Page i →

The Taiwan Voter

The Taiwan Voter examines the critical role that ethnic and national identities
play in politics, illustrated by the case of Taiwan. That country’s elections often
raise international tensions, and they have sometimes led to military
demonstrations by China, as in the 1995–96 Taiwan Strait Crisis. Yet no
scholarly books have examined the ways in which Taiwan’s voters make their
electoral choices in such a dangerous environment. Critiquing the conventional
interpretation of politics as an ideological battle between liberals and
conservatives, The Taiwan Voter demonstrates that in Taiwan the party system
and the voters’ response to it are instead shaped by one powerful determinant
of national identity—the China factor. The book also takes up Taiwan’s voter
turnout, “pocketbook voting,” and the effects of the new electoral system
adopted in 2004.

Taiwan’s electoral politics draws international scholarly interest because of the

prominence of ethnic and national identification in its politics. Of course,
identities matter almost everywhere. In most countries, though, the many
tangled strands of competing identities present a daunting challenge for
scholarly analysis. Taiwan, by contrast, is a country where the cleavages are
both powerful and limited in number, so that the logic of the interrelationships
among issues, partisanship, and identity are particularly clear. In this book,
Christopher H. Achen and T. Y. Wang bring together experts on Taiwan to
investigate the ways in which social identities, policy views, and partisan
preferences intersect and influence each other. These novel findings have wide
applicability to other countries, and thus they will be of interest to a broad
range of social scientists interested in identity politics.

Christopher H. Achen is Professor of Politics and Roger Williams Straus

Professor of Social Sciences at Princeton University.

T. Y. Wang is Professor of Politics and Government at Illinois State

University.
Page ii → Page iii →

The Taiwan Voter

Christopher H. Achen and T. Y. Wang, editors

University of Michigan Press

Ann Arbor
Page iv → Copyright © 2017 by Christopher H. Achen and T. Y. Wang

All rights reserved

This book may not be reproduced, in whole or in part, including illustrations,

in any form (beyond that copying permitted by Sections 107 and 108 of the U.S.
Copyright Law and except by reviewers for the public press), without written
permission from the publisher.

Published in the United States of America by the

University of Michigan Press

Manufactured in the United States of America

A CIP catalog record for this book is available from the British Library.

Library of Congress Cataloging-in-Publication data has been applied for.

ISBN: 978-0-472-07353-5 (hardcover : alk paper)

ISBN: 978-0-472-05353-7 (paperback : alk paper)

ISBN: 978-0-472-12303-2 (ebook)
ISBN: 978-0-472-12303-2 (Open Access)

doi: http://dx.doi.org/10.3998/ mpub.9375036

The Open Access edition was funded by the Top University Project of
National Chengchi University, made possible by a grant from Taiwan’s Ministry
of Education.
Page v → Page vi →

Contents
Preface

Chapter 1. The Taiwan Voter

Chapter 2. Who Is the Taiwan Voter?

Chapter 3. Changing Boundaries

Chapter 4. Parties, Partisans, and Independents in Taiwan

Chapter 5. Issues, Political Cleavages, and Party Competition in Taiwan

Chapter 6. Economic Voting in Taiwan

Chapter 7. Cross-Strait Relations and the Taiwan Voter

Chapter 8. Evaluation of Presidential Candidates’ Personal Traits

Chapter 9. Political Left and Right in Taiwan

Chapter 10. Electoral System Change and Its Effects on the Party System
in Taiwan

Chapter 11. Political Participation in Taiwan

Chapter 12. Conclusion: The Power of Identity in Taiwan

Contributors

Index
Page vii →

Preface

This book began at a conference in Taipei, Taiwan, on March 24, 2013. The
Election Study Center of National Chengchi University hosted all the authors
for an intense discussion of how a book about the Taiwan voter might be
written. (A few contributors were “virtual attendees” via an Internet link.)
Preliminary versions of chapters were presented and thoroughly discussed.
After additional reviews and revisions, this book was accepted for publication
by the University of Michigan Press.

The Election Study Center has provided extensive administrative and

financial support to this project from the beginning. The book would have
been impossible without the research excellence and professional colleagueship
of the members of the Center. Their warm hospitality to each editor on
multiple occasions has been a highlight of our professional lives.

A great many people and institutions also helped make The Taiwan Voter
possible. T. Y. Wang’s research sabbatical was funded by the Chiang Ching-kuo
Foundation and Taiwan’s National Science Council (now called the Ministry of
Science and Technology). He also received a Summer Faculty Fellowship from
Illinois State University to support this project. A grant to Wang from the
Taiwan Foundation for Democracy paid for editing and other expenses of
preparing the book for publication review. The Open Access edition was
funded by the Top University Project of National Chengchi University, made
possible by a grant from Taiwan’s Ministry of Education. Princeton University’s
Politics Department and Illinois State University’s Department of Politics and
Government also helped with logistical support and research funding.

Larry Bartels, during his time as director of Princeton’s Center for the Study
of Democratic Politics, hosted a small conference that helped build a
foundation for this book. Michele Epstein and Helene Wood provided the Page

viii → administrative support to make the conference successful. We thank all

three of them.

Achen remembers with respect and gratitude his former colleague Tang Tsou
of the University of Chicago Political Science Department, now deceased, who
arranged a first visit to Taiwan. The trip was paid for by Taiwan’s Government
Information Office, which was then directed by Dr. Shao Yu-ming. That eye-
opening initial encounter with the people of the island led to the professional
relationships that have made possible Achen’s participation in this project.

The anonymous reviewers for the University of Michigan Press gave us two
rounds of thoughtful comments and criticism. The book is much better for
their efforts. Gail Schmitt and Carissa L. Tudor provided editing assistance,
and we are in their debt as well. We also thank our editors at the University of
Michigan Press, Melody Herr and Mary Francis, supported by Danielle Coty,
for their enthusiasm about our project and for their encouragement and
guidance through the publication process. We have also very much appreciated
the unwavering support of the series editor, Mic Laver.
 To analyze Taiwan voters’ political attitudes and electoral behavior, the
contributors to this volume have utilized multiple waves of survey data,
aggregate electoral data, and information gathered through focus group
interviews, all collected on the island during the past two decades. Most of the
data used in this book are taken from Taiwan’s Election and Democratization
Study (TEDS). This long-running series of representative national samples of
Taiwan voters is the gold standard for academic election studies in Asia. The
coordinator of the multiyear TEDS project is Chi Huang of the Election Study
Center and the Political Science Department at National Chengchi University,
Taipei, and the data are managed and distributed by the Election Study Center.
More information is available on the TEDS website (http://www.tedsnet.org).

Other data sources for the book include the Taiwan National Security Survey
(TNSS), sponsored by the Program in Asian Security Studies under the
directorship of Emerson Niou of Duke University, and also a number of
individual surveys conducted by the faculty of the Election Study Center, many
of whom are contributors to this volume. Focus group interviews conducted
by the faculty and staff of the Election Study Center also appear in subsequent
chapters. Su-feng Cheng is the principal investigator for those projects.

The Taiwan Voter would have been impossible without all these data, and we
very much appreciate the assistance that all these individuals have provided us.

Page ix → We also owe thanks to a great many students and colleagues for
advice and encouragement along the way. Achen is particularly grateful to the
other members of the “Gang of Four dinner group,” Da-chi Liao, Pei-shan Lee,
and Vincent Wang, for their support and inspiration. Wang extends his
gratitude to Ali Riaz for his encouragement and friendship through the years.
 We are grateful to all those who have helped us. However, our interpretations
of the TEDS, the TNSS, and other data are our own. Each author in this
volume takes sole responsibility for the remaining errors and misjudgments in
what he or she has written.

Most important, the editors thank our wives, Tena Achen and Christine Lee,
for their patience with our overseas absences and their tolerance of our
frequent lengthy work days while we finished editing this book. Our gratitude
for their love goes beyond anything words can say.
Page x → Page 1 →
 Chapter 1

The Taiwan Voter

An Introduction

Christopher H. Achen and T. Y. Wang

Taiwan’s recent history is a remarkable saga. During the two decades from 1970
to 1990, Taiwan underwent dramatic economic change, as its gross domestic
product grew at an average rate of 9 percent per year.1 The economic success
rapidly propelled Taiwan into the ranks of the newly industrialized countries.
Democratization arrived in the late 1980s, too, with robust electoral
competition between the two principal parties. The long-ruling, formerly
authoritarian party was beaten at the polls in 2000 and handed over power
peacefully, only to return in 2008 in an equally peaceful transition when its
opponents lost. A third peaceful transition between parties took place in 2016.
An impoverished Asian one-party authoritarian state had become in a few
decades a prosperous, vibrant democracy.

The Taiwan success story has generated considerable scholarly interest. The
initial studies focused on economic growth—the “Taiwan miracle” (e.g., Chan
1988; Clark 1987, 1989; Wang 2000). In the ensuing decades, scholarly research
shifted its focus to Taiwan’s political miracle, first the democratization and
then the establishment of stable party competition (for example, the chapters
collected in Tien 1996). As time has passed, however, Taiwan’s domestic
political economy has become a more typical example of the stresses, debates,
and achievements of a rich democracy. Economic growth and employment,
energy policy, the environment, public works, and many other familiar topics
are frequent subjects of debate in Taiwan. However, those concerns are not
unique to the island.

Taiwan’s uniqueness lies elsewhere—in its relationship with China. No Page 2

→ other topic or relationship plays so central a role in Taiwan’s politics. It

structures foreign policy; it structures the political party system; it structures
much of how ordinary citizens orient themselves to politics. The fundamental
role of “the China factor” will be explicitly or implicitly discussed in many
chapters of this book.

In The Taiwan Voter, we focus on ordinary citizens’ political preferences,

attitudes, and choices since the onset of democratization. To study the causes
of voting, we take up the conventional “big three”—party identification, issue
orientation, and candidate evaluation—since there is a general consensus in
political science literature that they exert influential effects on voting (Jacoby
2010). But to those three, we add a fourth—identity. In particular, we look
closely at national identity, since it looms so large in Taiwan voting.

Though our concerns are primarily with citizens rather than with elites and
institutions, we take note of institutional features of Taiwan democracy that
shape the choices presented to its citizens. One chapter is specifically devoted
to the recent change in the electoral rules that seems to have put Taiwan on its
way to a classic two-party system.2
 In this volume, we are reporting on one country, but our focus extends well
beyond it. The study of Taiwan politics leads rapidly to interesting, sometimes
difficult, theoretical puzzles. Why have citizen identities, usually taken as
relatively fixed features of people’s political lives, evolved so rapidly in Taiwan,
while simultaneously one particular identity cleavage centered on “the China
factor” has become increasingly consolidated as the most important political
division in the country’s politics? In the presence of this cleavage, how do
Taiwan citizens3 make their electoral decisions? We intend to go beyond the
usual country study to ask questions like these. Implicitly, and sometimes
explicitly, this book suggests comparisons with other countries that share one
or another of these unusual features of Taiwan’s politics, such as Japan, Canada,
Ireland, and Israel. More generally, our findings have implications for every
country in which national identities have large effects in electoral politics.
Thus, we attempt to contribute not only to making sense of Taiwan, but more
broadly to the theoretical understanding of democratic elections in general.

Using single-country studies to generate theoretical understanding has a long

tradition in political science, and in electoral research in particular. The
Columbia studies under Paul Lazarsfeld demonstrated the powerful role of
group memberships and loyalties in American electoral decisions (Lazarsfeld,
Berelson, and Gaudet 1948; Berelson, Lazarsfeld, and McPhee 1954). Angus
Campbell and his Michigan colleagues’ seminal work, The American Voter,
showed how partisan identity shaped not only the vote but also the Page 3

→ very way that citizens perceive the political world and form their political
opinions (Campbell et al. 1960). David Butler and Donald Stokes’s Political
Change in Britain (1974) traced class conflict in politics, the parental
transmission of partisanship, and the importance of local context in
determining vote choices. A great many other articles and books, far too
numerous to list here, have been devoted to such topics as voter turnout,
economic voting, electoral institutions, and strategic voting in particular
democratic systems around the world. No one would say of these contributions
that they taught us only about the country in which they were carried out.

This point is particularly relevant in East and Southeast Asia, where electoral
studies are few and our understanding is thin. It seems to us critically
important to build a knowledge base for each democracy of the region. At
present, no one could write a competent comparative volume called The Asian
Voter: when the foundation stones are missing, no structure can be erected.

The problem of too little Asian country knowledge is also visible in some
recent important contributions to comparative electoral behavior. For example,
Thomassen (2005) compares what is known about electoral behavior across the
European democracies. Evans and De Graaf (2013) study how class voting
varies across Europe and the Anglo-Saxon countries. Carlin, Singer, and
Zechmeister (2015) use the Latin American Public Opinion Project data at
Vanderbilt University, among other sources, to show how partisanship,
ideology, and economic factors have different force in different countries
across Latin America. We are impressed by the depth of scholarship amassed by
the many contributors to these edited volumes. But we are equally impressed
by how often Taiwan fails to fit the conceptual frameworks they put forward.
Of course, none of these three books deals with Asia; their agenda lies
elsewhere. But even when Taiwan is explicitly included in the set of countries
studied, as in Dalton and Anderson’s (2011) comparative investigation of how
the institutional and party-system features of each country create a context that
shapes how citizens think and act in electoral politics, the fit is odd. Too often,
Taiwan is assimilated conceptually to countries with a European-derived
culture that it does not share. We go deeper into this topic in chapter 9 and in
the concluding chapter 12. For the present, it suffices to say that we make no
apologies for focusing on Taiwan. Getting each country right is a prerequisite
to reliable comparisons, both within continents and regions and across them.

Thus we intend this volume to join the ranks of similar books on political
behavior, such as The Japanese Voter (Flanagan et al. 1991) and The Irish Voter
(Marsh et al. 2008). Like them, this book studies one country, but it does Page 4

→ so as part of an international effort to understand electoral behavior in a

variety of democratic systems. While we draw primarily on Taiwan’s politics as
observed in the multiyear nationwide surveys of the Taiwan’s Election and
Democratization Study (TEDS) and the Taiwan National Security Survey, we
aim to speak to those interested in elections everywhere.

The Relationship to China

Taiwan at its closest point is only a little more than 100 miles from the Chinese
mainland. Like any small polity next door to a powerful country, Taiwan
necessarily pays careful, even obsessive attention to its neighbor. Indeed,
Taiwan’s politics revolves around the relationship with China. For Taiwan’s
citizens, the relationship to the mainland is multifaceted. China represents
simultaneously a cultural heritage, a security threat, and an economic
opportunity.

Cultural Heritage
Taiwan’s cultural inheritance from the Chinese mainland is undeniable.
Although its first settlers seem to have come from what is now the Philippines
and Indonesia, Taiwan’s modern history can be traced to the mid-seventeenth
century, when residents from the coastal areas of China migrated to the island
to escape from the war and devastation on the mainland. In that same period,
the Portuguese arrived on the island, which they called “Formosa”
(“beautiful”). The Dutch and Spanish also founded settlements. In the Tainan
area, the oldest inhabited part of Taiwan, a temple still stands, dedicated to two
gods with unmistakable white features. By 1684, however, China’s Qing dynasty
had driven out the Europeans and established a local government on the
island. The Qing maintained at least a nominal administrative relationship with
the island for the subsequent two-plus centuries. Like many dynasties before it,
the Qing dynasty eventually collapsed under the weight of corruption and
administrative failure. During the collapse, and in the wake of the First Sino-
Japanese War, Taiwan’s sovereignty was ceded to Japan in 1895, beginning
another colonial experience, this one lasting half a century.

Although the initial stage of Japanese rule was marked by the island residents’
armed resistance, insurrection was soon quelled by the colonial authority’s
adroit combination of repression, co-optation, and assimilation. Page 5 → In

some respects, the Japanese treated Taiwan as their own province, and they
began its modernization. Certainly Taiwan was treated better than most other
countries conquered by Japan just before and during World War II. During the
colonial period, some Taiwan people adopted Japanese names and became
naturalized Japanese citizens.4 Many young people were drafted and served in
the Japanese military during the war.
 At the end of the war, Japan unconditionally surrendered to the Allied forces.
Taiwan’s sovereignty was returned to China, then led by the Nationalist
(Kuomintang or KMT) government under Chiang Kai-shek. Chiang and two
million of his mainland followers retreated to the island in 1949 after being
defeated by the Communist troops on the Chinese mainland. To maintain their
claim that they remained the sole legitimate rulers of the mainland, KMT
leaders preached “recovering the Chinese mainland” as the sacred national
mission. The eventual unification of Taiwan and China was taken as a given.

The Chiang regime also took a series of measures to “re-Sinicize” local

residents in order to foster a Chinese identity. A China-centered curriculum
was established in schools. Maps of the nation showed all of China, including
the mainland, Taiwan, and all areas claimed by China, such as Tibet.
Schoolchildren were taught in Mandarin, and their use of local dialects was
punished. Ethnic television and radio programs were restricted as well. These
measures, combined with the shared culture and languages stemming from the
ancestral homeland, led many Taiwan citizens to view China as the principal
source of their racial and cultural heritage. In some cases, they saw Taiwan and
China as a political unity as well. In one or another sense, they identified as
“Chinese” (Wang and Liu 2004). This ensemble of affective and historical ties
continues for many Taiwan people to the present day, though increasingly in
attenuated form.

Security Threat
The Chiang regime generally encountered a warm welcome from island
residents after the KMT forces first arrived on Taiwan in 1945. But the initial
enthusiasm of many local residents for returning to the ancestral “motherland”
was substantially dampened when they saw the mainland troops sent to take
control of the island. Impoverished and poorly educated, many of the soldiers
were seen as beggars and thieves, less disciplined and capable than the Japanese
they replaced. Nationalist officials in turn viewed the islanders with suspicion
due to the half century of Japanese colonial rule. By 1947, Page 6 → the

animosity between the KMT government and local residents culminated in the
bloody crackdown against Taiwanese elites by Chiang’s troops, a tragic event
known as the “2–28 Incident” (Kerr 1965; Lai, Myers, and Wei 1991). This
outbreak of hostility solidified the perception, especially in the older
generation, of the KMT government as simply a new foreign regime and
occupying force, this time from China rather than from Portugal or Japan.

While the Chiang regime actively advanced its claim to be the legitimate ruler
of China, Beijing’s leaders made the same claim, and they attempted to forcibly
unify Taiwan with the Chinese mainland. During the 1950s and 1960s, several
major battles were fought over Kinmen (also known as Quemoy) and Matsu,
two small islands near the mainland but occupied by Taiwan. Although the
1970s saw a shift of Beijing’s strategy away from a reliance on the “military
liberation” of Taiwan to a wave of “peaceful initiatives,” Chinese leaders have
refused to renounce the use of force to resolve cross-Strait disputes. To force
Taipei into acceptance of its unification formula, known as “one country, two
systems,” Beijing has isolated Taiwan internationally and has backed its claim
to the island with the threat of military action.
 As China increasingly won the diplomatic battle on the world stage, Taiwan
went through a series of political changes toward democratization since the late
1980s. These included lifting martial law, legalizing political parties, ending
restrictions on public assembly and freedom of speech, and the popular
election of the president. The pace of democratic reform quickened after the
pro-independence Democratic Progressive Party (DPP) was formed in 1986
and became the island’s first major opposition party. An atmosphere of political
tolerance emerged on the island as opinions different from the “one China”
principle were permitted. Blaming President Lee Teng-hui for condoning a
Taiwan independence movement on the island, Chinese leaders launched
several missiles into the seas around Taiwan in an attempt to influence, first,
the 1995 Legislative Yuan election, and then Taiwan’s 1996 presidential election.
The misguided effort backfired, with most Taiwan observers estimating that
Lee’s successful reelection effort was aided by the Chinese saber-rattling
(Garver 1997; Cooper 1998, chap. 4).

During the 2000 presidential election campaign, Beijing leaders employed the
tactic of a “paper missile,” a White Paper on cross-Strait relations that
promised military intervention if Taiwan moved toward independence. This
attempt to disrupt the momentum of the pro-independence candidate, Chen
Shui-bian of the DPP, probably backfired again. Contrary to China’s
expectations, these threats provided Chen with a late boost and a narrow
margin of victory. Chen’s election was a milestone in Taiwan’s political
development, Page 7 → and it illumined the sharp contrast with the mainland’s
political system, as political power was peacefully transferred from one political
party to the other on the island for the first time and indeed on any territory
where ethnic Chinese rule. While Beijing has subsequently moderated its
unsophisticated and ineffective attempts at intimidating Taiwan voters (Tung
2005), especially after the China-friendly Ma Ying-jeou of the KMT was elected
president in 2008, it has continued to aim more than a thousand short- and
medium-range missiles at the island. As many in the older generation saw the
authoritarian KMT as an illegitimate Chinese occupation force, younger
generations now often see China under Communist Party rule as an
authoritarian and less developed country with an aggressive military posture. In
consequence, much of the Taiwan population views China as a threat to their
economic prosperity and democratic way of life.

Economic Opportunity

Perhaps surprisingly, cross-Strait economic exchanges have flourished during

the past two and a half decades despite tense political and sometime military
relations between Beijing and Taipei. Since the ban on contacts between
Taiwan and China was lifted by the Taipei government in 1987, economic
exchanges between two sides of the Taiwan Strait have increased dramatically.
Cross-Strait trade rose from $3.9 billion in 1989 to $31.2 billion in 2000, and
further to $102.3 billion in 2007, despite the restrictive policies imposed by
Taiwan’s pro-independence Chen administration from 2000 to 2008. Taiwan
investment in the Chinese mainland also increased, from $421 million in 1991–
92, to $2.6 billion in 2000, and to about $10 billion in 2007.5 Along with these
investments, many Taiwanese investors have now relocated to the Chinese
mainland.
 The trade and investment flow seems unlikely to be slowed by political
intervention in the near future. Reversing its predecessor, the Taipei
administration under Ma relaxed trade restrictions. Since 2008, the Taipei
government has reached more than a dozen accords with Beijing and has also
signed the landmark trade deal known as the Economic Cooperation
Framework Agreement (ECFA).6 Nor is trade liberalization likely to be
substantially slowed now that the opposition has again taken power in 2016.
The pro-independence DPP has always had concerns about Taiwan’s increasing
economic integration with the Chinese mainland, worrying that expanding
cross-Strait exchanges would increase the island’s dependence on China and
endanger its national security. Yet the DPP leadership increasingly Page 8 →
recognizes the importance of the Chinese market to Taiwan’s economic growth
and prosperity, as demonstrated by the party’s ambivalent campaign platform
on cross-Strait relations during the 2012 and 2016 presidential elections and by
the frequent visits to the Chinese mainland by key DPP politicians.7 Even for
the most anti-China politicians and citizens on Taiwan, the trade and
investment opportunities across the Strait in the world’s most populous
country are too lucrative to ignore. Well-armed, sworn enemies can still do
business.

Taiwan thus has a complex and ambivalent relationship with China that is
characterized by cultural affinity, security menace, and opportunities for
economic prosperity. Precisely because cross-Strait ties are close but unsettled,
they have formed the basis of the key political cleavage on the island that has
effects on every aspect of Taiwan’s politics. In particular, this multifaceted
relationship forces island citizens to decide whether they are Chinese,
Taiwanese, or both. It also forces them to take a stand on Taiwan’s future
relations with China—the issue of “unification vs. independence”—also
known as the tongdu issue. In turn, both these decisions are deeply implicated in
voters’ partisan identifications.

The Central Political Cleavage in Taiwan

Taiwan is an immigrant society within the Chinese diaspora. The shared

language, culture, and ancestral homeland help shape the identities of Taiwan
citizens. For a small number, the Chinese identity may preclude any national
identification with Taiwan: they follow the KMT’s traditional one-country view
of China, of which Taiwan is a province. For these citizens, “I’m Chinese”
means a great deal, including adherence to the position that there is only one
China, that Taiwan is part of China, and that Chinese citizens are subject to the
authority of the legitimate rulers of China (who are in Taipei). This view is
essentially consistent with Beijing’s official position, the sole disagreement
being over the location of the legitimate all-China government.

At the other extreme, a Taiwan citizen may say, “I’m Chinese,” meaning no
more than an Irish-American does when she says proudly on St. Patrick’s Day,
“I’m Irish,” even though her ancestors have been in North America for nearly
200 years and her sole national loyalty is to the United States. She would think
it bizarre if Ireland claimed sovereignty over her. In short, among different
citizens of immigrant lands like Taiwan or the United States, identification
with the country of ancestral origin ranges from deep to superficial, and in
some cases may be equivalent to outright rejection. Page 9 → Indeed, the Allied
war on the Axis powers in Europe during World War II was led by a German
American, Dwight Eisenhower.
 Identities may be central to personality, but even the deepest remain
potentially malleable. Unexpected events or changes in an ancestral country’s
behavior can transform the complex of emotion and historical memory that
constitutes national identity. The transformation can go in either direction, as
Taiwan’s history illustrates. A pan-Taiwanese identity first emerged after the
Qing government ceded the island’s sovereignty to Japan (Ching 2001). This
development was soon impeded by the Japanese colonial authority’s
comprehensive assimilation policies (Brown 2004; Chu and Lin 2003). Then,
after Taiwan’s “glorious return” to Chinese rule might have provided an
opportunity for the consolidation of a Chinese identity, instead the 2–28
Incident created a deep divide, particularly between the resident Minnan and
the newly arrived mainlanders, often creating parallel differences in national
and political identity.8

As chapter 3 explains, the Minnan/mainlander divide became increasingly

blurry with the progression of time. Through politicians’ conscious efforts,
social contacts among various ethnic groups, interethnic marriages, and the
spread of education, understanding, and respect among the various ethnic
groups have been improved and ethnic conflict reduced. As a new generation
of islanders has matured into adulthood, many have become relatively
indifferent to the historical memories that had divided their communities in
the past.

Simultaneously, many Taiwan citizens have been increasingly frustrated by the

international isolation imposed by the Chinese government. Beijing’s
international application of its “one China principle” has significantly
compressed the island country’s international space. Chinese leaders
persistently maintain that the Republic of China, the official name of Taiwan,
has lost its legitimacy and that Taipei has no legal right to establish diplomatic
relations with foreign governments or to participate in any international
organizations with statehood as a membership requirement. The growing
importance of China in international affairs has led many countries to break
relations with Taipei as a prior condition for establishing official ties with
Beijing. The number of nations having official relations with Taiwan dropped
significantly in the 1970s and 1980s. Throughout the subsequent two decades,
only about 30 countries, most of them small, diplomatically recognized Taiwan,
and Taipei had membership in just a handful of international governmental
organizations.

Taiwan’s increasing diplomatic isolation coincided with a period of rapid

economic growth and democratization on the island, leaving many in Taiwan
with the sense that their lack of standing in the world community was Page 10

→ discordant with their economic and political achievements. China was seen
not only as a hypothetical military threat but also as a chronic irritant—an alien
economic and political force opposed to their interests and self-respect. The
Chinese missile tests during 1995–96 and military threats in 2000, both
intended to intimidate Taiwan, raised the issue forcefully: Surely, Taiwan
people said, no self-respecting country would treat its own citizens that way.
Increasingly, the island’s citizens began to question the Chinese aspect of their
identities. Were they still Chinese, or had they now become just Taiwan people?

Taiwanese identity rose in parallel with sympathy for Taiwan independence.

The two attitudes reinforced each other, with powerful implications for
Taiwan’s politics. The previously unquestionable “sacred mission” of unifying
Taiwan with the Chinese mainland during the island’s authoritarian era was
now reconsidered in the minds of many island citizens. An increasing number
of the island residents began to view Taiwan as a separate and independent
political entity, not part of China (Wang and Liu 2004). The way was prepared
for the pro-independence DPP to become what it had never been previously, a
serious contender for political power. The precise causal relationships are not
easily sorted out, but the association was clear and powerful: as Taiwan people
increasingly abandoned their identity as Chinese, they became ever more likely
to support the opposition DPP and the cause of Taiwan independence (Liu and
Ho 1999). All these forces, combined with a more moderate DPP campaign
platform and a split in the KMT, paved the way for Chen’s victory in the 2000
presidential election (Rigger 2001, chaps. 8 and 9).

During the subsequent eight years, the pro-independence DPP government,

like the Chiang regime before it, selectively endorsed some aspects of Taiwan’s
historical memories. To reinforce the idea that the island is a political entity
separate from China, the “greater China nationalism” promoted by the KMT
authoritarian regime came under severe attack during Chen’s administration
(Greene 2008). Chen’s de-Sinicization measures, along with Beijing’s tactics of
diplomatic isolation, have had substantial effects on citizens’ identity. By the
time President Ma took office in 2008, more than half of the island residents
were Taiwanese identifiers and very few of them subscribed to the Chinese
identity alone. In total, 90 percent of the island residents considered themselves
either purely Taiwanese or holders of a dual identity—regarding themselves as
both Chinese and Taiwanese. In the view of some observers, the combination
of Chinese threat and active governmental attempts to reinforce Taiwan
consciousness has made inevitable “the ongoing consolidation of the
Taiwanese nation” (Schubert 2008, 111).

Page 11 → The Taiwanese/Chinese divide has now gone beyond individuals’

attachment to ethnic groups and has become embedded in different political
communities at the national level. The contestation between the two identities
is now more than the debate over Taiwan’s sovereignty and has become the key
political cleavage of the society, commonly known as the issue of “unification
vs. independence.” While this political cleavage involves the debate over
Taiwan’s ultimate sovereignty, it also structures opinions about how to interact
with a rising China in the meantime. In all these ways, preferences about the
island’s relationship with the Chinese mainland have been and continue to be
the most important issue of the society, while all other cleavages are distinctly
secondary. Consequently, Taiwan can be characterized as a single-issue society,
since the main cleavage affects almost all aspects of the island citizens’ political
attitudes and behaviors. Most notably, it is reflected in the voters’ partisan
identifications.

Indeed, Taiwan’s political landscape has undergone significant change after

the formation of the DPP in 1986. During the ensuing process of rapid
democratization, the previous authoritarian system dominated by the KMT was
transformed with the emergence of multiple minor political parties as spin-offs
from the KMT. After Chen Shui-bian of the DPP was elected president in
2000, Taiwan’s multiparty system increasingly moved into two distinctly
separate and relatively equal political camps—the Pan-Blue Alliance and the
Pan-Green Alliance, with the KMT and the DPP being the two leading parties
in each, respectively. The Pan-Blue Alliance consists of the KMT, the People
First Party (PFP), the New Party (NP) and the Nonpartisan Solidarity Union
(NPSU), while the Pan-Green Alliance includes the currently ruling DPP and
the Taiwan Solidarity Union (TSU). Even though none of the political parties
advocates the island’s immediate unification with China, Taiwan voters perceive
the two political alliances as representing opposite positions on the issue of
unification/independence. The Pan-Blue Alliance is perceived as adopting
policies that move in a direction toward Taiwan’s eventual unification with
China, whereas the Pan-Green Alliance is seen as making plans that would lead
to the island’s ultimate independence. Thus, chapter 4 shows that citizens who
hold a strong Taiwanese identity tend to espouse the island’s de jure
independence and provide electoral support for the Pan-Green Alliance, while
those who continue to recognize the Chinese heritage as a part of their identity
are more likely to back the Pan-Blue candidates.

As Lipset and Rokkan’s (1967) “freezing hypothesis” famously claimed half a

century ago, party systems become stabilized when they reflect preexisting
cleavage structures in the society. Because Taiwan Page 12 → citizens’ partisan
identifications notably reflect the most important political cleavage in the
society, the island’s political landscape has been increasingly molded into two
distinctly separate alliances that are led by two major political parties. As
chapter 10 explains, the 2005 electoral reform that abandoned the single non-
transferable vote (SNTV) electoral rules in favor of the mixed-member
majoritarian system (MMM) has further consolidated Taiwan’s two-party
system.
 While the endogeneity of the key political cleavage and its effects on Taiwan
citizens’ political life form a crucial part of our story, we will also pay attention
to the ways in which candidates’ personal traits affect electoral behavior.
Chapter 8 shows that the importance of candidates’ traits diminished as Taiwan
democratized. Because political power was monopolized by the KMT during
the authoritarian era, most nominees for elected positions shared the same
party label and candidates’ traits were voters’ means of differentiation. As the
issue of unification/independence became prominent in a democratic Taiwan,
partisan identification increasingly had major effects on the island citizens’
electoral decisions.

In sum, the central issue that dominates Taiwan politics is the debate over the
island’s sovereignty and its long-term relationship with China. Individuals’
stands on the unification/independence issue are closely related to their
national identities and deeply reflected in their partisan identification. The
cleavage is so deep that it has profound impacts on every aspect of the island
citizens’ political life. All issues, including such minor policy questions as
absentee voting or the adoption of an English translation system, can become
entangled in the main cleavage and be examined through a partisan lens.9

Political Cleavage and Politics: Taiwan as a Test Case

Some countries are so consequential in the world that they have legions of
outside observers. The United States, China, Britain, France, Germany, Japan,
and Russia take up space in the sophisticated media of every country. Political
life in most small countries, however, is of interest only to their own citizens.
Taiwan is a small country, and its domestic politics typically makes
international news only as entertainment—when fistfights break out in the
parliament, known as the Legislative Yuan. Apart from specialists in East Asian
politics, no one outside Taiwan cares very much. Why should readers with an
interest in comparative electoral behavior want to read about Taiwan in
particular?

Page 13 → In our judgment, there are two good reasons to pay special attention
to Taiwan. First, Taiwan is one of the best places to investigate the multifaceted
effects of cleavage on politics. With cleavages defined as divisions that separate
members of communities into groups (Rae and Taylor 1970), empirical
research has variously linked political cleavages to conflict and partisan
differences (Bonilla et al. 2011; Dahrendorf 1959; Dunning and Harrison 2010;
Tilley, Evans and Mitchell 2008). Yet, cleavages are not all bad, as Lipset (1963,
78–79) argued half a century ago, because “[a] stable democracy requires
relatively moderate tension among its contending political forces.” However, if
key differences in a society are allowed to accumulate and reinforce each other,
cleavages may deepen, conflicts may grow intense, and extreme polarization
may be manifested in voting and partisan divisions.

It is often argued that cross-cutting cleavages can mollify the intensity of

conflict and reduce polarization. The overlapping cleavages make it difficult to
build a coalition as few people are solely associated with any given cleavage.
Conversely, this implies that conflict and political polarization are generally the
most extreme when there is just one major cleavage in the society (Simmel
1908; Rae and Taylor 1970; Zuckerman 1975).
 As we have noted, the current central axis of politics in Taiwan is the issue of
unification vs. independence. Citizen’s opinions on that issue are closely tied to
their identity—Do they think of themselves as Chinese, as Taiwanese, or as
both? Ethnic identity also plays a role: mainlanders are unlikely to define
themselves as purely Taiwan citizens, while a majority of Minnan do so, and
thus they have different positions on the unification/independence issue.
Ethnic identity, national identity, and preferences over how to deal with a rising
China are all interrelated and reinforce each other in Taiwan. Political parties
on the island necessarily must define themselves in a way that makes sense in
this context. They all have distinct and well-differentiated positions on the
unification/independence axis, with two small parties (the NP and the TSU,
respectively) holding down the extremes. This dominant issue structures party
ideology and electoral strategies, while citizen’s party identifications and vote
choices map closely onto those issue positions.

The nexus of causal relationships for Taiwan citizens is thus theoretically

complicated but empirically simple: ethnic identification, national
identification, party identification, and political opinions connect closely to
each other and reinforce each other. Each of them connects to vote choice.
This set of relationship is relatively tight in Taiwan. To our knowledge, only
Israel has a similarly close linkage among ethnic origins (Ashkenazic vs.
Sephardic), concepts of the state (Israeli vs. Jewish), the central political Page 14

→ dimension (relationship with the Palestinians), and party choice (Shamir and
Arian 1999). Where relationships are strongest, causal patterns are most easily
discerned. One can study Taiwan for the same reason Charles Darwin gathered
scientific information on the Galapagos Islands: that is where the evidence is
clearest.
 The second reason for studying Taiwan is its strategic importance. Taiwan is a
flashpoint in East Asia, one of the most likely places to set off a general Asian
war. Domestic politics in China and Taiwan have perhaps been the central
drivers of cross-Strait tensions. In Taiwan, the cross-Strait relationship is the
key cleavage that profoundly shapes the interest structure of its electorate. If
voting is a means for individuals to make collective decisions, it is crucial to
know what Taiwan citizens are asking for and how these demands are
manifested in their partisan support and vote choices. Thus, understanding
Taiwan voters is important not only in its own right as a key mechanism of a
young democratic polity but also for its impact at the collective level in
preserving peace in East Asia. Until the last decade or so, even China did not
pay close attention. The resulting misjudgments helped cause the missile crisis
in 1996, with costly consequences for the mainland (Garver 1997; Cooper 1998,
chap. 4).

In this volume, we make no pretense of sorting out the full set of causal
relationships among the various kinds of identity and the resulting vote
choices. Those relations necessarily differ among individuals: some will come
to their partisanship because of their ethnicity, some because of their national
identity, and some because of their policy views. Others will inherit a
partisanship and let that determine their other identities, while still others will
exhibit more complex patterns. We leave to others the full specification of all
those causal paths—an important topic, but one that we must set aside.

Our perspective is rather that the key political cleavage is so powerful that it
renders other differences among Taiwan voters secondary or unimportant.
When any one component is inconsistent with another, there will be cognitive
or psychological pressure to bring them into line. Therefore, the island citizens’
portfolio of identities, their political opinions, and their vote choices are also
“co-integrated” in the sense that time series analysts use the term.10 That
pressure may be strong or weak, it may be heeded quickly or slowly, it may be
set aside, or it may be resisted. In any cross-section, some individuals may
ignore their contradictory views or deliberately hold inconsistent views, and
some may hold them for a lifetime. But for most individuals in a political
system like Taiwan’s, where one central, vivid issue dominates political life,
where that issue is closely tied to important ethnic Page 15 → and national
identities, and where the parties’ stances reflect that issue, most individuals will
show substantial consistency. We will not find many DPP supporters who have
a Chinese identity and support Taiwan’s unification with China, nor many
mainlanders favoring immediate independence.

Demonstrating the power of a political cleavage manifested in many forms is

a central feature of this book. Indeed, Taiwan is a perfect case for analyzing the
effects of political cleavages and personal identities on politics. Many small
countries are out of the international fray, and it matters little how their politics
and policies evolve. Not Taiwan. Political identities are important in every
political system, but often in such variegated or muted form that their effects
are difficult to detect. Not in Taiwan. And institutions matter in every country,
but changes in them are so rare and glacial that no one can be sure how much
difference they make. Not in Taiwan. Succeeding chapters will demonstrate not
just how interesting Taiwan politics is in its own right but how much Taiwan
can teach us about how politics works in the many countries around the world
where ethnic divisions and contested national identities are central to electoral
politics.
 Plan of the Book

This volume has 12 chapters. To make the volume accessible to both scholars
and general audiences, contributors have intentionally avoided complicated
statistical analysis. Not until the final chapter do we build a comprehensive
statistical model based on the factors that previous chapters have identified.

After this introduction, the book proceeds to a chapter by Chia-hung Tsai,

designed to introduce Taiwan electoral politics to those who may have little
prior experience with the topic. It explores who Taiwan voters are and how
they have voted in presidential elections since the rapid democratization began
about two decades ago. Tsai finds that Taiwan’s young democracy has
increasingly consolidated into a competitive two-party system formed by the
Pan-Blue and the Pan-Green Alliances, which are dominated by the KMT and
the DPP, respectively. While the Pan-Blue Alliance has generally enjoyed the
electoral advantage during most lower-level elections, the two parties have split
the six presidential elections since democratization. The Pan-Green Alliance is
increasingly competitive at all levels.

Tsai notes that a regional divide in Taiwan is clearly recognizable, validating

the general view that the island consists of “a Blue North and a Green South”
(the Pan-Blue Alliance dominates northern Taiwan and the Pan-Green Alliance
dominates southern Taiwan.11) Due to the policies implemented Page 16 → by
the KMT government during the authoritarian era, Taiwan voters who are
mainlanders, well educated, more affluent, government employees, or in their
30s and 40s are more likely to support Pan-Blue candidates. The island citizens
who belong to the Minnan ethnic group and who are less educated, less well-
off, working in labor and agricultural sectors, or in their 20s tend to identify
with the Pan-Green Alliance.

After this first chapter explores who Taiwan voters are, successive chapters
take up the “big three”—party identification, issue orientation, and candidate
evaluation—and how they affect the island citizens’ political attitudes and
behaviors. To these three, we add a fourth—identity, particularly national
identity. As previously noted, the preference about Taiwan’s future relationship
with China presents a deep dividing line between the island’s citizens. Precisely
because the unification/independence issue is the key political cleavage, it
closely intertwines with the island citizens’ national identities, their partisan
attachments, their issue preferences, and their views of the candidates.

In that spirit, in chapter 3 T. Y. Wang examines the development of political

identity in Taiwan. The author shows that the “China factor” has been the
essential component of the changing boundaries in group membership that
shapes Taiwan voters’ identities. Immediately after the KMT’s retreat to the
island, the ethnic divide between local citizens and mainlanders became salient.
After democratization and other social changes, the ethnic cleavage was
replaced by the contestation between Chinese and Taiwanese identities. As the
majority of citizens now hold a Taiwanese identity and few are Chinese
identifiers, the boundary of the Chinese/Taiwanese divide has gradually lost its
political significance in domestic politics but has moved to a different level.
Due to Beijing’s forceful claim on Taiwan’s sovereignty, the term “Taiwan” is
no longer a purely geographic designation. Being “Taiwanese” increasingly
implies an identity with Taiwan as an independent state. The shadows of both
the ethnic cleavage and the Taiwanese/Chinese divide continue to be cast on
the island citizens’ partisan identifications.

Following the discussion of identity change on the island, in chapter 4 Ching-

hsin Yu takes up the second of the “big four”—partisanship. He explores the
trajectory of Taiwan voters’ attachments to parties and how they are
intertwined with their positions along the axis of unification/independence.
The author traces the development of the island citizens’ partisan identification
and concludes that nonpartisan voters during the authoritarian era were not
“independent” in its true sense. As the KMT monopolized political power and
banned the formation of political parties, those who did not identify with the
KMT were treated as “nonpartisan independents” lest they form a cohesive
political force. Many of these citizens later became the Page 17 → loyal

supporters of the newly established opposition DPP. Along with their junior
partners, the KMT and the DPP later formed the Pan-Blue Alliance and the
Pan-Green Alliance. Because each of the two political alliances has a distinctive
position on the issue of Taiwan’s future relations with China, which is the most
important political cleavage on the island, Taiwan voters have developed a clear
partisan attachment in the competitive two-party system. That said, a
considerable number of citizens are nonpartisan and they behave differently
according to their educational levels. Yu finds that highly educated
independents are less interested in politics and less likely to vote but they are
also more autonomous and politically moderate than their less educated
counterparts.
 The next three chapters turn attention to the third leg of the “big four”—
citizens’ issue orientation. Conventionally, political issues have been seen as
crucial to the functioning of democracy. Political parties are expected to
present policy options to electorates, allowing citizens to select those
candidates who correspond most closely to their own ideological positions.
The congruence of issue positions and ideology between citizens and political
parties has thus become an important topic in electoral studies (e.g., Adams,
Ezrow and Somer-Topcu 2011; Lachat 2011; Thomassen and Schmitt 1997; but,
for a critique of this framework, see Achen and Bartels 2016).

Do issue preferences play a role in Taiwan citizens’ voting calculus? If so,

what are the important issues? How do they relate to the key political cleavage
on the island? Chapter 5 by Shing-yuan Sheng and Hsiao-chuan Liao
demonstrates the impact of the central political cleavage on Taiwan politics.
Examining the evolution of four key political debates on the island, the authors
show that reform vs. stability has become an issue of the past. The second
issue, wealth distribution, fails to differentiate political parties because the
partisan elites of various political affiliations have converged on the same
ideological position and have attempted to outbid each other in order to win
electoral support. Third, while environmental protection has the potential of
becoming an important political issue after the disaster at the Fukushima
nuclear plan in Japan, it is still an issue in the process of development in
Taiwan. The most important political issue on the island, then, is and will
continue to be Taiwan’s future relationship with China, which is closely tied to
the island citizens’ identity as Chinese or Taiwanese.
 The salience of Taiwan’s relationship with China does not mean that such
issues as the economy are unimportant. Empirical studies on the American
presidency have long concluded that the state of the economy is an important
contributing factor to electoral decisions. An abundance of literature has
demonstrated the linkage between the electoral success of an incumbent Page 18

→ government and its economic performance.12 The logic is simple. When

economic conditions are good, voters are likely to reward the incumbent with
their vote. When the economy is bad, voters tend to switch their support to the
challenger. Jimmy Carter’s devastating defeat by Ronald Reagan in 1980 and
George H. W. Bush’s loss to Bill Clinton in 1992 are testimony to the
conventional wisdom that “economics is the fate of politicians” (Norpoth
1985, 167).

Following this reward-punishment model, in chapter 6 Chia-hung Tsai

explores the role of the economy in Taiwan citizens’ voting calculus. The
author finds mixed evidence for economic voting in the island country’s
presidential elections. While prospective economic evaluations are found to be
an important determinant of Taiwan voters’ electoral choices, the weight of the
economy is overshadowed by, again, citizens’ partisan affiliations. This finding
shows that responses to the economy affect Taiwan voters’ electoral calculus,
but primarily through the colors of partisan lenses, which are closely aligned
with their positions on the key political cleavage in the society.

Chapter 7 by Alexander Tan and Karl Ho examines the complex dynamics of

cross-Straits relations, in particular the burgeoning economic exchanges with
China after Ma Ying-jeou of the KMT was elected president in 2008. The
authors find that island residents exhibit an ambivalent view of Taiwan’s close
and intensive interactions with the Chinese mainland since 2008. Following
partisan lines, some of them feel that such ties are beneficial to the island’s
economy, while others express concerns about the security implications. In the
aggregate, Taiwan voters recognize that isolation from China is not viable or
even possible, yet getting too close to China also troubles them. This explains
why the public changed its mind about the engagement policies of the Ma
administration, initially favoring it after the DPP’s isolationist policies during
2000–2008, but then coming to distrust close ties as they seemed to bring few
benefits to most ordinary citizens. Thus, cross-Strait relations, or more broadly
speaking “the China factor,” affect not only how Taiwan voters see the future
but also how they view the current administration.

Following these discussions of the impact of national identities, partisan

attachments, and issue orientations on Taiwan voters’ electoral decision,
chapter 8 by Hung-chung Wang and Lu-huei Chen examines the last
component of the “big four”—citizens’ evaluation of candidates. Assessment
of candidates for public office attracts considerable attention in each campaign
season. Scholars and pundits analyze and dissect candidates’ backgrounds and
characters, and discuss how these personal traits affect voters’ evaluation of
candidates and their voting decisions. Citizens’ evaluation of candidates Page 19

→ is important because, it is argued, seeking for and digesting political

information is costly. The assessment of candidates offers the public a useful
shortcut as it seems to provide a clue as to how they will perform their duties
once elected. In particular, given that the president occupies the principal
position in the government, presidential traits and characters have important
symbolic meaning, which set public standards for all political behaviors
(Greene 2001; Kinder 1986; McCurley and Mondak 1995).
 Following this line of research, Wang and Chen find that in Taiwan’s
presidential elections, KMT candidates tend to be viewed as more capable of
dealing with cross-Strait relationships and economic development, while DPP
nominees are associated with the issue of eliminating corruption and initiating
political reforms. Perhaps surprisingly, the perception of candidates’ personal
traits has little effect on the island citizens’ voting behavior. Taiwan voters’
electoral decisions are largely determined by their partisan identifications,
which, again, are closely in line with citizens’ positions on the axis of
unification/independence and their national identities. Thus, candidate issue
ownership is applicable to Taiwan’s parties, but voters’ evaluation of
candidates’ personal characteristics and perceived competence are conditioned
largely by their partisan affiliations.

In the context of Taiwan’s electoral politics, it should occasion no wonder

that Western notions of political “left” and “right” have little relevance in
Taiwan, as Yi-ching Hsiao, Su-feng Cheng, and Christopher Achen explain in
chapter 9. The left-right scale is irrelevant because the main cleavage in Taiwan
is not the degree of government control of the economy, as previous chapters
emphasize. Nor are “left” and “right” used to describe other aspects of
political debates in Taiwan. Political elites make essentially no use of those
words in Taiwan, and ordinary citizens are mystified by references to them. The
confusions are exacerbated by particular connotations of “left” and “right” in
both Mandarin and the Taiwanese dialect, as the authors explain.

While the “big four” are the most significant factors in voters’ electoral
calculus, they do not operate in a political vacuum. Institutional structures set a
broad framework for the actions of both political parties and individuals. They
set an approximate upper limit to the number of parties, configure the choice
menu on the ballot, structure voters’ electoral calculus, and provide incentives
or disincentives for citizens to show up at the ballot boxes on election days
(e.g., Cox 1997; 1999; Engstrom 2012). Thus chapter 10 by Chi Huang
examines the effects of recent electoral system change on Taiwan’s party system.
As with the experience of Japan, which also adopted the mixed-member system
a decade earlier, Taiwan has witnessed a dramatic change in its party system in
the postreform era. The fast convergence toward Page 20 → two-party
competition on the island is due in part to the interactions between the
powerful presidency and political elites’ ambition of synchronizing presidential
and legislative elections. Because the issue of unification vs. independence is
the single most important political cleavage on the island, as discussed in the
previous chapters, it has played a hidden yet significant role in shaping the
postreform party structure toward a “Pan-Blue vs. Pan-Green” system.

Chapter 11 by Chung-li Wu and Tzu-Ping Liu examines Taiwan citizens’

political participation. Similar to the trend in many Western democracies, the
path of turnout rates in the island country’s presidential and legislative
elections has gone downward since the early 1990s. Three particular factors
affect individuals’ turnout and other forms of political participation—age,
party identification, and political knowledge. In general, older people,
individuals identified with main political parties, and those equipped with more
political knowledge tend to have a higher level of political participation.
Because this more engaged group generally has higher socioeconomic status,
these findings further confirm the conventional wisdom that citizens in the
upper and middle classes tend to be more interested and involved in public
affairs than individuals in the lower class. The chapter shows that the gap has
widened in Taiwan in recent years.

Each of these chapters identifies one or more important aspects of Taiwan

politics. The final chapter by Christopher Achen and T. Y. Wang brings
together all of them to examine vote choice on the island. We have built a
comprehensive statistical model based on the factors that previous chapters
have identified and discuss their implications. The central finding is that one
key dimension organizes Taiwan citizens’ vote choices—the China factor.
Should Taiwan accommodate itself to China’s ever more powerful presence in
Asia, or should it forcefully assert its status as a separate country and resist
integration with the mainland? The answer to that question largely determines
what Taiwan people think about related policy issues, which party they adhere
to, and how they vote.

Thus, in contrast to left-right economic disputes, which dominate politics in

many Western countries, politics in Taiwan is fundamentally about nationalism
and the future of Taiwan’s national identity. In that respect, Taiwan is one of
many countries around the world in which similar issues shape domestic
politics. In recent years, the world has seen a succession of crises and threats to
peace in which national identity was the central issue—Bosnia, Kosovo, and
the Ukraine, to list just three examples in which local wars have broken out and
the major Western powers have taken an interest in restoring the peace.
Taiwan, too, is an international hotspot with a potential Page 21 → for setting off
a major international conflict. Thus far, democracy and peace have prevailed on
the island. Hence, Taiwan has much to teach us, and we offer this book as a
contribution to the international dialog about how differing conceptions of
national identity can be managed peacefully.

Notes

1. National Statistics, Republic of China (Taiwan) at

http://statdb.dgbas.gov.tw/pxweb/Dialog/Saveshow.asp (accessed March
15, 2015).

2. As in Japan, factional politics are prevalent in Taiwan. The island

country’s two major political parties have employed patron-client
relationships for power distribution and voter mobilization. For a
discussion of Taiwan’s factional politics, see Batto and Huang (2016).

3. We use “Taiwan citizen” to mean any citizen of Taiwan because the

term “Taiwanese” has various political connotations for different people.
To some, “Taiwanese” refer to citizens living in the territory effectively
governed by the Taiwan government, while to others the term means an
ethnic designation opposed to “mainlander.” Because the majority of
support for the Pan-Green Alliance led by the Democratic Progressive
Party comes from those who self-identify as “Taiwanese,” a “Taiwanese”
or a “Taiwanese citizen” thus may have a narrower and potentially
partisan meaning in English.

4. One well-known example is former Taiwan president Lee Teng-hui,

who freely admitted that he had become a Japanese citizen with a
Japanese name during the colonial period. Throughout his life, his
Japanese language skills were better than his Mandarin.

5. Retrieved from Liang-An Jing-ji Tong-ji Yue-bao (Monthly Report on

Cross-Straits Economy) published by the Mainland Affairs Council, the
Republic of China, at http://www.mac.gov.tw (accessed March 15, 2015).

6. For a complete list of cross-Strait agreements, see the Mainland Affairs

Council website, the Republic of China, http://www.mac.gov.tw/ct.asp?
xItem=67145&CtNode=5710&mp=1.

7. Greater Tainan mayor William Lai’s visit to Shanghai is one of the most
recent high-profile visits by DPP politician (Tsao and Chung 2014).

8. There are four major ethnic groups in Taiwan: Minnan, Hakka,

mainlander, and aborigine. Minnan refers to island residents whose
ancestors migrated to Taiwan from the Chinese mainland several hundred
years ago. They are the largest ethnic group at 77% of the island’s 23
million people. About 10% of Taiwan’s total population is Hakka,
descendants of immigrants who came to the island at roughly the same
time as the Minnan from areas in central China. Both Minnan and Hakka
are generally grouped together as “Taiwanese” even though they have
different customs and habits and speak different dialects. Approximately
12% of the total population are mainlanders, those who arrived from the
mainland in the late 1940s after the Chinese civil war, and their
descendants. Aborigines, the original settlers, constitute less than 2% of
the total population in Taiwan.
 Page 22 → 9. Taiwan has no system of absentee voting (“postal voting”).
One of the main explanations often advanced is that China-based
Taiwanese businesspeople and their family members rely on good
relations with China and thus are potential KMT supporters. Hence,
absentee voting is expected to increase the KMT’s votes, and the DPP
opposes it. Similarly, the Ma administration declared more than five years
ago that Taiwan would start using the pinyin system for English
translation, which is the phonetic system used in almost every part of the
world, including the UN. Because the pinyin system was developed and
adopted in the Chinese mainland, however, its usage was resisted by many
DPP politicians (Economist 2014).

10. Some of the original applications were to currency exchange rates.

Political scientists have studied arms races using the same ideas. In both
instances, the complex causal details are much less important than
showing that the system constantly attempts, in the midst of continual
disturbances, to reach toward an equilibrium, even if it never attains it.

11. For a discussion of Taiwan’s “Blue North and Green South”

phenomenon, see Chou 2012.

12. For a concise review of the literature on economic voting, see Lewis-
Beck and Stegmaier 2007.

References
Achen, Christopher H., and Larry M. Bartels. 2016. Democracy for Realists.
Princeton: Princeton University Press.

Adams, James, Lawrence Ezrow, and Zeynep Somer-Topcu. 2011. “Is Anybody
Listening? Evidence That Voters Do Not Respond to European Parties’
Policy Statements during Elections.” American Journal of Political Science 55
(2): 370–82.

Batto, Nathan, and F. Shin-at Huang. 2016. “Executive Competition, Electoral

Rules, and Faction Systems in Taiwan.” In Elections in Taiwan and Japan under
the Mixed- Member Majoritarian System, ed. Nathan F. Batto, Chi Huang,
Alexander Ç. Tan, and Gary Cox. Ann Arbor: University of Michigan Press.

Berelson, Bernard R., Paul F. Lazarsfeld, and William N. McPhee. 1954. Voting:
A Study of Opinion Formation in a Presidential Campaign. Chicago: University of
Chicago Press.

Bonilla, Claudio A., Ryan E. Carlin, Gregory J. Love and Ernesto Silva
Méndez. 2011. “Social or Political Cleavages? A Spatial Analysis of the
Party System in Post-Authoritarian Chile.”Public Choice 146 (2): 9–21.

Brown, Melissa. 2004. Is Taiwan Chinese? The Impact of Culture, Power, and
Migration on Changing Identities. Berkeley: University of California Press.

Butler, David, and Donald E. Stokes. 1974. Political Change in Britain. 2nd ed.
London: Macmillan.

Campbell, Angus, Philip E. Converse, Warren E. Miller, and Donald E. Stokes.

1960. The American Voter. New York: John Wiley and Sons.
Carlin, Ryan E., Matthew M. Singer, and Elizabeth J. Zechmeister. 2015. The
Latin American Voter. Ann Arbor: University of Michigan Press.

Page 23 → Chan, Steve. 1988. “Developing Strength from Weakness: The State in
Taiwan.” In State and Development, ed. Cal Clark, 38–51. New York: E. J. Brill.

