 张謇, 132, 149, 178–9, 185, 342, 467 Zhu Rongji 朱镕基, 762, 765–8, 776, 810
Zhang Naiqi, 316 Zhu Xianfang, 131
Zhang Zhidong 张之洞, 113, 173, 178–9, 351, Zigong, Sichuan, 329, 331,
364, 401, 780 337, 339
Zhangjiakou, Hebei, 69 Zongli Yamen (Foreign Office), 254
Zhangqiu, Shandong, 55 Zuo Zongtang 左宗棠, 171, 172, 254
845


The Cambridge Economic History of China: Ebin A

Uploaded by

Copyright:

Available Formats

The Cambridge Economic History of China: Ebin A

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Cambridge Economic History of China: Ebin A

Uploaded by

Copyright:

Available Formats

the cambridge economic history of

China’s rise as the world’s second-largest economy surely is the

D E B I N M A is Professor of Economics at Hitotsubashi University,

Published online by Cambridge University Press

Published online by Cambridge University Press

Published online by Cambridge University Press

Cambridge University Press is part of the University of Cambridge.

© Cambridge University Press 2022

Published online by Cambridge University Press

China’s rise as the world’s second-largest economy surely is the

D E B I N M A is Professor of Economics at Hitotsubashi University,

Published online by Cambridge University Press

Published online by Cambridge University Press

Published online by Cambridge University Press

Cambridge University Press is part of the University of Cambridge.

© Cambridge University Press 2022

Published online by Cambridge University Press

China’s rise as the world’s second-largest economy surely is the

D E B I N M A is Professor of Economics at Hitotsubashi University,

Published online by Cambridge University Press

Published online by Cambridge University Press

Published online by Cambridge University Press

Cambridge University Press is part of the University of Cambridge.

© Cambridge University Press 2022

Published online by Cambridge University Press

China’s rise as the world’s second-largest economy surely is the

D E B I N M A is Professor of Economics at Hitotsubashi University,

Published online by Cambridge University Press

Published online by Cambridge University Press

Published online by Cambridge University Press

Cambridge University Press is part of the University of Cambridge.

© Cambridge University Press 2022

Published online by Cambridge University Press

List of Figures page viii

1 . Ideology and the Contours of Economic Change 15

2 . Economic Transition in the Nineteenth Century 48

4 . Handicraft and Modern Industries 124

5 . The State and Enterprises in Late Qing China 167

Published online by Cambridge University Press

7 . Money and the Macro-economy 208

8 . Public Finance 244

9 . Financial Institutions and Financial Markets 280

10 . Chinese Business Organization 324

11 . The Economic Impact of the West: A Reappraisal 354

12 . Foreign Trade and Investment 414

13 . Transport and Communication Infrastructure 457

14 . Education and Human Capital 496

15 . The Origin of China’s Communist Institutions 531

16 . China’s Struggle with the Soviet Growth Model, 1949–1978 565

17 . Living Standards in Maoist China 606

18 . The Political Economy of China’s Great Leap Famine 642

19 . China’s External Economic Relations during the Mao Era 685

Published online by Cambridge University Press

20 . The Chinese Economy in the Reform Era 722

21 . China’s Great Boom as a Historical Process 775

Published online by Cambridge University Press

Published online by Cambridge University Press

Published online by Cambridge University Press

19.1 Chinese total trade, 1950–2006 687