WoddDevelopnmt, Vol. 22, No. 1 pp. 1l-26.

1994
Copyright 0 1994 Elsevier Science Ltd
Pergamon
Printed in Great Britain. All rights reserved
0305-7501094 $6.00 + 0.00

Industrial Policy Reform in Six Large Newly

Industrializing Countries: The Resource Curse
Thesis

RICHARD M. AUTY*
University of Lancaster

Summary. - All six large newly industrializing countries (NICs) in this study favored an
autarkic industrial policy (AIP) in the 1950s. AIP required, however, the rapid build-up of
heavy and chemical industry (HCI) which caused foreign exchange shortages if the HCI
matured slowly. This is because. the import and subsidy demands of the growing protected
HCI sector became too burdensome for the competitive primary sector whose rents shrank
in relative importance as the economic structure evolved. The two least well-endowed NICs
abandoned AIP early in favor of a competitive industrial policy (CIP) which relied on labor-
intensive exports to earn foreign exchange and resulted in rapid economic growth. In con-
trast, the resource bonus of the four best endowed NICs sustained AIP longer, but the longer
AIP lasted the harder it became to reform and the slow-maturing HCI sector adversely
affected economic growth. In this way the resource bonus became a curse rather than a
blessing.

1. EXPLAINING DIVERGENCE IN relative success of Colombia in Latin America.

INDUSTRIALIZATION Hughes (1988), for example, mischievously assigns
the Philippines to Latin America and Colombia to
Asia. Or again, Summers and Easterly (1992) point
The successful industrialization of many East
out that today’s star performers sre yesterday’s bas-
Asian countries contrasts with the disappointing out-
ket cases, none more so than Korea and Singapore.
come in much of Latin America and sub-Saharan
Anomalies pose similar problems for those such as
Africa. Not surprisingly, given this geographical pat-
Andrain (1988) and Bello and Rosenfeld (1992) who
tern, explanations usually stress cultural factors (in
support the other important cultural thesis which is
the form of a work ethic or type of political regime)
based upon differing political regimes.
or environmental factors (the level of urbanization or
Sachs (1985) offers an “environmental” explana-
the natural resource endowment). Closely linked to
tion which attributes differences in economic success
this controversy is a second dispute about the mecha-
to variations in the degree of urbanization. He argues
nism by which the causal factor has an impact on a
that high levels of urbanization foster the entrench-
country’s industrialization. Specifically, some
ment of powerful rent-seeking interest groups which,
researchers emphasize macroeconomic policy (Lal,
once ensconced, oppose any shift toward market-
1983; Stem, 1990) as the mechanism while others
conforming policies. Sachs notes that the successful
cite industrial policy (Amsden, 1989; Wade, 1990).
East Asian countries had significantly lower levels
Returning to the cause, the cultural thesis sug-
of urbanization than did Latin American countries
gests that a Confucian factor accounts for the superi-
when key decisions were taken about the scale and
or Asian performance (Morishima. 1982). It argues
nature of state intervention in the initial postwar
that a respect for authority and a tolerance of hierar-
decades.
chy within Confucianism bring discipline and a
Unfortunately, however, Sachs’s comparison of
greater readiness to sacrifice present consumption
countries in Southeast Asia and Latin America is too
for future welfare in East Asian countries compared
with Latin America and sub-Saharan Africa. But the
cultural thesis is flawed by the fact that there are *The financial support of the Nuffield Foundation, the
many anomalies (Elkan, 1973) such as the underper- British Academy and the World Bank is gratefully
formance of the Philippines in Southeast Asia or the acknowledged. Final revision accepted: July 9, 1993.

11
12 WORLD DEVELOPMENT

Sector Primary MFG Services

S&sector Mining Agric LI HCI

75 50 2.5 0
% GDP

q % Nontradeable
Figure 1. Policy and sectoml competitivenem.

narrow: other countries with low levels of urbaniza- Moreover, like all unicausal explanations, it under-
tion in Asia as well as in most of sub-Saharan Africa states the role of other variables for clarity of
failed to prevent the entrenchment of such interest exposition.
groups. Yet although Sachs has selected the wrong The fmdings of recent intercultural comparisons
causal factor, he may have correctly identified an of the performance of the mineral economies, which
important part of the mechanism by which policy account for almost one-quarter of the developing
errors persist. The entrenchment of powerful urban countries and include both oil-exporters (Gelb, 1988)
rent-seeking groups is a part of the mechanism. But and ore-exporters (Auty, 1993), are consistent with
it still remains to correctly identify the cause and the resource curse thesis. The mineral economies
also to specify the additional components of the have underperformed compared with countries of a
mechanism. similar size and level of economic development
A second environmental factor, the natural which lack the mineral bonus. This is because all too
resource endowment, may furnish the causal vari- often, the rents from the mineral resource were used
able. Moreover, its effect on choice of macro and to evade the need to compete internationally in man-
industrial policy options may provide the missing ufacturing (and in some instances, in agriculture
parts of the mechanism. The environmental factor is also). In extremis, the mineral may become the only
invoked by Mahon (1992) who argues that the tradeable commodity (see Figure l), as in the cases
reform of Latin American economies was blocked of Venezuela and Mexico during the 1979-81 oil
because the deployment of revenues from their pri- boom.
mary products had initially been too successful in This paper applies the resource curse thesis to a
raising living standards and labor costs. He suggests comparison of the postwar industrialization strate-
that by the time the limits to the primary sector’s gies of the larger newly industrializing countries
capacity to generate foreign exchange became appar- (NJCs). Hong Kong and Singapore are excluded due
ent, the costs of making the required switch to a to their small size but they should reinforce the the-
competitive industrial policy (which he sees as nec- sis advanced. The paper takes country size as a sur-
essarily based upon low-wage manufactured exports) rogate for resource endowment and makes two
were too high. This is because the costs of reform assumptions: first, after Myint (1964). that the larger
cannot be recouped within a political cycle that will the population, the greater the potential domestic
make it worthwhile for politicians, whether authori- market and the easier the capture of scale economies;
tarian or democratic, to introduce reform. and second, after Ginsburg (1957), that the more
Mahon’s argument is a variant of the resource extensive the geographical area, the greater the like-
curse thesis which suggests that resource-rich coun- lihood that a wide range of natural resources will be
tries may squander their resource advantage because present with which to source domestic production or
an overly optimistic estimation of their prospects to export in order to earn foreign exchange.
leads to the pursuit of lax economic policies. The larger country’s greater market potential and
A corollary is that resource-poor countries, mindful broader resource base widens the developmental
of their marginal position, may compensate for their options while the smaller country’s options are more
disadvantage by adopting firmer and more farsighted constrained. Just as the’effect of the market factor,
policies. But the resource curse thesis is not an iron however, is moderated by variations in per capita
law, rather it is a strong recurrent tendency. income, so the natural resource advantage arising
 RESOURCE CURSE THESIS 13

Mukel-Rick. Resource-Rick Comtrks

I I I I I I
0.4 0.6 0.8 1.0 1.2 1.4
Cropland per Head 1980 (Ha)
Sources: WRI 1992; World Bank 1982

Empirical data suggest that in 1980 a dome&c ma&t generating $40 billion in GDP would support a
single integrated steel plant of minimal viable size whereas a GDP of $120 billion would support three
such units and so engender greatex competitiveness.