Ching, T. S. Leo. 2001. Becoming Japanese: Colonial Taiwan and the Politics of Identity
Formation. Berkeley: University of California Press.

Chou, Yin-lung. 2012. “A Study of the Pan-Green Political Landscape in

Southern Taiwan” [in Chinese]. PhD diss., National Chengchi University,
Taipei.

Chu, Yun-han, and Chia-lung Lin. 2003. “Consolidating Taiwan’s New

Democracy amid Competing National Identities.” In China Today: Economic
Reforms, Social Cohesion and Collective Identities, edited by Taciana Fisac and
Leila Fernández-Stembridge, 240–67. London: Routledge Curzon.

Clark, Cal. 1987. “The Taiwan Exception: Implications for Contending

Political Economy Paradigms.” International Studies Quarterly 31:327–56.

Clark, Cal. 1989. Taiwan’s Development: Implications for Contending Political Economy
Paradigms. New York: Greenwood Press.

Cooper, John F. 1998. Taiwan’s Mid-1990s Elections. Westport, CT: Praeger.

Cox, Gary W. 1997. Making Votes Count: Strategic Coordination in the World’s
Electoral Systems. Cambridge: Cambridge University Press.
Cox, Gary W. 1999. “Electoral Rules and the Calculus of Mobilization.”
Legislative Studies Quarterly 24 (3): 387–420.

Dahrendorf, Ralf. 1959. Class and Class Conflict in Industrial Society. Stanford,
Calif., Stanford University Press,

Dalton, Russell J., and Christopher J. Anderson, eds. 2011. Citizens, Context, and
Choice. Oxford: Oxford University Press.

Dunning, Thad, and Lauren Harrison. 2010. “Cross-cutting Cleavages and

Ethnic Voting: An Experimental Study of Cousinage in Mali.” American
Political Science Review 104 (1): 21–39.

Economist. 2014. “Lost in Romanisation: Ideological Warfare over Spelling.”

June 7. http://www.economist.com/news/asia/21603500-ideological-
warfare-over-spelling-lost-romanisation.

Engstrom, Erik. 2012. “The Rise and Decline of Turnout in Congressional

Elections: Electoral Institutions, Competition, and Strategic Mobilization.”
American Journal of Political Science 56 (2): 373–86.

Evans, Geoffrey, and Nan Dirk De Graaf, eds. 2013. Political Choice Matters.
Oxford: Oxford University Press.

Flanagan, Scott C., Shinsaku Kohei, Ichiro Miyake, Bradley M. Richardson, and
Joji Watanuki, eds. 1991. The Japanese Voter. New Haven: Yale University
Press.

Garver, John W. 1997. Face Off: China, the United States, and Taiwan’s
Democratization. Seattle: University of Washington Press.
Greene, J. Megan. 2008. “History, Identity, and Politics: The First Chen Shui-
bian Administration’s Efforts to Craft Taiwan’s History.” In Presidential
Politics in Taiwan: The Administration of Chen Shui-bian, ed. Steven M.
Goldstein and Julian Chang, 115–41. Norwalk, CT: EastBridge.

Greene, Steven. 2001. “The Role of Character Assessments in Presidential

Approval.” American Politics Research 29 (2): 196–210.

Jacoby, William. 2010. “The American Voter.” In The Oxford Handbook of

American Elections and Political Behavior, ed. Jan E. Leighley, 260–77. New
York: Oxford University Press.

Page 24 → Kerr, George H. 1965. Formosa Betrayed. Boston: Houghton Mifflin.

Kinder, Donald. 1986. “Presidential Character Revisited.” In Political Cognition:

the 19th Annual Carnegie Symposium on Cognition, ed. Richard R. Lau and
David O. Sears, 233–55. Hillsdale, NJ: L. Erlbaum Associates.

Lachat, Romain. 2011. “Electoral Competitiveness and Issue Voting.” Political

Behavior 33 (4): 645–63.

Lai, Tse-han, Ramon H. Myers, and Wei Wou. 1991. A Tragic Beginning: The
Taiwan Uprising of February 28, 1947. Stanford: Stanford University Press.

Lazarsfeld, Paul F., Bernard Berelson, and Hazel Gaudet. 1948. The People’s
Choice: How the Voter Makes Up His Mind in a Presidential Campaign. 2nd ed.
New York: Columbia University Press.
Lewis-Beck, Michael, and Mary Stegmaier. 2007. “Economic Models of
Voting.” In The Oxford Handbook of Political Behavior, ed. Russell Dalton and
Hans-Dieter Klingemann, 518–37. Oxford: Oxford University Press.

Lipset, Seymour M., and Stein Rokkan. 1967. “Cleavage Structure, Party
Systems, and Voter Alignments: An Introduction.” In Party Systems and Voter
Alignments: Cross-National Perspectives, ed. Seymour M. Lipset and Stein
Rokkan, 1–64. New York: Free Press.

Lipset, Seymour Martin. 1963. Political Man: The Social Bases of Politics. Garden
City NY: Doubleday.

Liu, I-chou, and Ho Szu-yin. 1999. “The Taiwanese/Chinese Identity of the

Taiwan People.” Issues & Studies 35 (3): 1–34.

Marsh, Michael, Richard Sinnott, John Garry, and Fiachra Kennedy. 2008. The
Irish Voter: The Nature of Electoral Competition in the Republic of Ireland. New
York: Manchester University Press.

McCurley, Carl, and Jeffery J. Mondak. 1995. “Inspected by #1184063113: The

Influence of Incumbents’ Competence and Integrity in U.S. House
Elections.” American Journal of Political Science 39 (4): 864–85.

Norpoth, Helmut. 1985. “Politics, Economics, and the Cycle of Presidential

Popularity.” In Economic Conditions and Electoral Outcomes: The United States
and Western Europe, ed. Heinz Eulau and Michael S. Lewis-Beck, 167–86.
New York: Agathon Press.
Rae, Douglas W., and Michael Taylor. 1970. The Analysis of Political Cleavages.
New Haven: Yale University Press.

Rigger, Shelley. 2001. From Opposition to Power: Taiwan’s Democratic Progressive

Party. Boulder, CO: Lynne Rienner.

Schubert, Gunter. 2008. “Taiwan’s Evolving National Identity since the DPP
Takeover: From Civic to Ethnic?” In Presidential Politics in Taiwan: The
Administration of Chen Shui-bian, ed. Steven M. Goldstein and Julian Chang,
185–213. Norwalk, CT: EastBridge.

Shamir, Michal, and Asher Arian. 1999. “Collective Identities and Electoral
Competition in Israel.” American Political Science Review 93: 265–78.

Simmel, Georg. (1908) 1964. Conflict and The Web of Group Affiliations. Glencoe,
IL: Free Press.

Thomassen, Jacques, ed. 2005. The European Voter. Oxford: Oxford University
Press.

Page 25 → Thomassen, Jacques, and Hermann Schmidt. 1997. “Policy

Representation.” European Journal of Political Research 32: 165–84.

Tien, Hung-mao. 1996. Taiwan’s Electoral Politics and Democratic Transition.

Armonk, NY: M. E. Sharpe.

Tilley, James, Geoffrey Evans, and Claire Mitchell. 2008. “Consociationalism

and the Evolution of Political Cleavages in Northern Ireland, 1989–2004.”
British Journal of Political Science 38 (4): 699–717.
Tsao, Po-yen, and Jake Chung. 2014. “William Lai’s Words Feed Online Feud.”
Taipei Times, June 9.
http://www.taipeitimes.com/News/taiwan/archives/2014/06/09/2003592
336 (accessed March 15, 2015).

Tung, Chen-yuan. 2005. “An Assessment of China’s Taiwan Policy under the
Third Generation Leadership.” Asian Survey 45 (3): 343–61.

Wang, T. Y. 2000. “State Policy and Industrial Promotion in Taiwan: An

Interrupted Time Series Analysis, 1956–1994.” International Journal of Public
Administration 23 (10): 1749–76.

Wang, T. Y., and I-Chou Liu. 2004. “Contending Identities in Taiwan:

Implications for Cross-Strait Relations.” Asian Survey 44 (4): 568–90.

Zuckerman, Alan. 1975. “Political Cleavage: A Conceptual and Theoretical

Analysis.” British Journal of Political Science 5 (2): 231–48.
Page 26 →
 Chapter 2

Who Is the Taiwan Voter?

Chia-hung Tsai

The purpose of this chapter is to provide an overview of recent Taiwan

elections for readers who may not be familiar with them. The growing
importance of the north-south divide in Taiwan’s politics is emphasized. The
chapter also explores how the supporters of the two party camps differ from
each other. In particular, standard variables that are important for
differentiating party supporters in many other countries, such as income and
occupation, turn out to have only modest effects in Taiwan. Instead, ethnicity
plays the most crucial role in shaping the voting and partisan identities of
Taiwan citizens, just as the theoretical framework of this book suggests.

Recent Taiwan Presidential Elections

Chapter 1 has given a brief overview of Taiwan’s history that highlights the
island country’s political development. In particular, it argued that the pivot of
Taiwan politics is the citizens’ political identities, which are both related to
their ethnicity and manifested in their partisan affiliation with the Pan-Blue
Alliance or the Pan-Green Alliance—the two major political camps that have
formed since the late 1990s. Presuming that background, the present chapter
begins with an exploration of Taiwan’s general voting patterns as seen in
election returns tabulated by the Central Election Commission. The scope of
the analysis comprises the five presidential elections since 1996 when the
island’s president was first popularly elected, a time span of nearly two
decades.1 Since then, four more presidential elections have been held, with two
peaceful transfers of power in 2000 and in 2008.

Page 27 → The rapid democratization on the island culminated in the 1996

presidential election, in which Taiwan voters exercised their political right to
popularly elect their national leader for the first time. In the 1996 presidential
election, two candidates—Lin Yang-kang and Chen Lu-an—were members of
the KMT who only broke away from the party after they failed to secure its
nomination. Chen ran as an independent and Lin was supported by the New
Party—a KMT split-off strongly advocating Taiwan’s unification with the
Chinese mainland. Similarly, in the 2000 presidential election, James Soong left
the KMT and ran as an independent against Lien Chan, the KMT nominee, and
Chen Shui-bian, the DPP-affiliated candidate. Although Soong and Lien
together polled nearly 60 percent of the votes, Chen won the election with
merely 39.3 percent under the first-past-the-post voting system. After the
election, Soong organized the People First Party and has since served as its
chair.

Recognizing that a divided KMT would only benefit the opponents, Soong
joined Lien as his running mate in the 2004 presidential election. However, a
successful DPP campaign and an apparent assassination attempt on Chen’s life
24 hours before the polls opened gave Chen the victory.2 Then, in both the
2008 and 2012 elections, it was the KMT-affiliated Ma Ying-jeou who scored
the victory. While the 2008 presidential election was a classic two-party race
between the KMT and the DPP, the 2012 election was a three-party race with
the PFP-affiliated Soong unsuccessfully attempting a comeback.

Thus, while there are multiple political parties on the island, even the smallest
are either allies of the DPP or are KMT splinter groups. Hence the political
landscape can generally be characterized by two political alliances—the “Pan-
Blue” and the “Pan-Green”—led by the KMT and the DPP, respectively. The
current chapter uses those terms for all the elections in the democratic era,
even though the terms themselves did not appear until 2000. To simplify the
analysis, vote shares of candidates in the five presidential elections are
aggregated into those two blocs.

The Consolidation of Two-Party System

Figure 2.1 shows the electoral support enjoyed by the two party groups during
the five presidential elections Taiwan has conducted since democratization,
1996–2012. A dominant electoral force in the mid-1990s, the Pan-Blue Alliance
shrank rapidly during the subsequent decade, only to recover somewhat in the
two most recent presidential contests. In recent years, the Page 28 → partisan

divide on the island has been approximately 50–55 percent for the Pan-Blue
Alliance and 40–45 percent for the Pan-Green Alliance, with the smaller parties
of decreasing importance. It appears that Taiwan’s political landscape has
consolidated as a genuine two-party system dominated by the KMT and the
DPP.
 Fig. 2.1. Vote shares of the Pan-Blue and Pan-Green camps in the presidential elections, 1996–

2012. Data: Election Study Center.

Many democracies in the world show strong regional differences in their
voting patterns. Well-known historical examples include the late nineteenth and
early twentieth-century “Solid South” in the United States (Archer and Taylor
1981; Tindall 1972), the traditional loyalty to the Labour Party in northern
England and Scotland, the east vs. west divide in Korean and Ukrainian
politics, and the strong support for the Liberal Party in Quebec, Canada, in the
immediate postwar period. Taiwan also has a sharp regional division. Since
2000, the island country’s electoral politics has increasingly displayed a strong
north-south divide. In part due to the fact that the national government in
Taipei has poured budget money into northern Taiwan, Page 30 → the disparity
between the north and the rest of the country in development and living
standard has grown. As cross-Strait relations become increasingly vital to
Taiwan’s economic development in the 2000s, voters in northern Taiwan, where
high-tech industries, banking businesses, and the service industry gather, are
more supportive of the KMT and its policies than those living in southern
Taiwan.
 Page 29 →

Fig. 2.2. Pan-Blue’s vote shares in the presidential elections, 1996–2012. Data: Election Study
Center.

The five choropleth maps in figure 2.2 set out the vote shares garnered by
Pan-Blue presidential candidates during the five presidential elections. They
show the emergence of the north-south divide. In the 1996 presidential
election, running as the incumbent and the first native son to be Taiwan’s
national leader, the KMT’s nominee, Lee Teng-hui, was widely popular. The
Pan-Blue Alliance won at least 70 percent of the votes in every city and county
on the island, in both urban and rural areas. In particular, voters in much of
north-central Taiwan, including Taoyuan County, Hsinchu City and County,
Miaoli County, and eastern Taiwan, were highly supportive of the three Pan-
Blue candidates and in particular of Lee. The regional divide was not much in
evidence.

As noted above, the 2000 presidential election was a three-way race among
Chen Shui-bian, the DPP nominee; James Soong, the party switcher from the
KMT to independent; and Lien Chan, the KMT nominee. Using the Jhuoshuei
River as the traditional demarcation line between northern and southern
Taiwan, figure 2.2 again shows that the regional disparity in electoral support
for the two political alliances began to emerge in this election. The two Pan-
Blue candidates received a combined vote share of 60 percent to 70 percent in
many parts of northern Taiwan, especially in Taoyuan County and in Hsinchu
City and County. The Pan-Green candidate Chen ran better in southern Taiwan
but had little support in cities and counties in the north. Overall, Chen had less
than 40 percent of the national vote. Had the Pan-Blue Alliance not been
divided and thus split its electoral support, Chen would not have been able to
win the 2000 presidential election.

The regional disparity in electoral support became even clearer in the 2004
election, as figure 2.2 shows. Pan-Blue candidates continued to garner large
electoral support in northern Taiwan but were not able to receive more than 40
percent of votes in cities in the south, including Yunlin County, Chiayi City and
County, Tainan City and County, Kaohisung City and County, and Pingtung
County. Cities and counties in the central part of Taiwan, including Taichung
City and Nantou County, Taichung County and Chunhwa County, became
battleground areas as the Pan-Blue candidates barely secured a majority of
votes.

The regional divide between a “Blue North” and a “Green South”

appeared Page 31 → to become consolidated in the 2008 and 2012 presidential
elections. In 2008, the Pan-Blue nominee, Ma Ying-jeou, won a landslide
victory with a 58.5 percent to 41.5 percent victory. As figure 2 shows, however,
while Ma was able to garner 60 percent to 70 percent of the vote in much of
northern and eastern Taiwan, the electoral support for him was considerably
weaker in southern cities and counties. Although Ma won his reelection bid in
2012, the regional disparity in electoral support continued. The Pan-Blue
candidate continued to draw 60–70 percent of the vote in much of northern
and eastern Taiwan, while southern Taiwan remained the stronghold of the
Pan-Green Alliance.

Regional disparities in party support are due partly to purely regional factors.
But northern and southern Taiwan also differ in ethnic composition, income,
and occupations, which together account for some of the regional disparity.
The remainder of the chapter takes up those forces accounting for the partisan
divide.

Demographic Characteristics and Electoral Behavior

As we have seen, the Pan-Blue Alliance has enjoyed a 5–10 percentage point
electoral advantage at the national level since 2000. This section will examine
where that support comes from.
 As chapter 1 pointed out, there are three major ethnic groups on the island—
Minnan, Hakka, and the mainlanders. While they were all immigrants from the
Chinese mainland, the Minnan and Hakka are the island residents whose
ancestors migrated to Taiwan several hundred years ago. Mainlanders were
originally largely composed of the followers of Chiang Kai-shek when the
KMT government retreated to the island in 1949. Mainlanders now are
primarily their descendants. Each ethnic group thus carries a distinct memory
of Taiwan’s modern history that may affect their partisan support. The ethnicity
of Taiwan voters is closely related to their voting behavior (Wang 1998; Lin
1989; You 1996).

Figure 2.3 shows that mainlanders have been highly supportive of the Pan-
Blue Alliance. Nearly 90 percent of mainlanders voted for its candidates in
every election since 1996. By contrast, support for Pan-Blue candidates from
the Minnan group has declined significantly since 1996 when the KMT
nominee, Lee Teng-hui, received more than 80 percent of the votes from his
fellow Minnan voters. In subsequent elections, the Minnan voters’ electoral
support for Pan-Blue candidates has fluctuated between 40 percent and 55
percent. Although the KMT-affiliated Ma won the election Page 32 → in 2012
with the majority support of the Minnan citizens, the gap between the
mainlander and the Minnan voters for the Pan-Blue Alliance remains as large
as 40 percentage points.
 Fig. 2.3. Percentages of three ethnic groups voting for the Pan-Blue. Data: See Appendix 2.A3.
Last, the Hakka are more centrist. The majority of the Hakka group generally
have provided their electoral support to Pan-Blue candidates. However, from
2000 to 2004, President Chen Shui-bian appointed many Hakka politicians as
cabinet ministers in his first term. The ministry-level government agency, the
Council for Hakka Affairs, and Hakka TV were also established in 2001 and
2003, respectively. Perhaps in consequence, Hakka voters swung to the DPP in
2004. However, the majority returned to the KMT in subsequent elections.
 The evidence thus shows that the Pan-Blue Alliance generally has been able to
garner majority support from the two minority ethnic groups on the island, the
Hakka and the mainlanders, and that mainlanders are its Page 33 → most loyal
supporters.3 The majority group, the Minnan, leans to the Pan-Green. The Pan-
Green alliance receives its electoral support primarily from the Minnan group,
with very little backing from the mainlanders. Thus the ethnic cleavage between
the mainlanders and the Minnan group is significant. Although mainlanders
make up only about 12 percent of the island’s population, their loyal support
for the Pan-Blue Alliance and their high turnout rate, together with typically
substantial Hakka backing for the KMT, suggest that the ethnic cleavage will
continue to play an important role in Taiwan’s politics.

In many countries, the parties are substantially based on income classes

because that is a central cleavage in the political system. As we have already
suggested in this book and as chapter 9 argues in detail, conventional left-right
economic divisions do not describe Taiwan politics well. However, income
might still matter to some degree. One of the legacies of the KMT’s
authoritarian rule has been Taiwan’s successful economic development in the
1970s. The land reform and the Ten Major Construction Projects, for instance,
helped bring rapid economic growth. While most Taiwan citizens have
benefited from these economic policies, it has been argued that the well-off and
the middle class have benefited most, and thus have generally been the most
loyal supporters of the KMT (Winckler 1992). Figure 2.4 confirms this
observation to some degree. Citizens of higher income appeared to support the
Pan-Blue candidates more than those of lower income by roughly 10
percentage points in the last three presidential elections.
 In addition, the figure shows a significant decline of support for the Pan-Blue
Alliance by low-income citizens, from about 90 percent in the 1996 election to
roughly 40 percent in 2004, and then stabilizing at about 55 percent. The
changing partisan support by less affluent voters may appear exaggerated due
to the “Lee Teng-hui Complex,” referring to citizens’ emotional attachment to
the island country’s first popularly elected president in 1996 (Hsu 1998, 2004).
His broad popularity was not likely to be repeated when the KMT returned to
mainlander candidates. Nonetheless, it is clear that in recent years, for whatever
reasons, less affluent Taiwan voters tend to be less supportive of the Pan-Blue
Alliance than their well-off countrymen. Yet as the graph shows, the impact of
income is much smaller than that of ethnicity. Taiwan politics is simply not
primarily about class conflict.

Another measure of class, closely associated with income, is respondents’

education. Figure 2.5 shows that Taiwan voters with college or graduate school
education tend to be more supportive of the Pan-Blue Alliance, while citizens
with a lower level of education tend to back Pan-Green candidates. Note that
the largest supporting gap among voters with different education Page 34

→ levels appeared in 2004, while partisan support for the Pan-Blue Alliance
nearly converged among educational groups in 2012. Again, the effects are very
small compared to those of ethnicity, nearly disappearing in the most recent
election.
 Fig. 2.4. Percentages of three income groups voting for the Pan-Blue. Note: The survey for the
2000 presidential election lacks the question on household income. See appendix 2.A1 for the
categorization of income groups. Data: See appendix 2.A3.
Figure 2.6 gives social class one last chance. It displays vote shares for Pan-
Blue candidates according to respondents’ occupation. The differences in
partisan support among different occupational groups appear to be rather
consistent during the past five presidential elections. Specifically, Taiwan voters
employed in the public sector, such as the military, governmental employees,
and public school teachers, are highly supportive of the Pan-Blue Alliance, by
roughly between 60 percent to 75 percent. This is followed by those in the
private sector, with a Pan-Blue support rate between 50 to 60 percent. Citizens
who earn their living by farming, fishing, and forestry Page 35 → are least likely
to support Pan-Blue candidates. The difference in support for the Pan-Blue
Alliance between government employees and citizens in the agricultural sector
are as high as 25 percentage points in the 2008 and 2012 elections. Since state
employees are devoted supporters of the KMT, the Pan-Green Alliance relies
on backing from voters in the agricultural and labor sectors. Thus there is
clearly an occupational difference here, but it is tied in part to traditional
mainlander dominance of government and to the reliance on fishing and
farming in southern Taiwan, where the population is disproportionately
Minnan. This raises the question of whether the effect of occupation is causal
or merely correlational. Chapter 12 sorts out whether occupation itself has an
impact on vote choice, or whether occupational differences simply proxy for
other electoral forces, especially ethnicity and national identity.
 Fig. 2.5. Percentages of three education groups voting for the Pan-Blue. Data: See appendix
2.A3.

Figure 2.7 raises the question of whether Taiwan politics is undergoing

generational change. It shows the vote shares for Pan-Blue candidates based
Page 36 → on respondents’ age. In general, the average vote and the over-time
trends of all age groups are strikingly similar. Voters in their 30s and 40s tend
to be a bit more supportive of the KMT and its political partners in recent
years. However, it is noteworthy that the younger generation in its 20s became
the age group least supportive of the Pan-Blue Alliance in 2012. While it
remains to be seen whether this represents a trend, it may be indicative of a
discontented younger generation that leaders of both the Pan-Blue and Pan-
Green alliances need to cultivate. In any case, the differences are quite small
and inconsistent in comparison with other factors examined in this chapter,
and the evidence best fits the notion that Taiwan’s electoral cleavages are
relatively stable and generally reproducing themselves in younger generations
thus far.

Fig. 2.6. Percentages of five occupation groups voting for the Pan-Blue. Data: See appendix
2.A3.
 Lastly, one can ask how the ethnic cleavages seen in this chapter manifest
themselves in partisanship. Party preferences are crucial: they structure not only
presidential choices but also votes for lower-level offices. They Page 37 → also

determine how voters see the political world. Thus it is important to ask how
different ethnic groups have evolved in their partisan attachments. For that
purpose, annual data are available, not just information from presidential years.

Fig. 2.7. Percentages of five age groups voting for the KMT. Data: See appendix 2.A3.
 Figure 2.8 show how KMT partisan loyalties of the three main ethnic groups
have evolved since 1992. Remarkably, KMT identification is nearly constant
among all three groups over a period of 30 years. And the differences among
them are very large, amounting to more than 30 percentage points. It is clear
that ethnicity is a key determinant of KMT partisanship.

Figure 2.9 gives the same information for DPP partisans. Prior to 2000, many
citizens disguised their anti-KMT leanings by calling themselves
“independents.” But since Chen Shui-bian’s first presidential race, DPP
partisanship, too, has been very stable within each ethnic group. Apart from
some disenchantment in the later years of the Chen presidency, the DPP has
held the same share of partisans it attracted in its first successful presidential
Page 39 → race. And, again, the differences between Minnan and mainlanders are
very large, amounting to nearly 25 percentage points in recent years.

Page 38 →

Fig. 2.8. Partisan identification with the Pan-Blue Alliance by ethnic groups, 1992–2012. Data
source: Core Political Attitudes among Taiwanese (Election Study Center, NCCU).
 Fig. 2.9. Partisan identification with the Pan-Green Alliance by ethnic groups, 1992–2012. Data
source: Core Political Attitudes among Taiwanese (Election Study Center, NCCU).
Thus in general, the evidence seems to suggest that Taiwan voters who are in
their 30s and 40s, with higher level of education, who are relatively well-off, or
are working in the public sector tend to be the most supportive of the Pan-
Blue Alliance. Candidates of the Pan-Green Alliance are more appealing to
citizens who are in their 20s, have low income, or are employed in the
agricultural or labor sectors. None of these factors has effects nearly as large as
ethnicity, which has by far the biggest impact on vote choices and on
partisanship. However, as it will be shown in chapter 3, with Taiwan’s
democracy becoming increasingly consolidated the ethnic cleavage among its
citizens has gradually been transformed and cloaked in their partisan
attachments.

Conclusion
Drawing on survey data about Taiwan voters’ electoral choices since 1996, this
chapter has provided a general characterization of Taiwan voters. It has shown
that the partisan preferences of citizens on the island are focused on the Pan-
Blue and the Pan-Green alliances. As Taiwan’s democracy has increasingly
consolidated into a competitive two-party system, each alliance has become
dominated by a single party, the KMT and the DPP, respectively.

The Pan-Blue Alliance has enjoyed a 5–10 percentage point electoral

advantage during the past three presidential elections. This electoral gap is not
constant across the island. For a variety of economic and historical reasons,
Taiwan’s political landscape is divided, with the Pan-Blue Alliance dominating
the northern part of the island and the Pan-Green Alliance receiving strong
support in the south.

The chapter has also shown that the two electoral alliances are based on
different parts of the voting population. At the risk of oversimplification, one
may say that citizens who are mainlanders, well-educated, more affluent,
government employees, or in their 30s and 40s are more likely to provide
electoral support for Pan-Blue candidates. On the other hand, voters who are
less educated, less well-off, working in labor and agricultural sectors, or in their
20s tend to identify with the Pan-Green Alliance. But all these differences are
dwarfed by the impact of ethnicity, which has very large impacts on both vote
choices and on partisan identity.

This description of the support base of each party group raises a host of
questions. Why are people in each of these sectors drawn disproportionately to
one party rather than another? Are the differences primarily ideological, Page 40

→ while demographic differences are merely incidental? Or do identity

differences shape both opinions and partisan loyalties? Or does everything
reduce to the economy and pocketbook voting? And why has popular support
come to be focused on just two parties? Subsequent chapters take up these
questions.

Page 41 →

Appendix 2.A1

Wording and coding of demographic variables

Age. In what year were you born? 20–29 years old=1, 30–39 years old=2, 40–49 years old=3, 50–59
years old=4, 60 years old and above=5.

Education. What is the highest degree that you finished? Elementary school and below=1, high
school=2, college/graduate school=3.

Occupation. What is your current occupation and position?

Income. What is your household’s total income? See appendix 2.A2 for classification.

Ethnicity. What is your father’s ethnicity? Minnan=1, Hakka=2, mainlander=3.

Region. Where do you live? Taipei City, Taipei County and Keelung City, Taoyuan County, Miaoli
County, Hsinchu City, Hsinchu County Taichung City, Taichung County, Chunghwa County, Nantou
City Tainan City, Tainan County, Kaohsiung City, Kaohsiung County, Pingtung County, Penghu
County, Ilan County, Hwalien County and Taitung County.

Appendix 2.A2

Classification of Three Income Groups

According to the yearly family income data published by the Directorate-
General of Budget, Accounting and Statistics, the monthly family income of
the 20th percentile and 80th percentile is calculated. Based on these two
numbers, respondent’s income is categorized as low, middle, and high. For
example, the 80th percentile of monthly family income was NTD 131,798 and
the 20th percentile was NTD 24,680. Respondents who answered “no income,”
“below NTD 15,000,” and “between NTD 15,000 and 30,000” fall into the low
income group. Respondents who answered “between NTD 100,000 and NTD
200,000” and “more than 200,000” belong to the high income group.
Respondents whose income is higher than NTD 30,000 but lower than NTD
100,000 are defined as the middle income group.

In 1998 and 2000, respondents were asked their personal income instead of
their household income. Because it is difficult to translate between these two
measures, we dropped those years from the tabulation.

Page 42 →

Appendix 2.A3

List of Datasets

An Interdisciplinary Studies of Voting Behavior in the Presidential Election in 2000 (ESC 1996).
Principal investigator: Yih-yen Chen.
An Interdisciplinary Study of Voting Behavior in the 1996 Presidential Election (ESC 2000). Principal
investigator: John F. S. Hsieh.
Taiwan’s Election and Democratization Study, 2002–2004 (III): The Presidential Election, 2004
(TEDS2004P). Principal investigator: Shiow-duan Hwang.
Taiwan’s Election and Democratization Study, 2008 (TEDS2008P). Principal investigator: Ching-hsin
Yu.
 Taiwan’s Election and Democratization Study, 2009–2012 (III): The Survey of the Presidential and
Legislative Elections, 2012 (TEDS 2012). Principal investigator: Yun-han Chu.
Page 43 →

Notes

1. Prior to 1996, Taiwan’s president was indirectly elected by its now

defunct National Assembly.

2. The reality of the assassination attempt is disputed between the parties.

3. Aboriginals also give the KMT majority support, so that the KMT is in
many respects a coalition of minorities, while the DPP largely represents
the majority subethnic group. Many other countries have a similar party-
bloc structure, including the United States, in which the Republicans get
heavy support from the white majority while the Democrats are
disproportionately a coalition of racial, ethnic, and religious minorities.

References

Archer, J. Clark, and Peter J. Taylor. 1981. Section and Party: A Political Geography
of American Presidential Elections from Andrew Jackson to Ronald Reagan. New
York: Research Studies Press.

Hsu, Huo-yan. 1998. “The Political Psychology of Lee Teng-hui Complex and
Its Effects on Voting Behavior” [in Chinese]. Xuan Ju Yan Jiu [Journal of
electoral studies] 5 (2): 35–71.
Hsu, Huo-yan. 2004. “Taiwan-China Complex and Taiwan-China Concern:
Symbolic Politics in Taiwan’s Elections” [in Chinese]. Xuan Ju Yan Jiu
[Journal of electoral studies] 11 (2): 1–41.

Lin, Chia-lung. 1989. “Opposition Movement in Taiwan under an

Authoritarian-Clientelist Regime: A Political Explanation for the Social
Base of the DPP” [in Chinese]. Taiwan She Hui Xue Ji Kan [Taiwan: A
radical quarterly in social studies] (Spring): 117–43.

Tindall, George B. 1972. The Disruption of the Solid South. Athens: University of
Georgia Press.

Wang, Fu-chang. 1998. “Ethnic Consciousness, Nationalism, and Party

Support: Taiwan’s Ethnic Politics in the 1990s” [in Chinese]. Taiwan She Hui
Xue Kan [Taiwanese sociological review] 2:1–45.

Winckler, Edwin A. 1992. “Taiwan Transition.” In Political Change in Taiwan, ed.

Tun-jen Cheng and Stephan Haggard, 221–59. Boulder, CO: Lynne
Rienner.

You, Ying-lung. 1996. Ming-yi yu Taiwan cheng-chi bien-chien [Public Opinion and
Changing Taiwan Politics]. Taipei: Yue-Tan Press.
Page 44 → Page 45 →
 Chapter 3

Changing Boundaries

The Development of the Taiwan Voters’ Identity

T. Y. Wang

The ubiquitous effect of identity in contemporary politics has been widely

recognized in the scholarship of the social sciences. Political instability within
societies or bitter conflicts between nations has frequently been attributed to
differences in identity politics (Horowitz 2000; Isaacs 1975), and Taiwan is no
exception to this pattern. Indeed, the issue of identity has important
implications not only for the island country’s domestic politics but also for its
relationship with China; hence, Taiwan’s identity politics may affect peace and
stability in East Asia.

Taiwan voters’ identities have undergone substantial changes since the

Nationalist (Kuomintang) government led by Chiang Kai-shek retreated to
Taiwan in 1949. Since then, the “China factor” has played an essential part in
the formation and change of the island citizens’ identities. The re-Sinicization
efforts by the authoritarian KMT government, the infusion of two million of
Chiang’s followers, the rapid democratization in the ensuing decades, and the
continuing rivalry between China and Taiwan have transformed the island
residents’ identity in form and in substance. This chapter aims to trace the
trajectory of Taiwan voters’ identity change. With the notion of “boundary”
framed in the theoretical frameworks of primordialism and constructionism,
this study employs quantitative survey data collected over two decades and
qualitative information gathered by focus-group interviews and in-depth
interviews.1 It argues that the substance of identity is relational because it
involves the distinction between what it is and what it is not. The comparison
and reference to other groups thus establish the Page 46 → boundary, and hence
the identity. As the boundary changes, identity evolves as a result. The political
environment and the policies adopted by political elites on both sides of the
Taiwan Strait since World War II provide the impetus to the formation and the
change of Taiwan voters’ identities. Specifically, the initial boundary along the
ethnic line between Taiwan’s local residents, known as “Taiwanese,” and
“mainlanders” during the authoritarian era, changed to the contestation
between “Taiwanese consciousness” and “Chinese consciousness” as the island
country entered the period of democratization. With China’s forceful claim
over the island, the Beijing government’s hostility became one of the main
driving forces in the growth of a distinct national identity. The boundary of
identity politics on the island has thus shifted from the ethnic divide to the
characterization of Taiwan as an independent state.

Boundary Change: The Development of Identity

Identity can be understood as a psychological attachment to a social/political

category. The substance of identity is relational because it involves the
distinction between what it is and what it is not (Abdelal et al. 2009; Barnett
1999). The comparisons and reference to other identities therefore sets up
“boundaries” between groups of individuals that facilitate the formulation of
the “we-group” and the “they-group.” In this sense, the development of
identity is dependent on individuals’ interactions with others as well as the time
and the place that the interaction occurs. As interactions and the context differ,
the boundary changes and the content of identity may vary as a result. Even
though a certain identity may bear the same label, its substance may be
different as boundaries shift. Research on identity thus needs to examine the
impetus to identity formation and change. Primordialism and constructionism,2
the two most commonly invoked identity perspectives, provide the theoretical
underpinning of the analysis.

In the discussion of the development of identity, proponents of primordialism

maintain that identities are deeply rooted in primordial attachments, which, as
Charles Cooley and Edward Shils conceptualized, are “not merely to the other
family member as a person, but as a possessor of certain especially ‘significant
relations’ qualities” (Shils 1957, 142). These qualities stem from “the givens . . .
of social existence” that involve “immediate contiguity and kin connection
mainly, but beyond them the givenness that stems from being born into a
particular religious community, speaking a particular language, or even a dialect
of a language, and following particular social Page 47 → practices” (Geertz 1994,
31). These primordial bonds, which are based on blood, speech, and custom,
define human relations and establish boundaries among groups of people.
These boundaries then have the power of generating psychological attachments
to groups. Proponents of primordialism thus maintain that identities are deep-
rooted in blood, culture, and language. Although the effects of such primordial
ties may vary from individual to individual, their impacts are long lasting. As
Geertz (1994) indicated, “for virtually every person, in every society, at almost
all times, some attachments seem to flow more from a sense of natural—some
would say spiritual—affinity than from social interaction” (31). As a result,
identities rooted in blood, culture, and languages are assumed to be given and
immutable.

The constructionist perspective refutes the primordialist view that identities

are inherited and underivable (Gellner 2008; Laitin 1998). Constructionists
maintain that identities can be fostered or intentionally constructed and that
when social interactions change, identity can alter as a result. In this
conception, boundaries can be artificially established. First, societies have
various socialization mechanisms, such as education and family. Through these
media, individuals are exposed to the customs and conventions of various
groups, cultures, and norms of the society as well as to the historical memories
of the nation. Feelings of distinctiveness from other groups of people are
formed, and psychological attachments to these entities are created.
Furthermore, most nations maintain an interpretative version of history that
frequently emphasizes the oppression and exploitation of the people. To form a
hostile “they-group” and unified “we-group,” political elites frequently employ
state institutions to exploit such unequal and exploitative historical memories.
With the objective of mobilizing their fellow citizens, this collective group
consciousness becomes an essential part of the public identity. Sometimes, in an
attempt to change an existing identity or create a new identity, the state
intentionally recognizes one or a few groups and ignores others. This is
frequently done by subsidizing or celebrating the chosen groups’ languages,
cultures, or historical memories through education, propaganda, and other
political means. As Gellner (2008, 54) succinctly points out in his discussion of
national identity, “nationalism uses the pre-existing, historically inherited
proliferation of cultures, or cultural wealth, though it uses them very
selectively, and it most often transforms them radically. Dead languages can be
revived, traditions invented, quite fictitious pristine purities restored.” As a
result, the celebrated group identity becomes the dominant one in the society,
and a public identity is likely to be constructed through the engineering of the
state. From the constructionist perspective, identities are more than the
function of race and ethnicity, Page 48 → and they are not unchangeable and
underivable. Because boundaries can be intentionally manipulated through
policies and socialization media, identity can be constructed through deliberate
efforts.

The China Factor and Taiwan’s Changing Political

Environment

The Boundary of Ethnicity

As noted in chapter 1, Taiwan’s modern history has been closely linked to

political development on the Chinese mainland. After the Qing dynasty
established administrative control of the island in the seventeenth century,
thousands of residents in coastal provinces migrated to Taiwan seeking a better
life in the subsequent decade. Taiwan’s sovereignty was later ceded to Japan by
China’s last imperial dynasty in 1895. Unlike the Chinese mainland, which was
plagued by wars and chaos in the ensuing half century, Taiwan advanced to
become a more modern society during the same period under the assimilation
policy of the Japanese colonial authority. Although many Taiwan residents
received a Japanese education, they still longed for their Chinese ancestral
roots. The population was overjoyed when the island’s sovereignty was
returned to the Chinese government, which was controlled by the KMT at the
end of World War II. The local residents’ enthusiasm for returning to the
“motherland” soon evaporated because the ruling authority dispatched to
Taiwan in 1945 was corrupt and its policies discriminated against the local
residents. As a result, many Taiwanese elites were filled with nostalgia for
Japanese colonial rule, which was remembered as being efficient and honest.
KMT officials in turn viewed the islanders with suspicion since they were
under the influence of the Japanese Empire while the KMT government was
fighting Japanese invaders on the mainland. The animosity between the KMT
government and local residents, particularly the Minnan group, finally
culminated in a bloody crackdown on the Taiwanese elites by KMT troops, a
tragic event known as the 2–28 Incident of 1947. This outbreak of hostility
solidified the local perception of the KMT government as simply a new foreign
regime and occupying force (Lai, Myers, and Wou 1991).

The ethnic division on the island deepened further after the KMT
government retreated to Taiwan after its humiliating military defeat on the
mainland. Along with the government came two million of its followers, who
were known as mainlanders. Insisting that the Republic of China (ROC) had
the legitimate and sovereign claim over China, including both Page 49 → Taiwan

and the Chinese mainland, Taipei’s ruling elites imposed harsh authoritarian
rule. As it was stated in chapter 1, a series of prejudicial measures were
implemented to “re-Sinicize” local residents to accept the view that Taiwan
was part of a greater China and to foster their Chinese identity. By recognizing
and celebrating China-centered historical memories, activities that might
arouse local identities and promote Taiwan independence were censored and
suppressed. Meanwhile, local residents’ freedom of expression and association
was substantially curtailed under martial law. Through interpretations of the
constitution by the Grand Justice Court, clauses relevant to presidential term
limits and regular parliamentary elections were suspended. Using all of China
as the “imagined community,”3 local residents’ participation in legislative
politics was restricted because Taiwan was deemed only one of the 36
provinces of the nation, and their representation in the parliament was treated
as such (Wang 2008). The net effect of these authoritarian measures was that a
small group of mainland elites effectively monopolized political power on the
island, especially at the national level. The perception that mainlanders were the
oppressor and local Taiwanese were the oppressed underlined the ethnic
cleavage on the island. This ethnic divide thus became the first boundary
among the island’s residents in Taiwan’s modern history. Many mainlanders
believed that the only way to survive in this unfriendly environment was to
support the KMT government. This rational calculation and their emotional
affinity to the Chinese mainland strengthened the mainlanders’ espousal of
Chinese consciousness, which also turned into a partisan identification with the
KMT.

Although the boundary between Taiwanese, particularly the Minnan group,

and mainlanders continued to exist during the subsequent several decades, the
ethnic division as a major social cleavage increasingly lost its political
importance. First, constant social contacts among various ethnic groups in
schools, workplaces, and other social settings help to narrow the differences
between them. Most islanders’ ability to speak the official language—Mandarin
—and the ability to converse in local dialects by some mainlanders also
enhanced mutual understanding among ethnic groups on the island.
Interethnic marriages over the past several decades have blurred ethnic lines
and produced a new generation that has been more accepting of different ideas
from all ethnic groups. Most important, the rapid democratization on the
island in the 1980s also contributed to weakening the significance of the ethnic
divide. With the lifting of martial law, freedom of expression and association
became constitutional rights protected by law. After Lee Teng-hui became the
first native-born president in 1988, and later with the passage of constitutional
amendments in the mid-1990s that stipulate Page 50 → popular elections of the
presidency and a new parliament, mainlanders could no longer monopolize
political power. Meanwhile, local politicians began to advocate such slogans as
“collectivity of common fate” (sheng-ming-gong-tong-ti), “the rising new nation”
(xin-xing-min-zu), and “the new Taiwanese” (xin-Taiwanren) in an attempt to
ease the ethnic tension between Taiwanese and mainlanders and garner
electoral support. As local residents’ sense of relative deprivation in the power
distribution dissipated, along with politicians’ conscious efforts to allay ethnic
tension, the antagonistic feeling between the mainlander and other ethnic
groups gradually attenuated.

Empirical research employing interpretative approaches and quantitative

analyses supports this conclusion (Chang 2006; Cheng 2009; Pao 2009; Wang
2008). Table 3.1 presents survey responses from the Minnan and Hakka groups
on their views of Taiwan’s ethnic relations. It shows that by 2006 about 40
percent of Minnan and Hakka respondents or their close relatives had
intergroup marriages with mainlanders, and 67 percent of them felt they got
along with mainlanders as equally well as with those of their own ethnic
groups. About 69 percent believed that they had the same social status as
mainlanders. As the boundary between local residents and mainlanders has
become increasingly blurry, ethnic identity has lost its power of differentiation.
That said, this does not mean that there is no political difference among
Taiwan’s major ethnic groups. As chapter 2 has shown, the Page 51 → ethnic

cleavage among Taiwan citizens has gradually been transformed and cloaked in
their partisan attachments. That is, the mainlanders as a group consistently
have a disproportional partisan attachment to the Pan-Blue political alliance,
whereas members of the Minnan group present the opposite pattern in their
party identification with the Pan-Green camp. Alternatively, a new boundary
between Taiwanese consciousness and Chinese consciousness has increasingly
become prominent on the island.

Table 3.1. Minnan and Hakka on Ethnic Tension

N = 934)
Are you or any of your close relativesmarried to a mainlander? ( %
Yes 41.8

No 56.4

No response 1.7

Who do you get along with? ( N = 1082) %

Minnan and Hakka 24.1

Mainlander 2.7

Both about the same 67.0

No response 6.2

Do you feel those mainlanders you know have higher social status than you? ( N = 934) %
Better 14.2

About the same 69.1

Worse 2.8

No response 13.9

Source: Chen 2006.

The Boundary of Chinese/Taiwanese Consciousness

The contestation between Chinese consciousness (zhong-guo-yi-shi) and

Taiwanese consciousness (Taiwan-yi-shi) concerns the interpretation of the
relationship between Taiwan and the Chinese mainland.4 In the conception of
Chinese consciousness, “China” is a term that encompasses both cultural and
political significance. To be Chinese implies belonging not only to a particular
ethnic and cultural group but also to the political identity of the Chinese state
known as zhong-guo, in which Taiwanese culture is a part of Chinese culture and
the Taiwanese people are a part of the Chinese population. Since the island of
Taiwan is an integral part of China, the eventual unification of Taiwan with the
Chinese mainland is regarded as both natural and inevitable. Those who
espouse Taiwanese consciousness, by contrast, challenge the idea that Taiwan is
a part of China. They argue that the one-hundred-year separation of Taiwan
from the Chinese mainland has created a Taiwanese culture distinct from that
of the Chinese mainland. The ideas that Taiwan and China are one nation and
that all Chinese must be ruled by a single government within the same state are
thus rejected. Some even assert that “Taiwanese are not Chinese” and have
opposed the assertion that Taiwan is a part of China in any sense (Shih 1992,
quoted in Huang 1993, 49). While advocates of Chinese consciousness behold a
China-centered interpretation of the island’s relations with the Chinese
mainland, proponents of Taiwanese consciousness take a Taiwan-centered view
of the relationship.

As indicated earlier, during its authoritarian rule of the island, one of the
major tasks of the KMT government was to foster Chinese consciousness and
its accompanying Chinese identity so that local residents would accept the view
that China was their motherland. Throughout the several decades after 1949,
the leaders of the KMT upheld the principle of “one China,” of which Taiwan
was considered to be a part. Even after Lee Teng-hui became president, he was
careful not to challenge the One China principle for fear of offending the
party’s old guard. Presiding over the newly established Page 52 → National

Unification Council, Lee promulgated the National Unification Guidelines in

March 1991, symbolizing the commitment of his administration to Taiwan’s
eventual unification with the Chinese mainland. The rapid democratization
that was set in motion in the late 1980s nevertheless later made the advocacy of
Taiwan independence into a constitutional right protected by law. The previous
China-centered school curriculum was also revised and shifted to a new
emphasis on the island’s history and culture (Wang 2001). In particular, local
politicians’ advocacy of such slogans as “the new Taiwanese” had the effect of
fostering a new identity on the island. By the mid-1990s, Taiwanese
consciousness has become the strongest alternative to Chinese consciousness.

The rivalry between the two perspectives rose to its zenith during the eight-
year presidency of Chen Shui-bian, of the DPP, from 2000 to 2008. Chen has
strong pro-independence credentials, and his affiliated DPP is the only major
political party on the island that advocates the island’s de jure independence. To
raise the local population’s Taiwanese consciousness and identity, the Taipei
government under Chen’s leadership launched a series of de-Sinicization
initiatives on the island. As a result, Chinese consciousness and its related
Chinese identity were under severe attack during Chen’s presidency. Like the
Chiang regime, the Chen administration selectively recognized a portion of
Taiwan’s historical memories, in particular those related to the 2–28 Incident.
Other measures were adopted to promote an image that Taiwan is a separate
political entity from China, including adding the phrase “Issued in Taiwan” to
the island country’s green passport cover and dropping the national emblem of
the ROC as the official logo of Taiwan’s overseas missions. With the revision
of history textbooks, education has been used to change the collective
memories of the Taiwan’s citizens (Wang 2001; Wang and Chang 2005). The
Chen administration’s efforts to strengthen Taiwanese identity reached a climax
when the Rectification Movement (zheng-ming-yung-dong) was launched in 2007
by replacing “China” in the names of all relevant government and state-run
agencies with “Taiwan.”5 It also renamed the Chiang Kai-shek Memorial Hall
and proposed a referendum for UN membership that was held during the 2008
presidential elections (Enav 2007; Mo and Shih 2007). All of these measures
aimed to celebrate Taiwanese consciousness for its own sake and as a way of
rejecting Chinese consciousness. As the contestation between the two
perspectives became increasingly prominent on the island, the divide was
manifested in the changing patterns of the island residents’ self-identification as
a Taiwanese or a Chinese.

During the period between 1992 and 2012, surveys conducted by the Page 53

→ Election Study Center, National Chengchi University, in Taiwan, asked

respondents the following question:
 Fig. 3.1. Identity change by ethnicity, 1992–2012. Data source: Core Political Attitudes among
Taiwanese (Election Study Center, NCCU).
In our society, some people consider themselves Taiwanese, and others view themselves as Chinese,
while still others see themselves as both Taiwanese and Chinese. What is your view on this matter?

Figure 3.1 shows that the island citizens’ identity has undergone significant
change over the course of past two decades, signifying the vicissitudes of
Taiwanese consciousness and Chinese consciousness. Indeed, as figure 3.1.a
demonstrates, the overall ratio of Taiwanese identifiers has increased
substantially since 1992, from about 20 percent to 40 percent in 2000, and has
continued rising. When President Ma Ying-jeou of the KMT was inaugurated
in 2008, more than half of the island’s citizens considered themselves
Taiwanese. Four years later, the percentage of Taiwanese identifiers increased
to 56 percent, an increase of more than 45 percent from two decades ago.
During the same period, the proportion of Chinese identifiers on the island
Page 54 → declined significantly, from close to 30 percent in 1992 to about 4
percent in 2012, a decline of about 25 percent. Interestingly, the percentage of
dual identifiers on the island—those who see themselves as both Chinese and
Taiwanese—has been quite steady, hovering around 45 percent since 2001, even
though it also witnessed a slight decline by 2012.

The pattern of identity change appears to occur across ethnic lines for the
three largest ethnic groups on the island. Figures 3.1.b and 3.1.c show that the
changing patterns of both Minnan and Hakka groups resemble the overall
trend in Taiwan, as increasing proportions of the two groups consider
themselves Taiwanese while the number of Chinese identifiers has dropped
significantly, to less than 5 percent by 2012. Currently, a substantial proportion
of Minnan and Hakka, about 35 percent and 45 percent, respectively, hold dual
identity, considering themselves both Chinese and Taiwanese. For the
mainlanders, the ratio of Chinese identifiers also witnessed a significant decline
during the same period, from about 50 percent in 1992 to around 10 percent in
2012, as figure 3.1.d shows. Like the Minnan and Hakka groups, the number of
Taiwanese identifiers among mainlanders has increased during the same period,
but substantially less so by comparison. Since 2004, more than 60 percent of
them accepted that they are Chinese as well as Taiwanese.

Figure 3.2 also presents the distribution of identity change for the five
political generations in Taiwan.6 As with the overall trend, there was a steady
increase of respondents holding Taiwanese identity and a continual decline of
Chinese identifiers for each of the five generations. By 2012, more than half of
all age groups were Taiwanese identifiers, while the proportions of Chinese
identifiers fell to about 5 percent, except for respondents of the first
generation, who were at least 81 years old and increasingly dying off. What is
worth noting is that more than 60 percent of the fifth generation are holders of
Taiwanese identity while very few of them are Chinese identifiers. That 40
percent to 50 percent of the third and the fourth generations are holders of
dual identity probably is because these respondents received the China-
centered education during their formative years under the KMT authoritarian
rule.

What are the island citizens’ views on the meaning of being

Chinese/Taiwanese? An examination of the qualitative data from focus-group
interviews conducted since 2000 shows that Taiwan voters tend to take a
primordial view of being “Taiwanese.” A recurrent statement made by all
interviewees is that “I was born here, I grew up here, and therefore I am a
Taiwanese,” even though they do not reject their Chinese cultural origin. In
response to questions on the definitions of Chinese and Taiwanese, a third-
generation Page 56 → interviewee stated in 2000 that “Taiwanese are those who
live in Taiwan. We have the same customs . . . and we are ‘made in Taiwan’”
(2000SB05). For the fourth- and fifth-generation interviewees, the answer is
more straightforward: “Chinese are those who were born in China and
Taiwanese are those who were born in Taiwan” (2011TSQI06) and “those who
were born in Taiwan are Taiwanese” (2011TSQI07).
 Page 55 → Fig. 3.2. Chinese/Taiwanese identity by generations, 1992–2012. Data source: Core
Political Attitudes among Taiwanese (Election Study Center, NCCU).
Such a primordial conceptualization of Taiwanese identity appears to cross
the ethnic line. Some mainlanders recognized that their immediate ancestors or
even themselves were born on the Chinese mainland, but the very fact that they
had been living in Taiwan for half a century led them to espouse Taiwanese
identity. A first-generation mainlander contended in 2002 that “Taiwanese are
those who live in Taiwan. Even though I came from the mainland, I am a
Taiwanese as I have lived here for half a century” (2002–31). Another
mainlander stated that “I am a Taiwanese. I was born here and grew up here. I
drink Taiwan’s water and eat Taiwan’s rice. There is no mistake that I am a
Taiwanese” (2011CUSQD06).

This primordial conceptualization also crosses the generational line, and in

fact the tendency becomes stronger as interviewees get younger in age. For the
fifth-generation respondents, to define who they are by birthplace is the most
natural and straightforward approach. For instance, a 2011 fifth-generation
interviewee defined “Taiwanese” in a blunt manner: “those who were born in
Taiwan are Taiwanese” (2011TSQI07).

While there appears to be a consensus about the meaning of Taiwanese,

respondents of different ethnic groups and generations have defined “Chinese”
in two different but noncontradictory ways. Some conceptualize the term in a
primordial sense, treating the association of culture and ethnicity as the
defining characteristic of being Chinese. Since residents in Taiwan and on the
Chinese mainland share many cultural and ethnic traits, “Taiwanese” is
considered a subgroup of “Chinese,” and thus logically “Taiwanese” is
“Chinese,” broadly defined. Others define the term with a political meaning by
equating “Chinese” with “the citizens of the People’s Republic of China
(PRC)” and “China” with “the Chinese state” or “the PRC regime.” In this
conceptualization, they maintain that since Taiwanese are not citizens of the
PRC, they are not “Chinese.” Thus, in the mind of Taiwan voters, the term
“Chinese” may have separate cultural and political connotations. Along with
the quantitative survey data presented earlier, it appears that the vast majority
of the island’s citizens have employed the term of “Chinese” with a political
meaning. This may explain why few of them are Chinese identifiers: they do not
consider themselves citizens of the PRC. If they have emotional affinity with
the country of China (zhong-guo), many respondents adopted Page 58 → dual

identity, of which its Chinese element has a cultural meaning. In other words,
when the term “Chinese” (zhong-guo-ren) is applied to Taiwan voters, it has less
of the political designation that is traditionally associated with a Chinese polity
and more of the cultural and ethnic connotations that should be more
appropriately understood as “ethnic Chinese” (hua-ren).
 Page 57 →

Fig. 3.3. Identity and partisan support, 1992–2012. Data source: Core Political Attitudes among
Taiwanese (Election Study Center, NCCU).
Thus, the contestation between Taiwanese consciousness and Chinese
consciousness has increasingly tilted toward the former as more and more
citizens on the island become Taiwanese identifiers. Such a trend applies to all
three major ethnic groups in Taiwan, including mainlanders, who traditionally
espouse Chinese consciousness. While many mainlanders continue to view the
word “China” as a term representing a culture, a nation, and a state and are
holders of a dual identity, the majority of them have become indigenized to the
point of accepting a part of their identity as being Taiwanese. However, like the
ethnic divide on the island, Chinese/Taiwanese consciousness has also been
cloaked by partisanship. Figure 3.3 shows that the island citizens holding
Taiwanese identity tend to be supporters of the Pan-Green Alliance, led by the
DPP, whereas those with Chinese identity and dual identity are more likely to
develop a partisan attachment to the Pan-Blue Alliance, led by the KMT.
Indeed, as subsequent chapters show, the KMT tends to be perceived by
Taiwan voters as a party holding Chinese consciousness with a pro-unification
stand, and the DPP as a party holding Taiwanese consciousness with a pro-
independence position. Meanwhile, as the divide of Chinese/Taiwanese
consciousness is increasingly cloaked by partisanship within a domestic context,
the public discourse on the cleavage continues on the island but has shifted to
a new boundary.7

The Boundary of National Identity

In addition to its domestic element, the rivalry between Chinese consciousness

and Taiwanese consciousness also has an external dimension due to Beijing’s
forceful claim of the island. After being defeated by Communist troops in
1949, the KMT government, led by Chiang Kai-shek, retreated to Taiwan to
continue what was perceived as the sacred mission of “recovering the
mainland” from the Chinese Communists. By the late 1960s, the Chiang
family’s ambitious goal of national unification by the ROC, headquartered in
Taipei, had become increasingly unlikely. The passage of UN Resolution 2758
in 1971, which recognized the Beijing government as “the only lawful
representatives of China to the United Nations,” symbolized the decisive
victory of the PRC in the cross-Strait competition for international Page 59

→ legitimacy. Since then, Beijing leaders have steadfastly applied their version
of the One China principle in the international community: that “there is only
one China in the world, Taiwan is a part of China and the government of the
PRC is the sole legal government representing the whole of China” (TAO
2000). Refusing to recognize the legitimate existence of the ROC on Taiwan,
Chinese leaders maintain that the island has no legal right to establish
diplomatic relations with foreign countries or to participate in any
international organizations requiring statehood as a condition of membership.
Over the years, the increasing importance of China in international affairs has
led many countries to break diplomatic relations with Taipei as a necessary
condition for establishing formal ties with the Beijing government. The
number of states recognizing the ROC dropped significantly in the 1970s and
1980s. To participate in international organizations, most of which are
nongovernmental in nature, the Taipei government frequently had to
compromise on its name, flag, and national anthem in order to meet Beijing’s
demands (Lee and Wang 2003; Wang, Lee, and Yu 2011). As of 2013, only
twenty-three countries officially recognized the Republic of China, and Taipei
was in just a handful of international governmental organizations. Beijing’s
tactic of diplomatic isolation is humiliating to Taiwanese citizens and seriously
hurts their dignity because they believe that Taiwan is more qualified to enjoy
the legitimacy as a state than such UN members as Nauru and the Marshall
Islands.
 Indeed, Taiwan witnessed significant economic growth and political
development even as its international space was being squeezed by Beijing’s
tactic of diplomatic isolation. The island’s manufacturing sector went through
rapid expansion, with the country’s GDP growing at about 9 percent per year
during the two decades starting in 1970.8 As one of the largest trading nations
in the world and ranking among the top holders of foreign exchange reserves,
Taiwan was considered one of the newly industrialized countries. In addition
to its economic achievements, Taiwan also witnessed rapid democratization in
the 1990s. With political power being peacefully transferred between political
parties through regular elections since 2000, Taiwan has become a genuine
democracy. The island country’s lack of international standing simply does not
match its political achievement and status as a world-class economy, and
Taiwanese citizens blame Beijing for the country’s lack of international
legitimacy (Wang 2006).

Meanwhile, Beijing has also employed the tactic of military coercion. Refusing
to renounce the use of military force to resolve the cross-Strait dispute,
Chinese leaders have deployed more than 1,000 short-range ballistic missiles
along China’s coast area, targeting them at Taiwan. In order to weaken the Page

60 → island’s defensive capability, Beijing has deliberately prevented any country

from selling weapons to Taiwan. When Chinese leaders launched the eight-
month-long missile test and military exercises in 1995–96 and issued saber-
rattling threats in 2000, China became a genuine menace to the island citizens’
democratic way of life and economic prosperity (Garver 1997; Wang 2001).
Thus, Beijing’s tactics of diplomatic isolation and military coercion have invited
Taiwanese citizens’ resentment. As a hostile “they-group” bearing the name
“China” has formed in the mind of the island’s citizens, the boundary of
Chinese/Taiwanese consciousness that was largely confined to domestic
politics has increasingly shifted to the level between a “Chinese state” and a
“Taiwanese state.”

Fig. 3.4. Taiwan voters’ views on Beijing’s friendliness toward the Taipei government, 1996–2012.
Data source: Core Political Attitudes among Taiwanese. (Election Study Center, NCCU).
Figure 3.4. shows that a large proportion of Taiwanese citizens have
consistently sensed an “unfriendly” Chinese government since 1996. The
feeling of Beijing’s antagonism toward the Taipei government reached its
zenith during the period 2000–2008, when President Chen was in office, as 60
percent to 70 percent of Taiwan voters perceived the animosity from across the
Taiwan Strait. Data in table 3.2 demonstrate that there is a strong association
between the island citizens’ feeling of Beijing’s enmity and their holding of
Taiwanese identity. Qualitative data from focus-group interviews provide an
interpretative understanding of this relation. In addition to the primordial ties
mentioned in the last section, many focus-group respondents also justified their
holding of Taiwanese identity by their frustration at Beijing’s “suppression”
(da-ya) of Taiwan in the international community. One of the respondents
commented that “I was born here. . . . China continues bullying Page 61 → and
suppressing us, which makes me feel that Taiwan is not like a country. When I
travel abroad, I feel no country likes us [to be treated unfairly by Beijing]. . . . I
am a Taiwanese. . . . I was born here. . . . I never feel that I am Chinese. . . . I am
totally different from them” (2010S401). Some of them cited Taiwanese
athletes’ inability to carry the island country’s national flag and to sing the
national anthem in international sport competitions as examples of Beijing’s
“bullying practices” (ya-ba). As another respondent complained, “Because
Chinese doesn’t allow us to carry our national flag in international baseball
games or Olympic competitions, . . . I feel they are always harassing us. . . . I
feel they are bullying us. . . . [Interviewer: Do you think you are both Chinese
and Taiwanese?] . . . No, I am just a Taiwanese!” (2010N403).

The resulting humiliation and resentment prevented Taiwan citizens from

identifying with a hostile “China.” At the same time, the loss of
international Page 62 → legitimacy by the ROC has made the sense of belonging
to “Taiwan” as a state the only distinction between them and citizens of China.
The island residents’ bitter feelings of Beijing’s diplomatic “bullying” has thus
hardened their identity with an independent Taiwan that is separate from
China. As one respondent commented, “I feel I am a Taiwanese. . . . Because I
was born here and grew up here . . . I feel that we really should tell people that
[we are Taiwanese]. If we do not speak out, we would disappear . . . then when
we go abroad, foreigners will not respect you and you will feel that you don’t
have a country any more” (2010N302). Another stated that “I will always
emphasize that I am a Taiwanese when I go abroad” (2011TSQF04).
Quantitative survey data in table 3.3 confirm that this sentiment is widespread,
as about 80 percent of Taiwanese identifiers and 50 percent of dual identifiers
now view Taiwan as an independent state. As the boundary between Taiwanese
and Chinese gradually loses its power of differentiation within a domestic
context, the external dimension of the boundary becomes increasingly
prominent. The public discourse on the Chinese/Taiwanese distinction has
shifted to the level of national identity.

Table 3.2. Perceptions of Beijing’s Friendliness and Chinese/Taiwanese Identity

Friendly Beijing Unfriendly Beijing Total

21.2 78.8 61.9

Taiwanese identity
13.1 48.8 (826)

49.9 50.1 34.4

Dual identity
17.2 17.2 (459)

64.0 36.0 3.7

Chinese identity
2.4 1.3 (50)

100% 100% 100%

Total
(436) (899) (1335)

Source: Cheng 2013.

Note: Row percentages on top, total percentages underlined, and numbers of cases in parentheses.
Table 3.3. Identities and Views on Taiwan as an Independent State

Taiwan and China are two Taiwan and the Chinese mainland are
Total
independent states parts of China

Taiwanese 80.2 19.8 58.8

identity 47.2 11.6 (637)

48.2 51.8 37.6

Dual identity
18.1 19.5 (407)

Chinese 20.5 79.5 3.6

identity 0.7 2.9 (39)

100% 100% 100%

Total
(715) (368) (1083)

Source: Cheng 2011.

Note: Row percentages on top, total percentages underlined, and numbers of cases in parentheses.

Independence or Unification?

With a Taiwan-centered political identity becoming the overriding view on the

island, one might conclude that Taiwan voters would lose their emotional
attachment to China and support the island’s de jure independence at all costs.
Figure 3.5 presents the island residents’ positions on the issue of independence
versus unification. A quick glance at figure 3.5.a seems to confirm the above
conclusion; that is, the overall support for a cross-Strait unification has
declined precipitously. The declining support for unification is clearly visible
across all generations, as figure 3.5.b shows, and there is little sign of it
rebounding any time soon. While support for independence has been
increasing since 2000, figures 3.5.a and 3.5.c indicate that its overall support has
stabilized to about 20 percent between 2003 and 2012, although the fifth
generation tends to be more supportive of independence. After the eight-year
rule by the DPP, the only major political party on the island that has called for
the establishment of a Republic of Taiwan, support for de jure independence
has only increased slightly by comparison. It is important to note that,
throughout the last two decades, 35 percent to 45 percent of respondents have
remained undecided regarding their preferences over Taiwan’s long-term status.
Such a tendency is stronger for younger generations, Page 63 → as figure 3.5.d
shows, since 40 percent of the fourth and fifth generations have taken a wait-
and-see position since 2008.

Fig. 3.5. Independence/unification by generations, 1994–2012. Data source: Core Political Attitudes
among Taiwanese. (Election Study Center, NCCU).
 Previous analyses have shown that preferences for the island’s future relations
with China are conditioned by perceived costs (Hsieh and Niou 2005; Niou
2004; Wang 2005; Wu 1993). The majority of the respondents expressed a
preference for independence if there would be no war with China, but they
were not willing to if Beijing would respond violently. They rejected unification
if there were great disparity between the two sides of the Taiwan Strait but
were divided on the issue of unification if the social, economic, and political
conditions of the two systems were relatively compatible. In particular, those
respondents who took a wait-and-see position appear to be more sensitive to
future circumstances than others (Wang 2009). Supporters of independence or
unification are more likely to continue to back their respective positions even if
the preferred positions are associated with high costs. The “undetermined”
islanders, however, tend to switch their positions from one preference to the
other when favorable conditions are present and Page 64 → thus serve as swing
voters. Given their sizeable number, their decisions in either direction could
form a majority that will determine Taipei’s course of action regarding its
future relations with Beijing. Overall, these findings show that Taiwan voters
are pragmatic and risk averse on cross-Strait relations and that they are not
willing to sacrifice their hard-won democratic way of life and economic
prosperity for such radical political changes as declaring de jure independence
or unification. Despite the rising Taiwanese identity on the island, Taiwan’s
future relations with China remain undetermined because the island citizens
are pragmatic and risk averse.

Conclusions
As with many countries in the world, the formation and change of citizens’
identities have played a major role in Taiwan’s contemporary politics. To
examine the trajectory of Taiwan voters’ identity change, this study employs the
notion of “boundary” framed in the theoretical frameworks of primordialism
and constructionism. By analyzing quantitative survey data and qualitative
information gathered through focus-group interviews over the course of two
decades, it shows that the “China factor” has been the essential component of
the changing boundaries that shape Taiwan voters’ identities. Immediately after
the KMT’s retreat to the island, the identity boundary was set at the ethnic
divide between local citizens and mainlanders because a small group of
mainland elites monopolized political power. With the rapid democratization
since the late 1980s, along with other social changes on the island, the ethnic
divide has been replaced by the contestation of Chinese/Taiwanese
consciousness manifested in the island citizens’ holding of Taiwanese, Chinese,
or dual identities. Because the majority of Taiwan voters have now adopted
Taiwanese identity and few are holders of Chinese identity, the boundary of
Chinese/Taiwanese consciousness has gradually lost its political significance
within the domestic context. That said, the shadows of both ethnic cleavage
and the Taiwanese/Chinese divide continue to be cast over Taiwan’s politics
because it is cloaked in the island citizens’ partisan attachment. Meanwhile, due
to Beijing’s tactic of imposing diplomatic isolation on Taipei in the
international community, a new boundary is increasingly apparent on the
island. The term “Taiwan” is no longer a geographic designation, and
“Taiwanese” now assumes the new meaning of an identity with Taiwan as an
independent state.
 These findings have both theoretical and policy implications.
Theoretically, Page 65 → this study vindicates the constructionist argument that
identity can be intentionally or unintentionally fostered through socialization
media such as family and education. Both the KMT authoritarian regime ruled
by the Chiang family and later the pro-independence DPP government led by
Chen adopted a series of Sinicization and de-Sinicization measures,
respectively, to promote the “desired” identities. However, the validity of
constructionism does not reject the primordialist claim that identities are deep-
rooted in blood, culture, and language and are assumed to be given and hard to
change. The fact that a substantial proportion of mainlanders continue to hold
Chinese identity, along with the justifications employed by members of the
Minnan and Hakka groups for their holding of Taiwanese identity,
demonstrates the power of primordial bonds in shaping individuals’ identities.
The empirical findings presented above suggest that both primordialism and
constructionism are useful in understanding the development and the
formation of identities.

This chapter also shows that identity is a relational concept that defines an
individual psychological attachment. Because identity delineates what is and
what is not, it sets boundaries and forms a “they-group” and a “we-group.”
When the boundaries change, the substance of the identity may be altered even
though it may bear the same label. While the public discourse on
Chinese/Taiwanese identity continues, the substance of Taiwan’s identity
politics has increasingly shifted to the relationship between a unified Taiwan
and a hostile China. A new boundary of national identity has thus emerged on
the island.
 The empirical findings have important policy implications for cross-Strait
relations. One of the major reasons for many island residents to switch their
identities was Beijing’s diplomatic isolation of Taiwan and military coercion of
the island country. While successful in making Taiwan a pariah state in the
international community, the Chinese leaders’ forceful measures only invited
resentment from Taiwan voters and hardened their identity with Taiwan as an
independent state. The tendency also applies to mainlanders who have been
strong holders of Chinese identity and supporters of cross-Strait unification.
Since there is a strong association between holding Taiwanese identity and
recognition of the island’s independent and separate status from China,
Beijing’s forceful claim is counterproductive to its cause of unification. Given
that the rising Taiwanese identity has yet to be translated into pursuit of the
island’s de jure independence, some creative thinking needs to take place by
Beijing’s new leadership in order to attract the active support of Taiwan voters
so that cross-Strait disputes can be resolved peacefully.

Page 66 →

Appendix 3.A1.

List of Data Sources

Survey Data:

1. Core Political Attitudes among Taiwanese. Election Study Center, National

Chengchi University.
2. Lu-huei Chen. 2006. The Origin and Political Effect of Taiwanese Identity.
Election Study Center, National Chengchi University.
 3. Sufeng Cheng. 2011. Taiwan Identity: Formation Typology and Its Political
Consequences (II). National Science Council Research Plan, NSC 99–2410-
H-004–113.
4. Sufeng Cheng. 2013. A Study of Presidential Popularity and Its Political Effects
(II). National Science Council Research Plan, NSC100–2410-H004–086-
MY2.

Focus Group Interviews:

1. I-chou Liu. 2000. A Study in Major Political Identification Concepts of Taiwan

Public (I). National Science Council Research Plan, NSC89–2414-H-004–
022-SSS.
2. I-chou Liu. 2002. A Study in Major Political Identification Concepts of Taiwan
Public (II). National Science Council Research Plan, NSC 89–2414-H-004–
049.
3. Su-feng Cheng. 2010. Taiwan Identity: Formation Typology and Its Political
Consequences (I). National Science Council Research Plan, NSC99–2410-H-
004–113.
4. Su-feng Cheng. 2011. Taiwan Identity: Formation Typology and Its Political
Consequences (II). National Science Council Research Plan, NSC99–2410-
H-004–113.
5. Sufeng Cheng. 2013. A Study of Presidential Popularity and Its Political
Effects (II). National Science Council Research Plan, NSC100-2410-H004-
086-MY2.

Notes
1. A list of data sources is presented in the Appendix. The author would
like to thank Professors Lu-huei Chen, Sufeng Cheng, I-chou Liu and the
Election Study Page 67 → Center, National Chengchi University in Taipei,
for making the data available. All errors are my own.

2. Some scholars distinguish three theoretical approaches for identity

research: primordialism, constructivism, and instrumentalism. Both
constructivism and instrumentalism agree that identity is malleable by
social linkages and not biological givens. Their difference lies in how these
linkages are created. Thus, the instrumentalist approach is essentially “a
variant of constructivism” (Dawisha 2002, 6).

3. For the meaning of “imagined communities,” see Anderson 1991.

4. Group consciousness connotes identification with a group as well as an

awareness of the relative position of the group in a society. Some scholars
describe the rise of Taiwanese identity and its related awareness as the
emergence of Taiwanese nationalism (e.g., Wu 2004). Because nationalism
generally refers to a belief aiming to achieve and maintain national
independence, it is frequently used for the analysis of colonial movements
and ethnic conflicts that yearn to break free from domination (e.g., Fox
2004; Varshney 2003). As will be discussed below, this is not the case for
Taiwan since the rise of Taiwan-centered awareness does not imply that
the island’s citizens are actively pursuing Taiwan’s de jure independence
and permanent separation from China. The island citizens’ positions on
the issues of independence and unification are more pragmatic and risk
averse (see Hsieh and Niou 2005; Niou 2004; Wang 2005; and Wu 1993,
1996). Some Taiwanese scholars have thus preferred to use the term
“consciousness” (e.g., Wang 2001; also see Rigger 2006). This study
follows that reasoning.

5. For instance, the Chen administration renamed the state-run China

Petroleum Corporation as CPC Corp, Taiwan. Other renamed agencies
included the island country’s shipbuilding corporation and its central bank
(Bishop and Dickie 2007).

6. All respondents in the surveys are divided into five political

generational groups based on the following four birth years as cutoff
points: 1931, 1953, 1968, and 1982. The classification is based on the
premise that respondents of each generation would be 18 years old when
one of four significant political events occurred: (1) the retreat of the
KMT government to Taiwan in 1949, (2) Taipei’s withdrawal from the
UN in 1971, (3) the establishment of the Democratic Progressive Party in
1986, and (4) the peaceful transfer of political power in 2000. Such a
classification is based on the conceptualization that a political generation
is a group of people who share common experiences and historical
memories due to the fact that they are born in a same time period and live
through the same social and economic environment (Mannheim 1952;
Neumann 1965). For justifications on the significance of the four political
events, see Chang and Wang (2005).

7. A private polling organization recently released a report that 60 percent

of Taiwan citizens possess Chinese identity (China Times, March 5, 2013).
The report generated some controversy on the island. It also invited
concerns from foreign governments. Diplomats from the American
Institute in Taiwan and Interchange Association, Japan, the “unofficial”
 representative offices of the United States and Japan, respectively, visited
National Chengchi University’s Election Study Center to discuss the
findings.

8. National Statistics, Republic of China (Taiwan) at

http://statdb.dgbas.gov.tw/pxweb/Dialog/Saveshow.asp (accessed March
15, 2015).

Page 68 →

References

Abdelal, Rawi, Yoshiko Herrera, Alastair Iain Johnston, and Rose McDermott.
2009. “Identity as a Variable.” In Measuring Identity: A Guide for Social
Scientists, ed. Rawi Adbelal, Yoshiko Herrera, Alastair Iain Johnston, and
Rose McDermott, 17–32. New York: Cambridge University Press.

Anderson, Benedict. 1991. Imagined Communities: Reflections on the Origin and

Spread of Nationalism. New York: Verso.

Barnett, Michael. 1999. “Culture, Strategy, and Foreign Policy Change: Israel’s
Road to Oslo.” European Journal of International Relations 5 (1): 5–36.

Bishop, Mac William, and Mure Dickie. 2007. “Taiwan Drops ‘China’ in
Identity Move.” Financial Times, February 12. www.ft.com/cms/s/581a6f9a-
ba96–11db-bbf3–0000779e2340.html (accessed February 12, 2007).
Chang, Andy, and T. Y. Wang. 2005. “Taiwanese or Chinese? Independence or
Unification? An Analysis of Generational Differences in Taiwan.” Journal of
Asian and African Studies 40 (1–2): 29–49.

Chang, Yu-Tzung, 2006. “Campaigning Events on Vote Choice: A Study of the

2004 Presidential Election in Taiwan.” Soochow Journal of Political Science
22:121–59.

Cheng, Su-feng. 2009. “Ethnicity, Identity, and Vote Choice in Taiwan.” Journal
of Electoral Studies 16 (2): 23–49.

Dawisha, Adeed. 2002. “Nation and Nationalism: Historical Antecedents to

Contemporary Debates.” International Studies Review 4 (1): 3–22.

Enav, Peter. 2007. “Taiwan Leader Riles China, U.S.” Washington Post,
September 9. www.washingtonpost.com/wp-
dyn/content/article/2007/09/09/AR2007090900579.html (accessed
September 9, 2007).

Fox, Jonathan. 2004. “The Rise of Religious Nationalism and Conflict: Ethnic
Conflict and Revolutionary Wars, 1945–2001.” Journal of Peace Research 41 (6):
715–31.

Garver, John W. 1997. Face Off: China, the United States, and Taiwan’s
Democratization. Seattle: University of Washington Press.

Geertz, Clifford. 1994. “Primordial and Civic Ties.” In Nationalism, ed. John
Hutchinson and Anthony D. Smith, 29–34. New York: Oxford University
Press.
Gellner, Ernest. 2008. Nations and Nationalism. 2nd ed. Ithaca, NY: Cornell
University Press.

Horowitz, Donald L. 2000. Ethnic Groups in Conflict. 2nd ed. Berkeley:

University of California Press.

Hsieh, John Fuh-sheng, and Emerson M. S. Niou. 2005. “Measuring Taiwanese

Public Opinion on Taiwanese Independence.” China Quarterly 181 (1): 158–
68.

Huang, Zen-Jei. 1993. “Tai-du Yun-dong yu Tai-hai Liang-an Guo-jia Tung-yi

Zheng-cezhi Yan-Jiu” [The Study of the Taiwanese independence
movement and the reunification policies of both sides of the Taiwan Strait].
Master’s thesis, East-Asia Institute, National Cheng-chi University, Taipei.

Isaacs, Harold R. 1975. Idols of the Tribe: Group Identity and Political Change.
Cambridge: Harvard University Press.

Lai, Tse-han, Ramon H. Myers, and Wei Wou. 1991. A Tragic Beginning: the
Taiwan Uprising of February 28, 1947. Stanford: Stanford University Press.

Page 69 → Laitin, David. 1998. Identity in Formation: The Russian-Speaking

Populations in the Near Abroad. Ithaca, NY: Cornell University Press.

Lee, Wei-chin, and T. Y. Wang. 2003. Sayonara to the Lee Teng-Hui Era: Politics in
Taiwan, 1988–2000. Lanham, MD: University Press of America.

Mannheim, Karl. 1952. Essays on the Sociology of Knowledge. Ed. Paul Kecskemeti.
London: Routledge and Kegan Paul.
Mo, Yan-chih and Shih Hsiu-chuan. 2007. “Inscription Goes Up at Democracy
Hall.” Taipei Times, December 9
www.taipeitimes.com/News/front/archives/2007/12/09/2003391782
(accessed December 9, 2007).

Neumann, Sigmund. 1965. Permanent Revolution: The Total State in a World at War.
2nd ed. New York: Frederick A. Praeger.

Niou, Emerson M. S. 2004. “Understanding Taiwan Independence and its

Policy Implications.” Asian Survey 44 (4): 555–67.

Pao, Cheng-hao. 2009. “Party Identification and Vote Decision: An Over-Time

Comparison for the 2000–2008 Presidential Elections.” Tamkang Journal of
Humanities and Social Sciences 40: 68–89.

Rigger, Shelley. 2006. Taiwan’s Rising Rationalism: Generations, Politics and

“Taiwanese Nationalism.” Policy Studies 26. Washington: East West Center.

Shih, Ming. 1992. Taiwan Bu-shi Zhong-guo de Yi-bu-fen [Taiwan is not a part of
China]. Taipei: Vanguard.

Shils, Edward. 1957. “Primordial, Personal, Sacred and Civil Ties.” British
Journal of Sociology 8 (2): 130–45.

Taiwan Affairs Office of the State Council of the PRC (TAO). 2000. “The
One-China Principle and the Taiwan Issue.” People’s Daily Online [Renmin
Ribao], February 21.

Varshney, Ashutosh. 2003. “Nationalism, Ethnic Conflict, and Rationality.”

Perspectives on Politics 1 (1): 85–99.
Wang, Fu-chang, 2001, “National Imagination, Ethnic Consciousness, and
History: Content and Context Analyses of the ‘Getting to Know Taiwan’
Textbook Disputes.” Taiwan Historical Research 8 (2): 145–208.

Wang, Fu-chang, 2008, “The Role of Ethnic Politics Issues in Taiwan’s

Democratization Transition.” Taiwan Democracy Quarterly 5 (2): 89–140.

Wang, T. Y. 2001. “Cross-Strait Relations after the 2000 Election in Taiwan:

Changing Tactics in a New Reality.” Asian Survey 41 (5): 716–36.

Wang, T. Y. 2005. “The Perception of Threats and Pragmatic Policy Choice: A

Survey of Public Opinion in Taiwan.” Issues & Studies 41 (1): 87–111.

Wang, T. Y. 2006. “Taiwan’s Bid for UN Membership.” In China’s Rise, Taiwan’s

Dilemmas, and International Peace, ed. Edward Friedman. New York:
Routledge, 174–92.

Wang, T. Y. 2009. “Taiwan: Rising National Identity and the Challenges to

Cross-Strait Relations.” Presented at the Conference on State Secession and
Separatism in Europe and Asia, Hong Kong Baptist University and
University of Macau, December 7–9.

Wang, T. Y., and G. Andy Chang. 2005. “Ethnicity and Politics in Taiwan: An
Analysis of Mainlanders’ Identity and Policy Preference.” Issues & Studies 41
(4): 35–66.

Wang, T. Y., Wei-chin Lee, and Ching-hsin Yu. 2011. “Taiwan’s Expansion of
International Page 70 → Space: Opportunities and Challenges.” Journal of
Contemporary China 20 (69): 249–67.
Wu, Naiteh. 1993. “National Identity and Party Support: The Social Basis of
Party Competition in Taiwan.” Bulletin of the Institute of Ethnology 74:33–61.

Wu, Naiteh. 1996. “Liberalism and Ethnic Identity: Searching for the
Ideological Foundation of Taiwanese Nationalism.” Taiwanese Political Science
Review 1 (1): 5–40.

Wu, Yu-Shan. 2004. “Taiwanese Nationalism and Its Implications: Testing the
Worst-Case Scenario.” Asian Survey 44 (4): 614–25.
Page 71 →
 Chapter 4

Parties, Partisans, and Independents in

Taiwan

Ching-hsin Yu

Taiwan is a new democracy in which elections have played a significant role in

citizen political participation in the past decades. Mainstream electoral studies
have paid much attention to voters’ decision-making processes in elections.
Factors that influence voters’ electoral choices, such as partisanship, economic
evaluation, candidate qualifications, and policy preferences, are often included
in the discussions of voter behavior. Among those factors, citizen partisanship,
or party identification, is often cited by scholars as the most consistent and
influential one in Taiwan. However, the concept has met with confusing
definitions and implications. On the one hand, Taiwan’s political history has
shaped a distinct developmental pathway of citizen partisanship. As Taiwan
was undergoing a significant democratic transition in the mid-1980s, the party
system evolved from a one-party hegemonic system to a two-party system, then
to a multiparty system in the 1990s, followed by the formation of two major
camps, the Pan-Blue and the Pan-Green, in the 2000s. Citizen partisanship also
changed dramatically. On the other hand, as more research was devoted to the
exploration of the behavior of partisan voters, political independents were of
less concern. In addition to the differences between partisans and
independents, there are various types of political independents, whose attitudes
and behaviors are different from each other, so study of citizen partisanship in
Taiwan calls for examination of these different kinds of independents.

This chapter intends to explore the evolution and distribution of citizen

partisanship in Taiwan with a special interest in political independents.
After Page 72 → a brief literature review, it will examine the historical origins of
citizen partisanship and the evolution of the party system in Taiwan. It will
then compare attributes and attitudes of different types of partisans and
independents, such as political interest, preferences on the issues of unification
and independence, support for democracy, voting, and vote choice. In
conclusion, a reconsideration of the development of citizen partisanship and its
impact in Taiwan is discussed.

Studies of Partisans and Independents

A system of voter identification was developed from the group theory since
Campbell and his associates published their classical works on American voters
(Campbell, Gruin, and Miller 1954; Campbell et al. 1960). It maintains that
citizens tend to psychologically identify and behaviorally support a political
group, or party, based on their personal experiences and preferences. The link
to a given party will further shape citizens’ political attitudes and issue
positions. Party identification is not only a psychological attachment to a
political party but also a cue for a citizen’s political actions. Because citizens
have different experiences with a political party their identification may have
varying strength as a result. Some citizens may develop a strong party
identification, others may have a weak attachment to a political party, and still
others may possess a neutral feeling toward any groups. Therefore, American
voters are typically categorized in one of five categories: Strong Democrat,
Weak Democrat, Independent, Weak Republican, and Strong Republican; or in
seven categories with the addition of Independent Democrat and Independent
Republican (Weisberg 1993, 684).1

Regardless of its popularity, the concept of party identification has

nevertheless suffered from the problems of dimensionality and transitivity
(Converse 1966; Petrocik 1974; Weisberg 1980; Niemi, Wright, and Powell
1987; Bartle and Bellucci 2009). Generally speaking, the discussion of attitudes
and behaviors of citizens with strong partisanship are less controversial. How
to assess independents or citizens with weak partisanship raises a crucial issue
so that they will not be treated as a residual category.

Political independents are generally characterized as having positive attributes

like prudent judgment and adequate political involvement (Bryce 1929).
However, the normative virtues of the political independents have been
critically challenged by empirical studies (Campbell et al. 1960, 143). While the
debate over what constitutes adequate qualifications of a nonpartisan voter will
continue for a long period of time, the increasing popularity Page 73 → of survey
data has enabled scholarly research to reexamine the nature of political
independents (Burnham 1970; Dennis 1988; Keith et al. 1992; Magleby, Nelson,
and Westlye 2011).

Studies of voters’ party identification in Taiwan, beginning in the early 1970s,

have primarily followed the research on American voters. As partisan
attachment is widely recognized as one of the most influential factors affecting
a voter’s political decisions (Chen 1986, 1994; Chu 1996; Liu 1996, 1997),
empirical studies have found that independents in Taiwan are largely female,
less educated Minnans and Hakkas (in contrast to the mainlanders) with low
socioeconomic status. They have less political interest and pay little attention to
political affairs. More recent studies show that the number of independents had
increased over time and the political attitudes and voting behaviors of less-
educated independents differ from those of highly educated independents (Yeh
1994; Chu 2004; Wang 2010). The empirical evidence confirms the “revisionist”
image of the political independent as characterized by the authors of The
American Voter. It also demonstrates that a clear differentiation between pure
independents and partisan leaners in Taiwan is sometimes difficult (Wang and
Yu 2011).

Precisely because nonpartisan voters have varying degree of political

attributes, their existence has crucial implications to a functioning party system
and the health of democracy. Indeed, political independents frequently play a
pivotal role in deciding the final outcome of a close election. They also
function as a vital balance in an otherwise polarized society. This is especially
important for a nascent democracy like Taiwan, where the party system has not
yet stabilized. As the current research analyzes Taiwan voters’ partisan
identification, it will pay particular attention to political independents in
Taiwan.

Partisanship before 1986

The development of citizen partisanship had a unique history in Taiwan in its

early periods. The tragic incident that occurred on February 28, 1947 (the 2–28
Incident), created a deep antagonism between the incoming mainlanders and
the Minnan. The latter learned from the incident that politics could be
dangerous, and as a consequence, political apathy began to take root in their
society. When the Kuomintang government moved to Taiwan, an authoritarian
regime was quickly and firmly established. Constitutionally, the Temporary
Provisions were implemented to establish martial law, to provide the president
with tremendous discretionary power, and to prohibit the emergence of
political opposition. No new political parties were allowed Page 74 → to be
established under martial law. Any dissident opinions targeted at the
government were also subject to repression. Intense political socialization
aimed at enhancing citizen loyalty to the government, which was in reality
loyalty to the ruling KMT, was widespread in the school curricula (Wilson
1970). With the assistance of state-controlled mass media, only selective
information beneficial to the KMT regime was permitted to circulate. Equally
important, the introduction of a political commissar further ensured the KMT’s
unchallenged status in regard to the military. A party-state run by the KMT, like
that of many communist countries, became firmly established on the island.
Under the authoritarian party-state structure, the KMT was therefore the only
significant political party.2 Citizens in Taiwan were either KMT members or
not. No other political parties were able to compete for popular support
against the KMT, so the situation was similar to that of the one-party
hegemonic system as described by Sartori (1976).

Furthermore, under the name of Fa-tong (the sole legitimate government

representing China), the KMT regime maintained an extensive political
structure as it had done on the Chinese mainland. As “one China” was taken as
given and Taiwan was regarded as one of China’s 36 provinces, the political
connection between the island and the Chinese mainland was emphasized by a
Mainlander dominated regime. The political representation of Minnans and
Hakkas was intentionally suppressed. As a result, a political division rooted in
ethnicity emerged on the island (Wang 1993; Wachman 1994).

Despite the ethnic division, support of the KMT came from two sources. As
indicated in chapter 3, the first group of supporters were the mainlanders who
fled to Taiwan with the KMT government in the late 1940s. Because they
followed the party leaders in the Sino-Japanese War and the Chinese Civil War,
mainlanders have become the core supporters of the KMT (Wu 1995; Shyu,
1997). The second group of supporters has been the Minnan and Hakka
through political indoctrination by the regime. After the disastrous defeat on
the Chinese mainland, it was crucial for the KMT regime to establish strong
local support in Taiwan. Unlike mainlanders who shared a similar history with
the KMT leaders, Minnans and Hakkas were connected to the party by a mix of
political, economic, and ideological incentives. Politically, a Minnan or Hakka
with KMT membership had a better chance of being promoted in the
government. Party membership was a proxy criterion for national loyalty and
carried political expedience for the Minnan and Hakka who intended to have a
career in the government. Economically, maintaining a close tie with the KMT
would benefit businesses. By foregoing some prerogative interests, the regime
effectively exchanged economic benefits for the political support of the Minnan
and Hakka.

Page 75 → One noteworthy way the KMT regime built grassroots support was
through the implementation of local elections. Since the early 1950s, local
elections had been the KMT’s main locus for recruiting local elites, distributing
economic prerogatives, and marketing the regime’s image of a “free and
democratic China” in opposition to totalitarian Communist China. Of course,
the outcomes of elections were managed by the KMT regime. Nonetheless, it is
also reasonable to argue that local elections gave the Minnan and Hakka
experience with party politics and the electoral process. Under the façade of a
free and democratic China, the authoritarian KMT regime did provide a certain
degree of pluralism in society. Non-KMT independent candidates found some
room for political participation in elections. In spite of being outnumbered by
the KMT candidates, the non-KMT candidates won seats in various local
elections. The appearance and activities of these non-KMT candidates were
typical symbols of electoral competition under the one-party hegemonic
system. Indeed, there were some non-KMT candidates who consistently won
elections in certain regions. The non-KMT elected politicians provided
alternative choices for the Taiwan electorate, in particular, those Minnan and
Hakka who had no partisan affiliation.

The non-KMT forces continued to grow during Taiwan’s rapid economic and
sociopolitical transition. The KMT regime found it more and more difficult to
curb the expansion of the non-KMT forces in elections. Worse still for the
KMT regime, diplomatic setbacks in the 1970s, such as losing its seat at the
United Nations and the termination of formal relations with Japan and the
United States, had facilitated further expansion of the electoral arena. The non-
KMT candidates found more room for collective action in elections. For
example, the group Dang-wai (meaning “outside the KMT”), which appeared
in the early 1970s, unified the non-KMT candidates during elections. Members
of Dang-wai were able to share common platforms and manifestos without
being penalized by electoral regulations. Although Dang-wai members did not
formally organize as a political party, they worked together in elections as if
they were members of one. The Minnan/Hakka electorate was able to make a
choice between the KMT candidates and the Dang-wai candidates in elections.
Of course, although they competed against KMT candidates in elections, not
all of the non-KMT candidates maintained similar political stances. Nor were
all of the non-KMT candidates anti-KMT. Therefore, it would be premature to
suggest that a quasi two-party system had taken shape in Taiwan. Nonetheless,
the non-KMT forces did grow in one election after another. It was not until
the Democratic Progressive Party was formally established in 1986 (although it
was still illegal at that time) that a new party Page 76 → system with two
meaningful political parties emerged in Taiwan (Cheng 1989; Lu 1992; Hsieh
2005).

The evolution of the party system before 1986 resulted in a peculiar pattern of
citizen-party connectivity in Taiwan. First, a clear and stable partisanship
between the mainlanders and the KMT had emerged in the early period when
the KMT government moved to Taiwan. The mainlanders maintained a strong
affiliation with the KMT due to their close dependence on the KMT in all
respects. As the KMT regime made every effort to consolidate its control over
the island, material enticements and purposive incentives provided by the KMT
were used to facilitate its popularity among the Minnan and Hakka. For
example, the KMT regime adopted several liberalization measures in the early
1970s, such as recruiting Minnan and Hakka elites into the party and gradually
opening some electoral posts for public contestation, had effectively expanded
its connection with the Minnan and Hakka. Together with the mainlanders, the
more extensive connections between the KMT and Minnan and Hakka thus
helped the KMT to maintain a consistent advantage in elections. However,
except for those mainlanders and KMT-friendly Minnan and Hakka, a majority
of the citizens in Taiwan were not registered KMT members.3

Second, the historical legacy had a significant effect on the development of

citizen partisanship in Taiwan. Due to the unique sociopolitical development
in the 1950s and 1960s, the label of “political party” acquired negative
implications for some Taiwanese. The incident of February 28, 1947, the
discriminatory political structure in place since the early 1950s, and the white-
terror mentality prevented the Minnan and Hakka from embracing political
affairs. Even though they had participated in local elections, their connection
to the ruling KMT was weak. If they supported a KMT candidate in elections,
it did not mean that they identified with the KMT. In fact, voters generally
placed candidates above the party. It was the candidates, not the voters, who
were directly connected to the KMT. Strong popular support for KMT
candidates did not equate to strong support for the KMT. The candidate’s
personal image and connections generally were more important than the party
label in elections (Chen 1986). Since the image of political parties was not that
welcome, maintaining their stance as nonpartisan could be a good choice for
many Minnan and Hakka voters. Also due to the fact that the non-KMT
candidates were not allowed to engage in any organized campaign activities
under martial law, the Taiwan electorate supporting non-KMT candidates was
unable to form a normal partisanship, which KMT supporters could do.

Third, the Dang-wai represented a partially united front of non-KMT Page 77

→ candidates in elections. At the beginning, the name of Dang-wai was merely

an expedient way of differentiating non-KMT candidates from their KMT
opponents. However, the term gradually acquired special political and
organizational meanings. As the political system increasingly liberalized,
members of the Dang-wai regularly engaged in island-wide electoral campaigns.
Since martial law banned the formation of political parties, Dang-wai members
used various names in different elections as common symbols to distinguish
themselves. Consequently, the emergence of the Dang-wai provided Taiwan
voters with a unique form of two-party competition.

The election of the Taiwan Provincial Assembly is a typical example that

illustrates the development of political independents. The elections were
implemented in 1951 and ended in 1994. Only two minor political parties, the
Chinese Youth Party and the Democratic Socialist Party, which had been
established in China during the Second World War, were allowed to present
their candidates under party labels. However, as shown in table 4.1, the two
minor parties were too weak to win a meaningful number of seats in the
elections. Compared with those two parties, independent candidates performed
much better. Even though the winning number varied from elections to
elections due to a lack of organized campaigns, they managed to win from 15
percent to 27 percent of the electoral vote. They were thus a significant non-
KMT force. The majority of Dang-wai independents unified under the DPP
after 1986, which fared well in the elections of 1989 and 1994, garnering about
21 percent and 29 percent of the popular vote. The establishment of the DPP
also suggested that there was a split among political independents. Those
independents who did not join the DPP continued to run for election under a
nonpartisan label. Popular support for non-DPP independents did not
disappear, although it declined significantly.
 The growth of political independents in local elections implies the existence
of political space for non-KMT forces. The election results of the Provincial
Assembly provide a larger picture in which the KMT was strongly Page 78

→ dominant, although at the same time the number of political independents in

Taiwan’s political landscape was not insignificant. Unfortunately, other than
these macro electoral results, there are little empirical data showing individual
voters’ partisanship. Micro studies of individuals’ vote choices were not feasible
until the mid-1980s, when academics began to introduce the concept of party
identification and its relevance to Taiwan.

Table 4.1. Election Results: Taiwan Provincial Assembly (percentage of vote)

Year 1951 1954 1957 1960 1963 1968 1972 1977 1981 1985 1989 1994

KMT 78.18 84.21 80.30 79.45 82.43 84.51 79.45 72.73 76.62 76.62 70.13 60.08

DPP 20.78 29.11

New Party 2.53

Other parties 1.82 1.52 1.35 1.30

Independent 20.00 15.79 18.18 20.55 16.22 15.49 20.55 27.27 23.38 22.08 9.09 7.59

Sources: Data are from the ROC Central Election Commission, the Taiwan Provincial Election
Commission, and the Election Studies and Survey Data Archive of the Election Study Center of National
Chengchi University.

In a strict sense, there were only KMT partisans and it was difficult to identify
the non-KMT supporters in elections. Although the non-KMT candidates
often called themselves independents, their supporters were not exactly the
same as “independent voters” according to definitions of party identification.
They were voters who supported non-KMT candidates, regardless of how
persistent their support would be. True partisan support did not develop until
the establishment of the DPP in 1986, when the partisanship of Taiwan voters
began to take on a different form than in previous decades. In particular, those
who were non-KMT supporters before 1986 would not only support the
individual candidates as they had before but also began to adjust to the new
party label. Thus, it is difficult to provide a detailed portrait of non-KMT
supporters because the fact that they were non-KMT supporters does not
necessarily make them independents. Since there was a ban on the formation of
political parties, supporters of non-KMT candidates (independent candidates)
generally characterized themselves as independents (voters supporting
independent candidates) even though they in fact were non-KMT supporters.
Consequently, the concept of “independent” that is often cited as a counterpart
of party identifiers carries a different connotation.

Equally important, the evolving party system consistently shaped citizens’

partisanship differently than in the previous decades. In particular, Taiwan
experienced a new political landscape after the KMT’s disastrous defeat in the
2000 presidential election as some members left the party and organized the
People First Party and the Taiwan Solidarity Union. Along with the New Party,
a KMT split-off of 1994, they form a part of Taiwan’s multiparty system. The
island citizens’ partisanship has become diversified as a result, which can be
characterized as multipartisan including identification with the KMT, DPP, NP,
PFP, and TSU.4 Yet in addition to the increase in political parties, a significant
characteristic of the party system after 2000 was the formation of the Pan-Blue
and Pan-Green coalitions. The Pan-Blue includes the KMT, the NP, and the
PFP, while the Pan-Green includes the DPP and the TSU. The key departure
point of the Pan-Blue and Pan-Green mainly rests on their different stances
with regard to cross-Strait policy. These alliances also have strong impacts on
citizens’ vote choice in Page 79 → elections. Hence, citizen’s partisanship can be
also understood in terms of their allegiance to the Pan-Blue Alliance or the
Pan-Green Alliance, which presents a bipartisan structure.

Partisans and Independents since the 1990s

Thanks to the rapid advancement of electoral studies since the 1990s, more
sophisticated findings about citizens’ partisanship have appeared in Taiwan’s
academic community. This section will examine the attitudes and behaviors of
partisans and independents in recent decades. Instead of using voter
identification with an individual political party, it will use voter identification
with the two major political camps mentioned above. Hence, following the
prescriptions provided by The American Voter (Campbell et al. 1960), there will
be seven categories of partisan: Strong Pan-Blue, Weak Pan-Blue, Leaning Pan-
Blue, (pure) Independent, Leaning Pan-Green, Weak Pan-Green, and Strong
Pan-Green. In order to provide a more concise meaning of “Independent,” this
chapter also uses education as a criterion to divide the Independents into three
categories: low-educated, middle-educated, and high-educated. Consequently,
there will be comparisons among party identifiers and political independents in
terms of political interests, cross-Strait relations, support for democracy,
voting, and vote choices. These variables are selected for comparison because
they have been frequently identified as important aspects of the study of
Taiwan voters. Through the application of these variables with cumulative
survey data collected over the past two decades, it is expected that the analyses
will contribute to the study of partisanship and the impacts of political
independents in Taiwan. Descriptions and measurements of these variables can
be found in appendix 4.A1.
General Distribution

The empirical findings regarding citizen partisanship shown in table 4.2

demonstrate certain consistencies and changes during the past two decades. In
the early 1990s, when citizens started to experience contested elections between
the KMT and the DPP, a majority of the voters identified themselves as
nonpartisan. The number of independents was as high as 34 percent in 1992.
Next in predominance were the Weak Pan-Blue and Strong Pan-Blue
identifiers. As for the newly established DPP, its identifiers numbered far less
than those of the Pan-Blue Alliance. This suggests that the KMT Page 80

→ continued to enjoy a clear electoral advantage, as it had before. The number

of independents declined in the legislative election of 1995 as only 22.5 percent
of the electorate considered themselves as independents. Both the KMT and
the DPP had more identifiers. In particular, around 32 percent of the electorate
labeled themselves as Weak Pan-Blue. The presidential election of 1996 showed
a similar pattern as in 1995 as the percentage of those identifying with the
KMT and with the DPP rose. Only slightly less than one-quarter of the
electorate identified themselves as nonpartisan. In particular, the KMT had
enjoyed a strong surge of identifiers such that the combination of strong Pan-
Blue, Weak Pan-Blue, and Leaning Pan-Blue comprised more than 50 percent
of the electorate. The number of independents rose again in the elections for
legislators in 1998. It seems that the surge of independents resulted from the
decline of KMT identifiers. Meanwhile, the number of identifiers with the DPP,
either Strong Pan-Green or the Weak Pan-Green, had increased slowly but
steadily.
 The independents reached their peak in the presidential election of 2000 and
the legislative election of 2001. One key reason for the increase of independents
had to do with the dramatic change in the party system at that time. On the
one hand, the KMT suffered from an internal split before the presidential
election, which led to a disastrous defeat in the presidential election and the
DPP became the ruling party, 14 years after its establishment. On the other
hand, new political parties, such as the PFP and the TSU, emerged in the 2001
legislative election. The unprecedented face of a multiparty system appeared in
Taiwan. As political parties exhibited varying degree of quality, many citizens
became reluctant to identify themselves as leaning toward a particular party.
Thus, being an independent was the optimal choice for many voters. After
2001, the number of independents became somewhat variable in different
elections, ranging from slightly more than one-third in the 2008 to about one-
quarter in 2012.

One noticeable development was the gradual increase of partisan leaners since
2001. The percentages of both the Leaning Pan-Blue and Leaning Pan-Green
categories, with some minor ups and downs, showed an increasing trend
through 2012. The increase in partisan leaners resulted mainly from the
emergence of coalitional electoral competition at that time; for example,
members in the Pan-Blue camp (the KMT, the NP, and the PFP) cooperated in
the presidential elections in order to defeat the Pan-Green candidates.
However, in the legislative elections, each party in the Pan-Blue camp
nominated its own candidates and competed against each other. The
coexistence of cooperation and competition in these elections encouraged
voters to swing their partisanship between pure independents and leaners.
 Page 81 → Last, the percentages of strong party identifiers have been stable
over the past two decades. The percentage of Strong Pan-Blue identifiers
decreased when the KMT lost the presidential election in 2000, and the number
continued to decline in 2001 and 2004 but then showed a slight surge from
2008 to 2012. As for the Pan-Green, winning the presidential election in 2000
did bring growth in the percentage of strong identifiers. The trend did not
continue but maintained an average of between 4 percent and 6.5 percent from
2001 to 2012.

Political Interest

The general trend of partisan distribution for the past decades, as shown in
table 4.2, is rather stable. However, unlike the partisans who have a clear
political orientation, it would be imprudent to regard the political independent
as a unified subset of the electorate. Many studies have provided a more
detailed examination of the political independent, and, among the variables,
educational level seems to be the most cited demographic criterion that
differentiates subtypes of political independents. These studies argue that Page

82 → differences in other demographic characteristics, such as gender, age, and

career, do not reveal clear and consistent implications about the attitudes and
behavior of the political independent, and the educational differences have
continuously drawn researchers’ attention in studies of the political
independent. Therefore, this paper will also investigate the educational
differences among political independents in Taiwan.
Table 4.2 General Distribution of Partisanship

1992 1995 1996 1998 2000 2001 2004.03 2004.12 2008.01 2008.03 2012

Strong Pan-Blue 15 13.4 15.3 10.4 8 4.5 5.2 5.1 8.6 7.1 7.9

Weak Pan-Blue 18.4 31.8 31.6 21.5 13.7 9 11.6 13.4 14.3 15.9 17

Leaning Pan-Blue 7.2 2.3 4.2 8.1 5.4 12.7 13 14.4 13.8 13.3 15.1

Independent 34 22.5 22.1 28.2 35.4 35.9 32.7 27.6 34.6 30.7 26.5

Leaning Pan-Green 3.2 1.2 1.9 5.5 4.4 13 11.5 13.5 10.5 11.8 11.2

Weak Pan-Green 6.9 12.2 11.2 13.6 15.9 13.1 13.6 13.7 8.7 11.1 11.3

Strong Pan-Green 3.2 3.2 3.7 7.2 8 5.7 5.8 5.9 4 5.2 6.5

NR 12.1 13.5 10 5.5 9.2 6.1 6.7 6.4 5.4 4.9 4.5

Sources: Data for 1996–2000 are from the Election Study Center at National Chengchi University, and
data for 2001–12 are from the Taiwan’s Election and Democratization Study Project (see List of Data
Sources in appendix 4.A2).
Notes: 1: Numbers in the table are percentages of respondents in each election survey. 2: The elections in
1992, 1995, 1998, 2001, 2004.12 (December), 2008.01 (January), and 2012 are legislative elections. The
elections in 1996, 2000, 2004.03 (March), 2008.03 (March), and 2012 are presidential elections. In 2012,
the election for president and legislators were held at the same time.

Fig. 4.1. Educational distribution among independents. Data Sources and Notes: Same as table 4.2.
As indicated in figure 4.1, political independents in Taiwan have shown
different patterns during the past two decades. In the early 1990s, trends in the
percentages of political independents with low-level and middle-level
educations were similar until early 2000s. Both categories also accounted for
the majority of political independents at that time. Moreover, the percentage of
moderately educated political independents has been the highest, while the
numbers of least-educated political independents have decreased. By contrast,
the highly educated political independent did not account for a very high
percentage. Notably, the number of highly educated political independents
increased after 2000 and exceeded that of the least-educated political
independent. The different developments among the least, moderately, and
highly educated political independents should enable researchers to get a more
precise understanding about political attitudes and behavior.

The concept of political interest has been considered a driving force of

political participation since the publication of The People’s Choice (Lazarsfeld,
Berelson, and Gaudet 1948). Those voters with a greater interest in politics
would concurrently have more political information and would be more
attentive to political activities than those less interested. The literature also
suggests a close relationship between political interest and partisanship.
Identifying with a political party would provide voters convenient and reliable
cues for elections. Moreover, a political party continues to convey political
messages to voters all the time. It is anticipated that citizens with a close
relationship to a political party would have more political interest than those
citizens who do not maintain such a party affiliation. The results in table Page 83

→ 4.3 partially confirm these findings. The general pattern indicates, first, that
independents have been the most uninterested in politics during the past two
decades. Except for the presidential election in 2000 and the legislative election
in 2004, independents were more likely to say that they do not have an interest
in politics. Second, partisan voters, regardless of their strength of identification,
did not have clear associations with political interest. Partisan leaners in the
Pan-Blue camp were generally less interested in politics, yet even the stronger
partisans in the Pan-Green camp showed an unanticipated low interest in
politics. The overall trend suggests a moderate relationship between political
interest and partisanship. Even though independents have shown a relative lack
of interest in politics, the relationship between partisanship and political
interest is not consistent.

Nonetheless, the relationship between independents and political interest has

rather consistently and systematically followed educational differences. Figure
4.2 shows a different distribution of political interest among the three types of
Independents. The least-educated independent has the lowest political interest
compared to the other types of Independent. The percentage of lack of political
interest in the least-educated independent exceeds 80 percent. By contrast, the
highly educated independent consistently maintains a certain degree of political
interest. The political interest of the moderately educated independent, as
expected, lies in between. The graph clearly suggests that the least-educated
independents have the least interest in politics.
Table 4.3 Distribution of Political Interest among Partisans

1992 1995 1996 1998 2000 2001 2004.03 2004.12 2008.01 2008.03 2012

Strong Pan-Blue 31.4 38.1 25.4 42.1 13.5 42.4 38.9 10.8 53.3 38.9 34.5

Weak Pan-Blue 34.6 35.3 24.9 35.9 7.7 42.9 34.8 14. 56.5 39.5 34.1

Leaning Pan-Blue 34.9 38.2 29.8 48.4 13.1 46.4 37.4 10.1 63.3 37.7 34.6

Independent 46.7 50.8 40.7 58.8 22.3 66.3 54.2 23.7 72.5 52.3 48.5

Leaning Pan-Green 21.3 5.9 26.9 34.9 12 58 38.2 15.2 58.9 37.2 43.5

Weak Pan-Green 33 40 22.6 40.3 13.9 50.4 42.3 13.7 49.5 41.5 40.7

Strong Pan-Green 28.6 25.5 10 36.9 1.1 52.7 43.3 19.4 69.4 53.6 45.1

Source and Note: Same as table 4.2.

Fig. 4.2. Lack of political interest among independents with different educational levels. Data
Sources and Notes: Same as table 4.2.
Page 84 →

On Cross-Strait Relations
As mentioned above, the point of disagreement between the Pan-Blue and the
Pan-Green camps has been their different positions on the issue of unification
with China versus Taiwan independence. The choice is not an easy one for the
majority of Taiwan citizens to make, and the preference of maintaining the
status quo has become the more preferred alternative during the past decades.
On average, up to 55 percent of Taiwan voters have chosen neither unification
nor independence. This significant number has also implied that support for
maintaining the status quo comes from all types of partisan citizen, and
particularly from the independents. As indicated in table 4.4, the partisan
difference of the two political camps is significant only between the strong
party identifiers and the rest of the categories. Both Page 85 → the Strong Pan-
Blue identifiers and Strong Pan-Green identifiers (and relatively, the Weak Pan-
Green identifiers) have lower support for maintaining the status quo. The
Weak Pan-Blue, Leaning Pan-Blue, Independent, and Leaning Pan-Green, by
contrast, are strong supporters of the status quo. In particular, Leaning Pan-
Blue identifiers are the most likely to choose maintaining the status quo than
all other types of partisans. While independents also choose maintaining the
status quo, they do not present significant differences from party leaners.
Table 4.4 Distribution of Support for Maintaining the Status Quo

1992 1995 1996 1998 2000 2001 2004.03 2004.12 2008.01 2008.03 2012

Strong Pan-Blue 55 46 35.2 41.6 37.9 28.3 48.9 44.6 62.3 52.2 64.1

Weak Pan-Blue 60.9 52.5 51.1 44.4 46.6 54.6 60.4 68 67.4 67 66.9

Leaning Pan-Blue 68.6 62.9 57.6 56.1 53.1 50.2 65 56.9 73.7 71.3 69.9

Independent 51.2 48.5 51.3 54.3 51.2 55 57 58.6 55.9 64.4 66

Leaning Pan-Green 56.2 41.2 63 37.9 55.8 52.7 45 54.1 46.2 50.2 50.2

Weak Pan-Green 48.6 41.8 42.7 48.2 54.3 51.3 41.3 44.2 43.1 42.5 42.7

Strong Pan-Green 32.7 29.8 28.3 30.7 48.4 29.6 20 29.7 31.4 23.2 35.6

Data Sources and Notes: Same as table 4.2.

Fig. 4.3. Support among independents with different educational levels for maintaining the status
quo. Data Sources and Notes: Same as table 4.2.
If we take a closer look at independents’ preferences on maintaining the status
quo, figure 3 shows that there are noticeable differences among the three types.
On the one hand, the highly educated independents have been very supportive
of maintaining the status quo. The upward trend also suggests that the highly
educated independents have become core supporters for maintaining the status
quo. On the other hand, the least-educated independents have been less likely
to support maintenance of the status quo. Yet the trend after early 2008 did
show a strong increase up to 2012. As for the moderately educated
Independents, their preference for maintaining the status quo resembles that of
highly educated independents. Also, the attitudes of the moderately educated
independents are also more stable than the other two types of independents.

Support for Democracy

As a member of the Third Wave democracies, Taiwan has made an admirably

smooth and peaceful political transition over the past decades. However, those
achievements in democratic transition did not automatically bring about good
governance in Taiwan. Like many other Third Wave democracies, Page 86 → the
challenges of political transition may not have been associated with the
establishment of democratic institutions but, rather, with the new
socioeconomic issues occurring at the same time. In the past ten years, citizens
in Taiwan have encountered sluggish economic development, repeated political
corruption, and worsening social inequality. Consequently, the poor economic
performance and ineffective governance could have disillusioned citizens about
democracy and reminded them of the “good old days” in the authoritarian era
(Chang, Chu, and Park 2007).