Figure 2. Countty sizJruourcc cmstraints on akvelopmm!optim.

from large geographical size is moderated by the the three pairs of newly industrializing countries’
degree of population pressure on resources. beginning with Korea and Taiwan. then turning to
Figure 2 uses GDP as a measure of market poten- India and China before tackling Mexico and Brazil.
tial (the choice of class boundaries is given in the The fifth and final section summarizes the policy
footnote to the figure) and cropland per head as a implications.
proxy for resource endowment. Xtidentifies four cat-
egories of “large” country, three of which are of
interest here: the market-modest, resourcedeficient 2. RESOURCE ENDOWMENT, INDUSTRIAL
case; the market-rich, resource-pressured case and POLICY AND ECONOMIC GROWTH
the market-rich, resource-rich case. (The fourth case,
the market-poor, resource-rich country identifies the Despite the potential benefits for successful industri-
typical mineral economy.) Three pairs of countries alization of a large domestic market and a diversified
are studied here which correspond to the key group- natural resource endowment, a recent examination of
ings, namely: market-modest, resource-deficient the effect of country size on economic performance
Taiwan and Korea; market-rich India and China and (Perkins and Syrquin, 1989) reveals no clear-cut
market- and resource-rich Brazil and Mexico. It is relationship between greater size and economic per-
hypothesized, first, that the more constrained the nat- formance. Larger countries did grow slightly faster
ural resource endowment the less the scope for post- than smaller ones in 1960-82 (which Perkins and
poning the development of competitive mantiactur- Syrquin attribute to higher productivity), but income
ing: and, second, that such postponement makes distribution shows little correlation with size
reform increasingly costly (as the weakness of the (although it might be expected to be less equal in a
growing and overly protected manufacturing sector larger and more diversified environment). The main
impairs economic performance) and also difficult to finding of the study is that the industrial structure of
achieve. larger countries tends to diversify earlier than the
The analysis proceeds in five steps, beginning average and large countries tend to be more self-suf-
with the construction of a simple model of the rela- ficient or autarkic.
tionship between the resource endowment, industrial All six NICs studied here opted for autarkic
policy choice and sectoral and macroeconomic per- industrial policies (AIPs) in the 1950s. The motiva-
formance. In the next three sections, the model is tion for AIP was overwhelmingly strategic in the
deployed to explain the postwar industrialization of case of the four Asian countries, but was associated
14 WORLD DEVELOPMENT

Table 1. Chenety and SyrquinStructuml ChangeNomu (%GDP)

GDP/Capita(1980 $) 300 500 1,000 2@0 4,000

Pri&eonsumption 73.3 70.2 66.4 63.1 60.3

Public consumption 13.6 13.5 13.7 14.4 15.4
Gross investment 18.4 20.8 23.3 25.0 25.9
Total 105.3 104.5 103.4 1025 101.6

Value Added
Agriculture 39.4 31.7 22.8 15.4 9.7
Mining 5.0 6.6 7.7 7.5 6.1
Iuanufacturing 12.1 14.8 18.1 21.0 23.6
Light industry 7.6 8.4 9.6 10.1 10.4
Heavy industry 4.5 6.5 9.2 11.7 14.0
con.¶nKtion 4.4 4.9 5.5 6.1 6.7
Utilities 6.7 7.4 8.1 8.8 9.3
Services 32.4 34.6 37.8 41.2 44.7

HCI Value Added

chemiads and rubber 2.4 3.0 3.6 4.0 4.2
Non-metallic minerals 0.5 0.8 1.1 1.2 1.3
Basic metals 0.5 0.9 1.7 1.9
Machinery,engineering 1.0 1.9 ::; 4.8 6.6

Source: Syrquin and Chenery (1989).

more with ECLA-inspired import substitution in the Within such an orthodox framework, however,
two Latin American cases (Mahon, 1992). Autarky the purists urge minimal state intervention which
stresses heavy and chemical industry (HCI) because, precludes an active industrial policy. Such a stance
as the Syrquin and Chenery (1989) norms show (see eschews sectoral targeting and leaves financial mar-
Table l), that is the least developed component in a kets to determine the pattern of investment. Some
country’s industrial structure in the early stages of theorists such as Stern (1990) assume that macroeco-
&nomic development. Its implementation demands nomic management is the key to successful industri-
three critical policy decisions which concern: the alization and that the sectoral interventions associat-
overall macroeconomic framework within which ed with a competitive industrial policy (CIP) have
greater autarky is to be pursued, the degree of infant had at best a neutral impact.
industry protection to be given; and the sequence in Similarly, Ranis and Mahmood (1992) conclude
which HCI is to be expanded. from the economic performance of six developing
economies during 196&85 that an economic policy
(a) Macroeconomicpolicy which maintains a commitment to “linear liberaliza-
tion” will depoliticize the transfer of resources asso-
Macroeconomic policy profoundly affects industrial- ciated with development and thereby sustain rapid
ization. The postwar dispute over what constitutes a economic growth. Ranis and Mahmood extol Taiwan
prudent macroeconomic policy has been fought and Korea, but chastize the latter for backtracking on
between advocates of orthodox and structuralist poli- liberalization under the Big Push. They are skeptical
cies and by the 1990s it had been largely resolved in of industrial policy and assume that the persistence
favor of the former. The macroeconomic policy con- into the 1980s of Korean and Taiwanese trade and
troversy has subsequently narrowed to a debate over industrial policy intervention is less critical than
the scope for beneficial intervention within an ortho- their adherence to monetary restraint and a realistic
dox policy (Auty, 1993). There is, however, basic exchange rate. Ranis and Mahmood contrast the
agreement on the need for fiscal prudence and an gains from linear liberalization with the damage
outward-oriented trade policy which require a twin inflicted by switches between doctrinaire orthodoxy
commitment to a competitive exchange rate (to facil- and structuralist intervention on Latin America.
itate exports) and medium-term fiscal balance (to Structuralist policies assume that the global econ-
curb inflation). In this way orthodoxy reduces the omy works against the interests of developing coun-
risk of inflationary spirals, cumulative economic tries and that extensive state intervention is therefore
deterioration and prolonged deflationary corrections. required. As advocated by Prebisch, reformist struc-
 RESOURCE CURSE THESIS 15

turalism repeals open trade in or&r to build intema- adjustments that enhance the productivity of the
tional industrial competitiveness, using regional trad- newly installed plant
ing blocs to surmount any scale barriers to entry. (c) Sectoral sequencing
State intervention via multiple exchange rates and
subsidies is required to assist the growth of targeted The third decision concerns sectoral sequencing. In
sectors. Structuralist policy in its more radical mode, line with the four-stage East Asian development
minii external trade and uses central planning in model (Kuznets, 1988) CIP switches its targeted
place of the price mechanism (Chilcote, 1974). The manufacturing sectors as per capita income rises
resulting command economy may take the form of from primary import substitution in stage one to
command capitalism, as in India, or command labor-intensive manufacturing in stage two, to capi-
socialism, as in China (Lucas, 1988). tal-intensive and then to skill-intensive manufactnr-
ing in stage three (the drive into heavy industry), and
(b) Indu.strial policy fmally to research-intensive sectors. Both AIP and
CIP seek to speed up industrial diversification by tar-
The second critical policy decision concerns geting HCI. but AIP leapfrogs the second (labor-
industrial policy about which there remains much intensive export-manufacturing stage) so that its
controversy. The controversy persists because entry into HCI is more likely to be premature.
irrefutable evidence that an industrial policy can The sequencing of diversification into HCI
have a positive impact on an economy is still lack- requires a choice between a Drive and a Big Push.
ing. Recent institutionalist studies by Amsden (1989) An HCI Drive accelerates the backward integration
on Korea and by Wade (1990) on Taiwan, while which economic development entails from simple
making a plausible case for the benefits of a compet- consumer goods to ever more sophisticated interme-
itive industrial policy, do not provide definitive diate and capital goods. Au HCI Big Push is more
proof. There is. however, widespread agreement that ambitious, however, and seeks to maximize the
an ill-judged industrial policy can have a negative economies of scale through simultaneous entry into
impact on an economy (Hughes, 1988). several HCI sectors which have complementary
An autarkic industrial policy (AIP) which exces- demand. For example, a steel mill is established
sively prolongs HCI protection leads to slow HCI along with downstream steel-consuming industries
maturation and the neglect of sectors of existing so that all the new plants can be larger and therefore
comparative advantage such as agriculture and light more competitive than would otherwise be the case.
manufacturing. The persistence of such sectoral Murphy et al. (1989) argue that such complemen-
trends brings mounting distortions throughout the tary expansion carries substantial economy-wide
economy so that an inadequately competitive trade- benefits. The gains from the simultaneous pursuit of
ables sector damages m acroeconomic performance. industrialization in many sectors can substantially
The principal macroeconomic consequences are: an boost welfare and income even when investment in
overvalued exchange rate, recurrent chronic foreign one sector seems unprofitable. This is especially true
exchange and fiscal gaps, growing income inequality where high freight costs or trade restrictions limit
and erratic economic growth. access to foreign markets. The coordinated invest-
A competitive industrial policy (CIP) creates a ment of an HCI Big Push may reduce the explicit
positive relationship between structural change and costs of subsidies and tariffs because large implicit
macro policy, as Korea and Taiwan illustrate below. subsidies flow across sectors. This theoretical argu-
CIP is market-conforming because it repeals compe- ment, however, neglects implementation problems.
tition only temporarily. It is sometimes called dualis- Empirical evidence suggests that HCI Big Pushes in
tic (Ohno and Imaoka, 1987) because it simultane- the larger middle-income NICs (Brazil, Mexico and
ously maintains the competitiveness of established Korea) breach implementation capacity and depress
industrial sectors, while encouraging sectors of economic performance, especially under an autarkic
emerging comparative advantage. CIP provides a policy (Auty. forthcoming).
package of market information, assistance with tech-
Briefly. an HCI Big Push triggers a distinctive
nology acquisition. subsidized credit, tax breaks and
three-stage macroeconomic sequence (Auty. 1992a)
trade incentives for new entrants to set up infant which comprises:
industries. But it demands that favored firms rapidly -The HCI Big Push Construction Boom which
achieve economic aud technological maturity. hits the economy like a commodity boom and
Economic maturity implies that firms compete triggers inflation and an exchange rate apprecia-
internationally with no subsidies other than those tion along with macro imbalances (widening
required to compensate for the use of high-cost trade and fiscal gaps) which after 4-5 years leads
inputs from the next wave of protected industrial to:
sectors. Technical maturity requires that erstwhile -A GDP Slowdown as stabilization is required
infant firms must be able to make in-house technical which depresses domestic demand below its
16 WORLD DEVELOPMENT

long-term trend just as many HCI project8 are exchange requirements of an economy pursuing an
struggling on stream, causing low capacity use industrial policy must be generated by the resource-
and problem8 over debt service. This gives way based (primary) sector, by light manufacturing or by
to: HCI. Syrquin and Chenery (Table 1) show that a8
-An HCI Rebound - provided stabilization is development pmceeds the primary sector contracts,
successful - as demand rises and excess HCI so fist light industry and subsequently, HCI become
capacity i8 activated. the dominant tradeable sector.
Figure 3 outlines the sequence with reference to If HCI maturation is delayed and the light manu-
the Korean HCI Big Push during 1975-79 and shows facturing stage is neglected, as with AIP. then the
the extent of the growth likely to be foregone in a burden of generating foreign exchange remains with
well-managed economy. In economies which are less the primary sector whose relative size shrinks a8
effectively managed than that of Korea, the spill-over development proceed8 (Table 1 and Figure 1). A
effect8 on economic performance are likely to be resource-rich country can tolerate the slow matura-
more damaging, as subsequent sections show. An tion of HCI for longer than a resource-deficient
HCI Big Push is a more feasible option for the large country because it8 primary sector generates greater
country because a large domestic market reduces the rents. It can also more easily cope with a weak con-
degree to which the economies of scale are a barrier tribution from light manufacturing. Moreover, the
to entry. This is one way in which the sequencing primary sector may even encourage subsidization
decision is affected by the natural resource endow- of manufacturing where the rents are sufficient to
ment. Similarly, the presence of large market poten- cause a real appreciation of the exchange rate, as
tial makes autarky more feasible whereas limited during a commodity boom for example. Such subsi-
market potential renders it far less attractive. dies may appear to be temporarily justified in order
to offset what is, in effect, Dutch disease (the prema-
ture shrinkage of agriculture and underexpansion of
(d) The mechanism competitive manufacturing).
The primary sector of a richly endowed country
The size of the competitive tradeable sector (a.8 dis- may initially be able to carry the burden of an
tinct from the protected tradeable sector) vis-d-vis uncompetitive manufacturing sector, but as the rela-
the rest of the economy lies at the root of the diver- tive size of the primary sector declines, the manufac-
gent economic performance under alternative indus- turing sector must increasingly compete, unless total
trial policies. Assuming that services are all non- autarky is sought (Figure 1). A weak manufacturing
tradeable and that the learning curve makes instant sector will be ill-placed to do so and the longer it ha8
autarky impractically costly, then the foreign been cossetted. the more resistant to change become
the entrenched interest groups which benefit from
protection and block reform. In this way a rich natur-
al resource endowment can trigger a set of policy
200 choices which, within little more than a decade
190 (Gelb. Knight and Sabot, 1986). can transform the
180
resource bonus into a curse.
170
8 160
150
3. COMPETITIVE INDUSTRIAL POLICY
140
IN TAIWAN AND KOREA
130
120
(a) Early industrial policy liberalization
110

100 Countries at low levels of development are espe-

Re-msh HCI Big Stabilization Rebound
Push- cially deficient in HCI (Table 1) so that if they are
7273 74757677787980818283 8485868788
to become self-sufficient, a8 autarky requires, they
The
must concentrate their resources on building up the
HCI sector. This demands capital and skill -
thereby causing low-income developing countries
to deviate from their comparative advantage in
labor-intensive activity. Both Taiwan and Korea
FtIcll cap (a (PIP) (1.7) (1.9) (2.6) (9.U pursued autarkic industrial policies through the
C~llsnl Accw”t (% QDP) (5.6) (4.9) WI (3.4)
1950s behind high tariff barriers, with the usual
outcome of a high-cost, low-capacity use, rent-
Figure 3. HCI Big Push sequence, Korea. seeking industrial sector.
 RESOURCECURSE THESIS 17