Fortunately, whether citizens lost their confidence in democracy under poorly

performing governments is not clear. Table 4.5 shows that regardless of
partisan differences, the majority of citizens have maintained rather strong
support for democracy. Relatively speaking, citizens with stronger party
identification are also more supportive of democracy. Pan-Green camp
identifiers are more supportive of democracy than are Pan-Blue identifiers. The
Strong Pan-Green identifiers, in particular, have been the most stable believers
in democracy compared with other partisans. As for independents, their belief
in democracy is not significantly different from the Leaning Pan-Blue
identifiers but is slightly lower than for Leaning Pan-Green identifiers.
Independents who are not supportive of democracy present a different picture
when the level of education is taken into consideration. As figure 4.4 shows,
support for democracy by independents displays a surge as well as a decline
over the past decade. Independents’ support for democracy increased
significantly from 2000 to 2004, then dropped sharply from 2004 to 2012.
Figure 4.4 further indicates that the highly educated independents have been
slightly more supportive of democracy than both the moderately and least-
educated independents. Yet the three types of independent have moved closer
to each other in 2008 and 2012.

Table 4.5 Distribution of Support for Democracy

2000 2001 2004.03 2004.12 2008.01 2008.03 2012

Strong Pan-Blue 51.8 81.7 72.9 88.3 54.7 44.5 59

Weak Pan-Blue 48.4 89.1 83.9 85.2 53.6 46.3 57.9

Leaning Pan-Blue 54.2 84.6 74.1 88.7 53.1 49.2 47

Independent 52 82.5 82.7 87.7 56.5 43.1 47.1

Leaning Pan-Green 53.3 89.7 89.3 93.6 54.5 48.8 60.6

Weak Pan-Green 61.2 92 93.1 91.4 61.2 59.4 63.7

Strong Pan-Green 70.2 90.9 89 94 81.6 75.8 73.9

Data Sources and Notes: Same as table 4.2.

 Fig. 4.4. Support for democracy among independents with different educational levels. Data
Sources and Notes: Same as table 4.2.
Page 87 →

Voting

One of the bedrock arguments made about party identification is the party’s
strong impact on citizen’s voting. Citizens with stronger party identification are
supposed to have stronger intentions of supporting the party with which they
identify in elections. Partisan voters are therefore more likely to vote the way
parties request. Table 4.6 presents the trends in not voting among different
partisan voters. It shows, first, that stronger party identifiers are far more likely
than others to cast their ballots in elections. Except for the Strong Pan-Green
voters who had a higher percentage of not voting in 1996, 200412 and 200801,
the number of those not voting in both the Strong Pan-Blue and Strong Pan-
Green camps has been less than 10 percent. Table 4.6 also shows that weak
party identifiers are less likely to vote than strong party identifiers. Both the
Leaning Pan-Blue and Leaning Pan-Green identifiers are also less likely to vote
than the weak party identifiers. Fourth, except for some elections, political
independents tend to be the most unlikely to vote in elections. The general
pattern of not voting and partisanship seems to reasonably confirm the
conventional wisdom that partisanship does matter to a citizen’s intention to
vote or not.

Figure 4.5 provides additional information about different types of political

independents and their intention to vote. In the early 1990s, there was no clear
difference among the highly, moderately, and least-educated political
independent in voting. Noticeably, this pattern changed in 1998, when highly
educated political independents refrained from voting, as did the moderately
educated political independent, though with a moderate change. This new
pattern continued until early 2008. Also noticeably, the least-educated political
independents were by no means absent from voting. They were more likely to
vote than their more educated counterparts.

Page 88 →

Vote Choice

Based on the perspectives of party identification, citizens with a certain

partisanship would be expected to be more likely to support a given party in an
election. In Taiwan, citizen partisanship is effectively associated with vote
choice in elections. As indicated in table 4.7, different types of Pan-Blue
identifiers revealed consistent support for their party candidate. Moreover, the
strength of identification was also in agreement with the assumption of
transitivity, in which stronger identifiers tend to vote for their partisan
candidate more than weaker identifiers do. As for independents, in the 1996
election the Pan-Blue presidential candidate, Lee Teng-hui, dramatically
outperformed his competitor, Peng Ming-min. More than 41 percent of
Independents voted for Lee, while only 4 percent voted for Peng. This sharp
difference also accounts for Lee’s strong victory.

As discussed above, the KMT suffered a serious internal split during the
presidential election in 2000. This internal split also led to a split among the
Pan-Blue identifiers. Only 37.8 percent of Strong Pan-Blue identifiers and 34.8
percent of Weak Pan-Blue identifiers voted for their partisan candidate. The
number was even smaller for those Leaning Pan-Blue identifiers. Independents
were in favor of the Pan-Green candidate in 2000, which contributed to the
electoral success of the DPP presidential candidate, Chen Shui-bian. As for the
Pan-Green identifiers, they did not vote for the Pan-Blue candidate. Unlike the
Pan-Blue identifier, the partisan boundary for the Pan-Green identifier was
more consistent and predictable.

The presidential elections in 2004, 2008, and 2012 were typical Pan-Blue Page 89
→ versus Pan-Green competitions, in which there were neither internal splits
nor strong independent candidates. It is obvious that the impact of
partisanship on elections became even stronger over time. Distribution of party
identification in the three elections looked reasonable with no obvious
deviations. Also, the strength of party identification was relatively consistent.
Interestingly, unlike the 1996 and 2000 elections, in which independents one-
sidedly favored either the Pan-Blue or the Pan-Green candidate, their vote
choice seemed to be more divided. They had a higher tendency to support the
Pan-Green candidate in 2004 but were more likely to back Pan-Blue candidates
in the 2008 and 2012 elections. As the electoral competition between the two
political camps became intense, the influence of independents has also
increased. Interestingly, sizeable numbers of Taiwan voters continue to self-
identify as independents, as table 4.7 shows. This could be due to the historical
memories of one-party authoritarian rule that makes island citizens reluctant to
report partisanship. It could also be the unsatisfactory performance of political
parties as a whole, which leads to an Page 90 → unwillingness of identifying with
any party. For whatever reasons, political independents in Taiwan deserve
further analysis in future research.

Table 4.6. Distribution of Not Voting

1992 1995 1996 1998 2000 2001 2004.03 2004.12 2008.01 2008.03 2012

Strong Pan-Blue 3.1 6.5 6.1 6.3 4.3 9.8 2.2 9.2 8.5 3.7 4.2

Weak Pan-Blue 14.3 13.6 5.4 11.2 5.6 14.8 4.2 17.1 20.9 5.6 5.1

Leaning Pan-Blue 22 8.8 10.2 14.1 9.2 20.3 7.6 28.7 30.6 9.8 14.8

Independent 13.1 14.7 13 17.7 8.6 20.4 15.3 30.2 36.4 19.3 20.2

Leaning Pan-Green 14.6 5.6 7.4 28.4 7.5 22.5 10.5 29.8 25.4 16.4 12.2

Weak Pan-Green 9.5 11.6 6.3 12.7 8 14.8 4 15.7 18.5 9.4 11.2

Strong Pan-Green 4.2 4.3 11.5 8.1 2.1 7 0 10.8 12 2 2.5

Data Sources and Notes: Same as table 4.2.

 Fig. 4.5. Not Voting among Independents with Different Educational Levels. Data Sources and
Notes: Same as table 4.2

Table 4.7 Percentage of Votes for Pan-Blue Presidential Candidate

1996 2000 2004.03 2008.03 2012

Strong Pan-Blue 87.6 37.8 87.4 94.9 95.1

Weak Pan-Blue 75.1 34.8 87.9 91.6 91.3

Leaning Pan-Blue 56.9 22.4 76.3 85.3 79

Independent 41.6/4.0 11.1/30.0 19.3/34.4 39.3/16.0 35.6/20.7

Leaning Pan-Green 28 7.8 4.2 9.3 10.8

Weak Pan-Green 29.5 8.2 0.8 5.2 3.4

Strong Pan-Green 21.2 4.3 0 1.9 1.7

Data Sources and Notes: Same as table 4.2.

Notes: (1) The two numbers for the Independent category represent the vote for the Pan-Blue candidate
and the Pan-Green candidate, respectively. (2) Since the Pan-Blue and Pan-Green alliances were not
formed until after 2000, vote shares of Pan-Blue candidates in 1996 and 2000 elections refer only to those
of KMT candidates.
 Fig. 4.6. Vote choice among independents with different educational levels. Data Sources and
Notes: Same as table 4.2.
Figure 4.6 shows that highly educated independents were less likely to vote
for the Pan-Blue candidate in the presidential election of 1996. In contrast,
least-educated independents were strongly supportive of the Pan-Blue
candidate. The support for the Pan-Blue candidate in the 2000 election
declined concurrently among the three types of independents. The differences
in vote choice among independents with different educational levels seen in
1996 disappeared. In 2004, highly educated independents were more willing to
vote for the Pan-Blue candidate than their less educated counterparts. The
election of 2008 was similar to that of 2000, with the three types of
independents not showing much difference. Yet all the independents tended to
show considerable support for the Pan-Blue candidate. The highly educated
independents turned their backs on the Pan-Blue candidate in 2012. Both the
moderately educated and least-educated independents had higher percentages
of support for the Pan-Blue candidate. In addition to the changing patterns of
vote choice among independents, figure 4.5 also reveals one significant feature:
there is a correspondence between the elections that lead to power change and
the convergence of vote choice among the three types of independents. The
KMT’s loss of power in the 2000 elections was accompanied by unified,
declining support of independents. A similar instance occurred in 2008, when
the Independents almost unanimously supported the Pan-Blue candidate who
won the election.

Concluding Remarks

The evolution of the party system in Taiwan has been closely tied to the
development of democracy. The development of citizen partisanship has Page 91

→ taken a distinct route following the process of democratic opening. The

long-standing KMT had monopolized the political marketplace early on,
relegating citizen partisanship to nonpartisan status if they were not supporters
of the KMT. Nevertheless, those nonpartisan citizens participated in elections,
supported certain non-KMT candidates, and paved the way for alternative
partisanship later on. The establishment of the DPP represents a watershed
development in citizen partisanship, having made possible competing
partisanship. Additionally, the development of a multiparty system and the
formation of Pan-Blue and Pan-Green camps in the 2000s significantly
transformed citizen partisanship. There have been both changes and continuity
in the party system, as in citizen partisanship. Individually, each political party
has certain social bases, yet those bases are somewhat shared by allied parties.
That means a citizen might change his or her support from election to election.
The difference on the policy of cross-Strait relations has separated the Pan-
Blue and Pan-Green as well as supporters of each camp. Citizens seem to be in
the situation of a two-party system, which has a distinct partisan boundary that
is unable to be crossed. Therefore, a citizen in Taiwan could be either a
supporter of the KMT, NP, and PFP, or a supporter of the DPP and TSU.
 The distinct historical development of the political situation in Taiwan has
also given birth to a considerable number of political independents. Repressive
measures by the KMT in the early period generated a hostile political
atmosphere for many Taiwanese, so maintaining a politically independent
status was a naturally safe and expedient choice, and the implementation of
local elections provided these non-KMT citizens channels of participation
without committing to any partisanship. The existence of the independent thus
went hand in hand with the development of the party system. The independent
would later turn into a strong power base for the first meaningful opposition
party, the DPP. As Taiwan entered a period of rapid political transition in the
late 1980s, Taiwan citizens began to experience true two-party partisanship.
This new situation soon transformed again as new political parties continued to
emerge in the early 1990s and 2000s. Multipartisanship appeared following the
establishment of the NP, the PFP, and the TSU. Nonetheless, the formation of
multiple partisanship does not imply the eclipse of the independent, a
considerable number of whom still exist in the political marketplace. Also
notably, the multiparty system has been in fact more like two-party camps
within which citizens are more likely to cast their ballots according to party
coalition, especially in presidential elections.

Thus, citizen partisanship has been both continuous and changing. Strong
partisan citizens are more consistently affiliated with a given party, Page 92

→ but that is not the case for weak partisans and party leaners. The latter is
obvious in the Pan-Blue camp. In contrast, the Pan-Green camp has enjoyed a
more stable partisan affiliation with citizens. As for the independents, they
generally resemble the conventional picture depicted in The American Voter
(Campbell et al. 1960). Given the significant number of independents voting,
they are certainly not to be ignored in elections, but they do behave differently
according to their different educational levels. Though highly educated
independents are less interested in politics and less likely to vote in elections,
they do maintain a rather supportive attitude toward democracy and
maintaining the status quo in cross-Strait relations. It is likely that these highly
educated independents are comparably more autonomous and prudent than the
least-educated independents, who are more likely to be mobilized in elections.

Modern democracy needs political parties. Even though the normative

functions of the political party might change over times, the political party
continues to be indispensable for democracy. More specifically, the formation
and operation of the party system significantly affect the functioning of
democracy. One of the core assumptions of party politics is a stable connection
between citizens and political parties; the majority of the citizens in society are
able to voice their preferences through political parties. Yet if a considerable
number of citizens do not maintain a regular connection with a political party,
it is therefore implied that the political party does not matter much in society.
Worse still, a lack of partisan connection also implies a lack of important media
transmitting citizens’ political information. Citizens would drift around the
political world without an anchor (Wattenberg 1986, 130). Given the evidence
of a high degree of overlap in attributes between the pure independent and
partisan leaners in Taiwan, a clear classification of leaners is important both
theoretically and practically. If we treat leaners as extended partisan voters, then
the number of independents is less significant. It also leads to the conclusion
that a relative majority of the electorate is partisan and that political parties
have penetrated the major part of society. The party system is then accordingly
stable. On the other hand, if we treat leaners as equal to independents, then
more than one-third of the electorate will be labeled as independents; thus
there exists a strong segment of uncommitted nonpartisan voters in elections.
The formation and essences of party competition would also be affected
because more centrist party appeals, rather than purely partisan-oriented ones,
would become dominant in elections. Therefore, the changing partisan alliances
of these independents, somewhat including the leaners, continue to play a
crucial role in Taiwan’s electoral politics.
Appendix 4.A1. Measurements of Key Variables

Variable Description and Measurements

Before 2004, respondents were asked the following questions:

i. Do think of yourself as close to any particular party? When respondents answer yes, then

they are asked,

Directions (1a) Which party do you feel close to? and
and strength (1b) Do you feel very close to this party, somewhat close, or not very close?” (strong, weak).
of party ii. When respondents answer no to question (i), then respondents are asked, “Do you feel
identification yourself a little closer to one of the political parties than to the others?” If respondents
answer yes, then they are asked (1a) again (leaners).
iii. When respondents answer “no” to question (ii), then respondents are regarded as
independents.

After 2004, respondents were asked slightly different questions.

i. Among the main political parties in our country, including the KMT, DPP, PFP, NP, and
TSU, do you think of yourself as leaning toward any particular party? (yes, then ask (iii) and
(iv); no, then ask (ii)).

ii. Do you feel yourself leaning a little more to one of the political parties than the others?
(yes, then ask (iii); no will be classified as independent)
iii. Which party is that?
iv. Do you lean very strongly, somewhat, or just a little to this party?

Pan-Blue
Citizens who identify with the KMT, the NP, and the PFP are classified as Pan-Blue
and Pan-
identifiers; citizens who identify with the DPP and the TSU are classified as Pan-Green
Green
identifiers.
identifiers

More generally, would you say that you are very, somewhat, not very, or not at all interested
Political
in politics? (not interested, very little/not much interested, somewhat interested, very
interest
interested)
Maintaining Respondents are asked: “Concerning the relationship between Taiwan and mainland China,
the status which of the following six positions do you agree with: (1) immediate unification; (2)
quo immediate independence; (3) maintain the status quo, move toward unification in the future;
(4) maintain the status quo, move toward independence in the future; (5) maintain the status
quo and decide later; (6) maintain the status quo forever. (Items 1 and 3 are classified as pro-
unification; items 2 and 4 are classified as pro-independence; items 5 and 6 are classified as
maintaining the status quo.)

Respondents are asked, “Some say that ‘Democracy might have some problems, but it is still
Support for
the best political institution.’ Do you agree or disagree?” (strongly agree, agree, disagree,
democracy
strongly disagree)

Page 93 →

Page 94 →

Appendix 4.A2.

List of Data Sources

Yih-Yen Chen. 1992. Electoral Behavior and Democratization in Taiwan: A

Study of Legislative Election in 1992. National Science Council Research
Plan, NSC 82–0301-H004–034.

Yih-Yen Chen. 1995. Electoral Behavior and Democratization in Taiwan (iv): A

Study of Legislative Election in 1995. National Science Council Research
Plan, NSC 84–2414-H-004–053 B2.
Hsieh, John Fu-sheng. 1995. An Interdisciplinary Study of Voting Behavior in
the 1996 Presidential Election. National Science Council Research Plan,
NSC 85–2414-H-004–017 Q3.

I-chou Liu. 1998. Constituency Characteristics and Voters’ Behavior: An

Integrative Study of Legislative Election in 1998. National Science Council
Research Plan, NSC 88–2414-H-004–017.

Yih-Yen Chen. 1999. Integrative Study of Voting Behavior in the 2000

Presidential Election. National Science Council Research Plan, NSC89–
2414-H-004–021-SSS.

Chi Huang. 2001. Taiwan’s Election and Democratization Study, 2001 (TEDS
2001). National Science Council Research Plan, NSC 90–2420-H-194–001.

Shiow-duan Huang. 2004. Taiwan’s Election and Democratization Study, 2002–

2004(III): The Presidential Election, 2004 (TEDS 2004P). National Science
Council Research Plan, NSC 92–2420-H-031–004.

I-chou Liu. 2004. Taiwan’s Election and Democratization Study, 2002–2004

(1/4): The Legislative Election, 2004 (TEDS 2004L). National Science
Council Research Plan, NSC 93–2420-H-004–005-SSS.

Yun-han Chu. 2008. Taiwan’s Election and Democratization Study, 2005–

2008(III): The Legislative Election, 2008 (TEDS2008L). National Science
Council Research Plan, NSC 96–2420-H-002–025.
Ching-hsin Yu. 2008. Taiwan’s Election and Democratization Study, 2008
(TEDS 2008P: Presidential Election). National Science Council Research
Plan, NSC 96–2420-H-004–017.

Yun-han Chu. 2012. Taiwan’s Election and Democratization Study, 2009–2012

(III): The Survey of the Presidential and Legislative Elections,
2012(TEDS2012). National Science Council Research Plan, NSC 100–2420-
H-002–030.

Notes

1. Mainstream wisdom concerning party identification in the tradition of

the Michigan school and American National Election Studies have
postulated the concept as a long-term psychological attachment to a given
group (a specific political party). For a party identifier, party identification
is a sense of group identity or belonging. Also by definition, party
identification should be stable, not having frequent fluctuations.
Therefore, data collection begins with a nonspecific statement asking the
respondent: Page 95 → “Generally speaking, do you usually think of
yourself as a Republican, a Democrat, an Independent, or what?” The
wording of this question also probes the respondent’s endurance of
staying with a political party. Those respondents who answer with either
“Republican” or “Democrat” are then asked, “Would you call yourself a
strong Republican (or Democrat) or a not very strong Republican (or
Democrat)?” This follow-up question allows the researcher to explore
respondents’ intensity of party identification. So, the two questions
 produce four types of party identifier: Strong Republican Identifier, Weak
Republican Identifier, Weak Democrat Identifier, and Strong Democratic
Identifier.

2. Although minor political parties, such as the Chinese Youth Party and
the Democratic Socialist Party, came to Taiwan with the KMT
government, they were merely window-dressing since the KMT
monopolized all political resources (Tien 1989). Thus, both parties were
too weak to challenge the KMT.

3. The actual number of registered KMT members has been in dispute.

Some estimate that the party has more than 2.5 million members while
others claim that it has had less than one million. The KMT has been
reluctant to release its membership. The electoral defeat in the 2000
presidential election has led to a sharp decline in party membership. It is
estimated that the KMT now has about 850,000 members (cited from Yu
2002).

4. There were other minor parties during this period. Because they either
did not participate in any elections or failed to generate significant
political impacts, this research excludes them from the discussion.

References

Bartle, John, and Paolo Bellucci, eds. 2009. Political Parties and Partisanship: Social
Identity and Individual Attitudes. London: Routledge.

Bryce, James. 1929. Modern Democracies. London: Macmillan.

Burnham, W. D. 1970. Critical Elections and the Mainsprings of American Politics.
New York: W. W. Norton.

Campbell, Angus, Philip E. Converse, Warren E. Miller, and Donald E. Stokes.

1960. The American Voter. Chicago: University of Chicago Press.

Campbell, Angus, Gerald Gruin, and Warren E. Miller. 1954. The Voter Decides.
Evanston, IL: Row Peterson.

Chang, Yu-tzung, Yunhan Chu, and Chong-min Park. 2007. “Authoritarian

Nostalgia in Asia.” Journal of Democracy 18 (3): 66–80.

Chen, Yih-yan. 1986. “The Study of Voting Behavior in Taiwan: Reviews and
Prospects” [in Chinese]. Thought and Words 23 (6): 557–85.

Chen, Yih-yan. 1994. “Predicting Voter’s Choice in the 1992 Legislator

Election: A Cluster Analysis” [in Chinese]. Journal of Electoral Studies 1 (1): 1–
37.

Cheng, Tun-jen, 1989. “Democratizing the Quasi-Leninist Regime in Taiwan.”

World Politics 41 (4): 471–99.

Cheng, Tun-jen, and Stephan Haggard, eds. 1992. Political Change in Taiwan.
Boulder, CO: Lynne Rienner.

Chu, Shiau-yu. 2004. “The Research of Independent Voters’ Voting Behavior:

Comparison Page 96 → and Analysis of the 2000 and 2004 Presidential
Elections” [in Chinese]. Master’s thesis, Ming Chuan University.
Chu, Yun-han. 1996. “The Partisan Factor in Electoral Choice: A Case Analysis
of the 1991 National Assembly Election” [in Chinese]. Journal of Electoral
Studies 3 (1): 17–48.

Converse, Philip E. 1966. “On the Possibility of Major Political Realignment in

the South.” In Elections and Political Order, ed. Angus Campbell, Philip E.
Converse, Warren E. Miller, and Donald E. Stoke, 212–44. New York:
Wiley.

Dennis, Jack. 1988. “Political Independence in America. Part I: On Being an

Independent Partisan Supporter.” British Journal of Political Science 18 (1): 77–
109.

Hsieh, John Fuh-sheng. 2005. “Ethnicity, National Identity, and Domestic

Politics in Taiwan.” Journal of Asian and African Studies 40 (April): 13–28.

Keith, Bruce E., David B. Magleby, Candice J. Nelson, Elizabeth Orr, Mark C.
Westlye, and Raymond E. Wolfinger. 1992. The Myth of the Independent Voter.
Berkeley: University of California Press.

Lazarsfeld, Paul F., Bernard Berelson, and Hazel Gaudet. 1948. The People’s
Choice: How the Voter Makes Up His Mind in a Presidential Campaign. New
York: Columbia University Press.

Liu, I-chou. 1996. “The Behavior of Taiwanese Voters in 1992: Consolidation

of Partisan Ties.” In Taiwan’s Electoral Politics and Democratic Transition: Riding
the Third Wave, ed. Hung-Mao Tien, 226–41. Armonk, NY: M. E. Sharpe.
Liu, I-chou. 1997. “Difficulties of Concept Measurement in Political Science:
The Case of Party Identification” [in Chinese]. Presented at the Conference
of Research Methods of Political Science, National Chengchi University,
Taipei.

Lu, Alexander Ya-li, 1992. “Political Opposition in Taiwan: The Development

of the Democratic Progressive Party.” In Political Change in Taiwan, ed. Tun-
jen Cheng and Stephan Haggard, 121–46. Boulder, CO: Lynne Rienner.

Magleby, David B., Candice J. Nelson, and Mark C. Westlye. 2011. “The Myth of
the Independent Voter Revisited.” In Facing the Challenge of Democracy:
Explorations in the Analysis of Public Opinion and Political Participation, ed. Paul
M. Sniderman and Benjamin Highton, 238–63. Princeton: Princeton
University Press.

Niemi, Richard G., and Herbert F. Weisberg. 1993. Classics in Voting Behavior.
Washington, DC: Congressional Quarterly.

Niemi, Richard G., Stephen Wright, and Lynda W. Powell. 1987. “Multiple
Party Identifiers and the Measurement of Party Identification.” Journal of
Politics 49 (4): 1093–1103.

Petrocik, John R. 1974. “An Analysis of Intransitivities in the Index of Party

Identification.” Political Methodology 1:31–47.

Sartori, Giovanni. 1976. Parties and Party Systems: A Framework for Analysis.
Cambridge: Cambridge University Press.
Shyu, Huo-yan. 1997. “Elections and Social Cleavages in Taiwan: Party
Competition and Democratization” [in Chinese]. Paper presented at the
conference “Elections on Both Sides of the Straits,” Fairbank Center for
East Asian Research at Harvard University, Cambridge, May 8–9.

Tien, Hung-mao. 1989. The Great Transition: Political and Social Change in the
Republic of China. Stanford, CA: Hoover Institution Press.

Page 97 → Wachman, Alan M., 1994. “Competing Identities in Taiwan.” In The

Other Taiwan: 1945 to the Present, ed. Murray Rubinstein, 17–80. Armonk,
NY: M. E. Sharpe.

Wang, Fu-chang. 1993. “Causes and Patterns of Ethnic Intermarriage Among

the Hokkien, Hakka, and Mainlanders in Postwar Taiwan: A Preliminary
Examination” [in Chinese]. Bulletin of the Institute of Ethnology Academia Sinica
76 (Autumn): 43–96.

Wang, Jong-tian. 2010. “The Types of Independent Voters and Their Voting
Behavior: Some Observations from Taiwan’s 2008 Presidential Election” [in
Chinese]. Journal of Electoral Studies 17 (2): 35–70.

Wang, T. Y., and Ching-hsin Yu. 2011. “Independents and Voting Decisions: A
Test of James Bryce’s Hypothesis.” Presented at the 2011 International
Conference on Taiwan’s Election and Democratization Study, National
Chengchi University, Taipei.

Wattenberg, Martin P. 1986. The Decline of American Political Parties, 1952–1984.

Cambridge: Harvard University Press.
Weisberg, Herbert F. 1980. “A Multidimensional Conceptualization of Party
Identification.” Political Behavior 2 (1): 33–60.

Weisberg, Herbert F. 1993. “Political Partisanship.” In Measurement of Political

Attitudes, ed. John P. Robinson, Phillip R. Shaver, and Lawrence S.
Wrightsman, 681–736. San Diego, CA: Academic Press.

Wilson, Richard W. 1970. Learning to Be Chinese: The Political Socialization of

Children in Taiwan. Cambridge, MA: MIT Press.

Wu, Nai-Teh, 1995. “Ethnic Consciousness, Political Support, and National

Identity.” Taiwan Studies 1 (2): 45–66.

Yeh, Ming-yan. 1994. “The Research of Independent Voters in Taiwan” [in

Chinese]. Master’s thesis, National Chengchi University.

Yu, Chilik. 2002. “New Issues in the Study of Public Administration in Taiwan:
The Impact and Influence on the Civil Servant System by Alternation of
Political Power” [in Chinese]. Presented at the Conference of Professor
Chin-Chien Chang’s Centennial Anniversary, Chinese Society for Public
Administration, Taipei.
Page 98 →
 Chapter 5

Issues, Political Cleavages, and Party

Competition in Taiwan

Shing-yuan Sheng and Hsiao-chuan (Mandy) Liao

Taiwan experienced rapid socioeconomic and political changes in the 1960s

and 1970s and evolved from an authoritarian to a democratic political system
beginning in the mid-1980s.1 Since the early 1980s, many issues have emerged
in Taiwan’s political arena. Some of them quickly disappeared, some
temporarily attracted the attention of Taiwan citizens but gradually declined in
importance, and still others evolved into highly salient ones and have had
deterministic impacts on party competition and, hence, party turnover. Unlike
most industrial democracies, in which issues of wealth distribution and
materialism/postmaterialism (or fundamentalism/postmodernism) create the
most important political cleavages (Dalton 1988; Inglehart 1977, 1990, 1997;
Lipset and Rokkan 1967; Moreno 1999), those issues have not evolved into
prominent ones for party competition in Taiwan. Instead, during the 1980s, as
Taiwan transitioned to democracy, parties competed on the issue of reform
versus antireform (stability). The partisan elites of the Democratic Progressive
Party framed the reform issues and attracted the attention of Taiwan voters. As
a result, the DPP gradually transformed itself into a strong opposition party
with the support of about 30 percent of the electorate at the beginning of the
1990s. After political reforms and the resolution of unfair political practices,
the party faced the challenge of searching for a new issue to appeal to a larger
share of the electorate. So, it shifted the battlefield and framed the
independence/unification issue in order to connect with the enduring social
cleavages—ethnicity and Chinese/Taiwanese identity. Soon, the issue not only
dominated the political discussions of the elites and the public but it also
shaped party competition and affected elections.

Page 99 → This chapter, which is based on observation of the period from 1996
to 2012, endeavors to answer the following question: Why have some issues
evolved to create important political cleavages that have shaped party
competition in Taiwan, whereas other issues have not? Four issues are
discussed in this chapter: wealth distribution, environmental
protection/economy, reform/stability, and independence/unification. We
explore the answer to the research questions from the perspective of the elites
and that of the public. From the viewpoint of the elites, we show how they
frame and manipulate different issues in the political arena to attain their
political goals. From the viewpoint of the public, we show how the voters’
positions on issues coincide with manipulation by the partisan elites and how
the voters perceive the importance of the issues.

Issues and Political Cleavages in Taiwan

The importance of issues in politics has long been recognized and emphasized
by political scientists. Carmines and Stimson (1989, 3) describe it well in their
classic book Issue Evolution: “To speak of politics is to speak of political issues.”
As they argue, there are many issues in a political system, but the majority of
them lie dormant most of the time. Only a few issues occasionally “rise from
partisan obscurity and become so contentious, so partisan, and so long lasting
that they come to define the party system in which they arise, to transform the
grounds of debate which were their origin” (Carmines and Stimson 1986, 901).
According to Carmines and Stimson, which issues become salient and which
remain dormant depend on the actions of partisan elites and the responses of
the mass electorate. First, elites instinctively know that some issues may benefit
them, and they work to clarify those issues and frame them in partisan terms.
Second, the mass public must alter its cognitive perception of the parties with
respect to new issues, care about the differences among parties, and even
change its political attitude and vote choice according to its opinions about
new issues. When a substantial part of the mass public starts to change its
cognition, attitude, and behavior in this way, all parties in the political system
are necessarily forced to take a position along the dimensions of the new issue.
At this moment, the issue experiences an evolution and affects the agendas of
parties, the discourses of the elites, and party identification and vote choice of
the mass electorate. Carmines and Stimson illustrate an issue evolution in
which the party elites grew increasingly polarized on civil rights in the 1960s
and 1970s, leading the mass electorate to become similarly polarized on that
issue.

Page 100 → Furthermore, Lipset and Rokkan (1967) argue that political
cleavages reflect social cleavages, and social cleavages are determined
dramatically by the historic conditions of national and socioeconomic
development. Accordingly, specific national conditions lead to a variety of
alliance patterns among leaders of various social groups. As a result, these
cleavages define the potential social bases of political conflicts. Lipset and
Rokkan’s analysis offers a reasonable explanation for the rise of the party
systems and voter alignments in advanced democracies. In most advanced
democracies, the wealth distribution issue based on class is the most common
factor and perhaps the most important political cleavage. However,
approximately from the beginning of the 1970s, economic factors and
traditional class-conflict models fail to explain contemporary political
phenomena. Those who are materially better off protest the most, rather than
those who are materially disadvantaged. Traditional political cleavage and
theory of class conflict cannot explain this new political trend. Inglehart (1990,
1997) adds a new aspect—postmaterial or postmodern values based on
noneconomic issues—to the formation of political cleavages. This new cleavage
and the old wealth distribution cleavage are the most important issues in most
advanced democracies (Moreno 1999).

Taiwan experienced rapid socioeconomic and political changes in 1960s and

1970s, and evolved from an authoritarian to a democratic political system
beginning in the mid-1980s. A great deal of evidence suggests that from the
mid-1980s to the beginning of the 1990s, when Taiwan experienced a transition
to democracy, reform versus antireform (stability) was the most salient political
issue (Cheng and Hsu 1996; Chu 1994; Shyu 1998; Wu 1993). As many new
democracies in East Europe and Latin America, the democratic-authoritarian
or reform-antireform is the most salient issue (Moreno 1999). Chu (1994)
argued that the Taiwan public focused more on issues of political reform than
on issues of economy and wealth distribution during this period because the
Taiwan government’s economic policies had successfully resolved the problems
of wealth accumulation and distribution during the decades that economic
development heated up. Hsieh and Niou (1996a) examined the 1992 legislative
election and found that the reform/stability issue had the greatest impact on
voters’ evaluations of parties, followed by the wealth distribution and the
independence/unification issues, with the environment/economy issue having
the smallest impact. However, when Hsieh and Niou reviewed the 1993 local
elections for county magistrates and city mayors, they found that the
independence/unification issue and two valence issues—public work and
anticorruption—had important impacts on voters’ evaluations of the parties.
Other positional issues such as Page 101 → wealth distribution, environment
protection, and reform had limited impacts (Hsieh and Niou 1996b). Because
the debate on independence versus unification was increasingly contested in the
1990s, when political scientists examined the 2001 legislative election, they
found that independence/unification was the most influential issue and had
considerable impact on voters’ party evaluations and vote choice. The
reform/stability issue had some impact and was the second most important
political cleavage. The wealth distribution issue also had some impact and a
significant effect on vote choice, but the environment/economy issue was still
not influential in elections (Hsieh 2005; Sheng and Chen 2003).

Given the research results in previous literature, we may suspect that the
impacts of different issues fluctuate in different elections. This chapter will
systematically answer why and how some issues have created important
cleavages that have shaped party competition in Taiwan while other issues have
not. As noted, four issues are examined in this chapter: reform/stability, wealth
distribution, environmental protection/economy, and
independence/unification. We choose the four issues for two major reasons.
First, the reform/stability issue and independence/unification issue have
occupied Taiwan election platforms and have polarized Taiwan politics for a
long time. The wealth distribution issue and the environmental
protection/economy issue marked the old and new left-right issues that shape
endurable political cleavages in advanced democracies although Taiwan is not
such a case (Hsiao, Cheng, and Achen, chapter 9 of this book; Norris 2004). As
parties and politicians in Taiwan raised these issues in elections and in the
legislature frequently and sometimes caught the public’s attention, these two
issues did not form political cleavages.

The other reason for choosing these four issues is that they are position
issues. A position issue is one on which the opinions among the electorate may
easily be divided, such as the extent to which we should empower the
government. On the contrary, issues such as economic development or
anticorruption are valence issues—these are issues that are uniformly liked or
disliked by the electorate (Fiorina 1981). Position issues are more likely to
develop political cleavages and shape party competition than valence issues
since parties can take opposing sides to mobilize the electorate (Stokes 1963).
Valence issues have less potential to form a long-term political cleavage because
they do not differentiate parties effectively.

Thus, this chapter will dissect the four issues in Taiwan and demonstrate their
evolution from the perspectives of both the partisan elites and of the public.
We selected the observation period from 1996 to 2012. Data are from two
major databases: data on the elections of 1996, 1998, and 2000, Page 102 → which

were collected by the Election Study Center at National Chengchi University,

and data on the elections of 2001, 2004, 2008, and 2012, which were collected
by the Taiwan Election and Democratization Study project. All of these data
are from face-to-face interviews with individuals of the Taiwan public aged 20
and above and are based on probabilistic sampling.

In the following, we first focus on the partisan elites’ role in issue evolution
and discuss how partisan elites frame and manipulate the four issues in order
to attain their political goals. We also demonstrate how Taiwan citizens
perceive the elites’ signals and update their positions on issues. Then, we
discuss mass perception of the importance of issues. Last, we conclude with the
issue evolution of the four focal issues in Taiwan.

The Framing and Manipulation of the Issues by the

Partisan Elites

In the process of issue evolution, the elites of the Dang-wai (the non-KMT
forces and the forerunner of the DPP before 1986; see chapter 4) and DPP play
an important role, not only because the Dang-wai and DPP elites are strategic
politicians but also because they are from the opposition forces when Taiwan
endured a transitional period from authoritarian to democracy. They are
ambitious to acquire governing power. Politicians of the opposition or
minority parties naturally turn to new issues to improve their political
situation, whereas politicians of the majority parties naturally seek to maintain
the salience of the current agenda (Carmines and Stimson 1989, 12–13). The
Dang-wai and DPP elites strategically search for issues that benefit their
group’s growth and strength. Furthermore, they frame and manipulate the
issues to mobilize the mass public. From the beginning of the 1980s, the Dang-
wai and DPP elites have capitalized on four major issues: political reform,
social welfare, environmental protection, and Taiwan independence. The issues
of political reform and Taiwan independence have been relatively more
effective than the other two in terms of issue evolution. In the early 1980s, the
Dang-wai and DPP elites used political reform issues to challenge the
established authoritarian Kuomintang government and attract voters. After
political reform was achieved, the DPP elites switched to the
independence/unification issue and acted as an advocate of Taiwan
independence from the beginning of 1990s. The success of manipulating issues
to mobilize the mass public extended the support base of the DPP in the 1980s
and 1990s and brought the DPP to the presidency in 2000 and 2004. However,
once the DPP captured power, its room for manipulating issues became
smaller for two reasons. First, once the DPP was in power, Taiwan Page 103

→ voters were no longer satisfied with only rhetoric during elections; instead,
they wanted to see the actual implementation of the DPP’s campaign promises.
Second, the DPP soon found that it was limited by fiscal difficulties; in
particular, Taiwan has been in an economic downturn since 2000.

In the following, we will describe issue by issue how the elites have framed
and manipulated them. We begin with the very first one that appeared in
Taiwan politics—reform versus stability—and end with the most significant
issue in Taiwan—independence versus unification.

The Reform/Stability Issue

Taiwan experienced rapid economic development from the beginning of the

1960s through the 1970s and 1980s. As a result, Taiwan society experienced
dramatic socioeconomic change and transitioned from an underdeveloped
country to an industrialized society in the 1980s. Accompanying this
transformation, there emerged a new middle class with more education and
sophisticated political skills. This new class and the politically suppressed
Taiwanese, whose parents or grandparents had lived on the island before 1948,
strongly pushed the dominant KMT government to undertake political
reforms. However, the KMT government was reluctant to respond to the
prodding. Utilizing a reform-oriented strategy, the Dang-wai gradually
attracted the electoral support of Taiwan citizens. Eventually they got enough
support to organize a formal party, the DPP, in 1986. Indeed, the Dang-wai got
just 13.0 percent of all votes in the 1980 legislative election; however, the share
of votes for the DPP (established in 1986) increased to nearly 30 percent by the
end of the 1980s (see chapter 4 for details).

The reform/stability issue was the most salient one in the 1980s, during
Taiwan’s transition to democracy. Some Taiwan citizens were worried about
instability because of the rapid and radical political reform. They tended to
identify with the KMT, whereas those taking political reform more seriously
tended to identify with the DPP (Sheng and Chen 2003). In a survey conducted
in 1991 about the most significant problem in the country, 36.4 percent of
respondents mentioned political structure, 14.2 percent mentioned national
status and national identity, while only 9.3 percent mentioned wealth
distribution (Wu 1993, 6).

The opposition forces made great efforts to promote political reform and
earned a reputation for being reformist. There was a considerable number of
issues on the reform agenda from the mid-1980s to the beginning of the 1990s,
including the lifting of martial law in 1987, termination of the Temporary Page
104 → Provisions Effective During the Period of National Mobilization for
Suppression of the Communist Rebellion, and with the return to a regular
constitutional structure in 1991, full-scale elections of national representatives
in 1992, a popular vote for the provincial governors and city mayors of Taipei
City and Kaohsiung City in 1994, and direct elections for president in 1996.
Even when the large and structural changes were achieved, the DPP continued
its reformist role and turned its focus to social and economic reforms. These
reforms include anticorruption, anticrime, constitutional reform, dealing with
the KMT assets, and social welfare policies. In the process, the DPP legislators
sometimes allied with members of the New Party, a newly established small
party that sometimes collaborated with KMT legislators. The DPP might not
be the owner of the particular reform issue as it was during the earlier period,
but it was more active in this respect than the KMT (Sheng 2001).

A survey conducted by the Election Study Center in 1993 asked respondents

about their impressions of the two major parties. The results showed that 35.8
percent of respondents perceived the DPP as a radical party, and 32.4 percent
of respondents perceived it as a violent party. In contrast, 43.0 percent of
respondents perceived the KMT as a conservative party (Liu 1994, 64). Even at
the end of 2000 and the beginning of 2001, after the DPP had won the
presidency and began to govern the country, the mass public’s principal
negative image of the DPP was still that it was violent and radical, whereas the
principal positive image of the DPP was its contribution to democratic reform
(Cheng 2004, 195–98). From figure 5.1, we can see that Taiwan voters on
average located the DPP at 3.9 on the reform/stability issue dimension in 1996
and at 4.6 in 2001.2 However, because of the fierce party competition in the
legislature and in the elections, the DPP returned to the reform-oriented
position that it had held in the past. The DPP government directed two
financial reforms and held referendums on national issues in 2004. The slogan
of President Chen’s reelection in 2004 was “Taiwan first, Reform first.” Not
surprisingly, the public located the DPP at 3.9 in 2004. In contrast, the public
perceived the KMT as taking stability more seriously and as being more
conservative in regard to political reform. The public located the KMT at 6.9 in
1996 and did not shift much over the years, until 2008.

In mid-2006, President Chen Shui-bian and his family members were accused
of improper trading of shares, misuse of government funds, and corruption.3
In 2008, after leaving office, Chen Shui-bian was convicted of corruption and
money-laundering, shocking the Taiwan public. When more evidence revealed
that ex-president Chen had accepted money from Page 105 → bankers during the
second financial and banking reform, the public became even more
disillusioned with both Chen Shui-bian and the DPP. In the 2008 election, the
KMT presidential candidate, Ma Ying-jeou, took the opportunity to promote
reforms in administrative ethics and social justice. He promised to initiate
legislation on the criminal liability of illicit wealth of public servants and to
comprehensively examine constitutional reform.4 In contrast, as the governing
party the DPP took social stability more seriously than when it was in
opposition. The DPP candidate, Hsieh Chang-ting, stressed the importance of
consensus on reform. The public’s perception of the two parties’ issue positions
along the reform/stability spectrum reflects this situation. In 2008 the public
perceived the DPP, which was rated at 4.7, as more centrist than in 2004, while
the KMT, at 6.1, was seen as leaning more toward reform than it had been in
2004. Along with this development, both the KMT and DPP have taken a more
centrist position since 2008. In such a case, there is less space for parties to
manipulate this issue, which leaves the possibility of convergence of their issue
position in the future.

Fig. 5.1. Respondents’ issue positions on reform/stability and their perceptions about the
positions of the parties (1996–2008). Source: Appendix 5.A2. Note: On the horizontal axis, “P”
indicates that surveys were conducted after the presidential election; “L” indicates that surveys
were taken after the legislative election.

Page 106 →

The Wealth Distribution Issue

Unlike most advanced democracies, in which issues of wealth distribution are
salient for party competition, issues of wealth distribution are relatively not so
important in party competition in Taiwan. The evolution of this issue in
Taiwan has gone through different phases. Before the mid-1980s, both the
successful economic policy and the weak consciousness of the working class
detracted from the attention given to wealth distribution issues (Chu 1994, 3).
At the end of the 1980s and the beginning of the 1990s, social welfare issues
became more prominent due to fierce party competition and the increasing gap
between the rich and the poor; however, such issues have become less
polarizing, since most parties strive to offer social welfare benefits and none of
them want to be seen as opposing benefits. More important, the
independence/unification issue has attracted most of the attention of the
parties and the electorate. Last, after party turnover in the presidency in 2000,
the differences between the parties’ issue positions diminished while the DPP
was in power and faced fiscal problems and the KMT stayed with a catch-all
strategy on this issue. Later, we will discuss this issue according to its
development over time and explain its ups and downs in Taiwan politics.

From the beginning of the 1980s, the parties and candidates emphasized
wealth distribution issues more actively than before. This was connected to
ethnicity, the most significant social cleavage in Taiwan.5 When the KMT
government was moved to Taiwan in 1949, many mainlanders who worked for
the military, public sector, and education sector immigrated to Taiwan with the
government; thus, a high proportion of public servants were mainlanders. So
the KMT’s social welfare policies in the early era focused more on these groups
in order to consolidate its ruling apparatus. In 1987, the newly founded DPP
set up a department whose purpose was forming an alliance with social-
movement organizations in order to confront the KMT government.
Responding to political reforms and to the challenge to its conservative welfare
policies, and hoping to obtain widespread supports from the Taiwan voters, the
KMT started a small-scale welfare program in the 1980s.

Before the 1980s, small-scale welfare programs did not arouse much
discontent because class consciousness was weak. Taiwan’s economy heavily
relied on exports and small-scale enterprises, which created many outsourcing
industries. When the big outsourcing factories could not accommodate all the
contracts, they would farm out their contracts to small subcontractors. In 1986,
63.3 percent of workers in manufacturing industries worked in a factory with
fewer than 10 employees (Hsieh 1989, 17).6 Although these Page 107 → small-

scale enterprises supported the economic development of Taiwan, they

hindered the formation of working-class consciousness and stimulated mobility
between classes (Chang 1987; Gates 1979; Hsieh 1989). The structure of these
small enterprises created more possibilities for Taiwan workers to start their
own businesses,7 and from 1979 to 1987, 35 percent of Taiwan citizens chose to
do so rather than be hired by others (Hsieh 1989, 12). Even among employees,
30 percent of them wanted to open their own business in the future (Stites
1985, 238). Moreover, the heads of the small enterprises, as well as society in
general, also manipulated this desire for business ownership as a way to deal
with unhappy workers (Hsieh 1989).8 Because the working class in Taiwan
prefers starting their own businesses to confronting their bosses, it has been
rather difficult to form strong class consciousness among workers.
 The issue of wealth distribution stirred up more controversy after the 1980s
because the disparity between the rich and the poor grew as a result of the
rapid economic development and social transformation. According to a survey
on family income, the average income of the top 20 percent of richest families
was as much as 4.1 times that of the bottom 20 percent of families in 1980. The
ratio of the average income between the top 20 percent and the bottom 20
percent was 5.2 in 1992 (Directorate-General of Budget 2012). From the
beginning of the 1990s, not only did the DPP promote wealth distribution
legislation, it also actively searched for a new battleground in wealth
distribution issues to attract voters. In the 1992 legislative election, one
candidate campaigned on the promise of a pension for senior citizens and
achieved a significant victory. Later in the 1993 elections for county magistrates
and city mayors, many DPP candidates advocated pensions for the elderly
(Wang 2003, 81–82). When the DPP candidates were elected, they kept their
promise and granted NT3000 dollars to senior citizens. Even though these
payments did not last long in many counties because of fiscal difficulties, the
DPP acquired a reputation for caring about the social welfare of the average
Taiwan citizen. In the DPP’s 1999 survey on the question of a party’s capability
to design a fair and reasonable social welfare system, 32.9 percent of
respondents considered the DPP qualified while 20.7 percent chose the KMT
(Fell 2005, 39). It may be plausible that it was at that moment the DPP
achieved ownership of the social welfare issue, especially that of care of the
elderly.

However, the DPP’s ownership of that issue did not last long, since the KMT
was also competing for it. Worried that the DPP might take all of the credit for
advocacy of legislation on social welfare, the KMT not only started to propose
a national health insurance program but also struggled Page 108 → to propose its
own version of policies on subsidies for seniors.9 Eventually, the KMT
cooperated with the DPP to pass legislation providing benefits to elderly
farmers. Also, in 1995, the KMT government implemented the National Health
Insurance program. Although the DPP initially gained the support of the
electorate in part from promoting social welfare, both the DPP and KMT
found their niches in the wealth distribution issue in the mid-1990s.

In other words, because of fierce electoral competition, a considerable

number of social welfare issues became prominent in the Taiwan political
arena. When a party or candidate advocates a policy that might benefit a
specific group, political opponents may do likewise, or even advocate a more
radical extension of the same policy aiming at the same target. Since most
Taiwan parties and candidates attempted to take credit for welfare plans and
avoid the blame for blocking such plans, the differences between the parties
have gradually decreased (Chu 1994; Fell 2005; Sheng 2002; Sheng and Chen
2003). Furthermore, once the independence/unification issue appeared in the
Taiwan political arena, it diverted much of the attention of the parties and the
electorate from other issues. As Norris (2004, 119) claimed, “In Taiwan the
parties were identified mainly by nationalist issues, about relationships with
mainland China, rather than by left-right ideology.”10 Hence, the disparity of
wealth distribution failed to form a strong political cleavage on the island.

Indeed, from 2000 to 2008, while the DPP was in power, the party became
more concerned about the government’s fiscal capacity and economic
development and was not as active in promoting social welfare as it had been
prior to that time. In September 2000, the newly elected DPP president, Chen
Shui-bian, announced at a press conference, “Social welfare can be put off, but
economic development can’t be” (United Daily News, September 17, 2000, 1).
Therefore, the differences on the social welfare issue between the parties
became smaller. As figure 5.2 shows, in 2000 Taiwan voters on average rated
the DPP’s position on the wealth-distribution issue dimension at 6.5, while the
KMT on average was at 4.8 on the same issue dimension (questionnaire shown
in a2 of appendix 5.A1). However, in 2004, after the DPP had been in power
for four years, the rating was 6.0, compared to 5.6 for the KMT. In 2012, the
DPP was at 5.4, while the KMT was at 5.1 along the spectrum of wealth
distribution; thus, the issue positions of the two major parties in promoting
wealth distribution had converged.

Figure 5.2 also demonstrates that the DPP became more centrist on wealth
distribution in the perception of Taiwan public. Thus, it is not surprising that
many Taiwan voters felt that the DPP had become closer to Page 109 → business

interests and played money politics soon after it was elected (Liu 2003).
According to a survey conducted in 2002, two years after Chen Shui-bian
became president, 42 percent of the respondents perceived that the DPP was
too close to big business and to consortiums and spoke for the wealthy (United
Daily News, July 29, 2002, 3).
 Fig. 5.2. Respondents’ issue positions on social welfare/tax raises and their perceptions about
the parties’ positions (2000–2012). Source: Appendix 5.A2. Note: On the horizontal axis, “P”
indicates that surveys were conducted after the presidential election; “L” indicates that surveys
were taken after the legislative election. In 2012 the presidential election and the legislative
election were held at the same time.

The small differences between the issue positions of the two major parties has
constrained the DPP’s advantage on the wealth distribution issue. A political
cleavage due to opposition positions in the wealth distribution issue has been
difficult to develop. Take regulations on subsidies for senior citizens as an
example. In 2002 the KMT, together with the People First Party, first proposed
and passed a regulation to provide subsidies to the elderly except for those
with retirement pensions. Instead of discussion on whether the subsidies were
affordable for the government and were fair to other minority groups,
legislators from different parties competed to propose their own versions of
subsidies. Some proposed to broaden the qualifications, while others Page 110

→ proposed to raise the amount of the allowance. Even though a less

disputable revision of the legislation was made on June 2003, the parties
continued to propose revisions to the qualifications and the amount of the
allowance. In total, there were 30 legislative proposals from different parties
and legislators in the Fifth Legislative Yuan (Sheng 2005a). From the
perception of the public, all of the parties seemed to converge at a neutral place
on the wealth distribution issue. Thus, it becomes more difficult for the
electorate to differentiate parties on the basis of the wealth distribution issue.

More important, in order to maintain its overwhelming dominance in Taiwan

politics, the KMT has adopted a catch-all strategy, not a one-sided one. It will
not give up the votes of laborers or farmers even though it may stand closer to
public servants and capitalists due to its historical background. It may give
wealth distribution issues less priority but will not oppose improving wealth
distribution. On the contrary, the DPP, which was established through the
strong support of disadvantaged and dissatisfied groups, focuses on social
welfare programs to gain votes; however, the KMT’s strategy makes it difficult
for the DPP to create confrontational situations effectively.

Overall, although the wealth distribution issue has been a point of contention
in elections for a long period of time, it did not emerge as a significant political
cleavage in the society. The disparity between the rich and the poor has
widened in recent years, however, due to the economic downturn and money
politics in Taiwan (Sheng 2013). According to a survey on family income, the
average income of the top 20 percent of richest families was as much as 6.2
times that of the bottom 20 percent of families in 2010 (Directorate-General of
Budget 2012). If we limit the observations to the top 5 percent of richest
families and the bottom 5 percent of families, the ratio in income between
them is even greater. It was 32.7 in 1998, 55.1 in 2005, and 93.9 in 2012
(Ministry of Finance 2013). Because of the worsening imbalance in wealth
distribution, social welfare issues have become more salient (as evidence will
show in the next section), and the parties have used this issue as a means to
gain the support of disadvantaged voters. Indeed, in the 2012 election, voters
concerned more about social welfare were more likely to vote for the DPP
(Sheng 2013). After the election, persisting disputes on wealth distribution
issues, such as minimum wage and maximum hours, labor pensions, and
pensions of retired public servants, have made the rising gap in wealth a
prominent issue in the Taiwan society. From the experience of advanced
democracies, wealth distribution becomes a salient issue when there are a large-
scale changes or economic depressions, or both Page 111 → (Dalton 1996; Lipset
and Rokkan 1967). Will the rising gap in wealth in the Taiwan society polarize
elites and voters and become an important political cleavage in the future? In
our view, the issue of wealth distribution may not be able to single-handedly
form a significant political cleavage due to the similar stands of political parties
in Taiwan. However, if it aligns with the identity issue, the scenario may be
different, a point that we will discuss in the conclusion.

The Environmental Protection/Economy Issue

Since the beginning of the 1980s, the Taiwan public has started to notice the
environmental deterioration resulting from rapid economic development.
However, the ruling KMT emphasized development, in which economic
growth was the top priority, at the expense of environmental protection (Tang
and Tang 1997). Protesting the KMT’s promotion and endorsement of heavily
polluting industries, several environmental protection groups were organized
to confront the KMT government. The DPP played an important role in the
protests. In 1986, Lukang residents were mobilized to oppose a DuPont
investment that had been endorsed by the KMT government. This protest
made DuPont withdraw its project of establishing chemical factories in
Lukang. This incident prompted many antipollution protests in the following
years. In 1991, when the KMT government decided to build a fourth nuclear
power plant in Kongliao, many Kongliao villagers were mobilized to oppose
this policy by demonstrations and sit-ins at the proposed nuclear power plant’s
location. A violent clash with the police occurred, and a policeman’s death
brought the incident to nationwide attention. From then on, demonstrations
accompanied the building project of the fourth nuclear power plant. The
growing polarization of the proenvironment and pronuclear groups brought
the former into closer alignment with the DPP (Ho 2005a, 405–7), which
energized its base of support by mobilizing people who cared about
environmental protection.

As shown in figure 5.3, when respondents were asked to locate the position of
the parties on the environmental protection/economy issue dimension (with
higher scores indicating economic development and lower scores as
environmental protection), the DPP was rated at 5.0 in 1998 while the KMT
had a score of 7.3 (questionnaire shown in a3 of the appendix 5.A1). However,
after the DPP became the ruling party, its desire to stimulate economic
recovery led it, like the KMT, to favor more developmentalist policies. Page 112
→ When it faced a trade-off between the environment and the economy, most
of the time the DPP favored the latter over the former, frustrating many
Taiwan voters who cared about environmental issues (Ho 2005b).

Take the construction of the fourth nuclear power plant as an example. After
President Chen Shui-bian came in power in 2000, he issued an executive order
to halt the construction of the plant in October. This decision brought about
severe political battles and resulted in a serious political crisis between Chen,
the Executive Yuan, and the Legislative Yuan.11 Later, the Council of Grand
Justices issued the Interpretation No. 520, which stated that the Executive
Yuan’s actions had a “procedural flaw” for not reporting to the Legislative
Yuan before making the decision to halt the construction. The value of stock
market went down at least 2.5 percent as a result. In the end, the Executive
Yuan and the Legislative Yuan reached a compromise to restart the
construction with a consensus that a “nuclear free homeland” was the objective
in the long run. The political compromise appeased the opposition parties, but
frustrated people who stood for environment protection and expected the DPP
would have made a difference. Although a few DPP leaders continued their
fight against the use of nuclear energy and proposed to hold a nationwide
referendum on the fate of the fourth nuclear plant, Chen did not endorse their
proposal and chose promoting economic development rather than
environmental protection as his top priority (Fell 2012, 187–88). The data in
figure 5.3 demonstrate that the public has gradually changed its view of the
DPP on environmental issues.
 In contrast, the KMT was rather stable at around 7.2 on the issue spectrum of
environmental protection/economy until 2008. In the 2008 presidential
campaign, Ma Ying-jeou promised to initiate land restoration and reductions in
carbon emissions. His platform also included the imposition of an energy tax
and establishment of green traffic networks and buildings. As for controversial
public works, such as the highway between Su-ao and Hua-lien, he promised to
respect the results of environmental reports. When Ma Ying-jeou repeated his
promises for legislation and policy at the National NGO Environmental
Forum (Green Party Taiwan 2009), environmental groups had great hope that
he would fulfill them.12 The perception of the electorate with regard to the
KMT’s position on environmental issues was at 6.8 in 2008, which reflected this
expectation to some degree. Meanwhile, Hsieh Chang-ting, the DPP
presidential candidate in 2008, held to the DPP’s traditional policy, which
included ceasing the construction of the highway, and the party was perceived
at 5.4 on the issue spectrum of environmental protection/economy.

Although antinuclear protests continued to occur in Taiwan after the Page 113

→ fiasco of halting the construction of the fourth nuclear power plant, this
issue failed to catch the public’s attention as it did in 2000/2001. As both the
KMT and the DPP are generally in favor of developmental policies, it was not
until the Fukushima nuclear incident of 2011 that the safety of nuclear energy
became politically significant again. The Fukushima incident, which occurred in
Japan on March 11, 2011,13 turned the public’s attention to environmental
protection. When the KMT government requested a budget increase in 2012 to
fund the ongoing project of building the fourth nuclear plant, the public loudly
and clearly voiced their serious concerns about the safety of nuclear energy.
Several environmental groups instituted protests again. A TEDS survey
conducted in 2013 showed that 60.1 percent of respondents believed the
government should halt constructing the plant, while only 27.3 percent
supported the project (Sheng 2014). The DPP seized the opportunity to put
forward the proposition of a nuclear-free homeland. The party’s elites also
joined in the 309 NO NUKE Parade with environmental groups. To defuse the
crisis, the KMT government announced in 2013 that a referendum on the fate
of the power plant would be held later. Under the growing pressure, the KMT
government finally announced in April 2014 Page 114 → that the construction of
the fourth nuclear power plant would be mothballed for three years.14 It is
worth noting that this move did not represent a KMT position shift because it
keeps the option of restarting the construction in the future.
 Fig. 5.3. Respondents’ issue positions on environmental protection/economy and their

perceptions about the parties’ positions (1998–2008). Source: Appendix 5.A2. Note: On the
horizontal axis, “P” indicates that surveys were conducted after the presidential election; “L”
indicates that surveys were taken after the legislative election.

If the KMT continues to emphasize the economy while the DPP stands for
environmental protection, this issue may provide a political environment in
which to start discussions and debates. However, since the Taiwan public is
much more concerned about economic prosperity than environmental
protection (as evidence will show in the next section) and the ruling party cares
more about economic growth, which is the case for both the KMT and the
DPP, the environmental protection issue has difficulty in evolving to become a
significant political cleavage (Sheng and Chen 2003). Further, as Inglehart
(1990) noted, the new political cleavages of postmaterialism (such as
environmental protection) do not necessarily attract votes because the
supporters of those new issues are more active on issue agendas rather than
being passively dominated by partisan elites. In other words, it is more difficult
for partisan elites to manipulate the environmental protection issue. Thus, even
when the issue of environmental protection becomes more salient in the
Taiwan political area, its influence on party competition and elections may still
be less than that of existing political cleavages.

The Independence/Unification Issue

After the achievement of political reforms and the reorganization of unfair

political structures, the DPP, with 30.0 percent support of the electorate at the
beginning of the 1990s, faced the challenge of finding a new issue to continue
its political life. Which issue was the DPP able to maneuver most effectively to
attract Taiwan voters in the next stage? Two major issues gradually drew
attention from the DPP elites at the beginning of the 1990s. One was the
pursuit of Taiwan independence, and the other was the pursuit of a welfare
state (as discussed above). When the issue of Taiwan independence was raised
in the political arena, it was connected to Taiwan’s most important social
cleavages, those of ethnicity and Chinese/Taiwanese identity, so that it was
easily perceived by voters and aroused the emotions of substantial portions of
the Taiwan public.

Even though the DPP is pro-independence, its position on the

independence/unification spectrum has shifted at different times to attain its
political goals. In 1991, when it was eager to claim a position for Taiwan
independence, the DPP passed the Taiwan Independence Clause just two Page

115 → months before the National Assembly election. The DPP headquarters
issued several full-page and article-style ads on why it advocated a sovereign
and independent Republic of Taiwan (Fell 2005, 99). However, it received only
23.6 percent of the vote in the 1991 election, worse than its usual outcome. In
the following legislative election in 1992, taking into account that Taiwan
voters were not so comfortable with a radical stand on Taiwan independence,
the DPP packaged Taiwan independence into a more diluted form and
deemphasized the issue. The term “Republic of Taiwan” vanished from its ads
and the more moderate “diluted Taiwan independence” replaced “pure Taiwan
independence” (Fell 2005, 100).

In 1996, an independence-oriented DPP presidential candidate, Peng Ming-

min, made “Want Independence, Oppose Unification, Love Peace” his
campaign slogan. Again, this pledge moved the DPP to a more radical
independence-seeking position. The public perceived the DPP as an extreme
party in this regard and placed it at 2.0 on the independence/unification
spectrum in 1996, whereas the public perceived the KMT at 6.1, and the
average position of the public’s own view was at 5.2 (figure 5.4, questionnaire
shown in a4 of appendix 5.A1). The DPP garnered only 23.1 percent of all
votes in 1996. This serious defeat continued in the 1998 legislative election.
While the public stood at 5.0 on the independence/unification spectrum, the
DPP was perceived at 2.3, far from the public’s average position. The DPP
received 29.6 percent of the vote, less than what an ambitious and energetic
party would expect.

The continuous electoral defeats prompted the DPP to reconsider its position
on the independence/unification issue. Before the 2000 presidential election,
the DPP tried to take a centrist stance on the independence issue and to
convince Taiwan voters that it had the ability to handle cross-Strait relations.
First, the DPP passed the Resolution on Taiwan’s Future, which returned to its
original principle of Taiwan self-determination. The second step was Chen
Shui-bian’s speech “The New Central Way,” which referred to a vague middle
way on the independence/unification issue (Fell 2005, 106). In the 2000
presidential election, Chen Shui-bian was elected with 39.3 percent of the vote.
Another key factor that contributed to this victory was that two candidates—
Lien Chan and James Soong—ran on the pro-unification side and split the
vote.

Chen Shui-bian knew well that he would not be so lucky in the next election.
To compete in a single plural electoral system with two parties, the median
position might be a better position strategically. From then until the 2001
legislative election, Chen hewed to a centrist position on the
independence/unification issue. The first move in this direction was his Page 116

→ inaugural speech, in which Chen declared the Four Noes plus One Without
pledge,15 emphasizing that the new government would take a centrist position
on the issue. Taiwan voters did receive this signal and thus revised their image
of the DPP. As shown in figure 5.4, the Taiwan public on average located the
DPP at 3.1 in this issue dimension. It was the first time that the Taiwan public
placed the DPP at this moderate position on the independence/unification
issue.
Fig. 5.4. Respondents’ issue positions on independence/unification and their perceptions about
the positions of the parties (1996–2012). Source: Appendix 5.A2. Note: On the horizontal axis, “P”
indicates that surveys were conducted after the presidential election; “L” indicates that surveys
were taken after the legislative election.

After Chen was in power, he still faced an opposition-controlled legislature.

Although some legislators left the KMT and followed James Soong to a new
party, the PFP, together the KMT and PFP still held more than half of the seats
in the legislature. The situation for Chen’s government was worse since the
KMT legislators were much more cohesive than usual because of the serious
partisan conflict during the elections and a divided government created after
the elections (Hawang 2003; Sheng 2003, 2008). In the meantime, the PFP
legislators cooperated with the KMT legislators on most substantial legislation
because they shared the same supporting groups (Yu 2005). On the other hand,
the DPP faced a challenge from a newly founded party, the Taiwan Solidarity
Union, a party more extreme than the DPP on Page 117 → the

independence/unification issue dimension. As a result, Chen’s government

faced a political dilemma.

In hoping to win a majority in the legislature to implement Chen’s campaign

promises, the DPP adopted two campaign strategies in the 2001 legislative
election. The first was to run a national campaign. Normally, candidates
running under a multimember district with a single nontransferable vote
electoral system try to distinguish themselves from other candidates of the
same party because they target the same bloc of party supporters. Thus, they
have incentives to run independent and local campaigns; that is, to focus on
serving constituents and bringing pork-barrel projects to their constituency,
rather than concentrating on national issues (Sheng 2005b). However, in 2001
the DPP ran its campaign by promoting national campaign ads, and it
prohibited its candidates from running independent campaigns. Expecting a
coattail effect from the national star, Chen Shui-bian, the DDP candidates
followed this policy.

The DPP’s second electoral strategy was to declare publicly that it intended to
be a government for all people; that is, the government would work for all
Taiwan citizens, not only for DPP voters. Therefore, it took a centrist position
on the independence/unification issue and did not emphasize either the
independence/unification issue or the Taiwanese/Chinese identity issue (Fell
2005, 140–41). Instead, the party emphasized other issues, such as social welfare
and the political corruption of the era of the old KMT government. In the
2001 legislative election, the only party emphasizing independence was the
TSU. The TSU won 7.8 percent of the vote and 13 seats. The only party
emphasizing unification was NP, which nearly disappeared, getting only 2.6
percent of the vote and one seat in 2001.

By pursuing these strategies, the DPP won 33.4 percent of the vote and 38.7
percent of all seats in the legislature. Although the DPP increased its number
of seats and became the largest party in the legislature, the 38.7 percent of seats
was still not enough to control the legislature. Even if it could get support
from the TSU on most important legislative roll-call votes, the total fraction of
seats in the Pan-Green bloc (44.5% of the total) still would be outvoted by the
opposition coalition, the Pan-Blue bloc, formed by the KMT, PFP, and NP.
After the DPP failed to win support from a majority of the voters, it faced a
severe fight with the Pan-Blue bloc in the legislature. Consequently, the DPP
government was able to accomplish little and could not implement its
campaign promises.

Soon after the 2001 legislative election, the DPP found that the TSU had
attracted voters who favored an extreme position on Taiwan independence.
Several substantial moves showed the TSU’s aggressiveness and steadfast
position. Page 118 → For example, the TSU initiated a proposal to revise the
President and Vice-President Election and Recall Law to limit the qualification
of a presidential candidate: only those born in Taiwan would have the right to
be a presidential candidate. Another obvious example was the firm position
TSU legislators took, while the Plebiscite Law was under debate, in favor of
having substantive issues such as Taiwan independence decided by plebiscite.
Also, when creating Regulations for Managing the Relations between Citizens
of the PRC and Taiwan, the TSU always stood for very strict restrictions on
people from China. Compared to the TSU’s firm stand on the independence
issue, the DPP appeared ambiguous and hesitant on the issue.

Fearing that they would lose the electoral support of pro-independence voters
to the TSU, the DPP started to shift to a more pro-independence stance after
the 2001 legislative election. Constrained by its ruling position, the DPP did
not declare for Taiwan de jure independence. Rather, it played a safe game;
namely, to consolidate its original voter base by resorting to Taiwan self-
determination and Taiwan nationalism. The 228 Hand-in-Hand Rally before
the 2004 presidential election was designed to present the DPP as the party that
loved Taiwan. Further, the DPP advocated Rectify the Name of Nation and
Drafting the New Constitution in the later legislative elections. All of these
actions contributed to the party’s position shift to an extreme pro-
independence stance in 2004. The Taiwan public discerned this shift and placed
the DPP at 2.2, almost the same as its position in 1996 (figure 5.4).

The DPP continued its pro-independence strategy even as President Chen’s

second presidential term was almost over. In the cover letter of application for
membership in the United Nations, Chen requested the admission of Taiwan
(Office of the President, July 20, 2007), which was different from the earlier
request, “Readmission of the Republic of China.” The DPP further proposed a
referendum that requested the government to continue to apply for
membership in the United Nations under the name of Taiwan. The DPP
presidential candidate for the 2008 election, Hsieh Chang-ting, supported this
referendum despite the opposition of the United States and China. The United
States and China worried that the next step of the DPP would be an
independence referendum. When Hsieh visited the United States during the
election, he stated that an independence referendum would not be necessary
since Taiwan was already substantially independent. He claimed that the “cross-
Strait common market” policy of his KMT opponent, Ma Ying-jeou, was a
preliminary step to unification with China. As a result, the Taiwan public rated
the DPP’s position at 2.2 and the KMT’s at 7.4 in 2008 (figure 5.4).

Page 119 → The DPP lost the 2008 election. This defeat led the party to
reconsider its claims on the issue of independence/unification. Hsieh had
linked an open economic policy with Taiwan nationalism but failed to provide
a better substitute policy for Taiwan’s economic recession. This strengthened
the public’s impression about the DPP’s lack of ability to deal with both
economic and cross-Strait issues. Over half of the Taiwan public (54.3%)
thought that the KMT performed better on cross-Strait issues in 2008 (Sheng
2013). Some moderate DPP elites, such as Tuan Yi-kang, claimed that the mass
public was tired of the DPP’s dogmatic views on the issue of
independence/unification. Furthermore, after 2001 the public in general placed
themselves at 4.5–4.6 along the spectrum of independence/unification.
According to Downsian median voter theorem, it is expected that parties will
seek a more centrist position when most of the public stands at the middle.
During the 2012 election, the DPP presidential candidate, Tsai Ing-wen,
adopted a moderate strategy on the issue of independence/unification. She
claimed that she would unconditionally carry on the existing cross-Strait
policies if she won the election. She also admitted the importance of an open
economic policy and trade with China and committed herself to dealing
pragmatically with the relationship with China. However, she denied the “1992
consensus”16 and made a vague assertion of “Taiwan consensus.” The mass
public still perceived the DPP at 2.6 on the spectrum of
independence/unification, even though this was the second closest placement
to the center for the DPP since 1996.

In contrast, the Taiwan public perceives the KMT as a pro-unification party.

This perception is based on the KMT’s long-standing political declaration in
support of eventual unification with China. In 1990, the KMT government set
up the National Unification Council. In 1991, the council drafted “Guidelines
for National Unification,” which called for a phased approach toward
unification. However, Lee Teng-hui, the KMT’s president who held office from
1988 to 2000, maintained a vague position on the independence/unification
issue. Early in 1991 and 1992, when the DPP passed the “Taiwan Independence
Clause” and advocated a radical Republic of Taiwan pledge in elections, Lee
was lenient toward the DPP. His blurred and ambivalent attitude on unification
with China made the New KMT Alliance, a hard-line faction on the unification
position, accuse Lee of being a supporter of gradual independence. Lee’s
attitude toward independence/unification contributed to the departure of the
New KMT Alliance from the KMT and its formation of the NP in 1993.

During the campaign for the 1996 presidential election, when facing Lin Yan-
kang and Chen Lu-an, two candidates who were strongly pro-unification, Page
120 → Lee Teng-hui was still vague about his position on
independence/unification. A survey conducted by the Election Study Center in
1996 asked respondents about Lee’s position on that issue. The results showed
that 23 percent of the respondents considered Lee as favoring unification, 22.1
percent considered him as favoring independence, 23.0 percent thought he
favored the status quo, and 31.9 percent were not aware of Lee’s position at all
(Sheng 2002). This explained why Taiwan voters located the KMT at 6.1 on the
independence/unification dimension in 1996.

From 1996 to 2000, when Lee was still the KMT’s leader, Taiwan voters
considered the KMT as a party inclined to the right of center on the
independence/unification issue. This perception changed when Lee left the
KMT and became the spiritual leader of the TSU. The KMT reserved the
option of unification with China and strongly stood for an open economy with
China. After 2000, Taiwan citizens located the KMT at around 7.0 on the
unification side of the dimension. In 2008 the KMT presidential candidate, Ma
Ying-jeou, announced the No Unification, No Independence, and No Use of
Force pledge on the independence/unification issue. Also, in his inaugural
speech, he promised to foster more direct exchanges in cross-Strait relations
and draft a peace pact with China. After Ma took office, he encouraged
friendlier and more peaceful relations with China, such as frequent Chiang-
Chen talks,17 opening direct investment in China, and allowing Chinese tourists
and students to visit Taiwan. He also advocated signing the Cross-Strait
Financial Supervision and Cooperation Memorandum and the Cross-Strait
Economic Cooperation Framework Agreement to strengthen economic
relations across the straits. Although the No Unification pledge seemed to
make him a little more centrist, his economic and cultural policies, which
encouraged closer relations with China, contributed to his pro-unification
position in the public’s perception. The public rated the KMT’s position at 7.4
along the independence/unification spectrum.
 Furthermore, the concurrence between the issue position of the KMT’s and
DPP’s identifiers and the issue position that those identifiers perceive their
party to have is noteworthy, as figure 5.5 demonstrates. The issue position of
KMT identifiers and their perception of the KMT’s issue position correspond
to each other. This situation is more apparent when we compare the issue
position of DPP identifiers with their perception of the DPP’s issue position.
The concurrence may not be coincidental. Brody and Page (1972) argue that a
voter may be persuaded by a candidate’s position and thereby consider that
position to be his own. The Page 121 → public’s perception on issues will be
“distorted” by “pre-existing partisan attitudes.” In order to maintain cognitive
consonance, party identifiers will selectively absorb information and form their
issue position close to their preferable partisan position (RePass 1971).
Research on American voters has shown this possibility. For example, Page and
Jones (1979), Markus and Converse (1979), and Markus (1982) specify their
models with reciprocal causal relationships among party, issue, and candidate
factors by a simultaneous-equation statistical method. The concurrence in
figure 5.5 implies that partisan elites have influenced Taiwan citizens in shaping
their issue position.

Specifically speaking, elites try very hard to attract Taiwan citizens to their
side. They frame and develop the issues. They persuade and convince the
Taiwan public that certain issues are important and that their position benefits
the public the most. The ups and downs of the issue positions of party
identifiers show that they recognize the shift of their party’s issue position and
that they will shift their issue position accordingly. The DPP identifiers are
more likely to be persuaded by the DPP elites, while the Pan-Blue identifiers
are less likely to be persuaded by the DPP elites. And the KMT identifiers are
more likely to be persuaded by the KMT elites, while the Pan-Green identifiers
are less likely to be persuaded by the KMT elites. In other words, when voters
pick up the signal of political elites’ rhetoric or action on the issues, or both,
they are selectively persuaded. In such a case, we should be able to find that an
individual’s party preference affects his issue position so that a concurrence
between his position and his perception of his preferable party’s issue position
appears, as in figure 5.5.18

For example, the DPP identifiers perceived the DPP’s issue position as more
centrist (3.2) on the issue spectrum in 2000, when Chen Shui-bian offered the
New Central Way policy. At the same time, the DPP identifiers also moved to a
more centrist position (4.3) in 2000. Another example occurred in 2008. When
the DPP strongly catalyzed the birth of the Admission of Taiwan to the United
Nations referendum and Hsieh Chang-ting claimed that he would not propose
an independence referendum because Taiwan was substantially independent,
the DPP identifiers placed the DPP’s issue position at 2.0, the most pro-
independence stance that they have ever considered for the DPP. Meanwhile,
they rated themselves at 2.9, also the most pro-independence placement that
they have ever had.

Another interesting phenomenon in figure 5.5 is the trend in the issue

position of KMT and DPP identifiers. Compared to their perception of the
parties, Taiwan voters tend to be more centrist than their preferred parties, Page

122 → even though their positions generally fluctuate with those of their parties.
This may imply that Taiwan citizens, although standing close to the position of
their preferred party, do not consider themselves as extreme as their parties on
this issue. In other words, the issue position of party identifiers seems not to
deviate from that of independent voters too much on average. Furthermore,
not only their perceptions on parties’ issue positions but also their issue
positions were prone to be centrist in 2012. Indeed, DPP identifiers perceived
themselves at 3.3 on the issue spectrum in 2012 (2.9 in 2008), while KMT
identifiers placed themselves at 5.4 in 2012 (5.8 in 2008). Fell argues that the
fluctuations of party position were determined by intraparty struggle in the two
main parties after 2008 and suggests that the new generation of leaders have
sought “more consensual politics” (2011, 93). Convergence to a more centrist
stance on the issue of independence/unification for both the DPP identifiers
and KMT identifiers may be possible in the future unless political parties on
the island become extreme.
 Fig. 5.5. Party identifiers’ issue positions on independence/unification and their perceptions
about the positions of the parties (1996–2012). Source: Appendix 5.A3. Notes: 1. IDKMT = KMT
identifiers. IDDPP = DPP identifiers. Here the placements of the KMT and DPP are the
average perception of KMT identifiers and DPP identifiers, respectively; that is, KMT=KMT as
perceived by KMT identifiers; DPP=DPP as perceived by DPP identifiers. 2. On the horizontal
axis, “P” indicates that surveys were conducted after the presidential election; “L” indicates that
surveys were taken after the legislative election.

Page 123 →

The Most Important Problem Facing Taiwan

In this section, we discuss how the public perceives the most important
problem the country is facing and whether the public’s perspective responds to
the elites’ issue agenda. Survey data were collected after presidential and
legislative elections from 1996 to 2012 (questionnaire wordings are in a5 of the
appendix 5.A1). Since this is an open-ended question, there are divergent
responses. However, we managed to subdivide respondents’ answers into eight
categories: economic prosperity, independence/unification/cross-Strait affairs,
wealth distribution/social welfare, party/politician’s ability and corruption,
social order and national security, political/social reform/stability,
environmental protection, and others. Table 5.1 shows the results.

Based on data from the table, we find that the Taiwan public was extremely
concerned about economic prosperity, especially in 2001 and 2008, when
Taiwan’s economy was in a downturn. Also, the public was concerned about
social order and national security, especially in 1996 and 2004, when tensions
with China occurred in the Taiwan Strait. Sometimes, the public takes the party
or politician’s characteristics (e.g., ability, integrity, and corruption) more
seriously. However, economic prosperity, social order, and national security, as
well as the party’s and politician’s ability or corruption, Page 124 → are valence
issues, in which there is near-universal agreement on the ends of policy (Fiorina
1981, 18). Since most partisan elites (and the public) take the same side on the
valence issues, those issues have little potential to develop into a long-term
political cleavages.19
Table 5.1. The Most Important Problem Facing Taiwan in Elections

Issue 1996P 2001L 2004P 2004L 2008P 2008L 2012PL

Independence/unification/cross-Strait affairs 32 7 26 25 27 9 22

Political or social reform/stability 0 0 3 3 2 3 2

Wealth distribution /social welfare 0 1 3 6 0 6 6

Environmental protection 0 0 1 2 0 0 0

Economic prosperity 23 55 35 21 15 63 38

Social order and national security 32 7 7 19 10 10 2

Party or politician ability/corruption 0 4 5 10 7 0 4

Others 3 13 3 4 10 1 5

Don’t know 9 12 16 9 28 8 20

Refuse to answer 1 1 1 1 1 0 1

Total 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%

Sources: Data for 1996 are from the Election Study Center at National Chengchi University, and data for
2001 and after are from the Taiwan’s Election and Democratization Study Project.
Note: After the year, “P” indicates that surveys were conducted after the presidential election; “L”
indicates that surveys were taken after the legislative election.
 Fig. 5.6. The proportion of voters defining the most important position issue in the presidential
elections. Source: Table 5.1. Note: Data is not available for 2000. On the horizontal axis, “P”
indicates that surveys were conducted after the presidential election; “L” indicates that surveys

were taken after the legislative election.

Figures 5.6 and 5.7 illustrate only the importance of position issues—
specifically, the four focal issues of this chapter—in the perception of the
public. Figure 5.6 presents the public’s perspective on those issues in the
presidential elections, and figure 5.7 shows the situation for the legislative
elections. Several points are noteworthy.

First, from the public’s perspective, the independence/unification issue has

been the most important position issue across years both in the presidential
and legislative elections. The proportion of respondents who consider this
issue most important is much greater than the proportion for other issues
(figures 5.6 and 5.7). As previously discussed, politicians have manipulated this
issue most of the time. When the Taiwan independence issue was raised in the
political arena, it was connected to Taiwan’s most important cleavages, that of
ethnicity and Chinese/Taiwanese identity, so that it was easily perceived by
voters and aroused the affections and disaffections of substantial portions of
the Taiwan public. Also, the independence/unification Page 125 → issue is
connected to relations with China, so that it is also related to national security
and economic prosperity. This makes the independence/unification issue even
more important.

Fig. 5.7. The proportion of voters defining the most important position issue in the legislative
elections. Source: Table 5.1. Note: On the horizontal axis, “P” indicates that surveys were
conducted after the presidential election; “L” indicates that surveys were taken after the
legislative election.
 Second, the independence/unification issue is of more importance in the
presidential elections than in the legislative elections. Presidential candidates’
election platforms target national voters, whereas legislative candidates target
their own constituents. In order to attract voters, it is possible that issues
involved in presidential elections tend to be national and important affairs,
while issues involved in legislative elections are diverse—from national to local
affairs. Even in the same year, the proportion of the respondents considering
the independence/unification issue the most important was only 9.0 percent in
the legislative election held in January 2008; however, it was 27.0 percent in the
presidential election held in March of the same year (table 5.1).

Moreover, the importance of the independence/unification issue fluctuates

more in the legislative elections than in the presidential elections. More than
25.0 percent of the respondents perceived the issue as the most important
position issue in almost all presidential elections. However, in the legislative
elections, only 7.0 percent and 9.0 percent of the respondents Page 126

→ perceived it as the most important position issue in 2001 and 2008 whereas
25.0 percent did in the 2004.20 In 2003, the TSU initiated the Campaign for
Rectifying the Name of Taiwan and invited the former president, Lee Teng-hui,
to be its spiritual leader. Later they sponsored the Call Taiwan Taiwan parade.
Afraid of losing too many pro-independence voters, the DPP and Chen Shui-
bian expressed their support for this campaign both explicitly and implicitly
(New Taiwan Weekly News, September 12, 2003).21 Chen Shui-bian also
requested national enterprises to change “China” to “Taiwan” in their names.
The independence/unification issue had occupied the spotlight in the 2004
legislative election. This further demonstrates that the public gets cues from
the parties and politicians in order to shape its perspective on politics. As in
the United States, Democrats mention prodemocratic issues more, whereas
Republicans uphold conservative values and issues (RePass 1971, 395). Elites’
emphasizing or deemphasizing of a certain issue may affect the public’s
perspective on the issue agenda (Chihibber and Torcal 1997, 31).

Last, issues other than independence/unification are not as important from

the public’s perspective, but the importance differs across issues. More people
consider the wealth distribution issue more salient in the legislative elections
than in the presidential elections. Also, as the economic situation has declined,
the economically disadvantaged strongly felt relatively exploited; thus, more
people paid attention to the wealth distribution issue. So from 2004, a number
of respondents (6.0%) considered wealth distribution the most important
problem facing Taiwan (figure 5.7). As for the political reform issue, few voters
took it as the most important issue in either the presidential elections or
legislative elections, although reform/stability had been the top issue in the
1980s and early 1990s (Hsieh and Niou 1996b). The DPP elites promoted many
reform programs, whereas the KMT stressed stability. Political stability
occupied the top 10 advertisements of the KMT during elections until the
party turnover (Fell 2005, 26). However, the importance of the reform/stability
issue faded after Taiwan’s transition to democracy.

The environmental protection issue has the least respondents who perceive it
as an important problem. Candidates’ campaign platforms emphasizing
economic growth usually get more attention than those emphasizing
environmental protection. Furthermore, the ruling party tends to stress
economic growth rather than environmental protection. An example is that
even though for a long time the DPP claimed to protect the environment, it
compromised in favor of economic development when it was in power (Ho
2005b). Thus, it is not surprising that relatively few politicians emphasize Page

127 → the environmental protection issue. As a result, few respondents consider

environmental protection to be an important problem compared to other
issues.

Overall, partisan elites’ manipulation of the issues, such as developing an

issue, shifting the issue position, and emphasizing one issue while
deemphasizing another, has an impact on how Taiwan citizens perceive the
issues.

Conclusions

This chapter has examined four issues: reform/stability, wealth distribution,

environmental protection/economy, and independence/unification. We have
shown how partisan elites frame and manipulate these issues to attain their
political goals and how Taiwan voters respond. Research findings show that for
all of the four issues, Taiwan voters are able to perceive the parties’ issue
positions, shape their own issue position, and make political judgments based
on the issues. This shows that issue politics has gradually evolved in Taiwan
politics, although the impacts across issues differ.

In the process of issue evolution in Taiwan, the Dang-wai and DPP elites
played an important role. They strategically searched for issues that benefited
the party’s growth and strength, then clarified and manipulated the issues to
mobilize the mass public. Having emerged in the authoritarian era, the DPP
emphasized the political reform issues first. After accomplishing political
reform, it moved to the issue of Taiwan independence. The DPP elites were
successful in raising this issue, so that the party grew gradually in the 1980s and
1990s and won the presidency in 2000 and 2004. However, it lost the 2008 and
2012 presidential elections, and it has not broken through the bottleneck of
seat shares (40.0%) in legislative elections. One of the major reasons is that the
DPP cannot achieve a breakthrough on issue games.

Following Taiwan’s gradual democratization and political reform in the 1980s

and 1990s, it has become more difficult for the DPP to identify new and
attractive reform issues. The DPP has tried raising several such issues after
2000, such as the return of the KMT assets obtained in the authoritarian era,
revision of the 18 percent preferential interest rates for the pensions of retired
public servants, the reform of the constitution, and the restructure of the
legislature. The DPP has attained its goal on some issues, but it cannot take all
of the credit because several parties were involved in the regulation-making
process. Some reform issues, such as preferential treatment for retired civil
servants, are still controversial and might only help to consolidate the DPP’s
original base of support but not broaden its voter base.

Page 128 → Regarding the environmental protection/economy issue, the debate

on whether to continue the construction of a fourth nuclear plant has stirred
up mass attention recently. The disaster at the Fukushima nuclear plant in
Japan has further stimulated the public’s concern about the safety of nuclear
energy. If the KMT government had not been able to defuse the issue, the DPP
could have taken advantage of it to attract the electorate. However, the KMT
government decided to mothball the construction of the fourth nuclear power
plant for fear of losing electoral support even though the option to use nuclear
energy was kept open. In addition, as a result of the economic downturn since
2000, the Taiwan public is more concerned about economic prosperity, and the
governing party has always considered economic growth as the top priority in
its agenda. Without polarizing elites and voters, the environmental
protection/economy issue has little chance of evolving into a salient political
cleavage.

Thus, the independence/unification issue remains the most influential issue in

Taiwan politics because all major parties have clear and polarized positions on
it and it is connected to enduring social cleavages—ethnicity and
Chinese/Taiwanese identity. As it was demonstrated in figure 5.4, the positions
of the parties, in particular those of the KMT and the DPP, are obviously
separated more than those in figures 5.1, 5.2, and 5.3. This situation provides
partisan elites with opportunities to manipulate the issue in order to attract
voters. The Taiwan public in turn responds to the elites, shapes its issue
position, and perceives the importance of the issues accordingly. Among
position issues, these types of interactions between partisan elites and the mass
public is especially effective with the independence/unification issue.
Moreover, this issue is relatively important also because Taiwan voters make
political decisions based on it (Cheng and Hsu 1996; Fell 2005, 2012; Hsieh and
Niou 1996a, 1996b; Shyu 1998; Sheng 2002, 2013; Sheng and Chen 2003). From
2000 to 2008, party positions on the independence/unification issue became
more polarized. The Pan-Green’s position on the issue dimension was moving
toward a more pro-independence stance. In the meantime, the Pan-Blue’s
position was becoming more pro-unification. In 2012, although both the KMT
and the DPP were perceived as a little prone to being centrist (the KMT
moved from 7.4 to 7.0 on the scale, while the DPP moved from 2.2 to 2.6), the
difference between the two parties was still very large. The continuing
polarization between the two parties has made the independence/unification
issue more salient and has had decisive impacts on the political attitudes and
behavior of Taiwan voters. It is quite possible that all parties will keep focusing
most of their efforts on framing and manipulating this issue.

Page 129 → Last, the wealth distribution issue in Taiwan has not become as
salient as it has been in other advanced democracies even though the disparity
between the rich and the poor has become more severe in recent years. Political
parties and candidates have frequently manipulated the issue of wealth
inequality to gain electoral support from disadvantaged voters. Since they
generally attempt to take credit for welfare programs, they tend to outbid each
other. The differences between the positions of political parties on wealth
distribution are thus small. Indeed, all major parties in Taiwan have little
differences on this issue and have difficulty in distinguishing themselves from
one another. As a result, the wealth distribution issue has not been polarized to
the extent of becoming a significant political cleavage.

Will this issue become an important political cleavage in the future? There are
two possible scenarios. The first one is that the welfare distribution issue is
absorbed into the dominant cleavage of independence/unification and
polarizes political elites and voters. There is evidence that elites of major
parties took a ride on the independence/unification issue to evoke the public’s
consciousness of wealth inequality in Taiwan. In 2008 and 2012 elections, the
KMT fielded the argument that Taiwan’s economic misfortune was related to
cross-Strait tension and campaigned on expanding economic relations with
China, such as signing the Economic Cooperation Framework Agreement.
Entrepreneurs of big businesses welcome this proposal and nearly unanimously
endorsed the KMT presidential candidate, Ma Ying-jeou. In contrast, the DPP
opposed the proposal and advertised ECFA’s negative effects with a dialect
limerick popular among disadvantaged groups: “Female cannot find good
husbands, male cannot find good jobs, and sons have to find a job in
Heilongjiang.”22 Evidence shows that working class and self-employed citizens
tend to oppose the ECFA, while business owners and people with managerial
responsibilities generally support the accord (Lin and Hu 2011). It is also worth
noting that Taiwan voters’ positions on cross-Strait economic exchange are
reflected in their partisan identification. Supporters of the Pan-Blue Alliance
generally consider the ECFA beneficial to Taiwan’s economy, while the Pan-
Green identifiers tend to view the accord negatively. If this situation continues,
the welfare distribution issue may be absorbed into the dominant cleavage of
independence/unification in such a way that economically disadvantaged
citizens are more pro-independence while the advantaged publics are more
pro-unification. In such a case, class politics may emerge and wealth
distribution may become more salient in Taiwan through the strength of
identity issue. Independence and unification cleavage along with class conflicts
may aggravate polarization in the Taiwan society.

Page 130 → However, another scenario is also possible that may not contradict
the first one. That is, the KMT will slow down its pace with China for fear of
losing electoral support due to its contentious nature, while the DPP will be
unwilling to sacrifice economic prosperity and will moderate its pro-
independence position. The DPP presidential candidate Tsai Ing-wen
announced in the 2012 election that she would accept all cross-Strait
agreements signed between the KMT government and Beijing if she were
elected. This shows that the DPP cannot stand against the wishes of big
businesses when it is in power. In that case, the elites of both parties may stand
close to the centrist position on the major issue dimension and bring a less
polarized society to Taiwan.

Appendix 5.A1

Questionnaire Wording

a1: On the reform/stability issue

Looking at Taiwan’s overall development, some people believe that large scale reform is the
most important thing, even if it means sacrificing some social stability. Other people believe
that stability is the most important and that reform should not be allowed to affect social
stability. On this card, the position that large-scale reform is the most important thing is at
0 on a scale from 0 to 10, and the position that social stability is most important is at 10.
About where on this scale does your own view lie? As you understand it, about where on
this scale does the position of the KMT lie? About where on this scale does the position of
the DPP lie? About where on this scale does the position of the PFP lie? About where on
this scale does the position of the TSU lie?

a2: On the wealth distribution issue

Regarding the question of social welfare, some people believe that the government should
merely maintain the current system in order not to increase people’s taxes. Other people
believe that the government should promote social welfare, even though it will lead to tax
increases. On this card, the position that maintaining the current system is the most
important thing is at 0 on a scale from 0 to 10, and the position that promoting social
welfare is most important is at 10. About where on this scale does your own view lie? As
you understand it, about where on this scale does the position of the KMT lie? About where
on this scale does the position of the DPP lie? About where on this scale does the position of
the PFP lie? About where on this scale does the position of the TSU lie?

Page 131 → a3: On the environmental protection/economy issue

Regarding the question of economic development versus environmental protection, some

people in society emphasize environmental protection while others emphasize economic
development. On this card, the position that emphasizes environmental protection is at 0 on
a scale from 0 to 10, and the position that emphasizes economic development is at 10.
About where on this scale does your own view lie? As you understand it, about where on
this scale does the position of the KMT lie? About where on this scale does the position of
the DPP lie? About where on this scale does the position of the PFP lie? About where on
this scale does the position of the TSU lie?

a4: On the independence/unification issue

In our society people often talk about the question of Taiwan independence from or
unification with China. Some people say that Taiwan should declare independence right
away. Other people say that Taiwan and China should unify right away. Yet other people
have opinions between these two positions. On this card, the position that Taiwan should
immediately declare independence is at 0 on a scale from 0 to 10, and the position that
Taiwan should immediately unify with the mainland is at 10. About where on this scale
does your own view lie? As you understand it, about where on this scale does the position of
the KMT lie? About where on this scale does the position of the DPP lie? About where on
this scale does the position of the PFP lie? About where on this scale does the position of the
TSU lie? The questionnaire wordings are identical in most years, except for 1996P and
2000P. However, the survey results of the two years with slightly different wordings did not
deviate from common expectation much; thus, the authors kept them in the discussion.

a5: The most important problem facing Taiwan

During the presidential (legislative) election campaign, many different problems faced by our
country were raised. What do you think is the most important political problem facing
Taiwan today? (Open-ended question)

Note: The wording of this question might be slightly different on cross surveys,
but the core element of the question is the same: that is, in the respondent’s
perspective, what is the most important political problem facing Taiwan today?
Appendix 5.A2. Respondents’ Issue Positions and Their Perceptions about the Positions of
Parties (1996–2012)

Issue Election Respondents KMT DPP NP PFP TSU

Reform/stability

1996P 7.1 6.9 3.9 5.6

1998L 7.2 6.6 4.0 5.2

2000P 7.2 7.0 4.4 5.6 5.9

2001L 7.2 6.8 3.6 5.7 6.1 4.8

2004L 7.3 6.8 3.9 6.1 3.7

2008P 7.0 6.1 4.7

Wealth distribution

2000P 5.6 4.8 6.5 5.2 5.4

2004L 6.3 5.6 6.0 5.5 5.2

2008P 5.9 5.6 5.4

2012PL 5.9 5.1 5.4

Environment/economy

1998L 5.3 7.3 5.0 4.9

2000P 5.2 7.2 5.2 5.0 5.4

2001L 6.1 7.2 5.6 5.6 6.3 5.8

2004L 5.7 7.1 5.6 6.4 5.4

2008P 5.9 6.8 5.4

Independence/unification

1996P 5.2 6.1 2.0 6.5

1998L 5.0 6.5 2.3 7.3

2000P 5.3 6.4 3.1 7.3 7.1

2001L 5.2 7.2 2.6 7.6 7.2 2.5

2004L 4.6 7.4 2.2 7.4 1.7

2008P 4.5 7.4 2.2

2012PL 4.5 7.0 2.6

Source: Data for 1996–2000 are from the Election Study Center at National Chengchi University, and
data for 2001–12 are from the Taiwan’s Election and Democratization Study Project.
Note: Entries are respondents’ means on a scale of 0 to 10. After the year, P” indicates that surveys
were conducted after the presidential election; “L” indicates that surveys were taken after the legislative
election.

Page 132 →

Appendix 5.A3. Party Identifiers’ Issue Positions and Their Perceptions about the Positions of
Parties on Independence/Unification (1996–2012)

KMT KMT (as perceived by KMT DPP (as perceived by DPP

Election DPPIdentifiers
Identifiers identifiers) identifiers)

1996P 5.7 6.3 3.7 2.3

1998L 6.0 6.7 3.8 2.5

2000P 5.9 6.5 4.3 3.2

2001L 5.9 6.9 4.2 2.8

2004L 6.0 7.0 3.5 2.2

2008P 5.8 6.7 2.9 2.0

2012PL 5.4 6.4 3.3 2.6

Source: Data for 1996–2000 are from the Election Study Center at National Chengchi University, and
data for 2001–12 are from the Taiwan’s Election and Democratization Study Project.
Note: After the year, “P” indicates that surveys were conducted after the presidential election; “L”
indicates that surveys were taken after the legislative election.

Page 133 →
Notes

1. The authors thank the Election Study Center at National Chengchi

University for providing the data from the 1996, 1998, 2000 elections, and
the Taiwan Election and Democratization Studies Project for providing
data from the 2001 to 2012 elections. The coordinator of multiyear
project TEDS is Professor Chi Huang (National Chengchi University).
The authors are alone responsible for views expressed herein.

2. The measurement of respondents’ positions on the reform/stability

issue and their perceived positions of the parties are based on the
questionnaire shown in a1 of the appendix 5.A1.

3. Shih Ming-te, a former colleague of Chen, initiated the “Million Voices

against Corruption, President Chen Must Go” campaign and appealed to
Chen to resign from office. Shih and his followers, all in red shirts, sat
outside the presidential office from September until the prosecutors
charged Chen and his family with corruption; however, the court did not
process the case against Chen in 2006. The reason is that the president has
immunity from criminal accusations when he is president, according to
the constitution.

4. Liou Hsiao-Hsia, “Ma versus Hsieh: Comparison of Their Pork

Barrels,” United Daily News, March 15, 2008,
http://mag.udn.com/mag/vote2007–08/storypage.jsp?
f_MAIN_ID=357&f_SUB_ID=2361&f_ART_ID=73230 (accessed May
8, 2013; in Chinese).

5. For more discussion on the ethnic cleavage, see chapter 3 of this book.
6. Even in 2011, in all industries, 78.6 percent of employees worked in
companies with fewer than 5 workers while 90.9 percent worked in
companies with fewer than 10 workers (Directorate-General of Budget
2012).

7. Since the subcontractors, who usually had worked for the big
outsourcing factories, still relied on contracts from the original factories,
the owners of small factories Page 134 → did not consider themselves to be
“capitalists” or even “bosses” (Hsieh 1989). Their relations with their
employees, who usually had been their former colleagues from the big
factories, were more like partnerships, not the confrontational
relationships between capitalists and labor.

8. A legislator from a business district once appealed to his electorate in

the working class that he had “many well-achieved friends in business.
They all used to work as apprentices or workers. They all claimed they
have been in difficulty with their bosses when they were hired. However,
they all opened their own business later with the assistance of their
original bosses. So, workers do not always work for others; someday you
will become employers” (Chang 1987, 21; in Chinese).

9. The KMT leaders did not plan full-scale subsidies to the elderly from
the beginning. However, via hearings, media attention, and direct petition
to President Lee Teng-hui and Premier Lien Chan some legislators from
the agricultural districts, such as Chen Chih-ping, Lin His-shan, and
Wong Chung-chun, appealed to the KMT leaders to support subsidies for
elderly farmers (Sheng 2001, 90–91).
10. One of the reasons for parties to hold onto the
independence/unification issue, rather than left-right ideology, may be the
failure of the Taiwan public to distinguish between the left and the right.
Chen (2003) found out that only half of Taiwan citizens can identify their
position along the left-right spectrum, a rather low rate compared to
citizens in most democracies. Hsiao, Cheng, and Achen also find that the
Taiwan public misunderstands the meanings of “left” and “right” (see
chapter 6).

11. The KMT, PFP, and NP united and proposed to impeach the newly
elected president. BBC News, “Taiwan’s Suspension of NUKE4 Causes
Political Crises,” October 30, 2000,
http://news.bbc.co.uk/hi/chinese/news/newsid_998000/9987432.stm
(accessed October 8, 2015).

12. Green Party Taiwan, “Ma Ying-jeou Do Not Forget Your Promise of
Carbon Tax on Earth Day,” December 18, 2009.
http://www.greenparty.org.tw/index.php/discuss/comment/233–2009–
12–18–04–29–18 (accessed May 8, 2013; in Chinese).

13. An earthquake and the following tsunami severely damaged the

Fukushima nuclear plant. The radiation leak endangered the
neighborhood, causing the Japanese government to order the evacuation
of residents in the area. For months, people were suspicious of food and
water from the area because of possible radioactive contamination. Even
now, Fukushima residents still cannot return to their homes.

14. Central News Agency, “Mothballed Nuclear Power Plant Can Be

Activated Anytime: President,” July 31, 2015.
http://focustaiwan.tw/news/aipl/201507030027.aspx (accessed October
9, 2015).

15. The Four Noes and One Without pledge is essentially that as long as
China does not have an intention to use military force against Taiwan,
Chen Shui-bian would not declare independence, change the name of the
nation, push for the inclusion of the so-called state-to-state description in
the constitution, or promote a referendum on the issue of independence
versus unification.

16. The 1992 consensus refers to a memorandum of a meeting between

the semiofficial representatives of China and Taiwan in 1992. It stated that
both sides recognized the principle of one China. More specifically, China
and Taiwan belong to one China but the definition of one China is based
on their own interpretation. However, the DPP denied the existence of
1992 consensus.

Page 135 → 17. The Straits Exchange Foundation (the chairman, Chiang
Pin-kung, represents Taiwan) and the Association for Relations across the
Taiwan Straits (the chairman, Chen Yunlin, represents the PRC) are in
charge of most communication and negotiation on nonpolitical issues.

18. To the contrary, a number of studies of Taiwan voters also have shown
that individuals’ issue positions may affect their party preferences and
party identification (Hsieh and Niou 1996a, 1996b; Hsieh 2005; Sheng and
Chen 2003). We do not disagree with this argument and accept the
possibility of the reciprocal relationship between issue position and party
preference.
 19. For a discussion on various issues in Taiwan, readers may refer to the
following chapters of this book: chapter 6 on economic issue, chapter 7
on cross-Strait relations, and chapter 9 on wealth distribution.

20. In 2012, the presidential election and the legislative election were held
together so that the proportion of the independence/unification issue
rose.

21. Tien Yu-bin, Hsu Gu-cheng, and Chang Kim-Guo, 2003, “Support for
Rectifying the Name of Taiwan, the Public Is the Evidence,” New Taiwan
Weekly News, September 12, no. 390 (in Chinese).

22. Heilongjiang is located in northeastern China and has frigid weather

in the winter.

References

BBC News. 2000. “Taiwan’s Suspension of NUKE4 Causes Political Crises,”

October 30.
http://news.bbc.co.uk/hi/chinese/news/newsid_998000/9987432.stm.
Accessed October 8, 2015.

Brody, R., and B. Page. 1972. “The Assessment of Policy Voting.” American
Political Science Review 66:450–58.

Carmines, Edward G., and James A. Stimson. 1986. “On the Structure and
Sequence of Issue Evolution.” American Political Science Review 80:902–21.
Carmines, Edward G., and James A. Stimson. 1989. Issue Evolution: Race and the
Transformation of American Politics Princeton: Princeton University Press.

Central News Agency. 2015. “Mothballed Nuclear Power Plant Can Be

Activated Anytime: President.” July 31.
http://focustaiwan.tw/news/aipl/201507030027.aspx. Accessed October 9,
2015.

Chang, Pen-tsao. 1987. “How to Improve Harmony between the Labor and the
Capitalists after Lifting Martial Law.” Industrial Fortnight 181:19–21.

Chen, Luhu, Keng Su, Tu Ping-Lan, and Huang Quan-Po. 2009. “Rational Self-
interest or Emotional identity? An Analysis on Factors Which Influence the
Positions of Taiwanese Public in Cross-Strait Economic and Trade
Relations.” Soochow Journal of Political Science 27 (2): 87–125.

Chen, Wen-chun. 2003. “The Blue and the Green: The Political Ideologies of
the Mass Public in the 2000 Taiwan Presidential Election” [in Chinese].
Journal of Electoral Studies 10 (1): 41–80.

Cheng, Su-feng. 2004. “Party Image in Taiwan—an Application of Focus

Group.” Journal of Electoral Studies 11 (2): 185–216.

Page 136 → Cheng, Tun-jen, and Hsu Yung-ming. 1996. “Issue Structure, the
DPP’s Factionalism, and Party Realignment.” In Taiwan’s Electoral Politics and
Democratic Transition: Riding the Third Wave, ed. Tien Hung-Mao, 137–73.
Armonk, NY: M. E. Sharpe.
Chihibber, Pradeep, and Mariano Torcal. 1997. “Elite Strategy, Social
Cleavages, and Party Systems in a New Democracy Spain.” Comparative
Political Studies 30 (1): 27–54.

Chu, Yun-Han. 1994. “Party Competition, Conflict Structure, and Democratic

Consolidation.” Paper presented at the conference on Democratization and
Party Politics and Elections, Taipei, July 8, 1994.

Cox, Gary. 1990. “Centripetal and Centrifugal Incentives in Electoral Systems.”

American Journal of Political Science 34: 903–35.

Dalton, Russell J. 1988. Citizen Politics in Western Democracies. Chatham, NJ:

Chatham House.

Dalton, Russell J. 1996. “Political Cleavage, Issues, and Electoral Change.” In

Comparing Democracies, ed. Lawrence LeDuc, Richard G. Niemi, and Pippa
Norris, 319–42. London: Sage Publications.

Directorate-General of Budget. 2012. Statistical Yearbook of National Economics:

The Republic of China 2005. Taipei: Accounting and Statistics, Executive
Yuan, Republic of China.

Fell, Dafydd. 2005. Party Politics in Taiwan: Party Change and the Democratic
Evolution of Taiwan. London: Routledge.

Fell, Dafydd. 2011. “The Polarization of Taiwan’s Party Competition in the

DPP Era.” In Taiwan’s Democracy: Economic and Political Challenges, ed. Robert
Ash, John W. Carver, and Penelope B. Prime, 75–98. London: Routledge.

Fell, Dafydd. 2012. Government and Politics in Taiwan. London: Routledge.

Fiorina, Morris P. 1981. Retrospective Voting in American National Elections. New
Haven: Yale University Press.

Gates, Hill. 1979. “Dependency and Part-Time Proletariat in Taiwan.” Modern

China 5 (3): 381–408.

Green Party Taiwan. 2009. “Ma Ying-jeou Do Not Forget Your Promise of
Carbon Tax on Earth Day.” December 18.
http://www.greenparty.org.tw/index.php/discuss/comment/233–2009–12–
18–04–29–18.

Hawang, Shiow-duan. 2003. “The Predicament of Minority Government in the

Legislative Yuan.” Taiwanese Political Science Review 7 (2): 3–49.

Ho, Ming-sho. 2005a. “Taiwan’s State and Social Movements under the DPP
Government, 2002–2004.” Journal of East Asian Studies 5: 401–25.

Ho, Ming-sho. 2005b. “Weakened State and Social Movement: The Paradox of
Taiwanese Environmental Politics after the Power Transfer.” Journal of
Contemporary China 14 (43): 339–52.

Hsieh, John Fuh-sheng. 2005. “Ethnicity, National Identity, and Domestic

Politics in Taiwan.” Journal of Asian and African Studies 40 (1–2): 13–28.

Hsieh, John Fuh-sheng, and Emerson M. S. Niou. 1996a. “Issue Voting in the
Republic of China on Taiwan’s 1992 Legislative Yuan Election.” International
Political Science Review 17:13–27.

Hsieh, John Fuh-sheng, and Emerson M. S. Niou. 1996b. “Salient Issues in

Taiwan’s Electoral Politics.” Electoral Studies 15: 219–35.
Page 137 → Hsieh, Kuo-hsiung. 1989. “Workers Become Bosses: Class Mobility
in Taiwan Manufacturing Industry.” Taiwan: A Radical Quarterly in Social
Studies 2 (2): 11–54.

Inglehart, Ronald. 1977. The Silent Revolution: Changing Values and Political Styles
among Western Publics. Princeton: Princeton University Press.

Inglehart, Ronald. 1990. Cultural Shift in Advanced Industrial Society. Princeton:

Princeton University Press.

Inglehart, Ronald. 1997. Modernization and Postmodernization: Cultural, Economic

and Political Change in 43 Societies. Princeton: Princeton University Press.

Lipset, Seymour Martin. 1981. Political Man: The Social Bases of Politics. Baltimore:
Johns Hopkins University Press.

Lipset, Seymour Martin, and Stein Rokkan. 1967. “Cleavage Structure, Party
Systems, and Voter Alignments.” In Consensus and Conflict: Essays in Political
Sociology, ed. Seymour Martin Lipset, 113–85. New York: Transaction Books.

Lin, Thung-Hong, and Alfred Ko-We Hu. 2011. “Cross-Strait Trade and Class
Politics in Taiwan.” Thought and Words 49 (3): 95–134.

Liu, Aron. 2003. “The Political-Business Relationship in Taiwan: Since the

1990s.” Journal of Culture and Society 16: 97–126.

Liu, I-chou. 1994. “Generational Difference of Party Image among Taiwanese

Voters.” Journal of Electoral Studies 1 (1): 53–73.
Markus, Gregory B. 1982. “Political Attitudes during an Election Year: A
Report on the 1980 NES Panel Study.” American Political Science Review
76:538–60.

Markus, Gregory B., and Philip E. Converse. 1979. “The Dynamic

Simultaneous Equation Model of Electoral Choice.” American Political Science
Review 73:1055–70.

Moreno, Alejandro. 1999. Political Cleavages: Issues, Parties, and the Consolidation of
Democracy. Boulder, CO: Westview Press.

Norris, Pippa. 2004. Electoral Engineering: Voting Rules and Political Behavior.
Cambridge: Cambridge University Press.

Page, Benjamin I., and Calvin C. Jones. 1979. “Reciprocal Effects of Policy
Preferences, Party Loyalties, and the Vote.” American Political Science Review
73: 1071–90.

RePass, David E. 1971. “Issue Salience and Party Choice.” American Political
Science Review 65: 389–400.

Sheng, Shing-yuan. 2001. “An Exploratory Study on Formal and Informal

Legislative Participation: The Case of the Third Legislative Yuan.” Issues and
Studies (Chinese edition) 40 (5): 81–104.

Sheng, Shing-yuan. 2002. “The Issue Taiwan Independence vs. Unification with
the Mainland and Voting Behavior in Taiwan: An Analysis in the 1990s.”
Journal of Electoral Studies 9 (1): 41–80.
Sheng, Shing-yuan. 2003. “The Influence of the Legislative Branch and the
Executive Branch in the Process of Lawmaking: A Comparison of the
Unified and Divided Governments.” Taiwanese Political Science Review 7 (2):
51–105.

Sheng, Shing-yuan. 2005a. “Legislative Proposals of Legislators: Analyses of the

5th Legislative Yuan.” Presented at the Annual Conference of the Taiwan
Politics Association, Taipei.

Sheng, Shing-yuan. 2005b. “Constituency Representative and Collective

Representative: The Representative Roles of Taiwanese Legislators.” Soochow
Journal of Political Science 21: 1–40.

Page 138 → Sheng, Shing-yuan. 2008. “Party Leadership and Cohesion in the
Legislative Yuan: Before and after the First Party Turnover in the Executive
Branch.” Taiwan Democracy Quarterly 5 (4): 1–46.

Sheng, Shing-yuan. 2013. “Issues, Party Performance and Voters’ Voting

Behavior.” In 2012 Presidential and Legislative Elections: Changes and
Continuation, ed. Chen Lu-huei, 203–27. Taiwan: Wunan Press.

Sheng, Shing-yuan. 2014. “Why Do They Oppose to Continue Constructing

the Fourth Nuclear Power Plant?” Presented at the international conference
on “The Maturing of Taiwan Democracy: Findings and Insights from the
2012 TEDS Survey,” National Chengchi University, Taipei, March 22–23.

Sheng, Shing-yuan, and Chen Yih-yan. 2003. “Political Cleavage and Party
Competition: An Analysis of the 2001 Legislative Yuan Election.” Journal of
Electoral Studies 10 (1): 7–37.
Shyu, Huo-yan. 1998. “Taiwan Election and Social Cleavage: Party
Competition and Democratization.” In Cross-Straits Local Elections and
Political and Social Changes, ed. Chen Ming-tong and Cheng Yuan-nien, 127–
68. Taiwan: Yuedan Press.

Stites, Richard. 1985. “Industrial Work as an Entrepreneurial Strategy.” Modern

China 11 (2): 227–46.

Stokes, Donald. 1963. “Spatial Models of Party Competition.” American Political

Science Review 57 (2): 368–77.

Tang, Shui-yan, and Tang Ching-ping. 1997. “Democratization and

Environmental Politics in Taiwan.” Asian Survey 37 (3): 281–94.

Wang, Shu-twu. 2003. “The Wealth Distribution Policy under the Party
Turnover: An Analysis of the Senior Citizens’ Life Allowance of Middle-to-
Low Income Households.” Journal of Taiwan Wealth Distribution 4:79–105.

Wu, Nai-teh. 1993. “Social Cleavage and Party Competition.” Paper presented
at the Democratization and Party Competition Conference, Academia
Sinica, Taipei, June 11, 1993.

Yu, Ching-hsin. 2005. “The Evolving Party System in Taiwan, 1995–2004.”

Journal of Asian and African Studies 40 (1–2): 105–23.
Page 139 →
 Chapter 6

Economic Voting in Taiwan

Micro- and Macro-Level Analysis

Chia-hung Tsai

In chapter 5, Sheng and Liao find that the Taiwan voters consider economic prosperity as the
most important problem facing the country, especially when Taiwan’s economy was in a
downturn. It is, therefore, important to examine the Taiwan voter in this light. Economic voting
theory posits that voters tend to cast their votes in elections based on their assessment of
government performance in regard to the economy. That is, if voters perceive that the current
government is doing a good job in handling the economy, they are likely to vote for the
incumbent party (or candidate) in order to have it remain in office. But if voters are dissatisfied
with their economic conditions, they may affect a party turnover by voting against the incumbent
party (or candidate). The foregoing can be regarded as the basic rationale of economic voting at
the micro level.