Taiwan transferred resources from its dynamic which the domestic labor market tightens and trig-
farm sector to support such industry (Amsden, 1985) gers wage increases which must be accommodated
while Korea used US aid. They both abandoned this by higher worker productivity (Kuznets. 1988). Such
strategy in favor of CIP much earlier than the larger a rap@ absorption of surplus labor out of the rural
NICs (Taiwan in 1958 and Korea in 1963). Overall, sector was assisted by initial opposition to unioniza-
both the m acroeconomic policy and CIP of Taiwan tion which kept wages low and labor flexible in both
were more cautious and less interventionist than countries.
those of Korea (Wu. 1988). The turning point heralds the next stage of CIP
Taiwan adopted its cautious macro policy in 1958 which sees the gradual abandonment of lagging,
by adding fiscal balance and a strongly devalued cur- low-productivity sectors to the next wave of newly
rency to the unfashionably high real interest rates industrializing countries and entry into sectors of
which it had adopted earlier in that decade (Tsiang, emerging competitive advantage. Whereas the initial
1984). It promoted labor-intensive manufactures phase of CIP is strongly export-driven, domestic
which increased their share of total exports from demand becomes a steadily more important source
10% to 46% during 1955-65. Korea made a shift to of economic growth after the turning point as
a more liberal policy in 1963 following a strong increases in real domestic wages boost purchasing
devaluation and it too experienced rapid export power (Chenery et al., 1986). Economic growth was
growth (Wu, 1988). In both countries, distinct phases rapid under both stages of CIP in Taiwan and Korea
in the industrial policy incentives can be recognized (see Table 2).
which followed the East Asian development model
as emphasis shifted from the labor-intensive exports
of the 196Os, to heavy and chemical industry in the (b) Economic pelformance
1970s (with a gradual move from capital-intensive
industrial intermediates to skill-intensive engineer- Structural change also proceeded apace with HCI
ing) and then to knowledge-based products in the outstripping light industry’s contribution to GDP by
1980s. 1973 in Taiwan and 1977 in Korea. Rapid growth
The positive contribution of CIP to economic resulted in both countries from high rates of produc-
growth can be demonstrated with reference to the tivity growth combined with high levels of invest-
crucial second and third stages of the East Asian ment, notably in Korea. Taiwan should have grown
development model. First, there is an expansion of faster than Korea because it secured a head start and,
labor-intensive products (much for export) - which according to the Chenery and Syrquin norms, growth
rapidly absorbs surplus labor so that the turning accelerates as development proceeds. In fact, this is
point is reached (in the case of Taiwan and Korea the case during 1961-90: GDP growth in Taiwan
around 12 years after the adoption of CIP). The tum- averaged 9.1% compared with 8.7% in Korea. If the
ing point identifies the stage of development at 1960s are excluded, however (since they include

Table 2. Investment, I&NIWZ &i&ncy a& GDP growthbypolicyphase, Tti and Koreu

Policy phase AutarkicindustrialPolicy Competitiveindustrialpolicy

Expat-w HCl-Led Liberalized

Dates:
TaiWall 1950-58 1959-71 1972-81 1982-91
Korea 195242 1963-73 1974-82 1983-91

Investment (96 GDP)

T&WI 12.2 17.2 27.4 21.4
Korea na 19.9 29.1 30.9

ICOR
T&WI 1.5 1.8 3.0 2.6
Korea n.a 1.9 4.1 3.0

GDP Growth (%/Ys)

TaiWll 7.9 9.5 9.1 8.2
na. 10.4 7.1 10.2

Sources: UN (1975); National Statistical office (1992); DGBAS (1992).

18 WORLDDEVELOF’MENT

Table 3. Composition qfexportr by counhy. I%0 and 1980

Country Shareof Tctal Expats (%) Exportgrowth 1960-80(Wr)

primary MIUlllfIUilRing Total Manufacturing
1960 1980 1960 1980

KOrea 86 11 14 89 24.4 43.4

Taiwan 12 2 88 98 25.9 28.8

china Il.& 51 n.a 49 ma. La.

India 55 39 45 61 3.3 5.5

Mexico 88 61 12 39 8.0 13.4

Brazil 91 61 3 39 6.4 21.3

Source:Cheneryet UI! (1986),p. 111, except Taiwan: Council for Economic F’lanaingand Development.

three years of premform performance in the case of 4. LOW INCOME ASIA: INDIA AND CHINA
Korea), there is little evidence of any significant dif-
ference in the overall economic growth rate during (a) Implications of autarky for low income countries
1970-90: the per capita income in both countries
grew with a less than eight-year doubling rate. AIP shields infant industries longer than CIP so that
Closer inspection reveals, however, that the HCI maturation takes several decades rather than the
more interventionist CIP of Korea was associated five to eight years which orthodox theory considers
with a macroeconomic performance which was the maximum if the benefits from the matured indus-
more erratic than that of Taiwan and made less effi- try are to compensate for the costs of protection
cient use of investment. Inflation in Korea averaged (Bell et al., 1984). AIP also imposes persistent bias-
12.7% during 1961-90 compared with only 5.6% in es against the unprotected sectors of established
Taiwan (Auty, 1992b). As for investment effrcien- comparative advantage, namely agriculture and light
cy, the Korean ICOR was higher (less efficient) industry. The neglected sectors therefore have difft-
than that of Taiwan, albeit only slightly so through culty in retaining or capturing export markets.
the export-led phase (Table 2). Thereafter it deteri- This combination of slow HCI maturation and
orated to 4.1 during the HCI-led phase (one-third neglect of the established sectors is the prime cause
higher than Taiwan) when, in contrast to Taiwan, of AIP’s foreign exchange shortages. The latter ham-
Korea ambitiously accelerated its HCI Drive into a per economic growth by restricting the import of
Big Push. both the capital equipment needed to build the pro-
The differences in investment efficiency reflect tected industries and also of the inputs which manu-
the diverging structure of the two economies, as facturing and other sectors cannot secure domestical-
that of Korea became more capital intensive (and ly. The results are: sluggish growth of labor-inten-
HCI dominated) than that of Taiwan, and also less sive employment, postponement of the turning point
flexible. But the broader picture is one in which and amplified income inequality. By the time that
both of the resource-poor, market-modest NICs the economic and social costs of AIP become evi-
industrialized more successfully than their larger dent the policy is difficult to modify because the
counterparts. Unlike the four larger countries, they interest groups which benefit from industrial protec-
abandoned an autarkic industrial policy (AIP) tion have become sufficiently entrenched to block
sooner and built on their comparative advantage in reform (Sachs, 1985). Such an outcome is clearly
labor-intensive exports before targeting HCI after illustrated by China and India even though both
the turning point (see Table 3). In the absence of countries experienced traumatic economic shocks
either high rents from the primary sector or foreign (China in 1958-61, India in 1965-67) in the after-
aid, the export of labor-intensive manufactures was math of ambitious HCI Big Pushes which called into
the principal means of generating foreign exchange question the wisdom of their autarkic policies.
to pay for imports of materials and the more In fact, China and India display some remarkable
sophisticated manufactured products. In contrast, similarities in their postwar economic development.
the two large market-rich Asian countries could At the sectoral level, both countries launched ambi-
expect to achieve a much higher degree of eco- tious HCI Big Pushes (China in 1953-57 and India
nomic self-sufficiency so that the rapid maturation in 1956-65) which, until the reforms of the late
of manufacturing to earn foreign exchange was a 1970s. were followed by a period which was charac-
less urgent priority. terized by: a slow rate of job creation in manufactur-
 RESOURCE CURSE THESIS 19

Table 4. Emnomic growth, savings and ICOR 1963-8.5

1963-73 1973-85
Growth Savh@ ICOR CkOWth Savings ICOR
(%/yr) @GDP) (%@) (%GDP)

Kaea 9.6 11.7 20 7.3 26.3 4.3

Taiwan 10.7 18.1 1.9 7.9 30.0 3.9

China n.a 7.2 30.4 4.2

India 3.5 5n:: f:t 4.4 15.3 4.7

Mexico 7.8 25.1 26 4.8 22.5 5.7

Brazil 8.3 20.3 2.8 4.3 26.5 5.7

Source: Balassa (1988), except: China 1973-85, World Bask (1991).

ing, especially in light industry; neglect of productiv- required high rates of investment which, since the
ity improvement (and therefore of income growth efficiency of that investment was low (the incremen-
and purchasing power) in the agricultural sector; and tal capital output ratios were typically more than
the dissipation of the HCI economies of scale double those of Taiwan and Korea, compare Tables
through slow maturation and efforts to disperse pro- 2 and 3) implied sizeable consumption foregone by
duction widely through the country. those who could least afford it Yet despite the emer-
This period between the end of the first Big Push gence of severe problems during each country’s first
and the onset of reforms was one in which: economic HCI Big Push, autarkic HCI-led growth proved diffi-
growth was sluggish in India (see Table 4) and erratic cult to reform.
in China. Perkins (1988) estimates Chinese growth at This is a particularly interesting outcome in the
2.2% during 195&-66 and 5.2% during 1966-76) - case of China because. its command economy placed
the turning point was delayed and income inequality it in a strong position to implement reform, once the
grew (Mundle, 1990; Perkins. 1988). More spectical- efficiency losses (arising from high investment and
ly, an initial period of rapid HCI-led growth which low GDP growth) associated with AIP were appreci-
began in the 1950s was followed by slower growth ated. Yet China underwent a protracted struggle
into the mid-1970s which reflected thwarted reform between efficiency-oriented and politically oriented
efforts and the negligible contribution of technical factions which placed many obstacles in the path of
advance. The economic performance of the two coun- reform, of which the cultural revolution was the
tries improved from the late 1970s. with higher most significant (Prybyla, 1987).
growth in India and more sustained growth in China The Chinese reforms did make remarkable gains
- as the reforms, some of which fmt appeared in the when they were at last applied to the rural sector dur-
196Os, became more effective. ing 1978-84. An important reason for this is that the
decentralization of decision making fmally released
(b) Persistence of autarkic policies the potential benefits that had been embodied in the
rural infrastructure build-up under the communal
The costs in terms of consumption foregone have system. Output from agriculture grew by 8% annual-
been very high, however, especially for countries ly while the role of HCI declined sharply. The share
with populations that could ill afford such loss. HCI of investment in GDP, which had exceeded 35% in

Table 5. China’s economic peflomtmc e 1981-89

Rural reform Unintended HCl Big Push Stabilization

1981 1982 1983 1984 1985 1986 1987 1988 1989

Irwstment (%GDP) 29.1 29.6 30.3 32.2 38.7 39.3 39.2 39.9 36.1
GDP growth (%/j~) 4.9 8.3 9.8 13.5 13.1 8.0 10.5 11.2 3.9
Consumer price index (%/yr) (0.4) 16.3 30.3 12.1
Budget balance (ZGDP) fll) (2.2) (2.5) n.a
Current account (%GDP) (0.1) (1.1) (1.2)
Foreign debt (%GDP) 2.9 10.6 12.2 11.3

Source: Wald Bank (1991).

20 WORLD DEVELOPMENT

the late 1970s shrank to 30% in the early 1980s (see rate made. through the 1970s and 1980s by resource-
Table 5) so that there was a corresponding rise in pressured India and China (Wood, 1988). The Latin
consumption. American real exchange rate overvaluation not only
But the reforms were again set back when they eroded the competitiveness of the primary sector but
were extended to the urban-based industrial sector in also hampered competitive industrial diversification.
1984. They unintentionally triggered a third HCI Big An overvalued exchange rate cheapens those import-
Push (China’s second push occurred during 1969-73 ed inputs used by the protected sectors which are not
under the defencedriven third front decentralization produced domestically. It also appears to justify
strategy). The reason for this is that the authority extending protection to infant industries. Over the
which the central government ceded to the enterprises long term such policies result in more erratic and
was intexcepted by intermediate tiers of government. slower economic growth than do more orthodox
The latter then used that authority to undertake local policies. This reflects the recurrence of foreign
import substitution which again stressed HCI. At the exchange shortages, excessive fiscal deficits and
national level, the 1978-84 decline in the share in bouts of intlation.
GDP of both investment and HCI was reversed. ‘Ihe Like their market-rich Asian counterparts, both
share of HCI in investment, having fallen from Mexico and Brazil sought high levels of autarky and
48.7% to 35.7% during 1978-85 climbed back to leap-frogged the second (labor-intensive export)
43.5% by 1988. Consistent with the stylizd facts of a stage of the East Asian development model. Not
Big Push outlined earlier, the thhd HCI Big Push led only did this entrench HCI as the leading sector of
to macroeconomic imbalances which threatened to the economy, but it also reinforced the already sub-
terminate the entim reform program (Table 5). In the stantial income inequality in the Latin American
third stage of the Big Push cycle, however, an HCI countries. By leap-frogging the labor-intensive
rebound appeared to be under way in the early 1990s. export stage Brazil and Mexico postponed the labor
Meanwhile. around two-thirds of Chinese industrial market turning point, perpetuating a labor surplus in
production came from HCI compared with the the rural and urban informal sectors whose poverty
Chenery and Syrquin norm of two-fifths for a country contrasted with that of those fortunate enough to
of similar size and level of development. work in the protected sectors. In contrast to the rela-
India attempted to downgrade the importance of tively equitable income distribution of Taiwan and
HCI in the mid-1960s but, as in China, reform was Korea, where the ratio of the income of the richest
repeatedly postponed. There is no evidence, howev- quintile to the poorest quintile is five and seven,
er, of any economic rebound after the collapse of respectively. the ratio for Mexico reached 15 and
the Big Push in 1965 nor of any pronounced change that for Brazil 30 (Sachs, 1989). Ironically, the inten-
in the overall allocation of investment away from sified income inequality was deployed as an argu-
HCI. Meanwhile, a steady deterioration occurred in ment for postponing reform of AIP on the grounds
the performance of the state-dominated HCI sector that orthodox policies require a deflation which
into the 1980s which the more reform-minded gov- would hurt the poor the most.
ernments of that decade found difficult to reverse, A recurrent policy response under reformist
even under intense pressure from the International structuralism has been to evade reform by “growing”
Monetary Fund (IMF). The experience of both China out of the problems through populist booms (Sachs,
and India with reform since the 1960s underlines 1989) which are designed to avoid the austerity asso-
first, the strong appeal of economic autarky to mar- ciated with orthodox reform. But such populist
ket-rich developing countries (despite the adverse booms are counterproductive: they bring a brief peri-
implications for resource efficiency and consump- od of rapid economic growth as real wage increases
tion) and second, the difficulty of diminishing the boost demand and use up idle capacity before they
role of HCI once it has become dominant accelerate inflation and cause ballooning trade and
fiscal deficits. The deficits presage a sharp economic
contraction. Sachs (1989) cites Mexico in 1978-82
5. PROTRACTED AUTARKIC INDUSTRIAL and Brazil in 1985-89 as important examples of the
POLICY REFORM: MEXICO AND BRAZIL populist cycle.