This chapter has found mixed evidence of economic voting in the 2008 and 2012 Taiwan
presidential elections. On the one hand, partisanship overshadows retrospective economic
assessment, but on the other, prospective economic evaluation is a major determinant of voting
behavior. Our aggregate-level analysis, however, suggests that disposable income per capita
(DIPC) explains the incumbent party’s vote shares. But the standard error on DIPC is larger than
the .05 significance level. It is argued here that people tend to make their choices based on what
the government will do in the near future, while macro economic conditions affect election
outcomes.

Page 140 →

Economic Voting Theory

In the 1970s, Kramer (1971) argued that a party’s vote share represents the sum of three parts:
party identification, past economic performance, and incumbent advantage. He found that
changes in real personal income during the election year explained more than half of the
variation in the vote. Since then, political scientists have developed a great number of models
explaining, or even predicting, electoral outcomes (Abramowitz 1996; Erikson 1989; Fair 1978;
Hibbing and Alford 1981; Lewis-Beck and Rice 1992; Rosenstone 1983). Tufte (1978) provided
an engaging analysis of the impact of economic factors on congressional voting. His analysis
shows that presidential popularity, along with yearly change in real income per capita, fit the
election results from 1948 to 1976 very well. MacKuen, Erikson, and Stimson (1989) regressed
presidential approval on political events and consumer sentiment and concluded that approval is
a function of economic evaluation. Prior research on economic voting also implied that the
president is largely held accountable for the national economy; thus, voters who are not satisfied
with the national economy will send a signal targeting the president’s performance (Erikson
1989).

Individual-level analysis of economic voting clarifies the mechanism by which the economic
situation can affect political behavior. Key (1966) emphasized that voters look at the past records
of the incumbent, and Downs (1957) argued that voters will base their preference on candidates’
past performance when considering candidates’ promises. Kinder and Kiewiet (1981) provided a
theoretical foundation for pocketbook and sociotropic voting. They argued that personal
finances may provide a shortcut to information but that sociotropic voting does not place higher
informational demands on voters. They claimed that “[r]ather, voters must only develop rough
evaluations of national economic conditions, and then credit or blame the incumbent party
accordingly” (132). Feldman (1982) maintained that pocketbook voting occurs only among
people who hold on to economic individualism. Fiorina (1978), however, endorsed the
retrospective voting theory that the vote for the incumbent president’s party is a function of
individual personal income. Markus (1988) pooled survey data from the National Election
Studies from 1956 to 1984 and found that both the aggregate-level economic indicators and
personal finances are significant predictors of presidential voting choices. Nadeau and Lewis-
Beck (2001) emphasized the influence of incumbency in the election: people choose retrospective
pocketbook voting when the incumbent president is running in the election.
 Page 141 → At the aggregate and individual levels, the logic of retrospective voting is
straightforward: voters opt to support the incumbent who has improved economic conditions.
The prospective voting theory instead assumes that people would choose the candidate who holds
out a better prospect. Lanoue (1994) found that the effects of retrospective economic judgment
are more consistent than prospective ones, whereas Lockerbie (1992) found that better prospects
are more important than backward judgments. One of the underlying differences between the
prospective and retrospective frameworks concerns the amount of information needed to make
the forward or backward judgment. Considering that the modern economic situation involves
many aspects, including employment, housing, food prices, and so on, voters’ memory and
understanding of past economic records may be challenging. Moreover, retrospective judgments
can be subject to partisan framing: the incumbent may stimulate prosperity, but the opposition
may paint it in dismal colors, for example, as causing less equality and greater inflation. Finally,
the incumbent can hardly claim credit or be blamed for short-term economic fluctuations.
Therefore, I assume that the reward model of economic voting is less effective than the
prospective framework.

The other dimension of economic voting studies is whether personal finances or the national
economy has more influence on voting. Duch and Stevenson’s content analysis (2008) showed
that citizens do indeed possess information regarding the macro economy. Gomez and Wilson
(2006) found empirical evidence that sociotropic voting is common among less sophisticated
people because they consider the national economy to be the president’s sole responsibility.
Economic voting, therefore, does not require sophistication. Understanding national economic
situations may not demand too much information because voters can make a rough evaluation of
the national economy.

Finally, there is the question about whether micro- or macro-level data analysis is superior
regarding economic voting. The problem of endogeneity in cross-sectional data can be avoided
with aggregate-level data analysis; however, aggregate data have fewer observations. Being aware
of the trade-off, I analyze both types of data in this chapter.

The Case of Taiwan

It is not a coincidence that Taiwan voters embrace the value of economic growth, due to a
shortage of natural resources. As in Japan and other East Asian countries, politics is geared to
economic growth (Lumley 1976). When Page 142 → the Kuomintang government moved from
Nanjing to Taipei in 1949, its priority was land reform, and land reform, economic planning, and
shrinking the financial gap between the rich and the poor were the KMT’s primary
accomplishments.1 In the 1950s and 1960s, the KMT government focused on economic growth
(Chang 1965). According to Tien (1989, 26), the average GNP growth rate during the decade of
1953–62 was 7.5 percent. In the next decade, the number increased to 10.8 percent. By 1991,
exports accounted for 47.1 percent of GDP (Wu 1995, 59). Unfortunately, Taiwan’s economy has
faltered since the 1990s. Not only did the growth of national income flatten, but disposable
household income also declined. Labor shortages, inadequate infrastructure, and concerns over
the environment slowed down the rate of growth. Although the KMT has been credited with
economic development, the issue of social inequality and a power struggle cost the KMT the
2000 presidential election.

Although the KMT lost the 2000 and 2004 presidential elections, it called for improving cross-
Strait relations, arguing that the future of Taiwan’s economy could hinge on economic
cooperation with China. Tan and Ho in chapter 7 show the rapid increase in the amount of
Taiwan’s investments in China after 2000. While more and more capital and technical expertise
flows to China in exchange for cheap labor, land, and low environmental regulation, the resulting
closer tie between Taiwan and China indeed stimulated Taiwan’s economy, a result credited to
the KMT even though it was not in power at the time. In contrast, the Democratic Progressive
Party government discouraged the growing economic ties to the mainland, citing the concern for
national security. The quest for economic growth shaped the 2008 presidential election, and the
KMT returned to power.

Scholars have found evidence of economic voting in Taiwan, but their findings are not
conclusive. Hsieh, Lacy, and Niou (1998) supports prospective sociotropic voting in Taiwan’s
first popular presidential election, but Sheng (2009) suggested that both prospective and
retrospective economic perceptions were important in the 2008 presidential election. Hsiao
(2013) also stressed the strength of prospective economic perception. Wu and Lin’s (2013)
analysis showed no effect of economic evaluation but found that voter perception of cross-Strait
relations is a powerful predictor. More important, Wu and Lin (2012) discussed the validity of
retrospective and prospective economic evaluations. Because the 2008 presidential election was
held in March and the postelection survey is conducted in July and August due to the availability
of student interviewers, they argued that responses to these questions might be endogenous to
voting choice. They suggested that prospective economic views are more likely to predict voting
behavior Page 143 → if the election result is known before the survey is in the field. Their theory
conforms to the findings of the American literature, yet they stop short of testing it with
aggregate-level data.

In chapter 7, Tan and Ho find that the Taiwan voters’ evaluations of the national economy and
their own personal economic situations are not independent from partisan identification and
preference on the independence/unification issue. More precisely, those who preferred
unification or the status quo were more likely to evaluate the state of the economy positively,
while pro-independence respondents had a more negative view. If the confounding effect of party
support on economic evaluations indeed exists, it is hard to verify economic voting.

To confirm that economic outcome matters in mass political decisions, I will attempt to analyze
both macro- and micro-level data. Since the Ma administration signed the Economic
Cooperation Framework Agreement with Beijing in 2010, which is expected to increase Taiwan’s
trade with the Chinese mainland, I will focus on the public’s assessment of the national economy.
Indeed, previous research has shown that the sociotropic consideration, not pocketbook
concerns, correlates with voting decisions (Kinder and Kiewiet 1979; Erikson 1989). Our
hypotheses are thus:

1. Prospective economic evaluation is a significant predictor of voting behavior when

partisan predisposition is controlled for.
2. The partial effect of retrospective economic evaluation on voting behavior is significant
when partisan predisposition is controlled for.
3. The percentage change in disposable income per capita is associated with an incumbent’s
vote share, all other things being equal.

Findings

Survey Data Analysis: The 2004, 2008, and 2012 Presidential Elections
The national postelection survey data conducted by Taiwan’s Election and Democratization
Studies was chosen for this analysis. These three surveys use a stratified sampling frame that
covers all of the cities and counties except two surrounding islands, Kin-men and Mat-zu.

Each survey asks respondents’ retrospective economic evaluation: “Would you say that over the
past year, the state of the economy of Taiwan has gotten better, stayed about the same, or gotten
worse?” For the prospective Page 144 → evaluation, the wording is “Would you say that in the
forthcoming year, the state of the economy of Taiwan will get better, stay about the same, or get
worse?” For both the retrospective and prospective economic evaluations, I code the response of
“worse” as 1, “about the same” as 2, and “better” as 3.

To increase the predictive power of our economic voting models, I code self-identification as
with either the Pan-Blue or Pan-Green camp. If respondents identified themselves with either the
KMT, the People First Party, or the New Party, they are categorized as in the Pan-Blue camp. If
respondents identified with the DPP or the Taiwan Solidarity Union, they are considered Pan-
Green supporters. More than 30 percent of respondents, however, responded that they identified
with neither camp. Therefore, I created two dummy variables representing support for the Pan-
Blue and Pan-Green camps. Table 6.1 shows the wording of the questions and the coding
schemes.

Economic Evaluations, Economic Records, and Incumbent’s Vote

Tables 6.2 and 6.3 present the distribution of retrospective and prospective evaluations of
national economic conditions in three elections. In 2004, only one-third of respondents said that
the economic situation had become Page 145 → worse than it had been one year before (table 6.2).
But in 2008, the percentage had almost doubled. Although the KMT government tried to
address the economic problems by building closer ties with China, in 2012 more than 40 percent
of those polled said that the economy was worse than before. As for the prospective economic
evaluation, the three elections have differing distributions. In 2004 and 2012, a plurality of the
respondents said that economic conditions would be the same in the future. In 2008, however,
the proportion of respondents who said that economic conditions would become worse is a little
higher than that of people who said economic conditions would remain the same.
 Before estimating the effects of national economic evaluation on vote choice, the bivariate
relationship between economic evaluations and vote choice is assessed. Table 6.3 shows the
strong association between retrospective evaluation and incumbent voting in 2004. About 70
percent of people who felt that the economic situation had improved voted for the incumbent
party, the DPP. On the other hand, nearly 75 percent of those who that that economic times were
bad supported the challenger.

Table 6.3 also shows that in 2008 approximately 65 percent of people answered that the
economic situation had gotten worse, and half of them voted for the challenger, the KMT.
People who said that the economy had Page 146 → stayed the same overwhelmingly supported the
challenger. Fewer than 50 percent of respondents considered that the economic situation was
getting better and they clearly voted for the KMT. As for 2012, most respondents who thought
that the economy had gotten worse voted for the challenger, but most respondents who
answered ”stayed the same” and ”better” voted for the incumbent party. In other words,
retrospective economic voting occurred in 2012.
Table 6.2. Retrospective and Prospective Evaluation of National Economy

2004 2008 2012

Retrospective Worse 34.89 66.67 43.82

Same 38.51 30.27 38.58

Better 26.61 3.07 17.60

Prospective Worse 25.07 38.29 30.00

Same 47.58 36.85 48.24

Better 27.35 24.87 21.76

Note: Column percentages are in cells.

Table 6.3. Retrospective Evaluation of National Economic Condition and Vote Choice, 2004–2012

Nonincumbent % Incumbent % Total

2004 Worse 251 70.51 105 29.49 356

Same 171 43.29 224 56.71 395

Better 82 25.55 239 74.45 321

2008 Worse 393 50.13 391 49.87 784

Same 309 83.29 62 16.71 371

Better 33 84.62 6 15.38 39

2012 Worse 322 64.40 178 35.60 500

Same 152 32.14 321 67.86 473

Better 35 14.23 211 85.77 246

The first panel of table 6.4 shows that most people who said that the future economy would
become worse voted for the challenger in 2004, and that 85 percent of people who thought that
the economy would improve voted for the incumbent party. This result confirms the prospective
economic voting theory.

However, the second panel of table 6.4 suggests the reverse pattern in 2008; more than 90
percent of people who felt that the economy would get better voted for the nonincumbent party,
whereas more than 60 percent of the people who were pessimistic about the future economy
voted for the incumbent. Wu and Lin’s (2012) theory applies here in that voters may use the
election outcomes to justify their economic responses. In 2008, it is the winning challenger rather
than the losing incumbent who benefits from a good economic prospect—a prospect judged
after the election.

As for 2012, an election won by the incumbent, table 6.4 shows that more than 80 percent of
people who were optimistic about the future economy voted for the incumbent in 2012 and that
more than 60 percent who thought the opposite voted for the challenger. This result suggests
that voter perceptions of the economic future have an impact on voting.

These contingency tables suggest that voting is conditional on both prospective and
retrospective concern about the national economy. To estimate the partial effect of retrospective
and prospective economic evaluation in Page 147 → 2004, 2008, and 2012, our economic voting
model is set up as follows:

Table 6.4. Prospective Evaluation of National Economic Condition and Voting for the Incumbent, 2004–2012

Nonincumbent % Incumbent % Total

2004 Worse 207 78.41 57 21.59 264

Same 249 50.30 246 49.70 495

Better 48 15.34 265 84.66 313

2008 Worse 156 34.14 301 65.86 457

Same 283 67.38 137 32.62 420

Better 296 93.38 21 6.62 317

2012 Worse 240 66.67 120 33.33 360

Same 233 40.59 341 59.41 574

Better 36 12.63 249 87.37 285

Pr(Y) = β0 + β1 × Retrospective evaluation + β2 × Prospective evaluation + β3 × Pan-

Green + β4 × Pan-Blue + u
Where Y = 1 when voting for the incumbent party and 0 otherwise.

I ran three logistic regression models for each of the three years so that I could assess the
consistency of the estimates. The outcome variable was whether voters chose the incumbent
party or not. Notice that the DPP was the incumbent party in 2004 and 2008, and the KMT was
running the government in 2012. Independent variables include evaluation of the economy and
party identification. Retrospective and prospective evaluations of the national economy are
estimated respectively and jointly for each election. Considering that the proportion of people
who identified themselves as independents was constantly around 40 percent, I classified
individual partisanship as Pan-Green, Pan-Blue, and independents. Tan and Ho’s chapter also
suggests that the partisans of both incumbent and opposition parties have different views on the
economy. Therefore, two dummy variables, representing Pan-Green and Pan-Blue, have been
included in the models.

Table 6.5 partially confirms the hypotheses. Either retrospective or prospective economic
assessment alone would increase the probability of voting for the incumbent in 2004 and 2012
when both Pan-Green and Pan-Blue identification is controlled for. When both retrospective
and prospective evaluations are included in the full models of 2004 and 2012, retrospective
economic evaluation has no influence on voting behavior in 2004 and 2012. Comparing the
measure of model fit by AIC and BIC, I find that the retrospective voting model has a larger
value (worse fit) than the prospective voting model and the full model in each year. The
retrospective model also does poorly if goodness of fit is also measured by -2×ln(likelihood),
where again, the larger the value, the worse the model fits the data.

The signs of the coefficients for both retrospective and prospective evaluations in the full
model were negative in 2008. As in tables 6.3 and 6.4, people who felt that the economy would
get better and people who felt that the economy had become worse voted for the nonincumbent
party. It seems that the electorate deserted the DPP because of poor economic conditions in
addition to the KMT’s economic promise, including the ”six-three-three” slogan.2

Our survey data analysis suggests that people tend to use the future economic situation as the
yardstick for voting. If the incumbent has failed to bring about a better economic situation, he
will lose many votes even though Page 149 → his performance in office was good. Nevertheless,
people may not track the incumbent’s past record if they find the economic prospect appealing.
Page 148 → Table 6.5. Logistic Regression Models of Voting for the Incumbent Party, 2004–2012

2004 DPP 2008 DPP 2012 KMT

Retrospective Prospective Full Retrospective Prospective Full Retrospective Prospective Full

Intercept −0.293 −1.336*** −1.313*** 0.524† 0.430 1.161** −0.725* −1.414*** −1.610***

(0.308) (0.368) (0.394) (0.312) (0.292) (0.364) (0.283) (0.318) (0.352)

Pro-Green 3.388*** 3.294*** 3.301*** 3.413*** 3.301*** 3.299*** −3.129*** −3.156*** −3.116***

(0.365) (0.365) (0.368) (0.244) (0.243) (0.246) (0.278) (0.281) (0.282)

Pro-Blue −3.860*** −3.771*** −3.774*** −3.249*** −3.011*** −3.036*** 2.840*** 2.849*** 2.794***

(0.334) (0.336) (0.337) (0.383) (0.386) (0.388) (0.251) (0.252) (0.255)

Retrospective 0.332* −0.028 −1.062*** −0.805*** 0.536*** 0.221

(0.149) (0.173) (0.221) (0.237) (0.150) (0.167)

Prospective 0.839*** 0.854*** −0.762*** −0.569*** 0.863*** 0.768***

(0.174) (0.198) (0.157) (0.165) (0.157) (0.172)

AIC 513.585 494.065 496.038 584.124 583.463 573.917 650.530 631.597 631.841

BIC 533.494 513.974 520.925 604.464 603.804 599.342 670.953 652.021 657.370

Log
−252.792 −243.032 −243.019 −288.062 −287.732 −281.958 −321.265 −311.799 −310.921
likelihood

Deviance 505.585 486.065 486.038 576.124 575.463 563.917 642.530 623.597 621.841

Number of
1,072 1,072 1,072 1,194 1,194 1,194 1,219 1,219 1,219
observations

Notes: *** p < 0.001, ** p < 0.01, * p < 0.05. Standard errors are in parentheses.

Why do people tend to use their perceptions of future economic situation as the basis for
voting? The first possible explanation is that people may link the current economic conditions
with partisan squabbling. In 2004 and 2008, the DPP cited the KMT’s obstructive legislature and
China’s threat for their mediocre performance. In 2012, the KMT argued that they had inherited
the economic problems caused by the DPP’s policy of “self-isolation” from China and the world.
The second explanation is that the incumbent government has the resources to set up the agenda
(Page and Shapiro 1992); the incumbent government may either stimulate the economy in the
election year or divert people’s attention from economic conditions to national security or
inequality. For instance, the DPP proposed an unprecedented national referendum in the 2004
presidential election. In consequence, the economic record did not stand out as a clear-cut issue.3

If retrospective economic voting does not take place, how can the future national economy drive
voting behavior, especially in 2012? One of the possible explanations is that the Taiwan voters
responded to the notion of “peace dividend” proposed by the Ma administration. Ma personally
has repeatedly emphasized the benefits of closer ties with China, including more trade surpluses,
faster integration with the East Asian economy, and stronger national security. The prospect of a
better economy pictured by the Ma administration thus may explain the prospective economic
voting in Taiwan.

Across the nine models, an individual’s attachment to either the pan-Blue or pan-Green camp is
a major predictor of incumbent voting. In 2004 and 2008, people who identified themselves as
Pan-Green supporters voted for the incumbent party, but Pan-Blue supporters did not do so. In
contrast, Pan-Blue supporters tended to choose the KMT in 2012, but Pan-Green supporters
voted the other way. While economic voting does occur, the partisan variables are consistently
much more powerful.

Although survey data analysis supports prospective sociotropic voting, it relies on voter
perceptions and self-reporting. Aggregate-level studies of economic voting instead focus almost
completely on the effect of macroeconomic performance on elections. To predict an election
outcome, there are a great number of economic indicators to choose from. Erikson (1989)
followed Hibbs’s (1987) and Tufte’s (1978) investigations of macroeconomic conditions and the
presidential elections, arguing that the relative growth of per capita income change has largely
determined the election outcomes. Tsai (2000) also found that real personal income per capita
explains presidential popularity well, in addition to political events and war. Therefore, Page 150

→I calculated the percentage change of disposal income per capita from the previous year. In
addition, I include the percentage change in the year before the election year.

Because there is a perfect correlation between the DIPC in 2002 and 2003, the change in DIPC
has no impact on the vote shares. Therefore, our analysis drops the 2004 presidential election.
The DPP was the incumbent in 2008, so its vote share is regressed on Chen Shui-bian’s vote
share in 2004. Likewise, Ma’s vote share in 2012 is regressed on his vote share in 2008 as he
challenged the DPP’s ticket. Our aggregate-level economic voting model is:

Yt = γ0 + γ1 × (DIPC change) t-1 +γ2 × (DIPC change) t-2 +γ3 × (Incumbent’s vote
share) t-4 + u

The dependent variable is the raw percentage of votes won by the DPP and KMT in each
city/county respectively; I divided the number of votes for each party by the number of valid
votes. The first predictor variable is the percentage change in the DIPC from the previous year.
Because the presidential election is held in March, people judge the president’s performance by
the change in their income during the year between the two years before the election to the year
before it. For the 2008 election, the first percentage change in DIPC was determined as
(DIPC2007 − DIPC2006)/DIPC2006, and for the 2012 election, as (DIPC2011 −
DIPC2010)/DIPC2010.

To capture the possible influence from the economic situation two years prior to the two
elections, I used (DIPC2006 − DIPC2005)/DIPC2005 for the 2008 model, and for the 2012
model, (DIPC2010 − DIPC2009)/DIPC2009. Finally, I included as a covariate the DPP or KMT
vote shares in the city/county four years earlier.

The personal income data came from the Directorate General of Budget, Accounting and
Statistics (DGBAS) of the Executive Yuan.4 The vote-share data comes from the Election Study
Center, National Chengchi University (http://esc.nccu.edu.tw). It was jointly collected by the
center and the Central Election Committee.

Our data are from 23 cities and counties for the 2008 election and from 20 for the 2012
election. That is because Taichung County, Tainan County, and Kaohsiung County were merged
with Taichung City, Tainan City, and Kaohsiung City, respectively, in 2010. Fortunately, DGBAS
provides the DIPC of the three counties as if they had been merged back to 2000. Therefore, I am
able to calculate the percentage change in the DIPC for those three metropolitan areas for the
2012 election model.

Before estimating the effects of personal income factors, I plotted the change in the vote from
2004 to 2008 against the percentage change in Page 151 → DIPC from 2006 to 2007 for the DPP
(fig. 6.1), and I constructed a similar scatter plot for the KMT in the 2012 election (fig. 6.2). The
prediction line in figure 6.1 rises as the change in DIPC increases, and the R-squared is 0.06. This
plot implies that the DPP lost votes everywhere but they gained votes in the cities or counties
where the average income increased. Figure 6.2 shows a flat line, which suggests that the KMT
lost votes even in the cities or counties where income per capita increased. Both plots suggest
that the aggregate economic indicator may not be able to predict the variation in incumbent’s
vote share very well.

Fig. 6.1. Percentage change in DIPC versus change in incumbent’s (DPP) vote share, 2008

Table 6.6 shows four models. The first model presents the effects of the DPP’s vote share in
2004 and percentage changes in DIPC from 2007 and 2006 on the variation of DPP’s vote share.
The second model drops the change in DIPC from 2006. The third model demonstrates the
effects of Page 152 → the KMT’s vote share in 2008 and percentage changes in the DIPC in 2011
and 2010. Finally, the fourth model drops the second lagged DIPC variable.
 Fig. 6.2. Percentage change in DIPC versus change in incumbent’s (KMT) vote share, 2012

The first column shows that a 1 percentage point increase in disposable income per capita in
the preceding year produces about a 2 percentage point gain in Chen’s vote share in 2008. When
the second lag variable is dropped, the coefficient increases to 3.76. As for the 2012 election, the
third column shows that Ma’s vote share would increase by 2 percentage points with a one unit
change of disposable income per capita. The fourth column displays a similar result as the two-
year lag variable is dropped. Because the variation in disposable income per capita across 20 or
23 cities is small, the effect of change between one year before the election and the preceding year
is not significant, but it is substantial. Bartels and Zaller (2001) similarly found Page 153 → that the
contribution of economic growth (GDP or real disposable income, RDI) to an incumbent’s vote
share is about 2 percent. However, the relatively large standard errors here limit our ability to tell
exactly how an incumbent’s vote share reflects the economic conditions in cities and counties.
 To improve the standard errors, the 2008 and 2012 data are combined as a dataset with 43
observations. In addition to the coefficients on the lagged vote and on the change in DIPC, a
dummy variable indicating the 2012 election is estimated. Table 6.7 shows that the finding in
table 6.6 still holds; the coefficient and its standard error of change in DIPC both shrink as the
number of observations double. Therefore, the t-ratio remains small.

The conclusion is that the Taiwan voters continue to vote for the incumbent regardless of their
disposable income one or two years prior. This result seems to validate the effect of partisan
identification shown in chapter 5. Both incumbents’ vote shares in 2004 and in 2008 predict the
outcome variables very well, which means that voters who voted for a party tended to vote for it
again, regardless of which party was in office. The coefficient of the incumbent’s vote share in the
second model is smaller than that in the third model, which suggests that some Pan-Blue voters
switched to the PFP in 2012.

Survey data analysis implies that some people participated in “negative” retrospective voting in
2008: people who thought that the economy had Page 154 → become worse may have voted for the
challenger (i.e., the KMT). Table 6.6 and 6.7, however, show that there is a positive association
between the DPP’s vote shares and income growth. The gap between the individual- and
aggregate-level data analysis needs more examination.
Table 6.6. Models of the Incumbent Party’s Vote Share in 2008 and 2012

2008 2008 2012 2012

(Intercept) −8.45*** −8.96*** −5.58*** −5.39***

(1.49) (1.36) (1.17) (1.15)

Chen’s vote share 1.00*** 1.00***

in 2004 (0.03) (0.03)

Percentage change in 2.48 3.76

DIPC from 2007 (3.60) (3.28)

Percentage change in −3.73

DIPC from 2006 (4.19)

Ma’s vote share 0.97*** 0.97***

in 2008 (0.02) (0.02)

Percentage change in 2.61 2.32

DIPC from 2011 (2.62) (2.59)

Percentage change in −2.86

DIPC from 2010 (3.15)

R-squared 0.99 0.99 0.99 0.99

Adj. R-squared 0.98 0.98 0.99 0.99

N 23 23 20 20

Note: ***p < 0.001, two-tailed test. Standard errors are in parentheses.

Conclusion

In this paper, I examined evidence of economic voting in the 2004, 2008, and 2012 presidential
elections. On the one hand, I found that partisanship is a better predictor of outcome than
retrospective economic assessment. On the other, prospective economic evaluation is a major
determinant of voting behavior when controlling for partisanship. Notice that in 2008,
prospective evaluation of the economy may reflect people’s judgment on the newly elected
government’s handling of the economy rather than a true preelection prospective judgment.
 Our aggregate-level analysis suggests that economic conditions may not explain the incumbent
party’s election results. Instead, only the previous election outcome is a good predictor. Our
economic indicator, DIPC of one and two years, could involve measurement errors because
some people hide portions of their real income, such as rent and stocks. More work is needed to
devise a good macroeconomic voting model.

One of our findings is that the electorate does not punish incumbents for their performance in
the first term. This is interesting because it implies Page 155 → that the incumbent has the
advantage. A possible cultural explanation is that people in Taiwan dislike sudden change.
Instead, they tend to wait before throwing the incumbent out. Certainly, I need more election
results to test this hypothesis.

Table 6.7. Model of Combined Incumbent Party’s Vote Share in 2008 and 2012

Incumbent’s vote share

(Intercept) −8.03***

(0.86)

2012 presidential election 1.78***

(0. 36)

Incumbent’s vote share 0.99***

in the last election (0.02)

Percentage change in 2.10

the previous year (2.14)

R-squared 0.99

Adj. R-squared 0.99

N 43

Note: ***p < 0.001, two-tailed test. Standard errors are in parentheses.

To be sure, as both macro- and micro-level data analysis show, retrospective or prospective
economic voting is shadowed by party identification and the underlying national identity. From
the normative perspective, it may impede political accountability; politicians can play identity
cards instead of handling the economy well. However, economic growth will remain one of the
main sources of legitimacy. As a new generation arises, new parties are emerging, and
independents are increasing (see chapter 4), socioeconomic voting may become as important as
national identity.

Notes

1. In Taiwan, economic development in the 1960s and 1970s had not widened income
inequality as it had in other developing countries (Ferdinand 1996). By the 1980s, Taiwan’s
disparity between the rich and the poor was one of the lowest in the world (Roy 2003).

2. “Six-three-three” means a 6% economic growth rate, under 3% unemployment rate, and

raising income per capita to USD $30,000.

3. In January 2004, President Chen initiated the defensive referendum under Article 17 of
the Referendum Act. Because the DPP’s party platform states that “any change of Taiwan’s
independence status quo should be decided via referendum,” the international community
harshly rebuked DPP’s referendum proposal (Lin 2004). The Pan-Blue camp blasted the
DPP for holding the referendum and presidential election on the same day, asking voters to
boycott the referendum. Tsai, Hsu, and Huang (2007) argued that the two political camps
polarized on the referendum issue more than on the independence/unification issue.

4. Regarding statistics by county or city, please look at this link

http://statdb.dgbas.gov.tw/pxweb/Dialog/statfile9.asp. Dozens of indicators by city and by
year can be assessed through this interactive web page. Unfortunately, it is only in Chinese.
On the English version of the DGBAS’s website (http://eng.stat.gov.tw/mp.asp?mp=5),
there is no such web page.

References

Abramowitz, Alan I. 1996. “Bill and Al’s Excellent Adventure: Forecasting the 1996 Presidential
Election.” American Politics Research 24 (4): 434–42.
Bartels, Larry M., and John Zaller. 2001. “Presidential Vote Models: A Recount.” PS: Political
Science & Politics 34 (1): 9–20.

Chang, David W. 1965. “U.S. Aid and Economic Progress in Taiwan.” Asian Survey 5 (3): 152–60.

Downs, Anthony. 1957. An Economic Theory of Democracy. New York: Harper and Row.

Page 156 → Duch, Raymond M., and Randolph T. Stevenson. 2008. The Economic Vote: How Political
and Economic Institutions Condition Election Results. Cambridge: Cambridge University Press.

Erikson, Robert S. 1989. “Economic Conditions and the Presidential Vote.” American Political
Science Review 83 (2): 567–73.

Fair, Ray C. 1978. “The Effect of Economic Events on Votes for the President.” Review of
Economics and Statistics 60 (2): 159–73.

Feldman, Stanley. 1982. “Economic Self-Interest and Political Behavior.” American Journal of
Political Science 26 (3): 446–66.

Ferdinand, Peter. 1996. “The Taiwanese Economy.” In Take-off for Taiwan?, ed. Peter Ferdinand,
37–65. London: Royal Institute of International Affairs.

Fiorina, Morris P. 1978. “Economic Retrospective Voting in American National Elections: A

Micro-Analysis.” American Journal of Political Science 22 (1): 426–43.

Gomez, Brad T., and J. Matthew Wilson. 2006. “Cognitive Heterogeneity and Economic Voting:
A Comparative Analysis of Four Democratic Electorates.” American Journal of Political Science
50 (1): 127–45.

Hibbing, John R., and John R. Alford. 1981. “The Electoral Impact of Economic Conditions:
Who Is Held Responsible?” American Journal of Political Science 25 (3): 423–39.

Hibbs, Douglas A., Jr. 1987. The American Political Economy: Macroeconomics and Electoral Politics in
the United States. Cambridge: Harvard University Press.
Hsiao, Yi-ching. 2013. “Economic Accountability and Voting Choice: An Analysis of the 2012
Presidential Election” [in Chinese]. In 2008 nian zong tong yu li fa wei yuan xuan ju: Bian qian yu
yan xu [The 2012 presidential and legislative election: Continuity and change], ed. Lu-huei
Chen. Taipei: Wunan.

Hsieh, John Fu-sheng, Dean Lacy, and Emerson M. S. Niou. 1998. “Retrospective and
Prospective Voting in a One-Party-Dominant Democracy: Taiwan’s 1996 Presidential
Election.” Public Choice 97:383–99.

Key, V. O., Jr. 1966. The Responsible Electorate: Rationality in Presidential Voting. Cambridge: Belknap
Press of Harvard University Press.

Kinder, Donald, and D. Roderick Kiewiet. 1979. “Economic Discontent and Political Behavior:
The Role of Personal Grievances and Collective Economic Judgments in Congressional
Voting.” American Journal of Political Science 23 (3): 495–527.

Kinder, Donald, and D. Roderick Kiewiet. 1981. “Sociotropic Politics: The American Case.”
British Journal of Political Science 11 (1): 129–61.

Kramer, Gerald. 1971. “Short-Term Fluctuations in U.S. Voting Behavior, 1896–1964.” American
Political Science Review 65 (1): 131–43.

Lanoue, David J. 1994. “Retrospective and Prospective Voting in Presidential-Year Elections.”

Political Research Quarterly 47 (1): 193–205.

Lewis-Beck, Michael S., and Tom W. Rice. 1992. “Presidential Popularity and Presidential Vote.”
Public Opinion Quarterly 46 (4): 534–37.

Lin, Jih-wen. 2004. “Taiwan’s Referendum Act and the Stability of the Status Quo.” Issues and
Studies 40 (2): 119–53.

Lockerbie, Brad. 1992. “Prospective Voting in Presidential Elections, 1956–1988.” American

Politics Research 20 (3): 308–25.

Page 157 → Lumley, L. A. 1976. The Republic of China under Chiang Kai-shek. London: Barrie and
Jenkins.
MacKuen, Michael B., Robert S. Erikson, and James A. Stimson. 1989. “Macropartnership. 83 (4):
1125–42.

Markus, Gregory B. 1988. “The Impact of Personal and National Economic Conditions on the
Presidential Vote: A Pooled Cross-Sectional Analysis.” American Journal of Political Science 32
(1): 137–54.

Nadeau, Richard, and Michael Lewis-Beck. 2001. “National Economic Voting in U. S.

Presidential Elections.” Journal of Politics 63 (1): 159–81.

Page, Benjamin, and Robert Shapiro. 1992. The Rational Public. Chicago: University of Chicago
Press.

Rosenstone, Steven J. 1983. Forecasting Presidential Elections. New Haven: Yale University Press.

Roy, Denny. 2003. Taiwan: A Political History. Ithaca: Cornell University Press.

Sheng, Shin-yuan. 2009. “The Impacts of Economic and Welfare Issues on Voting Behavior: An
Inquiry into the 2008 Presidential Election” [in Chinese]. 2008 nian zong tong xuan ju: Lun er ci
zheng dang lun ti zhi guan jian xuan ju [The 2008 presidential election: The critical election of the
second party turnover], ed. Lu-huei Chen, Ching-hsin Yu, and Chi Huang. Taipei: Wunan.

Tien, Hung-mao. 1989. The Great Transition: Political and Social Change in the Republic of China.
Stanford, CA: Hoover Institution Press.

Tsai, Chia-hung. 2000. “American Voter Responses to International Political Events and
Economic Conditions: 1920–1996.” EurAmerica 30 (3): 143–91.

Tsai, Chia-hung, Yung-ming Hsu, and Hsiu-tin Huang. 2007. “Bi-Polarizing Politics: Explaining
the 2004 Presidential Election in Taiwan” [in Chinese]. Xuan Ju Yan Jiu [Journal of Electoral
Studies] 14 (1): 1–31.

Tufte, Edward R. 1978. Political Control of the Economy. Princeton: Princeton University Press.

Wu, Chin-en, and Yi-tze Lin. 2012. “Economic Voting and Presidential Elections: An
Assessment of Validity and Endogeneity” [in Chinese]. Taiwan Zheng Zhi Xue Kan [Taiwan
Political Science Review] 16 (2): 175–232.
Wu, Chin-en, and Yi-tze Lin. 2013. “Cross-Strait Economic Openness, Identity, and Vote
Choice: An Analysis of the 2008 and 2012 Presidential Elections” [in Chinese]. Xuan Ju Yan Ji
[Journal of Electoral Studies] 20 (2): 1–36.

Wu, Jaushieh Joseph. 1995. Taiwan’s Democratization: Forces behind the New Momentum. New York:
Oxford University Press.
Page 158 →
 Chapter 7

Cross-Strait Relations and the Taiwan Voter

Alexander C. Tan and Karl Ho

The saying “politics stop at the water’s edge” probably is not applicable to the
case of Taiwan because “high” international politics and “low” domestic
politics converge at the island state. The most prominent of the factors is
Taiwan’s relations with China, which seem to penetrate to the core of Taiwan’s
domestic politics and especially its electoral politics. While elsewhere, electoral
politics tend to be defined by the prominence of national or domestic concerns,
we learn from earlier chapters, chapter 3 in particular, that the China factor (as
an external factor) affects Taiwan’s perception of its security, both political and
economic, and influences citizens’ identity formation and political preferences.
A complete picture requires us to examine the role of cross-Strait relations in
defining voter identity.

We will begin by briefly examining Taiwan’s economic relations with China

and suggesting that despite greater interaction, the perceptions of Taiwan
voters have not significantly changed in favor of China. In fact, we suggest that
empirical evidence shows that Taiwan voters are ambivalent about the
increasing cross-Strait economic interaction. In the following section, we
examine how this ambivalence and vulnerability has politicized Taiwan society
and how this politicization helps paint a picture of who the Taiwan voter is.
 Cross-Strait Relations, the Economy, and Security

One of the features of Taiwan’s economy after 1990 has been increasing
economic relations with China, which brought risks along with the profits. Page

159 → The obvious benefit was the relatively benign impact of the 2008 global
financial crisis on the Taiwan economy as the growth in demand from China
following the crisis alleviated the decline in orders from Europe and the
United States. The increase in cross-Strait economic ties predates Taiwan’s
inauguration of democracy, which began in the late 1980s. Although it was not
officially sanctioned at the time, Taiwan companies and businesses had been
trading and investing in China through third parties (mostly in Hong Kong).
The surge in economic interactions was brought about by the convergence of
several factors such as the relative political relaxation in Taiwan, but more
noteworthy was the complementary change occurring in the industrial
structures of the two economies, in which Taiwan’s industrial restructuring saw
the manufacturing industries move to China. As these sunset (and generally
more labor-intensive) factories migrated to China for production, they in fact
established and created integrated production networks, further accelerating
economic integration. How important is China to Taiwan’s economy? How
integrated are the two economies? To answer these questions, two economic
indicators are important—trade and investments. Figure 7.1 shows Taiwan’s
exports and imports with China as a percentage of total exports and imports,
while figure 7.2 shows the amount of Taiwan’s annual investments in China.

As figure 7.1 shows, exports to and imports from China have increased
substantially in the last 20 years. Since the first handover of executive office
from the Kuomintang to the Democratic Progressive Party in 2000, the annual
increase in the China trade as a proportion of total Taiwan trade has steeply
increased. Beginning at 2.9 percent of total exports in 2000, China became a
major trade partner of Taiwan in the span of a decade, and by 2013 it
accounted for 26.8 percent of Taiwan’s total exports. Although the import
figures are significantly smaller, by 2010 imports from China accounted for
about 15.8 percent of total Taiwan imports. Masked under these two statistics
is the more important statistic—the huge trade surplus in favor of Taiwan.
Since 1990, the trade surplus has favored Taiwan such that by 2013 it
amounted to about US$39 billion. To simply state that numerous Taiwan
businesses and companies are profiting hugely from the China trade is an
understatement.

As with the surge in trade with China, Taiwan’s investments in China have
also increased rapidly. Figure 7.2 shows the amount of investment in China on
an annual basis. Although these figures were approved officially, they were very
likely underreported. What is unmistakable is the rapid increase in the amount
of Taiwan’s investments in China, which grew from US$2.6 billion in 2000 to
more than US$14 billion in 2010 and US$9.2 Page 160 → billion in 2013. As
mentioned earlier, the surge in investment is partly a consequence of Taiwan’s
industrial restructuring, as labor-intensive industries looked to reduce their
production costs, which increased due to rising business costs in Taiwan.
However, as Taiwan businesses and the economy benefit from the China trade,
the rapid increase of Taiwan investment in China has constantly raised the
specter of industrial hollowing-out, whereby manufacturing industries move
out of Taiwan en masse, leaving only administrative or design facilities.
 Fig. 7.1. Taiwan’s trade with China. Source: Cross-Straits Economic Monthly Report, Mainland
Affairs Council, http://www.mac.gov.tw (accessed April 19, 2014).

Besides the perils of industrial migration, the increasing trade and investments
in China also means that Taiwan businesses are exposed to risk without any
legal protection, because the two governments do not recognize each other or
have any bilateral framework to address economic disputes. There is the fear
within Taiwan that, beyond the China economic factor, Taiwan’s constrained
international status is limiting Taiwan’s ability to be part of any burgeoning
regional economic integration groups and thereby affecting Taiwan’s future
economic performance. In this sense, the Ma administration’s push for the
signing of the Economic Cooperation Framework Agreement with China is
perceived as a way to provide a legal framework for cross-Strait trade and
investments, as well as to allow Taiwan to be included in future regional Page 161
→ economic integration. The ECFA, in one sense, can be seen as simply a de
jure recognition of what is a de facto burgeoning cross-Strait economic tie that
has been going on for decades.
 Fig. 7.2. Officially approved investments by Taiwan in China. Source: Cross-Straits Economic
Monthly Report, Mainland Affairs Council, http://www.mac.gov.tw (accessed April 19, 2014).

If we follow the neoliberal perspective in international relations that

increasing economic interdependence and functional relations will lead to
greater trust and to peaceful relations between interdependent states, it
provides us with only an incomplete picture of Taiwan’s perception of cross-
Strait ties, because the politics of the ECFA also signals Taiwan’s own
insecurity and sense of vulnerability.

When asked to rate the People’s Republic of China (PRC) government’s

attitude toward the Taiwan government and people, more than 40 percent of
Taiwan respondents believe that the PRC government is unfriendly to the
Taiwan people, and over half believe that China is unfriendly to the Taiwan
government (except for a dip in 2011; see earlier chapter). The height of this
hostility, from the Taiwan point of view, was greatest in the second-term
campaign of independence-minded President Chen Shui-bian in 2004, where
79.4 percent of respondents believed that China is hostile to the Taiwan
government. Juxtaposing the trade and investment statistics from 2000 to 2010
with the perceived hostility data is quite revealing. Contrary to Page 162 → the

expectations of neoliberal and interdependence theory at a time when trade and

investments are trending upward, the Taiwan people’s perception of hostility
has stayed stubbornly high and has not declined in any appreciable way.

The growing economic integration between China and Taiwan creates a threat
to Taipei, due to China’s sovereignty claims over the island, and makes Taiwan
vulnerable, due to its increasing economic dependence on China. There are
concerns within Taiwan society about whether this dynamic represents a risk or
an opportunity for Taiwan. This vulnerability affects Taiwan citizen’s
perceptions of national security, driving a wedge among the elites—which is
also reflected among Taiwan voters—and directly shapes contending strategies
of how to best handle cross-Strait relations. The contentious politics took
center stage during the negotiations and eventual signing of the ECFA. There
is a very clear political divide, which is reflected in the positions of the two
major parties. The KMT believes that expanding these ties is important for
Taiwan’s continued economic dynamism, while the DPP argues that increasing
economic ties with China threatens national sovereignty and security (Gold
2009).
 This heightened sense of vulnerability and the divided outlook on how best to
approach the cross-Strait economic policy is also reflected by citizens’ views on
the pace of cross-Strait interactions. At the height of President Chen’s
administration, a fairly large group of respondents believed that the pace of
cross-Strait interactions was too slow (see 2004 and 2006 in table 7.1) with
more than one in four stating so. By 2008, the Three Direct Links, which began
in November 2008, and then the broader ECFA in June 2010 triggered a
significant swing to the “too fast” category, with more than one in three
worried about the fast pace. This segmentation at the voter level is largely
reflected in the divided discourse of party politicians, with the resultant
political polarization at the elite and voter levels being duly noted by numerous
political observers (Clark and Tan 2012b; Huang 2008; Liao and Yu 2008). It is
fair, then, to infer that cross-Strait economic relations have become highly
politicized (Clark and Tan 2012b), a point that we will return to in the
following section.

Cross-Strait Economic Relations and the Voters

In the previous section, we pointed out the highly politicized nature of cross-
Strait economic relations (specifically the policy-making process in Taiwan)
resulting from the drastically contrasting visions of the proper strategy to Page

163 → deal with cross-Strait ties. So how have cross-Strait economic relations
become highly politicized? How does the Taiwan voter reflect the
politicization of cross-Strait economic policy making?
 The contradictory claims to sovereignty of the PRC and the ROC had
underlain the conflicting relationship between these two polities since the end
of the Chinese Civil War and the establishment of the PRC. Yet in the mid-
1980s and the mid-1990s, a seeming détente existed between the two states
until the missile crisis in 1995–96, during Lee Teng-hui’s presidency. The
détente broke down as a result of different and contradictory perceptions of
territorial sovereignty despite the growing social and economic ties between the
two countries. China calculated that growing ties would rein in separatism in
Taiwan, while Taiwan’s democratization and growing economic wealth
stimulated the growth of a Taiwanese identity separate from China (Clark and
Tan 2012a, 2012b). Indeed, it is not an exaggeration to Page 164 → state that
Taiwan’s policies toward cross-Strait relations over the last two decades
constitute a game in which domestic and foreign policy are closely intertwined
as relations with China dominate the domestic political discourse and form the
main political and social cleavage separating the two major parties—the KMT
and DPP (Clark and Tan 2012a, 2012b).
Table 7.1. Views on the Pace of Cross-Straits Interactions, 2001–2008

Just right Too fast Too slow Don’t know N

Mar-01 38.6 17.8 22.0 21.6 1,077

Jul-01 35.5 17.3 20.3 26.8 1,100

Feb-02 37.7 15.8 17.8 28.7 1,081

Apr-02 36.5 13.1 17.5 32.9 1,091

Jul-02 30.7 16.4 18.4 34.5 1,091

Dec-02 29.6 15.4 19.6 35.5 1,076

May-03 30.7 23.4 20.2 25.7 1,082

Aug-03 30.6 20.5 19.5 29.4 1,149

Nov-03 32.1 17.0 17.5 33.4 1,100

Apr-04 39.9 15.1 20.2 24.7 1,083

Jul-04 36.6 14.2 20.8 28.4 1,153

Dec-04 35.8 19.1 25.2 19.9 1,067

May-05 40.2 25.7 18.9 15.1 1,084

Aug-05 34.1 25.1 25.0 15.7 1,096

Nov-05 33.6 22.4 30.6 13.4 1,102

Apr-06 30.6 19.9 34.5 15.0 1,088

Sep-06 36.2 26.2 25.1 12.5 1,068

Dec-06 35.6 19.6 29.6 15.2 1,073

Apr-07 34.5 22.5 30.2 12.7 1,072

Aug-07 36.1 21.0 28.7 14.2 1,095

Dec-07 40.7 18.6 32.2 8.5 1,067

Mar-08 40.8 18.9 35.2 5.1 1,068

Aug-08 40.6 29.5 17.5 12.4 1,094

Oct-08 38.6 30.1 20.5 10.8 1,081

Dec-08 47.5 37.2 14.5 0.8 1,068

Source: Mainland Affairs Council, ROC, http://www.mac.gov.tw/ (accessed April 19, 2014).

In the last section, we noted that the growing economic linkages have not
lessened the feeling of hostility to China and vulnerability among Taiwan
citizens, as shown by public opinion surveys. In fact, as mentioned earlier,
Taiwan’s growing wealth and its democratization have stimulated Taiwanese
identity formation and consciousness despite the growing economic links with
China over the last two decades.

Do these preferences color the Taiwan voters’ perception of cross-Strait

economic relations and the benefits and costs that it entails? In other words,
how politicized is the cross-Strait linkage, and how does it reflect who the
Taiwan voter is? To answer these questions, let us look at how the evaluation of
the economy, as well as the benefits of the ECFA, is affected by the preference
for independence or unification and by party identification.

Table 7.2 shows the distribution of how citizens evaluate the benefits of the
ECFA based on their preferred position on the issue of
independence/unification. Ignoring the middle categories of “status quo and
decide later” and “status quo forever,” it is evident from the statistics shown in
the table that preferences on the issue of independence/unification are highly
correlated with respondents’ evaluations of the benefits of the ECFA. Citizens
Page 165 → who have a preference for unification are less likely to claim that the
ECFA is bad for Taiwan, whereas citizens who prefer independence are more
likely to point out the costs to Taiwan’s economy by evaluating the country as
worse off as a result of the ECFA. It is not difficult to infer from these
numbers that evaluation of the ECFA is very much influenced by the citizens’
views on the independence/unification issue rather than the objective criteria
of whether the economy is actually growing or not.

Table 7.2. Evaluations of ECFA on the Economy by Independence/Unification Preference and

Partisan Identification

Worse Same Better N

Unify now 3.7% 44.4% 51.9% 1,569

Maintain status quo, unification later 6.3% 32.8% 60.9% 1,569

Maintain status quo and decide later 8.5% 41.1% 50.4% 1,569

Maintain status quo forever 12.8% 43.0% 44.2% 1,569

Maintain status quo, independence later 20.6% 55.2% 24.2% 1,569

Independence now 38.8% 53.1% 8.2% 1,569

Nonresponse 25.0% 61.1% 13.9% 1,569

Worse Same Better

Kuomintang 3.5% 27.4% 69.1% 1,570

Democratic Progressive Party 27.0% 58.1% 14.9% 1,570

New Party 0.0% 50.0% 50.0% 1,570

People First Party 5.3% 42.1% 52.6% 1,570

Taiwan Solidarity Union 14.3% 78.6% 7.1% 1,570

Source: TEDS 2012, Election Study Center, National Cheng-chi University.

Another way to view the politicization of Taiwan’s cross-Strait economic

interaction is how different partisan identifiers evaluate the impact of the
ECFA on the economy in general. Ignoring identifiers of the minor parties
(regardless of whether they are Pan-Blue or Pan-Green), it is clear from the
data in table 7.2 that the respondents’ identification with the KMT or the DPP
colors their views of the impact of the ECFA on the state of Taiwan’s economy.
About 85 percent of DPP supporters claim that the ECFA has made Taiwan’s
economy worse or kept it the same as before, while more than 69 percent of
KMT supporters give a positive evaluation to the ECFA. Yet again, it is not too
huge a leap to infer that the voters’ evaluations of cross-Strait economic
relations are more probably based on their partisanship rather than on
objective measures of general economic performance and health.

The impact of partisan identification and preference on the

independence/unification issue goes beyond just the evaluation of cross-Strait
economic relations as symbolized by the ECFA; it also permeates and colors
the Taiwan voters’ evaluations of the state of the general economy as well as
their evaluations of their own personal economic situations. Table 7.3 shows
how different respondents divided along their independence/unification
preferences evaluate the state of Taiwan’s general economy. These figures
correlate well with the evidence presented in table 7.3 regarding the ECFA’s
impact on Taiwan’s economy. In general, voters who preferred independence
for Page 166 → Taiwan were more likely to evaluate the state of the economy
poorly, while unifiers and status quo respondents had a generally more positive
view.
Table 7.3. Evaluations of the Economy by Independence/Unification Preference

Worse Same Better N

Unify now 41.4% 27.6% 31.0% 1,783

Maintain status quo, unification later 34.6% 42.6% 22.9% 1,783

Maintain status quo and decide later 37.8% 43.9% 18.2% 1,783

Maintain status quo forever 43.3% 37.0% 19.6% 1,783

Maintain status quo, independence later 51.5% 34.0% 14.6% 1,783

Independence now 69.9% 25.7% 4.4% 1,783

Nonresponse 47.4% 47.4% 5.1% 1,783

Source: TEDS 2012, Election Study Center, National Cheng-chi University.

When approaching the voting booth, however, Taiwan voters take on the
China factor in a more subtle manner. While the evaluations of the ECFA can
be filtered through their stance for unification or independence in the future—
hence, generating favorable or unfavorable views on the economic pact,
respectively—a closer look at the data in table 7.4 reveals more about the
intricacy of the electoral calculus. For those who rated the ECFA as beneficial
for personal economy, three-quarters, or a majority, of respondents reported
voting for the pro-unification KMT candidates Ma Ying-jeou and Wu Den-yih.
Of those who viewed the ECFA negatively, only slightly half chose the DPP
candidates Tsai Ing-wen and Su Jia-chyuan. The deciding voter group
comprised those who believed that the economy is not affected either way by
the ECFA. In this group, the KMT solicited much stronger support (46.7%
versus 30.5% for the DPP). Even when including those who refused to
respond, the opposition DPP was clearly unable to mount enough support by
opposing the agreement (see chapter 6 on more analyses of ECFA and
economic voting in Taiwan).

The evaluation of economy and cross-Strait economic ties cannot be totally

separated from the preference regarding national status as well as from the
partisan identification that points to the obvious politicization and polarization
along this dimension. When using objective economic indicators, such as
economic growth and trade statistics, the current statistics show that cross-
Strait economic linkage is a boon for the Taiwan economy; for example, the
huge trade surplus in favor of Taiwan that translates to huge capital inflows for
Taiwan companies and to increasing foreign exchange reserves for Taiwan. It is
also evident that Taiwan businesses benefit from investments in China, because
the increasing amount of investment can be Page 167 → interpreted to mean that
there is still profit to be made despite some stories of failed investments and
corporate bankruptcies.
Table 7.4. Evaluations of ECFA on the Personal Economy by Presidential Vote Decisions

Worse Same Better N

TSAI Ing-wen and SU Jia-chyuan 50.8% 30.5% 5.4% 1,569

MA Ying-jeou and WU Den-yih 18.1% 46.7% 75.0% 1,569

James SOONG and LIN Ruey-shiung 4.0% 1.8% 1.8% 1,569

Nonresponse 2.3% 1.1% 0.0% 1,569

Refused to answer 11.9% 8.0% 5.4% 1,569

Nonvoter 13.0% 11.9% 12.5% 1,569

100.0% 100.0% 100.0%

Source: TEDS 2012, Election Study Center, National Cheng-chi University.

Is Taiwan’s economy more vulnerable and less secure as interaction with

China increases? Since politics is an interplay of subjective and objective
realities, the Taiwan voters’ responses to this question largely depend on the
ideological and partisan lenses they wear—where the voter stands depends on
where they sit regarding their partisan identification.

Some Concluding Thoughts

This chapter explores the complex dynamics behind the relationship between
cross-Strait economic ties and elections in Taiwan. Specifically, we examine
how the burgeoning cross-Strait economic transactions affect the perceptions
of the Taiwan voters on how they perceive the vulnerability and security of
Taiwan, as well as the perils and profits of this economic interaction for
Taiwan’s destiny (and for their own economic welfare).
 When addressing the question whether foreign policy has only negligible
impact on voting decisions, Aldrich and his associates emphasize the
accessibility of attitudes toward such policy and very importantly the distinct
positions between parties and candidates (Aldrich, Sullivan, and Borgida 1989;
Aldrich et al. 2006). Indeed, Taiwan voters are clearly concerned with the cross-
Strait relations (and relatedly the economic interactions) between the two sides.
The attitude toward this important factor is readily accessible to voters not
only in how they view their future but also how they choose their government.
That subtle calculus is reflected in party identification and perceptions of the
national economy. From the empirical evidence we present in this chapter, we
demonstrate that the connections between the issues of economic development
and fostering cross-Strait relations, and subsequently how the voters cast their
ballots, are convoluted. With the status quo option available, for instance,
voters evade directly factoring in the choice of unification or independence (at
least not immediately) when choosing the government but instead are inclined
to resort to a party that can negotiate a fine balance as far as dealing with the
Chinese government is concerned. For the Taiwan voters, being economically
and politically isolated are clearly not good for the economy, yet getting too
close to China could be inviting trouble. This view explains the shift in public
opinion after the pro-unification KMT returned to power in 2008. Since then,
more and more voters have come to believe that the pace of cross-Strait
interactions Page 168 → has been too fast and that the government should put a
brake on unification (see chapter 3). This is quite an about-face compared to
the DPP administration in the 2000–2008 period.
 In the course of our exploration, the evidence (at least in the current and
short term) points to the ambivalence and seeming bifurcation of the Taiwan
voters with respect to their assessment of the current state of cross-Strait
economic ties and the strategies to deal with continuing engagement with
China. These views are most definitely colored by how the voters place
themselves in the dominant political and social cleavages in Taiwan as
expressed by partisan identification and ideological positions. Consequently, as
Clark and Tan (2012a, 102) suggest, questions can be raised about the long-
term stability of closer economic relations while the views on sovereignty
remain incompatible, thereby creating the paradox that the existing stability in
cross-Strait relations may well be a portent of future tensions and instability.