The two market- and resource-rich Latin American (a) Mexican oil postpones reform
countries, Mexico and Brazil, drifted into a struc-
turalist macro policy and AIP. The rents from their Buffie and Krause (1989) dispute the need for
primary sectors raised their per capita income levels reform in Mexico. They argue that Mexico’s phase
significantly above the developing country average of Stable Development (1956-70). based on increas-
but the rents also brought exchange rate overvalua- ing autarky, was highly successful. It was character-
tion before the 1982 debt crisis in contrast to the ized by modest inflation and high GDP growth (6%
large cumulative real depreciations of the exchange annually). Buffie assumes that such a policy was
 Table 6. Mexican Big Pushes

Pre-Push construction Stahilixation

Echeverria Big Push 1970 1971 1972 1973 1974 1975 1976 1977

GDP growth (%) 12.9 4.1 8.3 5.9 5.6 3.8 3.6
Intlation (%) 10.9 5.9 13.2 22.1 15.6 19.5 29.9
Real exchange rate 96.9 95.2 90.7 88.3 98.4 101.2 94.9 77.3
(1965 = 100)

Fiscal balance (%GDP) (1.7) (2.5) (4.9) (6.8) (7.2) (10.1) (6.7)
Currentaccount @GDP) (2.9) (2.0) (2.0) (2.6) (4.3) (4.8) g; (1.9)
Debt/GDP rate 15.9 15.7 15.2 16.4 17.7 18.5 2414 32.0

Construction Protractedstabilization

Portillo Big Push 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988

GDP growth (%) 8.2 8.7 7.0 7.9 2.8 (5.0) 4.2 3.8 (4.9) 2.5 2.5
Intlation (%) 16.8 19.5 26.7 26.0 61.0 90.4 59.1 56.7 74.3 139.3 103.8
Real exchange rate 76.4 83.1 97.4 118.6 84.7 76.3 89.0 90.4 58.4 56.6 71.8

Fiscal balance @GDP) (6.7) (7.4) (7.9) (14.7) (17.6) (8.9) (8.7) (10.0) (16.3) (9.9)
Currentaccount @GDP) (3.0) (4.4) (6.6) (7.4) (3.0) 3.1 2.6 (1.1) 23
Debt/GDP rate 33.7 34.2 35.1 36.2 41.4 53.9 37.9 66.6 625

Source: World Bank (1991). Wood (1988), Beristain and Triguerce (1990).
22 WORLD DEVELOPMENT

sustainable indefinitely, but this is not the case lization and liberalization measures. The 1979 oil
because he neglects sectoral (micro) issues. shock was positive for Mexico, and by the equivalent
Mexico’s Stable Development was chara&&ed of an extra 3% of GDP annually (Gelb, 1988).
by rising levels of inflation and industry protection Mexico used its oil as collateral to increase the inflow
and by increasingly severe foreign exchange short- of resources by the equivalent of a extra 2% of GDP
ages (Auty, 1991). Consistent with the resource and invested them in an HCI Big Push. In effect,
curse thesis, its longevity was made possible by the Mexico used its oil resource to avoid the need to
Mexican Green Revolution during 1945-65 make manufacturing internationally competitive
(Reynolds, 1970) which transformed a food trade when agriculture proved no longer equal to the task
deficit into an export surplus, but which was running which it had performed through the 1950s and 1960s.
out of steam by the late 1960s. Land for large farms Consistent with stage one of the three-stage Big
became scarce and the Green Revolution techniques Push model, the Portillo HCI Big Push (1978-82)
performed poorly on small ejido farms. This falter- boosted inflation (Table 6) and halted Mexico’s non
ing ability of the primary sector to underpin industri- oil export growth. Such exports plateaued at 5% of
alization under AIP is consistent with Mahon’s GDP and the manufacturing gain shown in Table 3
(1992) observation concerning the role of primary was underpinned by energy subsidies. By 1982 virtu-
product exports in Latin American development. The ally the entire nonoil economy became nontradeable,
latter raise per capita incomes during the early phas- i.e. in need of total protection or subsidies. The
es of development but as structural change proceeds deployment of the oil windfall in an HCI Big Push
and the primary sector shrinks in relative importance therefore reversed the country’s hesitant progress
(Table 1) AIP’s overly protected manufacturing sec- toward competitive diversification and provides a
tor cannot then adequately substitute for the relative clear-cut example of the resource curse thesis. Only
decline in foreign exchange and tax revenues from when the sharp 1986 fall in oil prices finally con-
the primary sector. Figure 1 visually depicts the styl- vinced the government that market-conforming
ized facts for AIP and CIP. and it adds the mineral reform could no longer be postponed, was AIP aban-
economies (an important example of the market-con- doned. Then change occurred with little evidence of
strained, resource-rich category of country in Figure “cultural” obstacles (Loser and Kalter, 1992).
2) as a more extreme example of the failure to diver-
sify into competitive tradeables. (b) Brazil’s economic miracle
In the case of Mexico (where the importance of
oil from the late. 1970s gave the country some fea- Brazil also used its rich resource base to sustain
tures of a mineral economy) president Echeverria AIP long enough to render reform even more pro-
backtracked on reform efforts and attempted to use a tracted than in Mexico. The partial reforms associat-
public sector boom during 1972-75 to escape the for- ed with the 1967-73 “economic miracle” were over-
eign exchange constraint (see Table 6). His failure taken by an HCI Big Push under Geisel during
led his successor, Portillo, to prepare for AIP reform 1974-79. Kubitschek had earlier launched an HCI
in 1977. But the decision to accelerate the exploita- Big Push during 1956-60 based on steel and auto-
tion of Mexican oil reversed Portillo’s initial stabi- mobiles which, like similar Big Pushes elsewhere,

Table 7. The HCI Big Push sequences, Brazil

Stage Prspush 1 2 3
collsmldon Stabilization Rebound

Brazil 1ateJOspush
Time period 1951-55 1956-60 1%1-67 1968-73
GDP growth (%/yr) 6.7 8.1 4.6 11.2
Iuflation (%/yr) 16.1 21.8 53.7 21.0
Fiscal gap (%/GDP) IL& (3.5) (0.5)
Currentaccount gap (%/GDP) 1.3 &::; (0.3) (2.2)

Brazil late-1978 ph
Time period 1968-73 1974-79 198M4 1985-M
GDP growth (‘??Jy) 11.2 5.5 1.5 4.7
Inflation (%/yr) 21.0 41.4 132.1 317.8
Fiscal gap (%/GDP) (0.5) (1.5) (3.0) (12.4)
Currentaccount gap (%/GDP) (2.2) (4.9) (3.8) (0.3)

Source: World Bank (1991); UN (1970,1976); IBGE (1987).

 RESOURCE CURSE THESIS 23

outstripped domestic implementation capacity and adjustment to the fast oil shock.
triggered the distinctive three-stage macroeconomic Like Mexico, Brazil’s rich resomce base encour-
sequence of boom, stabilization and HCI rebound aged a drift into autarky and was then used to post-
(see Table 7). Brazilian stabilization was delayed by pone AIP reform. Much investment under the Geisel
weak government during 196043, and began only HCI Big Push was made via state firms which
under the reformist military government during declined in efficiency and were increasingly used as
1964-67. The economic miracle (10% annual GDP political and macroeconomic policy tools (with price
growth during 1968-73) represents the delayed HCI freezes to curb inflation and overmanning to lower
Rebound as initial HCI design capacity use rates unemployment). Moreover, AIP incentives favored
were reached and low-cost in situ expansions established firms. rather than new entrants, so that
became feasible. the former had little need to become competitive.
The reforms were partial: protection was still rel- The assisted private firms displayed considerable
atively high, rents remained substantial (Bergsman, skill in retaining their subsidies. There is clear evi-
1970). HCI sectoral maturation rates were orders of dence both from the early 1980s (Auty, forthcoming)
magnitude longer than those for Korea and Taiwan. and late 1950s (Rodr& 1989) that ownership deter-
Exports remained heavily resourcedependent, a fact mined rent-seeking ability. Private domestic firms
which is understated by the figure in Table 3 because were more successful than multinational companies
Brazilian manufactured exports include a large (MNCs) (such as Ford and GM) at defending sunk
resource-processing component (World Bank, 1982). investments who, in turn, did better than the state
Meanwhile, there is evidence of repressed inflation firms.
in the early 1970s and growing income inequality Brazil’s adjustment to the HCI downswing stage
(Baer, 1989; Simonsen, 1988). Reform was far from began in 1979 and lagged in implementing the
consolidated. changes required (Bacha, 1986). As with the
The Geisel government’s response to the 1973 oil Kubitschek Big Push, stabilization was once again
shock reversed the reforms and drew on the coun- protracted. The HCI projects encountered a similarly
try’s large natural resource endowment to launch a difficult start-up to those of the Kubitschek Big Push
second HCI Big Push. Much investment went into as construction costs overran, inflation took off, fis-
capital-intensive resource-based metal and hydro cal and trade gaps widened and markets flagged. The
projects in the Amazon region and the south. Brazil 1985-86 upswing of the Sarney populist boom
relied much more on import substitution to adjust to tapped the delayed HCI rebound, but, consistent with
the negative shock than 11 other semi-industrial&d Sachs’s (1989) prediction it was not sustained and
countries examined by Balassa (1985) which looked the adverse economic consequences of the tardy
more to slower growth or export expansion for maturation of HCI under AIP persisted (see Table 7).

Table 8. Features of competitive and autarkic industrial pohy

Competitive Autarkic

Industrial policy
Market impact Market-conforming Market usurping
Incentives duration Taper4 off Renewed
Economy openness High LOW

Sefiaal targeting Sequenced HCI Most HCI

Macro policy
Overall stance Fhgmatic orthodoxy StrucnualiSt
Budget balance Tight Deficit accumulation
Exchange rate. Competitive OVCH3h?d

Sane policyoutcomes
HCI maturation rate 5-10 years Decades
Non-HCI viability Strong Week
Foreign exchange UncOIL!GthlIed constrained
Turning point Rapid arrival Delayed arrival
Incomedistribution Equitable Skewed
GDP growth rate High Modest
GDP sustainability High Erratic
Inflation Moderate High
24 WORLD DEVFLOPMENT

Brazilian negotiators cited the size of the countries abandoned AIP before the mid-1960s.
Brazilian economy when bargaining to reduce the They did so because their resource base (relatively
onerousness of IMF-backed orthodox adjustment. small domestic market and limited primary product
This carried the implication that, if necessary, Brazil exports) could no longer support the slow-maturing
could “go it alone.” A decade after the end of the HCI sectors under AIP. They quickly adopted two
Geisel HCI construction boom had reversed the key measures: orthodox macroeconomic policies and
“economic miracle’* reforms, Brazil still lacked the CIP (which initially favored export-led light industri-
consensus needed to dismantle its AIP. al growth). They accelerated their HCI Drives in the
1970s as the arrival of the turning point pushed up
real wages. Rapid economic growth was sustained
6. CONCLUSION both before and after the turning point
The four largest NICs, whether market-rich or
Among six large NICs. the countries with the least market- and resource-rich, all pursued AIP for longer
market potential and the poorest resource endowments than the two smallest countries and subsequently had
(Korea and Taiwan) have industrial&d more success- great difficulty in implementing industrial policy
fully than either the market-rich countries (India and reform in the face of entrenched urban groups with a
China) or the market- and resource-rich countries vested interest in rent-seeking behavior. For the
(Mexico and Brazil). This is so, despite the fact that largest NICs. far from the virtuous growth dynamic
both the market-rich countries and especially the mar- of Verdoorn effects, a conflict arose between the key
ket- and resource-rich countries have a wider range of demands of macroeconomic policy (fscal prudence
industrialization options which they derive from their and current account viability) and the needs of the
large country size. Such a counter intuitive outcome slow-maturing HCI (continued transfers of foreign
reflects the resource curse thesis, at the heart of which exchange from the tradeable sector as its size shrank
lie four conclusions about the impact of a favorable in relation to the economy as a whole).
natural resource base on policy choice. These conclu- A recurring problem for the four large NICs in
sions am: the richer the natural resource endowment seeking to resolve these macro-micro tensions was
then, firsf the longer lax macro policies are tolerated; the desire to evade the austerity associated with
second, the less pressure to achieve rapid industrial orthodox reform and to maintain flagging economic
maturation; third, the longer rent-seeking groups are growth via Big Pushes. The latter, like Sachs’s
tolerated (and the more entrenched they become); and (1989) “‘populist booms.” invariably proved counter
fourth, the greater the likelihood of decelerating and productive under AIP. They only exacerbated the
more erratic economic growth. basic problem of AIPz the inability of the emerging
By the mid-1950s. all six NICs had adopted AIPs manufacturing sectors to earn foreign exchange. It
which stressed the creation of HCI. Yet by the late may be no coincidence that Mexico and Brazil, the
1950s the economic problems of AIP were becoming best endowed of the six countries, were the last to
widely appreciated. The contrasts between AIP and embark on AIP reform. Both used their natural
CIP are summarized in Table 8. The key flaw in AIP resources (oil in Mexico, minerals and hydro in
is that as the sixe of the protected industrial sector Brazil) to delay reform further in the 1970s while
grows in relation to the competitive tradeables, so India and China (for whom HCI-led growth involved
the failure to generate adequate foreign exchange a much larger sacrifice in terms of consumption fore-
(due to the leapfrogging of stage two of the East gone) began to shift toward CIP.
Asian development model) adversely affects macro Two important qualifications are in order. First,
performance. Figure 1 contrasts the size of the trade- the focus upon a single causal factor necessarily
ables sector under AIP and CIP and shows how AIP understates the role of other significant factors such as
overburdens the primary sector as the latter experi- the East Asian stress on education and early progress
ences relative decline with rising per capita incomes. toward land redistribution. Second, the resource curse
Figure 1 also includes another important group of the& is not a law, merely a strong tendency, so that
developing countries, the mineral economies, for exceptions are likely while, in addition, prudent poli-
which the resource curse thesis is examined else- cy can avoid the pitfalls. Lagging countries can absorb
where (Gelb, 1988; Auty, 1993). In Figure 2. most the lessons of the postwar decades and escape the
mineral economies would cluster with Argentina apparent determinism of the resource curse thesis.
toward the bottom right. Nevertheless, the res- ounce curse thesis appears to be
The two most resource-deficient and market-poor a relatively robust speculation.
 RESOURCE CURSE THESIS 25