References

Aldrich, John, Christopher Gelpi, Peter Feaver, Jason Reifler, and Kristin
Thompson Sharp. 2006. “Foreign Policy and the Electoral Connection.”
Annual Review of Political Science 9:477–502.

Aldrich, John H., John L. Sullivan, and Eugene Borgida. 1989. “Foreign Affairs
and Issue Voting: Do Presidential Candidates “Waltz before a Blind
Audience?” American Political Science Review 83 (1): 123–41.

Clark, Cal, and Alexander C. Tan. 2011. Taiwan’s Political Economy: Meeting
Challenges, Pursuing Progress. Boulder, CO: Lynne Rienner.
Clark, Cal, and Alexander C. Tan. 2012a. “The Paradoxes in Taiwan’s ‘Two-
Level Game’ concerning Cross-Strait Relations.” American Journal of Chinese
Studies 19:89–104.

Clark, Cal, and Alexander C. Tan. 2012b. “Taiwan’s Increasingly Boxed-in

Economy: Economic Performance and Democratization.” Korea Observer 43
(1): 113–43.

Gold, Thomas. 2009. “Taiwan in 2009: Eroding Landslide.” Asian Survey 50:
65–75.

Huang, Minhua. 2008. “Polarized Politics, Divided Perceptions, and the

Political Consequences in Taiwan.” Paper presented at the International
Conference on Polarized Politics in a Comparative Perspective—America,
South Korea, and Taiwan, Taipei, January 26.

Liao, Dachi, and Eric Chen-hua Yu. 2008. “Are Taiwanese Politics Polarized?
An Overview since 2000.” Paper presented at the International Conference
on Polarized Politics in a Comparative Perspective—America, South
Korea, and Taiwan, Taipei, January 26.

Niou, Emerson. 2011. “The China Factor in Taiwan’s Domestic Politics.” Paper
presented at the Conference on Democracy and Diplomacy in East Asia,
University of Tokyo, Tokyo, September 16.

Page 169 → Wang, T. Y. 2005. “The Perception of Threats and Pragmatic Policy
Choice: A Survey of Public Opinion in Taiwan.” Issues & Studies 41 (1): 87–
111.
Wang, T. Y., Lu-huei Chen, and Shu Keng. 2010. “Symbolic Politics, Self-
Interests and Threat Perceptions: An Analysis of Taiwan Citizens’ Views on
Cross-Strait Economic Exchanges.” In Taiwan’s Politics in the 21st Century:
Changes and Challenges, ed. Wei-chin Lee, 159–84. Singapore: World Scientific.
Page 170 →
 Chapter 8

Evaluation of Presidential Candidates’

Personal Traits

Hung-chung Wang and Lu-huei Chen

In the study of political behavior, which factors affect an individual’s voting

behavior attract the most scholarly attention, and political scientists have
expended much effort searching for the answer. Using a sociopsychological
approach, the authors of The American Voter proposed a model called the funnel
of causality to explain American voters’ decision making. That model makes a
distinction between the short-term factors—issue and candidate—and the
long-term factor—party identification—on voting. The authors concluded that
among all determinants, voters’ party identification plays the most important
role in determining vote choice (Campbell et al. 1960).

Although Campbell and his colleagues also suggested the importance of

candidate evaluation on the vote, they did not put much emphasis on it. The
candidate, in their model, is treated as an idiosyncratic and short-term factor
and is dominated by the more enduring factors of party identification and
issue. Nevertheless, their model is criticized for giving little credit to the
importance of the candidate, since the stability of party identification and its
impact on election results have diminished. Scholars have argued that
candidate evaluation needs to be viewed as more meaningful than had been
thought (Miller and Shanks 1996; Niemi and Weisberg 2001; Wattenberg 1991).

In Taiwan, the candidate has played a very important role because multiparty
politics did not emerge until the 1980s. In one-party politics, the electorate can
choose only individual candidates from the same party, so one must focus on
the candidate rather than on the party. And much research suggests that even
after Taiwan switched its party system from a one-party Page 171 → system to a
multiparty system in the late 1980s, the candidate was still a critical factor in
Taiwan’s elections and significantly affected electoral outcomes at both the
central and local levels (Cheng, Chen, and Liu 2005; Hawang 1996; Liang 1994;
Shyu 1995; Yu 2003).

Nevertheless, that research did not systematically analyze the candidate factor
to determine which factors influence the formation of people’s candidate
evaluations, whether the candidate factor still plays an important role in
people’s vote choice when party identification is also included in the analysis,
and whether party competition affects individuals’ political behavior. To
correct this deficiency, in this chapter we will address these questions and
attempt to provide answers.

We begin by reviewing how individuals form their candidate evaluation and

introduce the criteria used for measuring candidates’ qualities. Then we explain
the sources of the data and the measurement of candidate qualities employed in
this research, and then look at Taiwan voters’ evaluations of the presidential
candidates in the 1996, 2004, and 2012 elections. We then analyze the
relationship between the people’s perception of candidates’ personal traits and
their party identification, and whether candidate evaluation matters in vote
choice.

The Formation of Candidate Evaluation and the Measurement

of Candidates’ Personal Qualities

How do citizens form their candidate evaluations? This issue has attracted
substantial scholarly attention and two information-processing models are
widely employed to explain it. The first one is the online model, which is
impression-driven and contends that an individual’s opinion is composed of
evaluations formed as information is encountered. This opinion is then
updated as new information arrives (Kim and Garrett 2012; Lodge, McGraw,
and Stroh 1989; Zaller 1992). The second model is memory-based and
maintains that individuals form their own opinion when a judgment is needed.
Individuals then retrieve the relevant information from their long-term
memory, forming a judgment based on what comes to mind (Kim and Garrett
2012; Lodge, McGraw, and Stroh 1989; Zaller 1992). Both models are
considered valid, but there is no consensus on which model is more powerful
in explaining an individual’s evaluation formation. However, scholars generally
agree that the assessment of candidates’ professional and personal qualities has
a significant impact on voters’ final voting decision (Flanigan and Zingale 1998;
Niemi and Weisberg 1993; Rahn et al. 1990).

Page 172 → To be more precise, as suggested by Miller and Shanks, “evaluations

of a candidate concerning a specific personal quality presumably represent the
accumulation of many impressions, both positive and negative, of a candidate”
(1996, 417). In other words, candidate evaluation is a collection of an
individual’s perceptions of a candidate’s personal traits. This evaluation may be
affected by various influential sources and general attributions derived from
individuals’ own partisanship. In addition, candidate trait may be more salient
under certain conditions, such as in times of crisis like facing a terrorist threat
(Merolla and Zechmeister 2009).

Which personal traits are adopted by voters to evaluate a candidate? Previous

studies of U.S. voters may offer some ideas on this issue for the case of Taiwan.
The authors of The American Voter examined citizens’ evaluations of candidates
in the 1952 and 1956 U.S. presidential elections based on three types of
personal traits: record and experience, abilities, and personal characteristics
(Campbell et al. 1960). These measurements were also employed by Lewis-Beck
et al. (2008) to test peoples’ perception of the candidates running in the 2000
and 2004 U.S. presidential elections. Kinder and his colleagues categorized
respondents’ answers to the question about the best definition of an ideal
president into two kinds of “abstract qualities.” The first quality is personality:
citizens’ “judgments about what an ideal president should and should not be
like as a person.” The second kind is performance: “what an ideal president
should do or should avoid doing while in office” (1980, 317). Their research
findings demonstrate that competence and trust are the most important
qualities for an ideal president. Furthermore, the well-educated respondents
considered that an ideal president is competent, and they expected a president
to be an “exemplary manager,” whereas citizens with lower educational levels
claimed that “likeability and personal morality” are the most important
qualities that an ideal president should possess, and they expect a good
president to be an “exemplary friend” (Kinder et al. 1980, 320).
 Miller, Wattenberg, and Malanchuk (1986) proposed five generic dimensions
for candidate evaluation: competence, integrity, reliability, charisma, and
personal attributes. Their research suggests that these five dimensions influence
individuals’ candidate evaluation, and among them, competence plays the
largest role. In addition, Miller and Shanks (1996, 420–25) employed nine
survey questions to investigate U.S. voters’ perceptions of the two candidates,
George H. W. Bush and Bill Clinton, in the 1992 presidential election. These
survey questions examined respondents’ evaluations of the following candidate
qualities: “cares about people like me,” “inspiring,” ”compassionate,” “gets
things done,” “intelligent,” “a Page 173 → leader,” “knowledgeable,” “honest,”
and “moral.” However, according to their analysis of the election, Americans’
vote choice was not strongly determined by their perception of the candidates’
personal traits. Evaluation of candidate qualities made only a small
contribution to Clinton’s victory (Miller and Shanks 1996).

Although scholars may use different personal traits to measure citizens’

candidate evaluation, we consider that the measurement proposed by Kinder et
al. (1980) is a very reasonable one due to its generality. As noted earlier, Kinder
et al. simply divided the candidates’ personal traits into two types: personality
and performance. This abstract classification can include in the analysis
personal traits used by other scholars and is easier for conducting further
research. Therefore, by following the measurement established by Kinder et al.,
we used these two personal qualities—personality and performance—as the
major dimensions with which to investigate the Taiwan public’s evaluation of
candidates in the 1996, 2004, and 2012 presidential elections.

Data and Measurement for Taiwan’s Presidential Candidates

As mentioned earlier, the candidate factor has, for a long time, been considered
to be a crucial variable influencing Taiwan citizens’ voting behavior. Compared
to party identification and issue, systematic study of the candidate factor has
been rare in Taiwan. Moreover, previous research on candidate factors focused
on single elections (Cheng, Chen, and Liu 2005; Hawang 2005; Yu 2003). It
does not offer us an overall picture of candidate evaluation in Taiwan. Nor
does this literature suggest a pattern for how the island’s citizens evaluate their
candidates. Because Taiwan citizens generally pay more attention to candidate
traits in presidential elections, candidate-oriented voting tends to be more
significant, a political phenomenon echoing the observation of Tverdova
(2011). Therefore, instead of using the results of a single election, we will focus
on the candidates who ran for Taiwan’s presidency in 1996, 2004, and 2012, and
especially on the winners of these three elections—Lee Teng-hui, Chen Shui-
bian, and Ma Ying-jeou—observing the change and continuity in candidate
evaluation after Taiwan’s democratization.1

To achieve this goal, we used data from surveys conducted for the three
elections mentioned above.2 Nevertheless, we found that no survey question
dealing with candidate evaluation was employed consistently in public opinion
polls in Taiwan, so making a comparative study via identical survey Page 174

→ questions was impracticable. As a consequence, we decided to adopt the

measurement established by Kinder and his colleagues (1980) to investigate
citizen perceptions of candidates’ personal traits. We simply divided a
candidate’s traits into two groups—personal character and competence—and
used them to determine how individuals form their candidate evaluation.
Furthermore, some scholars have suggested that citizens’ candidate evaluation
is conditioned by their party identification (Baker, Lawrence, and Tavits 2006;
Campbell et al. 1960; Miller and Shanks 1996), so we investigated the
relationship between individuals’ perceptions of a candidate’s personal traits
and their partisanship to see whether that was true in this case. Furthermore,
we also examined whether this perception was significantly associated with
people’s vote choice in the presidential elections. The survey questions
employed from the 1996, 2004, and 2012 public opinion polls can be found in
appendix 8.A1. We employed these survey questions to determine which
candidate had an advantage in certain personal traits.

Candidate Evaluation in the 1996, 2004, and 2012 Presidential

Elections

The 1996 presidential election represents a very important milestone in

Taiwan’s process of democratization because it was the first direct election for
the national leader. The leading candidate in this election was the incumbent
president, Lee Teng-hui. He ran for president in 1996, after having held office
for eight and a half years, and chose Lien Chan, who was the premier during
that time, as his running mate.3 That Taiwan citizens could directly choose
their chief executive came about because Lee had decided to adopt direct
elections as the new electoral system for selecting the president.

Lee was born in Taiwan in 1923 and grew up while Taiwan was ruled by Japan
(1895–1945). He was educated in Japanese schools and was thus strongly
influenced by Japanese culture. He then studied in the United States, receiving
his master’s degree from Iowa State University in 1953 and his PhD from
Cornell University in 1968, both in agricultural economics. While Lee was an
economist with the Joint Commission on Rural Reconstruction in 1971, he was
recommended to President Chiang Ching-kuo for a position in the cabinet and
became its youngest member at that time. He then achieved remarkable success
in his political career, being appointed to several important positions, including
mayor of Taipei City in 1978 and governor of Taiwan Province in 1981. Most
important, President Chiang nominated Lee to be his vice president in 1984.

Page 175 → Lee is a unique political figure. After Chiang died in January 1988,
Lee succeeded him as president of Republic of China (Taiwan) and chairman of
the ruling party, the Kuomintang. Undoubtedly, Lee was an excellent political
strategist. He was low-key while serving as the governor of Taiwan Province
and as Chiang’s vice president, but as president Lee displayed his skill at
statecraft by gradually expelling his opponents—the nonmainstream faction of
the KMT—from certain important positions in the central government, army,
and ruling party, thus consolidating his power.

While Taiwan citizens possess various ethnic identities and have different
partisan affiliation, as chapter 3 points out, they commonly have had a unique
psychological connection with Lee. Members of the Minnan and Hakka groups
liked and felt close to Lee because he was the first president who had been
born in Taiwan. Mainlanders also supported him early in his term because he
was promoted by Chiang. This unique psychological attachment was commonly
known as the Lee Teng-hui Complex (Shyu 1995, 1998).

However, Lee’s promotion of Taiwanization provoked some segments of the

KMT who believe that Taiwan should maintain a close relationship with China.
Angered by Lee’s localization policy, these politicians left the KMT and formed
the New Party. In order to confront this nonmainstream faction, Lee needed
help from central government outsiders. He thus recruited certain local
politicians to fill the vacancies in the executive and legislative branches of the
central government; however, many of these local politicians were corrupt. As
a result, even though Lee was called the father of Taiwan democracy, he was
criticized for spreading “black and gold” politics in Taiwan.

In 1996, Lee was the official nominee of the KMT. Failing to secure the
support of the KMT, the former president of the Judicial Yuan, Lin Yang-
kang,4 and the former president of the Control Yuan, Chen Lu-an, both
decided to leave the party and run as independents in Taiwan’s first direct
presidential election. The Democratic Progressive Party, the first major
opposition party in Taiwan, nominated Peng Ming-min, an exile who had been
an opponent of the authoritarian regime of Chiang Kai-shek. Although Lee
had to compete with these very well known and capable opponents, he defeated
them by a huge margin.5 Indeed, Lee was the only presidential candidate in
Taiwan’s democratic history who was able to garner electoral support from
voters of different partisan affiliations and ethnic backgrounds. A survey shows
that, in 1996 presidential election, Lee received 70.6 percent of the votes from
members of the Minnan group and also won the majority support of the
mainlander (51.2%) and the Hakka (79.4%) voters (Hsieh 1995). While 94
percent of the KMT identifiers supported Lee, close to 30 Page 176 → percent of
the DPP supporters also voted for him in Taiwan’s first democratic election.

How did the Taiwan public evaluate the candidates running for the president
in 1996? As shown in table 8.1, candidates’ personal character was measured on
the basis of the respondents’ perceptions of the following: affinity with the
people, leadership, integrity, trustworthiness, and understanding the needs of
the people. The measures of competence were the candidates’ ability to deal
with economic development, ethnic harmony, law and order, and peaceful
development of cross-Strait relations. Respondents were asked to judge which
candidate had the advantage over the other opponents with regard to these
characteristics. Table 8.1 shows that then incumbent president Lee received the
most positive evaluation, both in terms of personal character and competence,
than the other candidates. Specifically, 66.2 percent of the respondents believed
that Lee was the candidate who had the strongest leadership qualities. In
addition, Lee also received higher ratings for trustworthiness, affinity with the
people, and understanding the needs of the people. The only exception was the
item of integrity because only 10.7 percent of respondents felt that Lee had a
better record than the other candidates. However, it is worth noting that most
respondents (50.5%) thought there was no difference among candidates
regarding their level of integrity.

Lee’s advantage was also demonstrated by respondents’ perception of

candidates’ competence. Among the measurements of competence, 63.9 percent
of respondents said Lee was the candidate who knew how to promote
Taiwan’s Page 177 → economic development. Moreover, Lee was also considered
the candidate who could deal with peaceful development of cross-Strait
relations, law and order, and ethnic harmony better than the other candidates.
In short, among the candidates campaigning for the 1996 presidential election,
Lee Teng-hui had a much higher positive evaluation than his opponents.
Table 8.1. Candidate Evaluation in the 1996 Presidential Election

Personal Character Chen Lee Peng Lin ND* Total N

Affinity with the people 24.2% 40.6 3.7 8.2 23.2 100.0 1,396

Leadership 1.5% 66.2 6.8 3.1 22.4 100.0 1,396

Integrity 24.9% 10.7 9.6 4.3 50.5 100.0 1,396

Trustworthiness 10.9% 43.6 6.7 5.7 33.2 100.0 1,396

Understands the needs of the people 13.2% 35.5 9.0 5.3 36.9 100.0 1,396

Competence Chen Lee Peng Lin ND Total N

Economic development 1.6% 63.9 3.9 3.3 27.4 100.0 1,396

Ethnic harmony 14.8% 39.6 5.7 6.8 33.2 100.0 1,396

Law and order 6.4% 41.4 4.8 10.9 36.5 100.0 1,396

Peaceful development in cross-Strait relations 7.4% 45.3 2.8| 8.0| 36.5 100.0 1,396

Source: Hsieh 1995.

Notes: 1. Entries are row percentages. 2. ND = no difference. 3. Chen = Chen Lu-an; Lee = Le Teng-
hui; Peng = Peng Ming-min; Lin = Lin Yang-kang.

Taiwan’s first party turnover took place in 2000 after the DPP candidate,
Chen Shui-bian, won the 2000 presidential election. Chen was born in 1950 to
a poor farming family in southern Taiwan. Although lacking family financial
support during his youth, Chen did very well at school and entered the law
school at National Taiwan University in 1970. He passed the bar examinations
even before completing the law school program and became the youngest
lawyer in Taiwan’s history. Chen had a very successful career as a lawyer, and
his first connection with politics was defending the participants in the Formosa
Incident in 1980.6 Afterward, he turned to politics professionally, winning a
seat on the Taipei City Council in 1981 and then becoming a member of
Taiwan’s parliament, the Legislative Yuan, in 1989. His victory in the first
mayoral election in the capital city, Taipei, in 1994 was a critical achievement
for Chen and the DPP because he was the first non-KMT mayor since 1972.

Chen was an extremely popular politician. He had the courage to do things

that traditional politicians had not dared to do. For instance, he appointed
young people as government officials and improved the administration of the
Taipei city government. In addition, he was also the first politician to dress up
as such characters as Superman and Santa Claus to show his affinity with the
people on certain occasions. In short, Chen was a new type of politician and
was regarded as the symbol of the opposition party, even though he was also
criticized by his opponents for his bold actions.

Although Chen was a quite popular mayor and his job approval rating
exceeded 70 percent in his first term, he was defeated by Ma Ying-jeou in the
1998 mayoral election. Chen then decided to run for president as the DPP
candidate in the 2000 election, defeating Hsu Hsin-liang, the former DPP
chairman, in the primary. Chen eventually took advantage of the split in the
KMT in the 2000 election, becoming the first non-KMT president in history.7
Chen, however, garnered a low approval rating in his first term as a result of
the economic downturn, the increasing tension between Taiwan and China, and
the partisan antagonism between the Pan-Blue and the Pan-Green camps.
Moreover, his opponents in the 2000 presidential election, Lien Chan and
James Soong, who were leaders of the Pan-Blue Alliance, formed a formidable
alliance as running mates in the 2004 presidential election. As a result, it was a
general belief that Chen would lose his reelection bid.
 Page 178 → However, a mysterious assassination attempt against Chen’s life
occurred 19 hours before the polls opened, giving him a needed boost, and
Chen was reelected by a historically small margin of 0.2 percent of the total
votes. Supporter of the Pan-Blue Alliance suspected that the assassination
attempt was faked by Chen’s followers in order to gain sympathy votes. Thus,
Lien and Soong refused to concede. Instead, they challenged the electoral
outcome in court.

To assess citizens’ candidate evaluation in the 2004 presidential election, five

questions in the TEDS2004P survey were used: getting things done, integrity,
sincerity and trustworthiness, understanding the needs of people, and affinity
with the people. In addition, 10 traits served as measurements of candidate
competence: ethnic harmony, political stability, economic development,
eliminating black and gold politics, Taiwan’s international status, democratic
reform, law and order, peaceful development of cross-Strait relations, the
unemployment problem, and education reform.

Table 8.2 shows how the Taiwan public evaluated the candidates running for
president in 2004. The incumbent president, Chen, received more positive
evaluations than Lien in terms of personal character except for sincerity and
trustworthiness. Among these indicators, Chen had a significant advantage Page
179 → over Lien for getting things done and for affinity with the people. The
results are different when we look at respondents’ perceptions of Chen’s and
Lien’s competence. Respondents tended to give Lien a more positive evaluation
than Chen, whose only advantages were in eliminating black and gold politics,
democratic reform, and law and order. Among these measurements,
respondents significantly agreed that compared with Chen, Lien was more
likely to promote peaceful development of cross-Strait relations, since they
knew that the DPP government had failed to maintain harmonious relations
with mainland China.

Table 8.2. Candidate Evaluation in the 2004 Presidential Election

Personal Character Chen Lien ND* Total N

Gets things done 48.4% 11.3 40.3 100.0 1823

Integrity 23.6% 14.0 62.5 100.0 1823

Sincerity and trustworthiness 21.0% 23.3 55.7 100.0 1823

Understands the needs of people 31.4% 12.7 55.9 100.0 1823

Affinity with the people 46.9% 9.8 43.3 100.0 1823

Competence Chen Lien ND Total N

Ethnic harmony 21.2% 28.6 50.3 100.0 1823

Political stability 19.5% 25.5 55.0 100.0 1823

Economic development 13.2% 32.6 54.2 100.0 1823

Eliminate black and gold politics 41.3% 8.0 50.7 100.0 1823

Taiwan’s international status 20.3% 21.8 57.9 100.0 1823

Democratic reform 31.2% 11.3 5.2 100.0 1823

Law and order 17.9% 16.4 65.7 100.0 1823

Peaceful development in
9.4% 39.5 51.1 100.0 1823
cross-Strait relations

Unemployment problem 11.4% 25.4 63.2 100.0 1823

Education reform 9.8% 24.6 65.6 100.0 1823

Source: Hawang 2003.

Notes: 1. Entries are row percentages. 2. ND = no difference. 3. Chen = Chen Shui-bian; Lien = Lien
Chan.
 The DPP’s eight-year dominance of the central government began with Chen
Shui-bian’s victory in 2000 and ended with the election of Ma Ying-jeou in
2008. Ma was born in Hong Kong in 1950 and then moved to Taiwan. Both of
his parents were officials of the KMT and public servants in the central
government. Ma is a superstar of the Pan-Blue camp and has several advantages
over most Taiwan politicians. First, Ma has had much experience as a civil
servant. Immediately after receiving his SJD from Harvard Law School in 1981,
Ma returned to Taiwan and served as deputy director of the First Bureau of the
Presidential Office and as President Chiang Ching-kuo’s English interpreter. In
1988 Ma was appointed to the chair of the Research, Development and
Evaluation Commission, becoming the youngest cabinet member in the ROC
government. He then was named to be deputy minister of the Mainland Affairs
Council, the institute in charge of cross-Strait relations, and minister of justice.
In 1998, he was elected mayor of Taipei City in his first election, in which he
defeated the incumbent, Chen Shui-bian, which was an indication of Ma’s
popularity. Second, like John F. Kennedy, Ma is a good-looking man and also is
an alumnus of Harvard University. His civility has won him the middle class’s
support, particularly that of women. However, Ma was accused of being a spy
for the KMT government while he was studying in the United States, a charge
he firmly denied, and has also been criticized by opposition parties for his
conservative political stance.

Although Ma was questioned for his status as a U.S. permanent resident

during the campaign period, he easily defeated the DPP candidate, Frank
Hsieh, in the 2008 presidential election. Ma’s landslide victory brought the
KMT back to power.8 However, he faced the same problem that had plagued
Chen Shui-bian in Chen’s first term: Taiwan’s economic downturn. Although
cross-Strait relations became more peaceful after Ma took office, that
achievement did not help his approval rating. In 2012, he was challenged by
two opposition-party chairpersons, Tsai Ing-wen of the DPP and James Soong
of the People First Party.

Table 8.3 shows the Taiwan public’s evaluation of candidates in the 2012 Page

180 → presidential election. In the TEDS2012 survey data, the measurements of

a candidate’s personal character included the respondents’ perceptions of
whether the candidate “understands the needs of people,” “ever made you feel
angry,” “ever made you feel afraid,” and “ever made you feel hopeful.” As for
candidate competence, we examined how respondents evaluated candidates’
capability, whether they are “able to protect Taiwan’s interests,” and whether
they can “maintain cross-Strait peace.”

As data in table 8.3 shows, respondents are more likely to see Ma as a

politician who knows citizens’ needs, acts in a way to not make citizens’ worry,
and presents the country with a hopeful future. While similar assessments in
these respects are also applied to Tsai, Ma had a significant advantage over Tsai
in terms of perceived competence. In particular, about 60 percent of the
respondents believed that Ma could better handle the issue related to cross-
Strait relations.

Even though most citizens were not satisfied with his performance as the
country’s president, Ma was able to defeat his opponents in the 2012
presidential election, Tsai and Soong. If candidate evaluation does influence
individuals’ vote choice, we may make the inference that Ma’s advantage in
dealing with the relationship between Taiwan and China was the determining
factor leading to his victory in 2012.9
 Perceptions of Candidate Personal Traits versus Party
Identification

After examining how Taiwan citizens evaluated the candidates running for the
1996, 2004, and 2012 presidential elections, we went further and
investigated Page 181 → the relationship between the perception of the
candidates’ personal traits and respondents’ party identification since the latter
is treated as the most crucial factor affecting people’s political behavior.

Table 8.3. Candidate Evaluation in the 2012 Presidential Election

Personal Character Ma Tsai ND* Total N

Understands the needs of people % 36.9 32.4 30.8 100.0 1,826

Not make you feel unhappy % 22.6 33.8 43.6 100.0 1,826

Not make you worry % 34.0 26.8 39.2 100.0 1,826

Makes you feel hopeful for the future % 16.9 26.3 56.8 100.0 1,826

Competence Ma Tsai ND Total N

Capability % 40.4 31.5 28.0 100.0 1,826

Protect Taiwan’s interests % 40.7 31.2 28.1 100.0 1,826

Maintain cross-Strait peace % 59.4 13.2 27.4 100.0 1,826

Source: Chu 2011.

Notes: 1. Entries are row percentages. 2. ND = no difference. 3. Ma = Ma Ying-jeou; Tsai = Tsai Ing-
wen.

Petrocik (1996) proposed the theory of issue ownership, suggesting that each
political party has its own advantages in handling certain issues. For example,
the Republican Party is positively identified with issues associated with taxes,
whereas the Democratic Party has a good reputation in dealing with issues
related to social welfare. According to Petrocik, candidates focus their
campaign efforts on these issues to win voter support, and voters will vote for
the candidate if they think the party’s issues are important to them.

Based on Petrocik’s theory, Hayes (2005) then developed the theory of trait
ownership to explain the origins of candidate trait perception, examining the
connection between candidate traits, party issues, and strategic candidate
behavior. Hayes pointed out that American voters tend to connect the issues
owned by a political party with their evaluation of the candidates’ personal
traits, suggesting that the evaluation of candidate’s personal trait is conditioned
by party label. For example, the Democrats tend to be regarded as more
compassionate and empathetic than their counterparts, whereas GOP
candidates are expected to be strong leaders with high moral standards. Thus,
the winning strategy for a candidate is to trespass on his or her opponents’ trait
territory (Hayes 2005). Although research shows that partisan identifiers tend
to evaluate candidates of their own party more favorably (Campbell et al. 1960),
Hayes’s study reminds us that some candidates may be viewed as superior or
inferior in certain personal traits because of their party labels. In short,
perceptions of candidate personal traits can be partisan.

Indeed, scholarly research has demonstrated that the theory of issue

ownership can be equally applied to the case of Taiwan because major parties
on the island have divergent positions on a number of important issues. For
example, because the KMT and the DPP have opposite views on the issue of
cross-Strait relations the KMT is generally perceived as a pro-unification party
and the DPP a party for Taiwan independence (Hsieh and Niou 1996; Lin,
Chu, and Hinich 1996). Similarly, the public tends to view the KMT as a pro-
stability party and the DPP as a party associated with political reform (Hsieh
and Niou 1996; Sheng 2007; and Wang 2012). In general, Taiwan citizens tend
to view the KMT as superior in managing issues related to the economy,
education, and cross-Strait relations. Alternatively, gender equality, social
welfare, and governance at the local level are the winning issues for the DPP
(Chang 2010). Interestingly, even though Taiwan citizens experienced economic
hardship during Ma’s first term, they still consider the KMT as competent to
manage Taiwan’s economy (Sheng 2013).

Page 182 → Why do the KMT and the DPP own these particular issues? In the
case of the economy, Taiwan enjoyed its longest economic boom when the
KMT presidents, Chiang and Lee, held office, but the country suffered an
economic downturn when the DPP was in power. As a result, the public tends
to think that the KMT is more competent to manage the national economy. On
cross-Strait relations, the DPP views China as the main threat to Taiwan’s
sovereignty and national security while Beijing leaders also strongly oppose the
DPP’s pro-independence stance. Cross-Strait relations were tense as a result
during the 2000–2008 period when the DPP was in power. Because the KMT’s
official position is for unification and the Ma administration adopted a policy
of rapprochement toward China, the relationship between Taipei and Beijing
was peaceful during Ma’s presidency. As a result, the KMT is considered as
more competent to handle Taiwan’s relationship with China. Regarding the
reform issue, the KMT, as a long-term ruling party, has an extensive record of
corruption. The DPP politicians, especially in their early stage of career as city
mayors or county magistrates, were known for their rectitude and
determination for reform. Hence, political reform is a DPP-owned issue.
 Therefore, we would like to investigate whether the theory of trait ownership
is also applicable in the case of Taiwan. We not only examine whether all
respondents, no matter their party affiliations, agree that certain candidates are
superior in some personal characteristics, but we also assess whether there is a
direct link between the parties’ issue ownership and the Taiwan citizens’
perceptions of candidates’ personal traits.

As noted above, Lee Teng-hui overwhelmed his opponents in the 1996

competition in all respondents’ evaluations of personal traits, both personal
character and competence. Among these personal character evaluations, some
of them are worthy of further analysis. Figure 8.1 illustrates how respondents
affiliated with different parties evaluated the candidates’ leadership in 1996.
Regardless of their party affiliation, all respondents thought that Lee had the
strongest leadership compared to the other candidates. It is not a surprise that
the KMT supporters (88.8%) endorsed their chairman’s leadership, but even
non-KMT supporters acknowledged that Lee was superior as a national leader:
49.0 percent of the DPP and 47.5 percent of the NP respondents had a positive
view of Lee’s leadership. This result demonstrates the advantage that the
incumbent chief executive has.

Respondents’ perceptions of the 1996 presidential candidates’ capability in

handling economic development are presented in figure 8.2. The pattern is very
similar to that for leadership. Lee’s capability in dealing with economic issues
was also approved by respondents from different parties. A Page 183 → majority

of the DPP (54.8%) and the NP (50.0%) supporters also thought that Lee was
the best candidate to handle the national economy, suggesting that the KMT
candidate, Lee, had the advantage on this issue. Lee also had the advantage in
regard to the issue of cross-Strait relations. According to figure 8.3, the
majority of the Taiwan public agreed that Lee was more likely to manage the
relationship between Taiwan and China well. Among all partisans, the KMT
and the DPP supporters and the independents reported that Lee would be
better able to manage peaceful cross-Strait relations than his opponents. The
NP partisans were the only exception: they claimed that Lin was the candidate
who would be more competent in this respect. The only personal trait for
which Lee did not have an advantage was integrity (figure 8.4). Chen won more
endorsements among all citizens in integrity. In the Taiwan public’s mind, Lee
was not the most incorruptible candidate. Furthermore, more KMT party
members gave Chen credit for this personal trait instead of Lee (21.4% vs.
19.2%), showing that respondents seemed to have a strong impression that Lee
was connected to black and gold politics while in office.

Fig. 8.1. Trait: Leadership in 1996. Source: Hsieh 1995.

According to the above analysis, the 1996 presidential election displayed Lee’s
dominant advantage in the Taiwan public’s perception of the candidates’
personal traits. However, the 2004 election is another story. Only two
candidates represented two party coalitions in the run for the presidency, and
those two candidates had their own advantages with regard to personal traits,
as we already mentioned above. As figure 8.5 shows, Chen had the advantage
for the trait “getting things done.” Regardless of their party identification,
respondents reported that Chen was more likely than Lien to get Page 185

→ things done. Most important, the percentage of Pan-Blue supporters who

thought that Chen was more likely to fulfill the president’s duty was higher
than that of those who thought that Lien was (31.9% vs. 26.5%).

Page 184 →

Fig. 8.2. Trait: Economic development in 1996. Source: Hsieh 1995.

Fig. 8.3. Trait: Peaceful development in cross-Strait relations in 1996. Source: Hsieh 1995.
 Fig. 8.4. Trait: Integrity in 1996. Source: Hsieh 1995.

Fig. 8.5. Trait: Get things done in 2004. Source: Hawang 2003.
Differential public assessment of economic development is reflected in figure
8.6, which shows that the Pan-Blue and the Pan-Green supporters gave their
candidates a more positive assessment on this issue (67.0% and 31.6%).
Independents gave Lien a higher rating than Chen (22.8% vs. 7.2%). In other
words, Lien was considered to be more capable of handling the issues
associated with the national economy. Respondents may have formed this
perception because Lien had been the premier of the Executive Yuan under Lee
Teng-hui and had served as the country’s vice president. Lien’s term of office
also coincided with the period during which Taiwan experienced a booming
economy. As a result, Lien had the advantage on this issue. Figure 8.7 illustrates
the respondents’ perception of the candidates’ ability to deal with cross-Strait
relations, showing a pattern quite similar to that of figure 8.6. With the
exception of the Pan-Green identifiers, the Taiwan public all agreed that Lien
owned this trait. Actually, among the Pan-Green supporters, the proportion of
those who endorsed Chen only slightly exceeded that who endorsed Lien
(22.7% vs. 21.5%). Obviously, this issue was a weak point for Chen, and
citizens did not believe he was capable of handling this issue.

The DPP candidate’s advantage shows up in matters of the elimination of

black and gold politics and democratic reform. According to figures 8.8 and
7.9, respondents generally thought that Chen was more capable than Lien in
dealing with these two issues (41.3% vs. 8.0% in elimination of black and gold
politics and 31.2% vs.11.3 % in democratic reform). Particularly among the
Pan-Blue supporters, the proportion of those who endorsed Chen as being
capable of eliminating black and gold politics is almost the Page 186 → same as
that of those who thought Lien was, suggesting that the KMT candidate’s
weakness was associated with the issue of corruption. The KMT candidate’s
predominance on the cross-Strait relations issue is also shown in figure 8.10. In
addition to Ma winning all respondents’ endorsements, both the KMT
partisans and independents agreed that he was more capable than Tsai of
handling peaceful relations between Taiwan and China. Tsai did not have an
advantage on this issue even among her supporters.
 Fig. 8.6. Trait: Economic development in 2004. Source: Hawang 2003.

Fig. 8.7. Trait: Peaceful development in cross-Strait relations in 2004. Source: Hawang 2003.
On the basis of the figures presented above, we find that party’s issue
ownership is highly associated with voters’ perceptions of candidate personal
traits. Respondents generally reported that the KMT candidates—Lee, Lien,
and Ma—were more capable of dealing with the KMT-owned issues—
economic development and cross-Strait relations—than their opponents. On
the other hand, DPP candidates have an advantage on the issue of reform,
including democratic reform and elimination of black and gold politics. In
short, our research findings suggest that on certain issues some Page 188

→ candidates are superior to their opponents because of their party’s ownership

of the issues.
 Page 187 →

Fig. 8.8. Trait: Eliminate black and gold politics in 2004. Source: Hawang 2003.

Fig. 8.9. Trait: Democratic Reform in 2004. Source: Hawang (2003).

Fig. 8.10. Trait: Cross-Strait peace in 2012. Source: Chu 2011.

Perceptions of Candidate Personal Traits vs. Vote Choice

Since the candidate factor is viewed as one of the three most critical factors
affecting people’s voting behavior, we next focus on whether these perceptions
of candidate personal traits have a significant impact on the Taiwan public’s
vote choice. Tables 8.4, 8.5, and 8.6 show the relationship between vote choices
and citizens’ perceptions of a candidate’s capability of handling peaceful
development of cross-Strait relations, which is the main political cleavage in
Taiwan. Data in these tables suggest that the perception of candidates’ personal
traits did not significantly correspond to vote choice in the 1996, 2004, and
2012 presidential election. For example, among the DPP supporters in table
8.4, a majority (50.0%) of those who considered Lee to be the best one to
handle the issue still voted for the DPP candidate, Peng. The vast majority of
KMT identifiers, no matter which candidate won their endorsement, decided to
support Lee in the 1996 election. The only exception was for partisan
independents: they were the only group whose vote choice corresponded highly
with their evaluations of a candidate’s capacity to handle issues related to cross-
Strait relations. Respondents of the 2004 and 2012 survey behaved in a similar
way, as data in tables 8.5 and 8.6 show as the vast majority of KMT and DPP
supporters made their electoral decisions based on partisan affiliations. The
results appear to show that party identification still played a more important
role than candidate evaluation in affecting vote choice. In fact, the relationship
between vote choice and such candidate traits as boosting economic
development, eliminating black and gold politics, and promoting democratic
reform are very similar. The empirical findings are not shown here due to space
limitation.

Moreover, as tables 8.5 and 8.6 show, Pan-Green supporters are more likely to
be split on the question of which candidate is most competent in dealing with
cross-Strait relationships than Pan-Blue identifiers. This can be explained by
the theory of issue ownership, which suggests that each political party has
advantages on certain issues. Since the KMT is perceived as being more
competent to handle Taiwan’s economic development and relationship with
China, it is not surprising that Pan-Green supporters are more divided in this
regard. However, they still make their vote choices based on partisan
affiliations even though they credit Pan-Blue candidates with a stronger ability
to deal with cross-Strait relationships.

Overall, the empirical evidence from the 1996, 2004, and 2012 presidential Page
189 → elections allows us to draw the following conclusion: the effect of
candidate evaluation on vote choice is not as significant as that of party
identification. Taiwan citizens’ electoral decisions are mainly guided by partisan
affiliation rather than by evaluation of candidate personal traits.

Conclusion

This chapter has explored how the Taiwan public evaluates candidate quality.
We focused on candidates running in the 1996, 2004, and 2012 presidential
elections, investigating people’s perception of the candidates’ personal traits.
Page 190 → Following previous research, candidate personal traits were divided
into two types: personal character and competence. Our empirical results reveal
some interesting trends.
Table 8.5 Perception of Peaceful Development of Cross-Strait Relations versus Vote Choice in
2004

Pan-Blue PID Vote choice

Chen Lien Total
Personal trait
Chen 2 (28.6) 5 (71.4) 7 (100.0)
preference
Lien 5 (1.4) 358 (98.6) 363 (100.0)

Pan-Green PID Vote choice

Chen Lien Total
Personal trait
Chen 123 (99.2) 1 (0.8) 124 (100.0)
preference
Lien 97 (94.2) 6 (5.8) 103 (100.0)

IND Vote choice

Chen Lien Total
Personal trait
Chen 21 (100.0) 0 (0.0) 21 (100.0)
preference
Lien 32 (34.8) 60 (65.2) 92 (100.0)

Source: Hawang 2003.

Notes: 1. Chen = Chen Shui-bian; Lien = Lien Chan. 2. IND = partisan independent.
Table 8.6. Perception of Cross-Strait Peace versus Vote Choice in 2012

Pan-Blue PID Vote choice

Ma Tsai Total
Personal trait
Ma 591 (98.3) 10 (1.7) 601 (100.0)
preference
Tsai 2 (66.7) 1 (33.3) 3 (100.0)

Pan-Green PID Vote choice

Ma Tsai Total
Personal trait
Ma 16 (10.1) 143 (89.9) 159 (100.0)
preference
Tsai 2 (1.2) 168 (98.8) 170 (100.0)

IND Vote choice

Ma Tsai Total
Personal trait
Ma 100 (72.5) 38 (27.5) 138 (100.0)
preference
Tsai 2 (7.7) 24 (92.3) 26 (100.0)

Source: Chu 2011.

Notes: 1. Ma = Ma Ying-jeou; Tsai = Tsai Ing-wen. 2. IND = partisan independent.

First, candidate evaluation appears to be highly associated with electoral

outcome. Lee Teng-hui’s overwhelming advantage in citizens’ perceptions of
candidate personal traits matches his landslide victory in the 1996 election.
However, that overwhelming advantage was not reproduced by the candidates
running in the 2004 election. Neither Chen Shui-bian nor Lien Chan Page 191

→ was able to win respondents’ endorsements in all personal traits. Instead, as

noted above, Chen had the advantage in personal character and Lien had the
advantage in competence. The intense competition between Chen and Lien in
citizens’ perceptions of candidate quality reflects the historically small winning
margin in the 2004 presidential election.
 Second, our research confirms that the theory of candidate trait ownership is
also applicable in Taiwan’s case. Although candidate evaluation is conditioned
by party identification, in that people are more likely to give their own party’s
candidate a higher evaluation, candidates representing different parties do own
certain personal traits that are highly correlated with the political issues owned
by their parties. In Taiwan, KMT candidates, in respondents’ perceptions, are
more likely to manage economic development and cross-Strait relations well,
whereas DPP candidates are considered a better choice to deal with reform
issues, such as elimination of black and gold politics.

However, our analysis shows that citizens’ vote choices do not necessarily
match their evaluations of candidate personal traits. Supporters of the two
major parties in general still vote for their party’s candidate even though they
think that their party’s nominee does not have the advantage in certain personal
traits. Party identification still plays a critical role in the Taiwan public’s vote
choice. Partisan independents were the only group whose candidate evaluation
matched their vote choice. This result does not mean that independents’
candidate evaluation single-handedly determine their electoral decisions, which
is arguably decided by many determinants.

In the funnel of causality model proposed by Michigan scholars, candidate,

issue, and party identification are three major factors determining vote choice.
We, in this chapter, begin with the analysis of how respondents evaluate their
presidential candidates’ personal traits and then include political issues and
party identification into our research. Our findings demonstrate that party
identification, due to parties’ issue ownership, conditions how respondents
evaluate candidates. In short, the result confirms that party identification is the
most important factor in Taiwan citizens’ voting behavior.

Do these results demonstrate that candidate evaluation has no impact on the

Taiwan public’s voting decisions? In our opinion, it may be inappropriate to
draw such a conclusion because here we have examined only the relationship
between citizens’ evaluations of single candidate’s personal traits and vote
choice, whereas citizens’ voting decisions may be determined by a
comprehensive evaluation of candidate quality. Unfortunately, the existing
survey data do not allow us to investigate the effect of comprehensive
candidate evaluation on vote choice. Furthermore, here we analyzed only
individuals’ candidate evaluation in presidential elections. Page 192 → The results
may be different if we focus on the effect of candidate evaluation at the local
level.

Last but not least, we believe that candidate evaluation is valuable in analyzing
the Taiwan citizens’ voting behavior, but data on this topic that is both more
concise and accurate awaits further research. For example, attempts can be
made to investigate which type of candidate trait, among personal character
and competence, is more likely to play the larger role in determining people’s
vote choice. Moreover, the relationship between citizens’ party identification
and evaluation of candidate personal traits should also receive more attention
from Taiwan scholars. Which candidate trait is more important than others? Is
there an evaluation gap between the two major parties’ supporters? For
instance, KMT supporters may place more weight on a candidate’s competence,
whereas DPP partisans may consider personal character to be more important.
There is still a need for more research on this subject.

Appendix 8.A1. Survey Questions Employed as the

Measurements of Candidate Evaluation

1996 Presidential Election

Personal Character

Which candidate do you think has the most affinity with the people?
Which candidate do you think has the strongest leadership?
Which one is the most incorruptible?
Which one is the most trustworthy?
Which one has the best understanding of people’s needs?

Competence

Which one is more likely to have the ability to promote economic development?
Which one is more likely to have the ability to promote ethnic harmony?
Which one is more likely to have the ability to improve law and order?
Which one is more likely to have the ability to promote peaceful developments in cross-Strait
relations?
Page 193 →

2004 Presidential Election

Personal Character
 First, let’s look at getting things done. How good is Chen Shui-bian (Lien Chan) at getting things
done?
Concerning incorruptness, how incorrupt do you think Chen Shui-bian (Lien Chan) is?
Concerning sincerity and trustworthiness, how sincere and trustworthy do you think Chen Shui-
bian (Lien Chan) is?
Concerning understanding the needs of the people, how well do you think Chen Shui-bian (Lien
Chan) understands the needs of the people?
Concerning affinity with the people, how much do you think Chen Shui-bian (Lien Chan) has affinity
with the people?

Competence

Do you think the candidates on the Chen Shui-bian (Lien Chan) ticket have the ability to promote
ethnic harmony?
Do you think the candidates on the Chen Shui-bian (Lien Chan) ticket have the ability to maintain
political stability?
Do you think the candidates on the Chen Shui-bian (Lien Chan) ticket have the ability to promote
economic development?
Do you think the candidates on the Chen Shui-bian (Lien Chan) ticket have the ability to eliminate
black and gold politics (gangster and money politics)?
Do you think the candidates on the Chen Shui-bian (Lien Chan) ticket have the ability to raise
Taiwan’s international status?
Do you think the candidates on the Chen Shui-bian (Lien Chan) ticket have the ability to promote
democratic reforms?
Do you think the candidates on the Chen Shui-bian (Lien Chan) ticket have the ability to improve law
and order?
Do you think the candidates on the Chen Shui-bian (Lien Chan) ticket have the ability to promote
peaceful developments in cross-Strait relations?
Do you think the candidates on the Chen Shui-bian (Lien Chan) ticket have the ability to resolve the
unemployment problem?
Do you think the candidates on the Chen Shui-bian (Lien Chan) ticket have the ability to handle
educational reform?
Page 194 →
2012 Presidential Election

Personal Character

How would you rate Tsai Ing-wen (Ma Ying-jeou; James Soong) using a 0 to 10 scale, if 0 means that
you think the candidate does not understand at all the needs of ordinary people and 10 means that
a candidate completely understands the needs of ordinary people?
Has Tsai Ing-wen (Ma Ying-jeou; James Soong), because of the kind of person she is or because of
something she has done, ever made you feel unhappy?
Has Tsai Ing-wen (Ma Ying-jeou; James Soong), because of the kind of person she is or because of
something she has done, ever made you worry?
Has Tsai Ing-wen (Ma Ying-jeou; James Soong), because of the kind of person she is or because of
something she has done, ever made you feel hopeful for the future?

Competence

I’d like to ask you to evaluate the capability of three presidential candidates, how would you rate Tsai
Ing-wen (Ma Ying-jeou; James Soong) using a 0 to 10 scale?
How would you rate Tsai Ing-wen (Ma Ying-jeou; James Soong) using a 0 to 10 scale, if 0 means
candidates are completely incapable of protecting Taiwan’s interests, and 10 means that candidates
are completely able to protect Taiwan’s interests?
How would you rate Tsai Ing-wen (Ma Ying-jeou; James Soong) using a 0 to 10 scale, if 0 means
candidates are completely incapable of maintaining cross-Strait peace, and 10 means that
candidates are completely able to maintain cross-Strait peace?

Notes

1. Regarding candidate evaluation, we focus here on people’s perception

of two candidates’ personal traits: personal character and competence.
However, no question was included in the 2000 and 2008 surveys that
examined respondents’ evaluation of these two dimensions. Since the
three waves of survey data we used already included all winners in the
direct presidential elections, we decided not to include the 2000 and 2008
data in our analysis.

2. The first dataset analyzed in this chapter is from “An Interdisciplinary

Study of Page 195 → Voting Behavior in the 1996 Presidential Election.”
That research project was conducted by the Election Study Center,
National Chenchi University, Taiwan, and the principal investigator was
Professor John Fu-sheng Hsieh. The second dataset is from “Taiwan’s
Election and Democratization Study, 2002–2004 (III): The Presidential
Election, 2004 (TEDS 2004P).” The principal investigator was Professor
Shiow-duan Hawang for TEDS 2004P. The third dataset is from
“Taiwan’s Election and Democratization Study, 2009–2012 (III): The
Survey of the Presidential and Legislative Elections, 2012 (TEDS2012).”
The principal investigator was Professor Yun-han Chu for TEDS 2012.
The coordinator of the multiyear-project TEDS is Professor Chi Huang
(National Chenchi University). More information is available on the
TEDS website (http://www.tedsnet.org). These survey data provide
information about the Taiwan citizens’ evaluation of presidential
candidates.

3. Before 1996, instead of being selected by direct election, presidents of

the Republic of China were elected by the members of the National
Assembly. Lee served as president in this way from 1990 to 1996. Lee
originally considered that the selection of the national leader was the duty
of the National Assembly, but he then changed his mind to promote
direction elections for the presidency.

4. Although Lin ran as an independent, he and his running mate, Hau Pei-
tsun, the former premier, were strongly endorsed by the New Party in the
1996 presidential election. Hence, Chen, in fact, was the only candidate
running for the election independently.

5. Lee won 54.0% of popular votes in the 1996 presidential elections;

Peng, Lin, and Chen received 21.1%, 14.9%, and 10.0%, respectively.

6. Formosa Magazine was a magazine created by certain anti-KMT

individuals. On December 10, 1979, a prodemocracy demonstration led by
members of Formosa Magazine took place in Kaohsiung and demanded
democracy in Taiwan. However, the demonstration was prohibited by the
KMT’s authoritarian regime and was suppressed by the police. The leaders
of the demonstration were arrested and accused of treason. The Formosa
Incident is regarded as a very important event in the history of Taiwan’s
democratization because of its huge impact on Taiwan society.

7. In the 2000 presidential election, Chen won 39.3% of popular vote. His
opponents, Soong and Lien, received 36.8% and 23.1%, respectively.

8. In the 2008 Taiwan presidential election, Ma won 58.4% of popular

vote whereas Hsieh received only 41.6%. Corruption in the Chen
administration and the economic recession were considered the main
factors contributing to Ma’s landslide victory.

9. Ma won 51.6% of popular vote, whereas Tsai and Soong received only
45.6% and 2.8%, respectively.
References

Baker, David C., Adam B. Lawrence, and Margit Tavits. 2006. “Partisanship and
the Dynamics of Candidate Centered Politics in American Presidential
Nominations.” Electoral Studies 25: 599–610.

Campbell, Angus, Philip E. Converse, Warren E. Miller, and Donald Stokes.

1960. The American Voter. New York: Wiley.

Page 196 → Chang, Ching-ching. 2010. “Political Advertising in Taiwan’s

Elections and Perception of Issue/Trait Ownership” [in Chinese]. Chuanbo
yu Shehui Xuekan [Communication and Society] 11: 31–69.

Cheng, Su-feng, Lu-huei Chen, and Jia-wei Liu. 2005. “The Candidate Factor in
Taiwan’s 2004 Presidential Election” [in Chinese]. Taiwan Minzhu Jikan
[Taiwan Democracy Quarterly] 2 (2): 31–70.

Chu, Yun-han. 2011. “Taiwan’s Election and Democratization Study, 2009–

2012 (III): The Survey of the Presidential and Legislative Elections, 2012
(TEDS2012).” NSC 100–2420-H-002–030. Taipei: National Science
Council.

Flanigan, William H., and Nancy H. Zingale. 1998. Political Behavior of the
American Electorate. Washington, DC: CQ Press.

Hawang, Shiow-duan. 1996. “The Importance of Candidate Images and

Capabilities in Presidential Election” [in Chinese]. Xuanju Yanjiu [Journal of
Electoral Studies] 3 (1): 103–35.
Hawang, Shiow-duan. 2003. “Taiwan’s Election and Democratization Study
2002–2004 (III): The Presidential Election, 2004 (TEDS 2004P).” NSC 92-
2420-H-031–004. Taipei: National Science Council.

Hawang, Shiow-duan. 2005. “Candidate Image, Feeling Thermometer, and

Presidential Voting Behavior” [in Chinese]. Taiwan Minzhu Jikan [Taiwan
Democracy Quarterly] 2 (4): 1–30.

Hayes, Danny. 2005. “Candidate Qualities through a Partisan Lens: A Theory

of Trait Ownership.” American Journal of Political Science 49 (4): 908–23.

Hsieh, John Fu-sheng. 1995. An Interdisciplinary Study of Voting Behavior in the

1996 Presidential Election. Taipei: National Science Council.

Hsieh, John Fu-sheng, and Emerson M. S. Niou. 1996. “Salient Issues in

Taiwan’s Electoral Politics.” Electoral Studies 15 (2): 219–35.

Kim, Young-mie, and Kelly Garrett. 2012. “On-Line and Memory-Based:

Revisiting the Relationship between Candidate Evaluation Processing
Models.” Political Behavior 34 (2): 345–68.

Kinder, R. Donald, Mark D. Peters, Robert P. Abelson, and Susan T. Fiske.

1980. “Presidential Prototypes.” Political Behavior 2 (4): 315–36.

Lewis-Back, Michael S., William G. Jacoby, Helmut Norpoth, and Herbert F.

Weisberg. 2008. The American Voter Revisited. Ann Arbor: University of
Michigan Press.
Liang, Shin-wu. 1994. “Research on the 1994 Taipei Mayor Election: The
Testing of the Predictive Model-Candidate’s Image Indicators” [in Chinese].
Xuanju Yanjiu [Journal of Electoral Studies] 1 (2): 97–130.

Lin, Tse-min, Yun-han Chu, and Melvin J. Hinich. 1996. “Conflict

Displacement and Regime Transition in Taiwan: A Spatial Analysis.” World
Politics 48 (4): 453–81.

Lodge, Milton, Kathleen M. McGraw, and Patrick Stroh. 1989. “An Impression-
Driven Model of Candidate Evaluation.” American Political Science Review 83
(2): 399–419.

Merolla, Jennifer L., and Elizabeth J. Zechmeister. 2009. “Terrorist Threat,

Leadership, and the Vote: Evidence from Three Experiments.” Political
Behavior 31 (4): 575–601.

Miller, Arthur H., Martin P. Wattenberg, and Oksana Malanchuk. 1986.

“Schematic Assessments of Presidential Candidates.” American Political
Science Review 80 (2): 521–40.

Page 197 → Miller, Warren E., and J. Merrill Shanks. 1996. The New American
Voter. Cambridge: Harvard University Press.

Niemi, Richard G., and Herbert F. Weisberg. 1993. Classics in Voting Behavior.
Washington, DC: CQ Press.

Niemi, Richard G., and Herbert F. Weisberg. 2001. Controversies in Voting

Behavior. Washington, DC: CQ Press.
Petrocik, John R. 1996. “Issue Ownership in Presidential Elections, with a 1980
Case Study.” American Journal of Political Science 40 (3): 825–50.

Rahn, Wendy M., John H. Aldrich, John L. Sullivan, and Eugene Borgida. 1990.
“A Social-Cognitive Model of Political Candidate Appraisal.” In Information
and Democratic Processes, ed. John Ferejohn and James Kuklinski, 136–59.
Urbana: University of Illinois Press.

Sheng, Shing-yuan. 2007. “Issue, Political Cleavage and Party Competition in

Taiwan: From the Angles of the Elites and the Public.” Presented at the
Annual Meeting of the American Political Science Association, Chicago,
August 30–September 2.

Sheng, Shing-yuan. 2013. “Issues, Party Performance and Voters’ Voting

Behavior” [in Chinese]. In 2012 nian zongtong yu lifaweiyuan xuanju: Bianqian
yu yanxu [2012 presidential and legislative elections: Changes and
continuation], ed. Lu-huei Chen, 203–27. Taiwan: Wunan Press.

Shyu, Huo-yan. 1995. “‘Lee Teng-hui Complex’ and Voting Behavior: A

Psycho-political Analysis of Vote Choice in the 1994 Taiwan’s
Gubernatorial and Taipei Mayoral Elections” [in Chinese]. Xuanju Yanjiu
[Journal of Electoral Studies] 2 (2): 1–36.

Shyu, Huo-yan. 1998. “The Political Psychology of Lee Teng-hui Complex and
Its Effects on Voting Behavior” [in Chinese]. Xuanju Yanjiu [Journal of
Electoral Studies] 5 (2): 35–71.
Tverdova, Yulia V. 2011. “Follow the Party or Follow the Leaders? Candidate
Evaluations, Party Evaluation, and Macropolitical Context.” In Citizens,
Context, and Choice: How Context Shapes Citizens’ Electoral Choices. ed. Russell J.
Dalton and Christopher J. Anderson. New York: Oxford University Press,
126–48.