REFERENCES

Am&n, A., “The state and Taiwan’s economic develop Gelb, A., Gil Windfallr: Blessing or Curse? (New York:
merit,” in P. B. Evans (Ed.), Bringbzg the State Back In Oxford University Press, 1988).
(Cambridge: Cambridge University Press, 1985), pp. Gelb, A. H., J. B. Knight and R. H. Sabot, “Lewis through a
78-106. looking-glass: Public sector employment, rent-seeking
Amsden, A., Asia’s Next Giant (New York: Oxford and economic growth,” Mimeo (Cambridge MA:
University Press, 1989). Employment and Enterprise Policy Analysis Project,
Andrain, C. F., Political Change in the Third World (Boston HIID, 1986).
MA: Unwin Hyman, 1988). Ginsburg, N. S.. “Natural resourm and economic develop
Auty, R. M., &my Indus@ Policy and Economic merit,” Annals of the Association of American
Performance in Five Large Newly I&ustriaBzinx Geo~Mphen,
- _ Vol. 47 (1957), pp. 197-212.
Co;ntries: Korea, Brazil, kxico, india and China Hughes, H., Achieving Indus%alizarion in East Asia
(London: Mansell, forthcoming). (Camlxidae: Cambridge University Press. 1988).
Auty, R. M., Sustaining Developme& in Mine& Economies: IBGE, Contebilidadae S&al (Rio de-Janeiro: IBGE, 1987).
The Resource Curse Thesis (London: Routledge, 1993). Kumets, P. W., “An East Asian model of economic develop-
Auty, R. M., “The macro impacts of Korea’s heavy industry ment: Japan, Taiwan and Korea,” Economic Development
drive re-evaluatfl Journal of Development Studies, Vol. andStructural Change, Vol. 36 (1988), pp. 1143.
29 (1992a), pp. 24-48. Lal, D., The Poverty of Development Economics (London:
Auty, R. M., “A comparison of the industrial policies of IEA, 1983).
Taiwan and Korea,” Working Paper No. 8 (Lancaster, UK: Loser, C., and E. Kalter, ‘Mexico: The strategy to achieve
Lancaster University Geography Department, 1992b). economic growth, ” Occasional Paper, No. 99
Auty, R. M., “RBI through boom, downswing andlie- (Washington, DC: IMF, 1992).
tion: Mexico,” Energy Policy, Vol. 19 (1991), pp. 13-23. Lucas, R. E. B., “India’s industrial policy,” in R. E. B. Lucas
Bacha, E. L., “External shocks and growth prospects: Case of and G. F. Papanek (Eds.), The InaXun Economy (Boulder
Brazil 1973-79,” World Development, Vol. 14,1986, pp. CO: Westview).
919-936. Mahon, J. E., “Was Latin America too rich to prosper?
Baer, W., The Bra&an Economy: Growth andDevelopment Structural and political c&a&s to export-lad industrial
(New York: Praeger, 1989). growth,” Joumal ofDevelopment Studies, Vol. 28 (1992).
Balassa, B., “‘Ihe lessons of East Asian development: An lqx 241-263.
overview,” Economic Development and Cultural Change, Morishima, M., Why Has Japan Succeeded? Western
Vol. 36 (Supplement), 1988, pp. S273-S290. Technology and Japanese Ethos (Cambridge: Cambridge
Balassa, B., “Adjusting to external shocks: Ihe newly indus- University Press, 1982).
trializing developing economies 1974-76 and 197%81,” Mundle, S., “Food, finance and foreign trade: ‘Ihe limits of
Weltwirtschaftesliches Archiv, Vol. 122 (1985), pp. high growth in India,” Mime0 (Cambridge: Cambridge
1141-1161. University, 1990).
Bell, M., et (II., “Assessing the performance of infant indus- Murphy, K. M. et al., “Industrialization and the Big Push,”
tries,” Journal of Development Economics, Vol. 16 Journal of Political Economy, Vol. 97 (1989), pp.
(1984), pp. 101-128. 1003-1026.
Belle, W., and S. Rosenfeld, Dragons in Dictress (London: Myint, H., The Economics of the Developing Countries
Penguin, 1992). (London: Hutchinson, 1964).
Bergsman, J., Brazil: Industtilizatiac and T&e Policie.r National Statistics Office, Major Statistics of the Korean
(Oxford: Oxford University Press, 1970). Economy (Seoul: National Statistics Oflice, 1992).
Beristain, J., and I. Trigueros, “Mexico,” in J. Williamson Ohno, K., and H. Imaoka, “The experience of dual industrial
(Ed.), Latin American Adjustment (Washington, DC: growth: Taiwan and Korea,” The Developing Economies,
International Institutefor Economics, 1990), pp. 156177. Vol. 35 (1987), pp. 31&324.
Buffie, E., and A. S. Krause, “Mexico 1958-86horn stabil- Perkins, D., “Refaming China’s economic system,” Journal
ising development to debt crisis,” in J. D. Sachs (Ed.), of Economic Literature, Vol. 26 (1988). pp. 601-645.
De&loping -Counrry Debt and the World Ecoh&y Perkins, D., and M. Syquin, “Large-coun&: The influence
(Chicago, II: University of Chicago Press, 1989), pp. of sixe,” in H. B. Chenery, H. B. and T. N. Srinivasan
141-168. (Eds.), Han&ook of Development Economics (New York:
Chenety, H. B. et al., Industrialization and Growth @Jew North-Holland, 1989), pp. 1691-1753.
York: Oxford University Press, 1986). Prybyla, J. S., Market and P&m Under Socialism (Stanford
Chilcote, R. H., “Dependency: A critical synthesis of the lit- CA: Hoover Institution Press, 1987).
erature,” Latin American Perspectives, Vol. 1, 1974, pp. Ranis, G., and S. Mahmood, The Polirical Economy of
4-29. Development Policy Change (Oxfad: Blackwell, 1992).
Council for Economic Planning and Development, Reynolds, C. W., l7z Mexican Economy: Twentieth Century
Taiwan Statistical Data Book 1991 flair&: . . CEPD. Structure and Growth (New Haven, CT: Yale University
1992). Press, 1970).
DGBAS, Stat&ical Abstract of National Income, Taimzn Rodrik, D., “Rent-seeking among Brazilian firms,” Mime0
Area, Republic of China (1951-91) (Taipei: DGBAS, (Williamrtown, MA: Williams College, 1989).
1992). Sachs, J. D., “Social conflict and populist policies in Latin
Elkan, W., Introduction to Development Economics America,” Working Paper, No. 2897 (Cambridge, MA:
(London: Penguin, 1973). National Bureau of Economic Research, 1989).
26 WORLDDEVELOPMENT

Sachs, J. D., “External debt xnd --nomic perfa- United Nations, UN Stahtical Yeaho& (New Yak: UN,
mancein LatinAnmica and hi&” Brookinxs Pmrs on various Yearsb
Economic Activity, No. 2 (1985), pp. 523-575. - Wade, R,~G&ming the Market (Princeton, NJ: Princeton
Simsonsen. M. II.. “Brazil.” in R Dombuh.xndF.LC.H. university Press, 1990).
Helmeri (Ed.%),The Ojen Economy (New Yak: Oxfad Wood, A., “Global trends in real exchange rates, 1960 to
University Press, 1988), pp. 285-306. 1984,” Discuhm Paper No. 35 (Washington, DC: World
Stem, J. J., “Kaea’s industrial policy and changing industri- Bxnk, 1988).
al structure,” HHD Development Discussion Pxpez, No. Wald Bxnk, World Development Report I992 (New Yak
343 (Cambridge, MA: HJJD, 1990). Oxfad University Pre.ss, 1992).
Summers, L. H., and W. B. hterly, “Culture is not to World Bank, World Bank Tubh 1991 (Baltimore, MD:
blame,” Pi& Times (April 15,1992). Johns Hopkins University Press, 199 1).
Syrquin, M., and H. B. Chenery, “Patterns of development World BanL, Brazil: Industrial Policia and Manujkciured
1950-83,” Discussion Paper No. 41 (Washington, DC: Exports (Wrphington, DC: World Bxnk, 1982).
World Bank, 1989). WRI, World Ruourcw 1992-93 (Oxfad: World Resources
Tsiang, S. C., “Tsiwxn’s economic miracle: Lcssom in eco- InstituteIOxfad University Press, 1992).
nomic development,” in A. C. Harberger (Ed.), World Wu, Q., “South Kaea and Taiwan: A comparative analysis
Economic Growth (San Francisco CA: ICS Press, 1984), of economic development,” Discussion Paper No. 252
pp. 301-331. (Brighton: IDS, 1988).