Wang, Hung-chung. 2012. “The Impact of Political Awareness and Elite

Discourse on Taiwanese Issue Positions: A Simple Test of Zaller’s
Mainstream Effect and Polarization Effect” [in Chinese]. Taiwan Minzhu
Jikan [Taiwan Democracy Quarterly] 9 (2): 71–124.

Wattenberg, Martin P. 1991. The Rise of Candidate-Centered Politics: Presidential

Elections of the 1980s. Cambridge: Harvard University Press.

Yu, Ching-hsin. 2003. “Exploring the Electorate’s Ideal Candidate: The Case of
the 2000 Presidential Election in Taiwan” [in Chinese]. Dongwu Zhengzhi
Xuebao [Soochow Journal of Political Science] 17: 93–120

Zaller, John R. 1992. The Nature and Origins of Mass Opinion. New York:
Cambridge University Press.
Page 198 →
 Chapter 9

Political Left and Right in Taiwan

Yi-ching Hsiao, Su-feng Cheng, and Christopher H. Achen

The left-right ideological dimension is an important conceptual tool for

understanding most European democratic countries and their former colonies,
such as those in North and South America. Party competition, the electorate’s
voting decisions, and governmental policy making can all be described in that
framework (for example, Barnes 1971; Bartle 1998; Dalton 2008; Dalton and
Tanaka 2007; Erikson, Wright, and Mclver 1993; Norris 2004, 97–125; Potrafke
2009).1 In its mathematical form, the “spatial model” of left-right voting has
been a favorite of theorists since Hotelling (1929) and before. Of course,
citizens in Western countries vary in how well they understand the dimension
(Stokes 1962; Converse 1964; Converse and Pierce 1986, 127–29; Fuchs and
Klingemann 1989; Inglehart 1990; Lewis-Beck and Chlarson 2002).
Nevertheless, political elites, scholars, and journalists make ready use of it to
describe their national politics.

The ubiquity of “left” and “right” in the elite discourse of their countries has
led some Western scholars to imagine that, in some form or another, those
terms must be meaningful political concepts in virtually every country. Thus
Sigelman and Yough (1978, 356) write that “party systems throughout the
world can meaningfully be profiled in terms of polarization along the left-right
continuum.” Similarly, Converse and Pierce (1986, 112) say, “This currency of
‘left,’ ‘center,’ and ‘right’ has of course been widely exported, and is a
commonplace for politically sophisticated observers around the world.”

This same logic is embedded in the Comparative Study of Electoral Systems

(CSES) international surveys, in which every participating country is required
to ask the following question:2

Page 199 → In politics people sometimes talk of left and right. Where would you place yourself on a
scale from 0 to 10, where 0 means the left and 10 means the right?

In Taiwan, however, left-right language is simply not used to describe the

current party system—not by ordinary people, not by journalists, not by
politicians, and not by Taiwan scholars. Asking them about it is like asking
them about sharia law or dancing the flamenco—cultural concepts that are
prominent elsewhere but not in Taiwan.

Since they lack the appropriate political context, how do Taiwan citizens
answer the CSES question? What do they understand by political “left” and
“right”? Scholars have occasionally remarked on anomalies in the use of left-
right language in Taiwan (for example, Chen 2003), but no one has focused
explicitly and in detail on how Taiwan citizens perceive “left” and “right” in
politics. The purpose of this chapter is to do so. We begin by reviewing the use
of left-right language in Western democracies. Then we proceed to the Taiwan
case.

The Concept of a Left-Right Dimension

The political concepts of left and right originated during the French
Revolution two centuries ago, when the more radical supporters of the
Revolution sat on the left in the Estates General, with their ideological
opponents on the right. Thus, from its beginnings the left-right distinction in
the West reflected ideological divisions over tradition and hierarchy in society.
“By left we shall mean advocating social change in the direction of greater
equality—political, economic, or social; by right we shall mean supporting a
traditional more or less hierarchical social order, and opposing change toward
greater equality” (Lipset et al. 1954, 1135).

In the modern era, the left-right dimension has referred primarily to

differences in the desired degree of government intervention in both society
and economy.3 Thus Laver and Hunt (1992, 12) write that

the left pole has in general become associated with policies designed to bring about the redistribution
of resources from those with more to those with less, and with the promotion of social rights that
apply to groups of individuals taken as a whole even at the expense of individual members of those
groups. The right pole has become associated with the promotion of individual rights, including the
right Page 200 → not to have personal resources expropriated for redistribution by the state, even at
the expense of social inequality and of poverty among worse off social groups.

Scholars interested in social class issues in politics often adhere closely to the
latter meaning. Thus Jansen, Evans, and De Graaf (2013, 54) say that “we
construct a left-right party position based on economic and welfare policy
issues,” and they explicitly set aside broader definitions proposed by other
scholars.

More loosely, other historical cleavages typical of Western societies (Lipset

and Rokkan 1967) have sometimes been subsumed under the left-right rubric
when they happened to line up with the views of left and right political parties.
Thus debates over divorce laws, abortion, supranational integration, and many
other issues are given “left” and “right” interpretations (Dalton 2012;
Zechmeister 2006; Zechmeister and Corral 2013). As Inglehart (1990, 292)
argues, “The Left-Right image is an oversimplification, but an almost inevitable
one, which in the long run tends to assimilate all important issues.” Hence the
left-right dimension is sometimes described as a kind of “super-issue,”
especially in Europe.

Issues unrelated to governmental intervention in the market or in society have

no persistent left-right meaning, however. There is nothing left or right about
disputes among ethnic, racial, or religious groups, for instance. Similarly,
neither the left nor the right has a monopoly on forceful assertions of
nationalism. Across Europe, conservative and socialist parties are found on
both sides of the debate over ceding some national sovereignty to the
European Union.

Particular leftist parties may adopt certain social views in a particular

historical period, of course, as may rightist parties. However, history
demonstrates that the issue packages that seem so coherent and inevitable to
partisans at the time have often varied dramatically in different times and
places. In practice, major party platforms are strategic documents cobbled
together to balance party factions and meet short-term electoral needs. They
are not ideologically coherent statements of a political vision (see, for example,
Bawn et al. 2012).

Thus we expect that when left-right language is extended beyond its central
modern meaning, all sorts of jumbled statistical patterns will result. That is
precisely what scholars have found (Fuchs and Klingemann 2009). Zechmeister
(2010) even finds some reversed signs for the correlation between economic
views and left-right self-descriptions in Latin American surveys Page 201

→ because “left” and “right” are being used to mean something else in certain
countries. “It is indeed a fact,” Zechmeister (2015, 199) writes, “that the
political significance of the left-right semantics varies across countries, across
time, and even across subgroups of a population.”

All these countries, however odd or broad the meaning they give to left and
right, agree in one respect: in their party systems, “left” and “right” are
meaningful political terms. The voters may understand them to a greater or
lesser degree, but the words themselves are meaningful. What scholars have not
discussed much at all, however, are countries in which those words are not
used in electoral politics, so that the concepts “left” and “right” have no
application to the party system. It is to such a case that we now turn.

Left and Right in Taiwan

Knowledgeable observers (Cheng and Hsu 1996; Rigger 2001, 39–41) are
agreed that for good historical reasons, conventional left-right issues do not
consistently differentiate the two main Taiwan parties, the Kuomintang and the
Democratic Progressive Party. According to Sheng and Liao’s longitudinal
study (see chapter 5 of this volume), the DPP has often been seen as somewhat
more favorable to “environmental protection,” to “reform,” and to “social
welfare programs,” leading some foreign observers to think of it as the left
party. But in practice, across issues such as pensions and medical care
programs, neither of the two parties has been consistently on the left or the
right. Early in his term in office, DPP president Chen Shui-bian set aside many
of his party’s social welfare promises in favor of promoting economic growth, a
typical right-wing choice. As Fell (2012, 199) remarks, “As with environmental
issues, welfare is not a core ideological issue for the party, thus could be
sacrificed.”

Typically, the two main Taiwan parties are flexible, low intensity, and
opportunistic on social welfare policies, differentiating themselves instead
along national identity lines instead, as we have seen repeatedly in this book
and as previous observers have noted (Cheng and Hsu 1996; Fell 2005b, chap.
4; Fell 2008, 69). Thus, elite politics in Taiwan is organized differently than in
most Western countries, and the parties have little incentive to use left-right
language in explaining themselves to voters.

Unsurprisingly, then, the left-right dimension does not predict voting in

Taiwan. Norris (2004, 110–11) utilized the Comparative Study of Electoral
Systems dataset to explore citizens’ voting decisions in a variety of
democracies Page 202 → and found that respondents’ left-right position was
significantly correlated with their voting decision—with the exception of
Taiwan and Belarus. All this is quite different from most Western countries,
where, in spite of considerable noise and misunderstanding by many in the
population, on average the party placements make reasonable sense, and where
individual citizens’ self-placements correlate at least fairly well with their voting
decisions (for example, on France, see Converse 1966 and Fleury and Lewis-
Beck 1993; more generally, Norris 2004, 110–11).

Left-right semantics do not predict the vote in Taiwan because, as we have

seen, those words are not used in the political culture to describe party
differences. Survey respondents are forced to guess their meaning. Thus, Chen
(2003) found that only about half of Taiwan citizens were able to specify their
position on a left-right dimension, a much higher failure rate than in most
democracies. Moreover, even for those who did place themselves, Chen’s study
of their unconventional responses led him to question whether respondents
really understood what “left” and “right” meant. In the same way, Jou (2010,
373) encountered very low left-right cognition in his study of Taiwan citizens:
fewer than 50% could place themselves on the scale. Taiwan’s left-right
placement rate was the lowest among 35 countries in CSES cases studied by
Russell Dalton, and the only country under 50 percent (Dalton 2011, 107).

The evidence that Taiwan voters do not understand left and right language is
strengthened when one looks at center self-placements by voters—a 5 on the
0–10 point scale. As Converse and Pierce (1986, 128–29) note, respondents
who choose the midpoint are often poorly informed and simply trying to
appear helpful to the interviewer: center placement “is an obvious selection for
a person who is neutral, uncommitted, and even thoroughly indifferent to or
ignorant about this generic axis of dispute.” (Similarly, see Lambert 1983 and
Ogmundson 1979.) In two different studies, Jou (2010, 373) found very high
center placement in Taiwan—among the minority of voters who could place
themselves at all, more than half chose the center position. Thus, altogether,
more than three quarters of the Taiwan respondents chose either the neutral
position or no position at all. The same finding appears in the 2012 TEDS
survey: 78 percent of Taiwan respondents were either neutral or
uncomprehending when asked the left-right question.4
 All these Taiwan anomalies raise several questions. Is there any sense of the
words “left” and “right,” conventional or not, that has meaning for the voters?
What do ordinary Taiwan citizens mean when they are asked the meaning of
those words? And how do they place the parties on that dimension?

Page 203 →

Taiwan Politics and the Cultural Connotations of “Left”

and “Right”

As we have noted, a conventional left-right dimension seems to play little role

in the vote choices of Taiwan’s voters. In this respect, the citizenry simply
reflect the nature of Taiwan politics. The voters see real differences between
the parties on the national identity issue, but few on secondary issues like social
welfare (Chu and Lin 1996, 92–95). As Sheng and Liao showed in chapter 5,
since the beginning of democratization in the 1980s, citizen preferences on
such issues as Taiwan independence vs. reunification with China,
environmental protection vs. economic development, social welfare vs. low
taxes, and reform vs. social stability have influenced party preferences and
voting decisions to some degree (Hsieh, Niou, and Lin 1995; Sheng and Chen
2003; Tsai 2008; Wang 2001; Wang 2003). But the first of these—independence
vs. reunification—is the most powerful issue, not only in locating the main two
parties on the political spectrum, but also in determining voter choice.
National identity concerns stemming from “the China factor” have been the
main political cleavage to discriminate between the Pan-Blue camp and the
Pan-Green camp (as chapter 4 reveals). However, national identity is not itself
conventionally left-right in character, and those words are not used to describe
the issue in Taiwan.5

Not only does left-right thinking fit Taiwan’s current elite and electoral
politics poorly, but in recent history that language was actively employed to
characterize something else—Taiwan’s foreign policy disagreements with
mainland China. Before democratization in the late 1980s, the Kuomintang
Party viewed the “leftist” Chinese Communist Party as the mortal enemy. With
its monopoly on political communications, the KMT made every effort to
suppress “left” views. Taiwan people were taught that the “left-side” was evil.
Expressing sympathy for the left was a form of rebellion. The KMT
emphasized that it was the “right” party. An element of negative evaluation still
attaches to “left-side” political views in Taiwan.

These connotations of “left” and “right” are enhanced by the two Chinese
dialects most used in Taiwan, Taiwanese (spoken by a majority of citizens) and
Mandarin (the language of instruction in schools). Just as some respondents in
English-speaking countries consider the political “right” to mean “correct” or
“in the right,” so also in the Taiwanese dialect the same word “right” is used to
mean both “the opposite of left” and also “correct” or “true.” “Left” in
Taiwanese connotes “bad” in some way. The heritage from the authoritarian
period enhances this identification.

Page 204 → An equally consequential factor for left-right usage in Taiwan is

that, apart from all political overtones, “left” in the Mandarin dialect connotes
deviousness, unorthodoxy, or heresy.6 The inference extends even to left-
handed people, who are often considered “alternative” or nonmainstream. In
everyday Taiwan life, the right side when walking (or the right-hand seat when
sitting) is reserved for elders, honored guests, or respected citizens.7 The
implication, then, is that one’s favorite political party should be placed on the
right and the disliked party on the left.

Thus we expect that placing Taiwan’s parties on a conventional left-right scale

will be difficult for ordinary citizens because that dimension is nearly irrelevant
for party choice in Taiwan. When respondents are forced to place the parties or
to define “left” and “right,” they will often fall back on other meanings familiar
from local culture, such as “Communist” vs. “anti-Communist,” “bad” vs.
“good,” or “wrong” vs. “correct.” To assess these propositions, we use both
cognitive interviews and public opinion surveys. The cognitive interview data
come from a project called “A Study of Major Political Identification Concepts
of the Taiwan Public.”8 In this project, which was carried out in 2001, 50
respondents were asked to define the left-right dimension in politics, and then
to place themselves and the major Taiwan political parties on the left-right
spectrum. The interviewers gave no examples or cues of how this task was to
be done. All the respondents had to define the concept and the positions
according to their own understandings.9

We also employ opinion survey data collected by Taiwan’s Election and

Democratization Study after the 2001 legislative election and the 2008
presidential election (hereafter TEDS2001 and TEDS2008P). The sampling
population is adult citizens in Taiwan. The number of successful interviews
was 2,022 and 1,905, respectively. In TEDS, a 0–10 scale was utilized for the
respondents’ placement of themselves and the major parties, as well as for
answers to issue questions.10
 Citizens’ Perceptions of “Left” and “Right”

We begin with a discussion of the cognitive interviews to give a sense of what

Taiwan respondents mean by “left” and “right.” In the end, all fifty
respondents managed some sort of definition, but the task was not easy for
most of them. When asked to give a definition, 18 respondents (36%) began by
asking what the question meant. Another five respondents (10%) said that the
left-right dimension was politically irrelevant in Taiwan. For example (our
translations):

Page 205 → I don’t think the concept of left-right exists in Taiwan. (No. 01)

I just don’t get it . . . basically, I think the concept is meaningless in Taiwan. (No. 41)

Thus nearly half the respondents could not or did not use the left-right
distinction in their thinking about Taiwan politics.

When pressed to a definition, respondents’ answers varied widely, and most

did not fit the customary meaning of left-right in Western democracies. We
have attempted to catalog the respondents’ answers into six different
categories, with the “left” answer listed first: liberalism vs. conservatism (4%),
doves vs. hawks (18%), Communism vs. democracy (18%), Taiwan
independence vs. reunification with China (22%), bad vs. good (18%), and
ruling party vs. opposition party (20%). Among the six different definitions,
the first category clearly conforms to the definition of left-right in Western
democracies, and the second might be generously interpreted to do so as well.
But those two groups comprise fewer than 25 percent of the respondents. The
remaining four categories reflect confusions of various kinds.
 To convey a sense of how the interviewees in the nonstandard categories
express themselves, we give a sample of their responses. We begin with the
third category of respondents, those who regard all democracies as “right.”

Communism vs. Democracy

Basically I think the left-wingers are closer to socialism, so socialists are counted as left. Closer to democracy and liberty
are the right-wingers. From my point of view, in present-day Taiwan, no matter whether it’s DDP, KMT, PFP, or
TSU [abbreviations of the Taiwan parties], they are basically on the democracy side, so all the Taiwanese parties
belong to the right. (No. 13)

The left reminds me of the Communist Party. . . . the left was referred to as the Communist Party . . . The right is the
more democratic party. (No.29)

I always think Communism vs. democracy. . . . It seems to me that the right side is democracy and the left side is
communism. I don’t really get it. (No.42)

Page 206 → The fourth category reflects the main political cleavage in Taiwan,
the orientation toward China’s claim of ownership of Taiwan. As noted above,
this policy dimension is not about economics or social class, but is rather a
dispute between two versions of national identity. Thus it is not conventionally
left-right.

Reunification vs. Independence

I don’t have the concept of the left-right. . . . it is made by [other] people. . . . The left in politics is [Taiwan]
independence, and the right is unification [with China]. (No. 06)

Generally speaking, the definition of the extreme right is strongly supporting unification, while the extreme left is strongly
supporting independence. This is how I see it. (No. 34)
 Finally, the last two categories, nearly 40 percent of the interviewees,
completely misunderstand left-right categories. The first group takes their cue
from the connotation in Mandarin and Taiwanese of “left” as deviant or
subpar.

Good vs. Bad

I feel the left seems to be negative . . . and the right is more positive. . . . I feel it is good vs. bad. . . . Because I think the
left means heresy in our old saying, that unorthodox ways are “left ways” [in Chinese]. . . . If I use “the left” to describe
something bad, then I think the opposite side of it should be something better. (No. 07)

Left-right in politics means who does things right. Those who do the right things are the rights, while those who do the
wrong things are the lefts. (No. 31)

The last group of respondents associates “right” with the ruling party,
regardless of its ideology.

Page 207 →

Ruling Party vs. Opposition Party

The left and the right? I think to say it in a simple way, it is the ruling party and the non-ruling party . . . I think the
left is the non-ruling party and the right is the ruling party. (No. 33)

Ruling party is counted as the right, and generally speaking the left is the opposition party. (No.08)

Note that at the time of the interviews, the presidency was held by the pro-
independence DPP. Categorizing them as “right,” as these respondents do,
contradicts the categorization given by those respondents who focused on
reunification vs. independence and thereby called the DPP “left.”
 In summary, these cognitive interviews display the great range of
interpretation of left and right among Taiwan’s citizens. Only a few use
European-derived interpretations to structure their dimensional thinking.
Some impose idiosyncratic understandings. Many do not make use of the
concept at all. Thus imposing Western left-right frameworks on Taiwan
respondents violates their understanding of the island’s politics and distorts the
analysis of elections there.

Survey Evidence

These conclusions are strengthened when we turn to nationwide opinion

survey data. We first explore how citizens locate themselves and the major
parties on the left-right dimension and on a variety of policy issues, including
Taiwan independence vs. reunification with China, environmental protection
vs. economic development, promoting social welfare vs. keeping taxes down,
and large-scale reform vs. social stability. The latter three items are
conventionally left-right in character, especially the social welfare question,
while the first issue concerns competing national identities and has no left-right
ideological content, as we have discussed. In each case, the scale runs from 0–
10 (question wordings are given in the appendix). These issues were discussed
by Sheng and Liao in chapter 5 to validate the importance of Taiwan’s principal
political cleavage. Here our purpose is different: we explore the correlations
between the left-right dimension and these four policy issue to verify that left-
right is an inapplicable instrument for interpreting Taiwan politics. We begin
with nonresponse rates.
 Fig. 9.1. Nonresponse rates for left-right self-placement and other issues. Data Sources: TEDS
2001 and TEDS 2008P.

Page 208 → Figure 9.1 reveals that across a variety of issues, the nonresponse
rate for the left-right question is by far the highest. In 2001 and in 2008, about
half the citizens could not respond when asked where they placed themselves
on the left-right dimension. By comparison, only about 10 percent failed to
provide their own opinions on standard Taiwan political issues. Many citizens
give middle scores for their position on the ideological spectrum, probably
because they are behaving cautiously in a task they did not fully understand, as
discussed earlier.11 Further investigation showed that, as expected,
knowledgeable or highly partisan respondents were more often able to give an
answer to the left-right question (as in Converse and Valen 1971, 131), while
party preference made no difference. Nevertheless, even among those well-
informed respondents who answered all five of the political knowledge
questions or missed only one, almost 40 percent could not place themselves on
a left-right scale in 2001. Failure rates were considerably higher among those
with less understanding. Altogether, for half of all Taiwan citizens, there is no
interpretation of left-right language that makes enough sense to allow them to
place themselves on the scale, and, as we have already noted, the other half
often manage the task only with idiosyncratic definitions of “left” and “right.”

The conclusion is much the same when we examine placement of the Page 209

→ two principal political parties on the same 0–10 scale. For each policy issue,
figure 9.2 shows the difference of the mean DPP placement from the mean
KMT placement, with positive numbers indicating that the DPP is closer to the
first option in each issue choice. For example, on the issue of independence vs.
reunification, a positive difference means that the DPP is seen (correctly) as
closer to the independence position. As figure 9.1 demonstrated, some
respondents could not place the parties on each issue. Hence the comparative
placements in figure 9.2 are based solely on those respondents who did so.

Fig. 9.2. Difference in average respondent placement of DPP vs. KMT (1–10 scale), Data Sources:
TEDS 2001 and TEDS 2008P
 As in Sheng and Liao’s results in chapter 5, figure 9.2 shows that in both 2001
and 2008 the DPP is considered closer than the KMT to the positions of
“independence,” “environment,” and “reform,” with the party difference by far
the largest on independence vs. reunification. Better informed and more
partisan respondents perceived somewhat larger party differences (not Page 210

→ shown but available from the authors on request). These findings are
completely consistent with contemporary interpretations of party competition
and elite political cleavages in Taiwan, as we noted above.

This optimistic view of the respondents is tempered by the one inconsistency

between 2001 and 2008, however, which occurs on the clearest and most
conventional left-right issue, social welfare vs. keeping taxes down. In 2001, the
DPP was considered closer to “promoting social welfare” than the KMT, but
this ordering was reversed in 2008. It may be that 2008 respondents were
simply expressing their more pessimistic view of the Chen Shui-bian
administration after its eight years in office, during which he deemphasized
social welfare, as we have seen. In any event, the instability and vanishingly
small party differences on this issue reinforce the point that Western notions of
“left” and “right” do not distinguish the two principal Taiwan parties.

A second and more striking anomaly is that the respondents give their
favorite party a more rightward (higher) score and their disliked party a more
leftward (lower) score. Figures 9.3 and 9.4 show this effect for 2001 and 2008,
respectively. The effect is visible in both figures, but is particularly dramatic in
2001. In that year, those respondents who liked the DPP (the proenvironment,
“proreform” major party) moved its average placement so far to the right that
it wound up to the right of the KMT. Here again is evidence that many Taiwan
citizens consider their favorite party as “right” and their disliked party as
“left,” regardless of its actual policy views.

The argument that “left” and “right” do not have conventional meanings in
Taiwan is further strengthened when left-right placement is correlated with
issue positions on the four items mentioned above. If ”left-right” in Taiwan
captured standard Western notions about the role of government in society, the
correlations should be strong and positive with the three domestic policy
issues, particularly so for the social welfare vs. low taxes issue, but small or
zero with the independence-reunification question. However, figure 9.5 reveals
that in both survey years, all the correlations are very small—none larger than
0.130. Worse yet, in both years, left-right position is slightly negatively correlated
with attitudes toward social welfare, just the reverse of what is required for
conventional ideological meaningfulness.12

Part of the explanation for the reversed correlation may lie in the survey
measurements. In conducting the TEDS questionnaire, interviewers show the
respondents cards with a 0–10 scale. The “0” is located the left side, signifying
“Taiwan independence,” “environmental protection,” “ lower taxes,” and
“large-scale reform.” On the other side, the “10” signifies “reunification with
China,” “economic development,” “ promoting social welfare,” and Page 212

→ “social stability.” (See figure 9.6) Thus, some respondents who did not
understand the meaning of left-right may simply have regarded all the issue
positions on the left side of the card as “left,” and all those on the right side as
“right.” Since among all four issues, only the “social welfare” question has the
“left” answer on the right-hand side of the card and the “right” answer on the
left (the bold-faced answers in the list above), this may account for the weak
(and slightly negative) correlation between it and the respondent’s left-right
position. Once again, this suggests that the left-right dimension is little
understood by Taiwan citizens and little related to their policy views.

Page 211 →

Fig. 9.3. Placement of the parties on left-right by favorability toward the DPP (2001). Data Source:
TEDS 2001.
Fig. 9.4. Placement of the parties on left-right by favorability toward the DPP (2008). Data
Source: TEDS 2008P.

Fig. 9.5. Correlation of citizens’ left-right position with issues. Data Sources: TEDS 2001 and
TEDS 2008P.
 Hard-core devotees of the conventional wisdom may yet have one final
objection. “All right,” they may say, “the left-right orientation is weak in
Taiwan. Previous scholars have found that it does not predict voting. But
perhaps by 2012, after several decades of democratization, the result is
different. Doesn’t everything in politics turn into left-right eventually?”

The answer, for the record, is no. As we have seen, a great many Taiwan
respondents have to be discarded to assess the relationship of left-right
position to the Pan-Blue vs. Pan-Green vote because they have no idea what
the left-right question means. But even in that heavily truncated sample, no
trace of causal importance appears. To give left-right orientation every chance,
we did not load up the explanatory equations with many different noisy
measures of related opinions, a tactic sure to reduce them all to statistical
insignificance. Nor did we control for Michigan-style party identification
questions, which have an overwhelmingly powerful impact in Taiwan (see Page

213 → Chapter 12). But even with all these biases in its favor, the left-right
position failed to show much sign of explanatory life. With the scale set to a
range of 0–1 and controlling only for dummy variables indicating the
respondent’s party preference, the probit coefficient was just .1 and far from
statistical significance. Taken at face value, that coefficient would imply at best
a 3 percentage point impact on a little more than half the sample (the
remaining group having zero impact because they do not recognize the terms
“left” and “right”). But even a very modest effect of that size is far from
reliable statistically.
 Fig. 9.6. Content and direction on TEDS show card

Some indication of why the left-right variable fails is given by a close look at
the party identifiers.13 Just 17 percent of KMT supporters placed themselves on
the far right (a score of 10). But fully 14 percent of supporters of the more
radical DPP respondents placed themselves there. Indeed, 58 percent of the
DPP sample placed themselves at one or another position on “the right”
(scores 6–10). More dramatically, there were just three supporters of the
strongly pro-independence TSU party in the sample, but they all placed
themselves on the right. Overall, 77 percent of the sample placed themselves on
the right.

As we have seen in Taiwan, “right” often means “correct.” No wonder, then,

that the left-right variable adds almost nothing to explaining vote choices once
partisanship is controlled. Once we know which parties the respondents belong
to, knowing that most of them also consider themselves “correct” adds no new
information.

Conclusion
Some readers of early versions of this chapter felt that our central point was
already well known, since many scholars have shown that the meanings of Page

214 → “left” and “right” differ across countries. Hence they felt that the Taiwan
case brings nothing new. In the light of how the topic is treated in much
previous literature, such a misconception is entirely understandable. But it
misses our point entirely.

In the conventional view, the left-right dimension is a cultural near-universal.

Whatever its meaning, every country should exhibit some form of left vs. right
in politics. When a party system fails, scholars often blame the parties: too
much clientelism, too little policy clarity, too little polarization, too little time
for the voters to learn—all these are put forward as causes when left-right
meanings are confused or impotent in voting decisions. (The European case is
treated in van der Eijk et al. 2005, 177–80 and Berglund et al. 2005, 116–22; see
Zechmeister 2015 for the Latin American counterpart.) If these party failures
were corrected, the argument goes, a strong left-right effect would make an
appearance. In Europe and Latin America and some other parts of the world
where that argument has been made, it is probably accurate. But to suppose
without close study that it applies everywhere would be reminiscent of those
American tourists in the 1950s who imagined that if only they spoke English
slowly and loudly enough, anyone around the world could understand them.

Our point is rather that research using left-right concepts may go seriously
wrong if the left-right dimension essentially does not exist in some countries.
For example, the idiosyncrasies of left-right language in Taiwan explain why
Dalton and Tanaka (2007) measured low party polarization in Taiwan, a result
that would surprise knowledgeable observers of Taiwan politics. As we have
seen, in Taiwan a majority of citizens cannot place either themselves or the
major parties on a left-right ideological spectrum. Even among those who can
do so, many appear to be guessing or using idiosyncratic definitions of “left”
and “right.” As we have also seen, many Taiwan citizens identify their favorite
party as “right” and consider the disliked party “left,” corresponding to the
Taiwanese or Mandarin connotations of “left” as “bad,” “devious,” or
“heretical.”

Now if many respondents favoring one major party put it on the right (“the
right side”) and the other party on the left (“the bad side”), while many
respondents favoring the other party do the reverse, the rights and lefts will
tend to average out in the mean placement of each party. Average party
placements will be pushed toward the center, making a highly polarized party
system appear convergent and consensual. But “low polarization” is simply
mistaken. As we have seen repeatedly in this volume, cross-Strait relations help
shape citizens’ national identity (chapter 3), determine people’s party
identification (chapter 4), set the core political cleavage for party Page 215

→ competition (chapter 5) and the party system (chapter 10), and thus are the
critical factor for vote choice (chapters 7, 9, and 12). In particular, the issue of
Taiwan independence has led to serious political conflicts and unusually bitter
party divisions since Taiwan’s democratization in the late 1980s and early 1990s,
including fistfights on the floor of the national legislature.14 Taiwan’s parties
are not ideologically similar, and polarization is not low. But the CSES survey
imposes left-right language on countries that do not use it, and scholars
thereby may be led astray.
 National identity issues are often poorly assimilated to left-right
categorizations. As Jou (2010, 371) insightfully phrased it, speaking of Taiwan,
“One may . . . hypothesize that an entrenched national identity cleavage leads
to the paradoxical scenario of greater polarization accompanied by lower left-
right identification.” It is precisely that sort of insight that is impossible to
grasp unless one breaks out of the notion that some version of left-right applies
everywhere.

The Taiwan voters’ difficulties in understanding left-right political concepts

are perfectly understandable. This volume’s theme is that party competition in
Taiwan is structured by the “China factor,” especially on the “independence vs.
reunification with China” issue, but this is clearly not a conventional left-right
dimension. Hence even at the political elite level, left-right distinctions are
generally irrelevant and a poor guide to sorting out the parties. Little wonder
that the voters do not use it.

This finding has important implications, not just for understanding Taiwan,
but for the study of other countries around the globe. The challenge of
comparative political research is the great diversity of political life and culture
in different countries. As electoral research becomes more truly international,
some Western concepts and frameworks will inevitably come under challenge
and be modified or set aside in many countries outside the West. We have
argued that “left-right” is one such example—a framework to be used where it
applies, but not elsewhere. Many countries, notably in Africa, have political
systems primarily shaped by racial, ethnic, or linguistic divisions, not class
conflict. Even in Western countries like Canada, with politics dominated by
religious and linguistic divisions, voters struggle with the left-right concept,
and some confuse “right” with “correct” (Ogmundson 1979, 800; Lambert
1983; Lambert et al. 1986). In Ireland, too, where divisions stemming from the
Civil War have defined the party system for a century, the left-right concept has
traditionally differentiated the main parties rather poorly, leaving the voters
confused by it as well (Sinnott 1995, 24–33, 74–78, 162; Marsh et al. 2008, 42).15
Left and right notions are rarely used in Africa either, apart from South Africa
Page 216 → (Jennifer Widner, personal communication). Taiwan provides an
insightful example of how much difference the absence of left-right language
may make in all these countries.

The “China factor” is central to understanding Taiwan politics. The parties

and the voters divide over this question: How is Taiwan’s nationhood to be
understood in light of China’s presence, its growing power, and its claims of
sovereignty over Taiwan? It is that political dimension that shapes both
Taiwan’s party system and the voters’ understanding of politics. Eurocentric
notions of “left” and “right” simply do not apply.

To understand Taiwan, one must stop seeing it through a European lens.

Conventional understandings of the importance of “left” and “right” in
politics have no claim to universality. Many countries are well described in that
way, but many others may not be, particularly in Asia and Africa. Here again,
the study of Taiwan rewards the scholar with a deeper understanding of
politics in many other places.

Appendix 9.A1.

Questionnaire Items and Operationalization of the Variables

Left-Right Ideology Question Wording:

In politics, sometimes people talk about the left and the right. This card lists
eleven positions from the left (0) to the right (10). Which position do you
occupy?

Taiwan Independence vs. Reunification with China Question Wording:

Sometimes people will talk about the question of Taiwan independence or

reunification with China. Some people say that Taiwan should declare
independence immediately. Other people say that Taiwan and China should
unify immediately. Other people have opinions between these two positions.
This card lists eleven positions from independence (0) to reunification (10).
Which position do you occupy?

Page 217 → Environmental Protection vs. Economic Development Question

Wording:

Regarding the question of economic development versus environmental

protection, some people in society emphasize environmental protection while
others emphasize economic development. On this card, the position that
emphasizes environmental protection is at 0 on a scale from 0 to 10, and the
position that emphasizes economic development is at 10. About where on this
scale does your own view lie?

Promoting Social Welfare vs. Lower Priority for Social Welfare/Increasing Tax
Question Wordings:
TEDS2001 On the question of social welfare, some people believe that people
should take care of themselves and the government should not get involved
while other people believe that the government should actively promote social
welfare and take care of all the people. This card lists eleven positions from
individuals should take care of themselves and the government should not get
involved (0) to the government should actively promote social welfare (10).
Which position do you occupy?

TEDS2008P Regarding the question of social welfare, some people believe that
the government should merely maintain the current system in order not to
increase people’s taxes. Other people believe that the government should
promote social welfare, even though it will lead to a tax increase. On this card,
the position that maintaining the current system is the most important thing is
at 0 on a scale from 0 to 10, and the position that promoting social welfare is
most important is at 10. About where on this scale does your own view lie?

Large-Scale Reform vs. Social Stability Wording:

Looking at Taiwan’s overall development, some people believe that large-scale

reform is the most important thing, even if it means sacrificing some social
stability. Other people believe that stability is most important and that reform
should not be allowed to affect social stability. On this card, the position that
large-scale reform is the most important thing is at 0 on a scale from 0 to 10,
and the position that social stability is most important is at 10. About where on
this scale does your own view lie?

Page 218 → Party Preference Question Wording:

Now we’d like to understand your opinions about each of the political parties.
If 0 means you dislike a party very much, and 10 means you like that party very
much, what number would you give the KMT? The DPP?

Notes

1. The cognitive interview data in this article come from a project entitled
“A Study in Major Political Identification Concepts of the Taiwan Public”
(NSC89–2414-H-004–022-SSS), whose principal investigator is Professor
I-chou Liu of the Political Science Department of National Chengchi
University. The survey data analyzed were collected by the Taiwan
Election and Democratization Studies, 2001: The Legislative Election
(TEDS2001) (NSC 90–2420-H-194–001) and 2008: The Presidential
Election (TEDS2008P) (NSC 96–2420-H-004–017). The coordinator of
the multiyear project TEDS is Professor Chi Huang (National Chengchi
University). Further information is available on the TEDS website
(http://www.tedsnet.org). T. J. Cheng gave us helpful comments. We also
thank Charles Witke for timely advice about Latin meanings. The authors
appreciate the assistance of each of these institutions and individuals.
However, the authors alone are responsible for the views expressed here.

2. In the early years of the CSES, no exceptions were allowed. It is now

possible to petition to be exempted if the question is meaningless in a
particular country.

3. More generally, perhaps, conservatism (the “right”) is the defense of the

established order. Historically, this meant a defense of hierarchy and a
preference for the status quo rather than for the untried and risky
proposals of reformers (Huntington 1957; Beer 1966, chap. 9). By
extension, “left” vs. “right” came to include divisions over change vs.
custom, reason vs. tradition, and other issues raised by the revolutionary
impulses of the eighteenth and nineteenth centuries. In practice,
conservatives usually defended the interests of those who were successful
and privileged within society. In the modern era of widespread
democratization, acceptance of capitalism, and rapid scientific innovation,
however, aristocratic views of society and defenses of stasis are out of
fashion. Contemporary success and privilege in the West derive primarily
from economic achievement within the capitalist order, and government
intervention is their greatest threat. Hence modern conservatism opposes
additional government intervention. Thus, in its central meaning, today’s
conservatism lacks the traditional emphasis on social hierarchy and
reverence for the past, but it is nevertheless the legitimate descendent of
its ideological ancestors.

4. Of course, some respondents may have had a legitimate thoughtful

position at the midpoint of the scale. But as the next paragraph shows,
many respondents at other scale positions were seriously confused. There
is no simple escape from the central finding that the overwhelming
majority of Taiwan respondents do not know what “left” and “right”
mean in their politics, nor should they.

5. Of course, an outside observer can call any issue “left-right.” Thus Fell
(2005a, 112) accurately remarks that “[some] analysts talk of a left and
right in Taiwan.” However, Page 219 → this occurs almost exclusively
among foreign scholars doing comparative work across many countries,
using cross-national datasets like the CSES that impose the same left-right
survey question on every country. It is quite rare for Taiwan scholars
themselves to mention the left-right as a description of contemporary
politics on the island, except to criticize it (Chen 2003).

6. English inherits a similar relationship from a now-dead language. In

Latin, “dexter” means the right side, from which English takes the word
“dexterous,” meaning “skillful.” The Latin word for “left” is “sinister,”
with overtones of bad omens from fortune tellers. The same word in
English has come to mean “evil” or “portending evil.”

7. One of us saw a Taiwan university official insist that a sign directing

visitors to a set of university offices be placed on the right-hand wall, even
though the offices were on the left side of the building. The reason given
was that the right side was “greater.”

8. The project is sponsored by National Science Council (Taiwan). The

principal investigator is Professor I-chou Liu at the Political Science
Department of National Chengchi University, who generously shared the
data.

9. These respondents were chosen to represent both genders and a variety

of ages and occupations. Their demographic profiles are available from
Su-feng Cheng.

10. For the question wordings, see appendix 9.A1.

11. Respondents were somewhat more responsive in 2008, raising the

possibility that experience with democracy is improving comprehension.
Of course, a comparison of two time points can be no more than
suggestive. However, similar gains in the coherence of political attitudes
over time have been noted in other new democracies (for example, Arian
and Shamir 1983, 150).

12. It is clear in figure 9.3 to figure 9.5 that Taiwan citizens locate party
positions on the left-right scale mainly on the basis of their party
preference, not the four traditional policy issues. This result also has been
validated by a multivariable model, but the model cannot be included in
the chapter due to space limitations. The model is available from Yi-ching
Hsiao on request.

13. Just under 61% of the sample placed themselves at a neutral 5 on the
0–10 scale, yet another reminder of the weak understanding of this
variable. Those respondents have been omitted in the percentage
calculations that follow.

14. Dalton and Tanaka (2007) recognize that the main axis of Taiwan
politics is the independence-reunification issue, not the European
versions of “left” and “right,” but they interpret their data as indicating
only a little differentiation between the two main parties on the national
identity issue.

15. Both Canada and Ireland have a social democratic party, so that
conventional left-right language plays some role in party politics.
However, both parties typically receive only modest proportions of the
vote, finishing well behind the two principal parties. Even in the special
circumstances of 2011, the best year ever for both, they finished in second
place, well behind the top party.
References

Arian, Asher, and Michal Shamir. 1983. “The Primarily Political Functions of
the Left-Right Continuum.” Comparative Politics 15 (2): 139–58.

Page 220 → Barnes, Samuel H. 1971. “Left, Right, and the Italian Voter.”
Comparative Political Studies 4: 157–75.

Bartle, John. 1998. “Left-Right Position Matters, but Does Social Class? Causal
Models of the 1992 British General Election.” British Journal of Political Science
28 (3): 501–29.

Bawn, Kathleen, Martin Cohen, David Karol, Seth Masket, Hans Noel, and
John Zaller. 2012. “A Theory of Political Parties: Groups, Policy Demands,
and Nominations in American Politics.” Perspectives on Politics 10 (3): 571–97.

Beer, Samuel H. 1966. British Politics in the Collectivist Age. New York: Knopf.

Berglund, Frode, Soren Holmberg, Hermann Schmitt, and Jacques Thomassen.

2005. “Party Identification and Party Choice.” In The European Voter: A
Comparative Study of Modern Democracies, ed. Jacques Thomassen, 106–24.
Oxford: Oxford University Press.

Chen, Wen-chun. 2003. “The Blue and the Green: The Political Ideologies of
the Mass Public in the 2000 Taiwan Presidential Election” [in Chinese].
Journal of Electoral Studies 10 (1): 41–80.

Cheng, Tun-jen, and Yung-ming Hsu. 1996. “Issue Structure, the DPP’s
Factionalism, and Party Realignment.” In Taiwan’s Electoral Politics and
Democratic Transition, ed. Hung-mao Tien, 137–73. Armonk, NY: M. E.
 Sharpe.

Chu, Yun-han, and Tse-min Lin. 1996. “The Process of Democratic

Consolidation in Taiwan: Social Cleavage, Electoral Competition, and the
Emerging Party System.” In Taiwan’s Electoral Politics and Democratic
Transition, ed. Hung-mao Tien, 79–104. Armonk, NY: M. E. Sharpe.

Converse, Philip E. 1964. “The Nature of Belief Systems in Mass Publics.” In

Ideology and Discontent, ed. David E. Apter, 206–61. Glencoe, IL: Free Press.

Converse, Philip E. 1966. ‘The Problem of Party Distances in Models of Voting

Change.” In The Electoral Process, ed. M. Kent Jennings and H. Zeigler, 175–
207. Englewood Cliffs, New Jersey: Prentice-Hall.

Converse, Philip E., and Henry Valen. 1971. “Dimensions of Cleavage and
Perceived Party Distances in Norwegian Voting.” Scandinavian Political
Studies 6 (6) : 107–52.

Converse, Philip E., and Roy Pierce. 1986. Political Representation in France.
Cambridge: Belknap Press of Harvard University Press.

Dalton, Russell J. 2008. “The Quantity and the Quality of Party Systems: Party
System Polarization, Its Measurement, and Its Consequences.” Comparative
Political Studies 4 (7): 899–920.

Dalton, Russell J. 2011. “Left-Right Orientations, Context, and Voting

Choices.” In Citizens, Context, and Choice, ed. Russell J. Dalton and
Christopher J. Anderson, 103–25. Oxford: Oxford University Press.
Dalton, Russell J. 2012. “Parties and Representative Government.” In Political
Parties and Democratic Linkage, ed. Russell Dalton, David Farrell, and Ian
McAllister. Oxford: Oxford University Press.

Dalton, Russell J., and Aiji Tanaka. 2007. “The Patterns of Party Polarization in
East Asia.” Journal of East Asian Studies 7 (2): 203–23.

Erikson, Robert S., Gerald C. Wright, and John P Mclver. 1993. Statehouse
Democracy: Page 221 → Public Opinion and Policy in the American States.
Cambridge: Cambridge University Press.

Fell, Dafydd. 2005a. “Measuring and Explaining Party Change in Taiwan:

1991–2004.” Journal of East Asian Studies 5: 105–33.

Fell, Dafydd. 2005b. Party Politics in Taiwan. Oxford: Routledge.

Fell, Dafydd. 2008. “Inter-Party Competition in Taiwan: Toward a New Party

System?” In Presidential Politics in Taiwan, ed. Steven M. Goldstein and Julian
Chang, 49–84. Norwalk, Connecticut: EastBridge.

Fell, Dafydd. 2012. Government and Politics in Taiwan. Oxford: Routledge.

Fleury, Christopher J., and Michael S. Lewis-Beck. 1993. “Anchoring the French
Voter: Ideology versus Party.” Journal of Politics 55 (4): 1100–1109.

Fuchs, Dieter, and Hans-Dieter Klingemann. 1989. “The Left-Right Schema.”

In Continuities in Political Action: A Longitudinal Study of Political Orientations in
Three Western Democracies, ed. M. Kent Jennings and Jan van Deth, 203–34.
New York: Walter de Gruyter.
Hotelling, Harold. 1929. “Stability in Competition.” Economic Journal 39 (153):
41–57.

Hsieh, John Fuh-sheng, Emerson M. S. Niou, and Huei-ping Lin. 1995. “Issue
Voting in the 1994 Gubernatorial and Mayoral Elections: An Application of
the Rational Choice Approach” [in Chinese]. Journal of Electoral Studies 2 (1):
77–92.

Huntington, Samuel P. 1957. “Conservatism as an Ideology.” American Political

Science Review 51 (2): 454–73.

Inglehart, Ronald. 1990. Culture Shift in Advanced Industrial Society. Princeton:

Princeton University Press.

Jansen, Guido, Geoffrey Evans, and Nan Dirk De Graaf. 2013. “Class Voting
and Left-Right Party Positions.” In Political Choice Matters, ed. Geoffrey
Evans and Nan Dirk De Graaf, 46–82. Oxford: Oxford University Press.

Jou, Willy. 2010. “The Heuristic Value of the Left-Right Schema in East Asia.”
International Political Science Review 31:366–94.

Lambert, Ronald D. 1983. “Question Design, Response Set and the

Measurement of Left/Right Thinking in Survey Research.” Canadian Journal
of Political Science 16 (1): 135–44.

Lambert, Ronald D., James E. Curtis, Steven D. Brown, and Barry J. Kay. 1986.
“In Search of Left/Right Beliefs in the Canadian Electorate.” Canadian
Journal of Political Science 19 (3): 541–63.
Laver, Michael, and W. Ben Hunt. 1992. Policy and Party Competition. New York:
Routledge.

Lewis-Beck, Michael S., and Kevin Chlarson. 2002. “Party, Ideology,

Institutions and the 1995 French Presidential Election.” British Journal of
Political Science 32 (3): 489–512.

Lipset, Seymour M., Paul F. Lazarsfeld, Allen H. Barton, and Juan Linz. 1954.
“The Psychology of Voting.” In Handbook of Social Psychology, vol. 2, ed.
Gardner Lindzey. Reading, MA: Addison-Wesley.

Lipset, Seymour Martin, and Stein Rokkan. 1967. Party Systems and Voter
Alignments. New York: Free Press.

Page 222 → Marsh, Michael, Richard Sinnott, John Garry, and Fiachra Kennedy.
2008. The Irish Voter. Manchester: Manchester University Press.

Norris, Pippa. 2004. Electoral Engineering: Voting Rules and Political Behavior. New
York: Cambridge University Press.

Ogmundson, R. L. 1979. “A Note on the Ambiguous Meanings of Survey

Research Measures Which Use the Words ‘Left’ and ‘Right’.” Canadian
Journal of Political Science 12 (4): 799–805.

Potrafke, Niklas. 2009. “Does Government Ideology Influence Political

Alignment with the U.S.? An Empirical Analysis of Voting in the UN
General Assembly.” Review of International Organizations 4 (3): 245–68.
Rigger, Shelley. 2001. From Opposition to Power: Taiwan’s Democratic Progressive
Party. Boulder: Lynne Rienner Publishers.

Sheng, Shing-yuan, and Yin-yan Chen. 2003. “Political Cleavage and Party
Competition: An Analysis of the 2001 Legislative Yuan Election” [in
Chinese]. Journal of Electoral Studies 10 (1): 7–40.

Sigelman, Lee, and Syng Nam Yough. 1978. “Left-Right Polarization in

National Party Systems.” Comparative Political Studies 11: 355–379.

Sinnott, Richard. 1995. Irish Voters Decide. Manchester: Manchester University

Press.

Stokes, Donald E. 1962. “Spatial Models of Party Competition.” American

Political Science Review 57 (2): 368–77.

Tsai, Chia-hung. 2008. “Making Sense of Issue Position, Party Image, Party
Preference, and Voting Choice: A Case Study of Taiwan’s 2004 Legislative
Election.” Journal of Social Science and Philosophy 20 (1): 1–24.

van der Eijk, Cees, Hermann Schmitt, and Tanja Binder. 2005. “Left-Right
Orientations and Party Choice.” In The European Voter: A Comparative Study
of Modern Democracies, ed. Jacques Thomassen, 167–91. Oxford: Oxford
University Press.

Wang, Ding-ming. 2001. “The Impacts of Policy Issues on Voting Behavior in

Taiwan: A Mixed Logit Approach” [in Chinese]. Journal of Electoral Studies 8
(2): 95–123.
Wang, Ding-ming. 2003. “Voting Utility and Choice Decision in 2001 Election:
The Application of Spatial Voting Theory in Different Electoral Systems”
[in Chinese]. Journal of Electoral Studies 10 (1): 171–206.

Zechmeister, Elizabeth J. 2006. “What’s Left and Who’s Right? A Q-Method

Study of Individual and Contextual Influences on the Meaning of
Ideological Labels.” Political Behavior 28 (2): 151–73.

Zechmeister, Elizabeth J. 2010. “The Varying Economic Meaning of ‘Left’ and

‘Right’ in Latin America.” AmericasBarometer Insights 38.

Zechmeister, Elizabeth J. 2015. “Left-Right Identifications and the Latin

American Voter.” In The Latin American Voter, ed. Ryan Carlin, Matthew M.
Singer, and Elizabeth J. Zechmeister, 195–225. Ann Arbor: University of
Michigan Press.

Zechmeister, Elizabeth J., and Margarita Corral. 2013. “Individual and

Contextual Constraints on Ideological Labels in Latin America.”
Comparative Political Studies 46 (6): 675–701.
Page 223 →
 Chapter 10

Electoral System Change and Its Effects on

the Party System in Taiwan

Chi Huang

On June 7, 2005, the ad hoc National Assembly of Taiwan ratified a

constitutional amendment to change the electoral rules of the Legislative Yuan
(the parliament) by halving the number of seats from 225 to 113, extending
legislators’ terms of office from three years to four, and adopting the mixed-
member majoritarian (MMM) system1 to replace the half-century-old single
nontransferable vote (SNTV) system for legislative elections. The new mixed-
member system in Taiwan consists of one tier of single-member districts
(SMDs) of 73 seats and a party list tier of 34 seats. In addition, there are 6 seats
reserved for highland and lowland aboriginals elected on the basis of the
SNTV system (Huang 2007, 2008a).

This chapter examines the significant changes in the legislative electoral

system in Taiwan and then evaluates their consequences to the political party
system. Taiwan’s electoral reform in 2005 is of great interest in itself because it
illustrates how the cleavages, institutions, parties, and voters interact to
produce election outcomes, both expected and unexpected. But more
important, it constitutes a critical case to reexamine the popular seat-
maximization approach to electoral reform (Benoit 2007), since in Taiwan it
was the ruling-party legislators’ own initiative to downsize the parliament,
which ignited a raging controversy and then backfired. Curiously enough, the
leaders of the two archrival parties, the Nationalist Party (Kuomintang) and the
Democratic Progressive Party, appeared as if they were silent partners pushing
through the same new MMM rule, although at the same time each was seeking
its own goals. Furthermore, Taiwan’s Page 224 → case fills a gap in the literature
concerning the “redistributive” type of electoral reform (Renwick 2010) in new
democracies and thus can be crucial for comparative studies of electoral
engineering (Ahmed 2013; Colomer 2004; Norris 2004; Sartori 1994).

The chapter begins with a review of the theoretical literature on the party
system followed by an outline of a comprehensive multilevel framework linking
cleavages with electoral systems, after which is a discussion of party politics
under the SNTV system. I then apply the framework to trace the process of
events and interactions between agents that led to the critical junctures of the
passage of the electoral reform proposal in August 2004 and its final
ratification in June 2005. Last is a discussion of the impact of the new MMM
system on Taiwan’s party system at the national, district, and voter levels.

Theoretical Perspectives

There have been two main theories that explain the party system and voting
behaviors in democracies: the cleavage structure and the electoral system. The
former is represented by Lipset and Rokkan (1967), who explained changes to
party systems, electoral realignments, and political mobilizations in Western
European countries through cleavages along the lines of groups, regions,
farmers, workers, laborers, and entrepreneurs. The Lipset-Rokkan “freezing
hypothesis” claims that these preexisting cleavage structures were then
“frozen” in the 1920s into party alignments through voter mobilization. That
is, party systems basically stabilize only when they reflect the fundamental
cleavages in societies. Once the party system is formed, it reinforces the
cleavage system in order to perpetuate itself. The second school is represented
by Duverger (1959), who believed it was the electoral system that principally
shaped the party system. For many scholars, Duverger’s “law” (that single-
member districts favor a two-party system) and “hypothesis” (that proportional
representation leads to a multiparty system; Riker 1982) still provide the
foundation of how the electoral system affects the party system, while the
relationship between district magnitude, M, and the effective number of parties
has been extended into the “M + 1 rule,” that is, voters will concentrate their
votes on the top M + 1 candidates (Cox 1997).

In spite of debates in the literature, these two theories are not necessarily
contradictory. While cleavage theory focuses more on the macro-level origins
of the party system, Duverger’s law focuses on the meso-level institutional
structures. Indeed, later development of the literature witnesses greater
appreciation Page 225 → of the interplay between social heterogeneity and
electoral rules (Clark and Golder 2006; Cox 1997; Neto and Cox 1997;
Ordeshook and Shvetsova 1994). In these interactive models, electoral systems
set an upper limit to the number of parties and work like filters of social
divisions. That is, within this upper limit, the more “permissive” an electoral
system is (such as the proportional representation system), the easier it is for
preexisting cleavages to manifest as political parties. The more “restrictive”
electoral systems (such as the SMD system), on the other hand, tend to
constrain the number of parties.
A Synthesized Framework

The literature on electoral rules is indeed impressive, and the area is often
revered as one of the most advanced in political science. But previous studies
of Taiwan’s electoral reform only either examined legislative elections or looked
at legislative and executive elections separately (e.g., Chang and Chang 2009).
Yet legislative electoral systems, important as they are in translating votes into
seats, do not operate in a vacuum. Their evolution and impact can be fully
understood only when they are embedded within broader social and
institutional contexts. Building upon the vast literature, Huang (2008a, 4–5)
developed a three-level analytical framework that incorporates the macro
perspective of social cleavages, the meso perspective of institutional structures
and electoral systems, and the micro perspective of voting behaviors. Following
Powell’s (1982) insight that constitutional settings have a substantial impact on
democratic performance, this general framework embeds the electoral systems
within constitutional systems at the meso level. It assumes that political elites
seek not only to maximize seats in the parliament but also to seize executive
offices. It is the combination of the legislative electoral system and the
constitutional setup that defines the payoffs of capturing executive offices and
the degree of cross-district coordination required to win the executive offices.
Hence “[t]o fully understand the effects of an electoral system, we must imbed
it within the broader political contexts, especially the constitutional framework,
of the country in question” (Huang 2016, 302).

In presidential and semipresidential systems, for example, the president

exercises the executive authority. The ultimate goal of most political parties and
their leaders is undoubtedly to control both the executive and legislative
branches in order to form a unified government, although the appeal of and
the competition for the presidency is often a higher priority (Batto and Page 226

→ Cox 2016; Curini and Hino 2012). This implies that presidential contests
often spill over to the legislative elections (Huang and Wang 2009, 2014).
Although parliamentary elections after a presidential one may allow the
president to consolidate his or her honeymoon (Samuels and Shugart 2010;
Shugart and Carey 1992), our general framework also reveals the possibility
that the parliament becomes a second battlefield for blocking government
policies or embarrassing the ruling party once a viable party loses the
presidential contest, or both. In the latter case, the temporal proximity of
presidential and legislative elections may well make campaigns appear to be
never-ending tournaments that escalate the already-fierce competition and leave
little room for party truces.

This synthesized framework, in contrast with the traditional legislature-centric

perspective, broadens our theoretical landscape by taking into account the
payoffs of holding executive office, the degree of coordination required to
capture the executive office, and the sustainability of the executive office (term
limit). Although Moser and Scheiner (2012) argue that strong presidentialism
hinders party-system institutionalization, my framework does not exclude the
possibility that the greater degree of national coordination required to win the
executive offices may well motivate elites to form coalitions with or to join and
stay in the major parties, or both (Hicken 2009; Hicken and Stoll 2008). I argue
that the joint effects of the presidential and legislative electoral systems, as well
as the temporal proximity of the two elections, exert pressure on elites and
parties to engage in cross-district coordination. Furthermore, social diversity is
more than a background condition waiting to be filtered by the electoral rules.
I argue that deep-rooted sociopolitical cleavages can act as latent yet powerful
forces structuring the speed and direction in which such coordination efforts
move. Finally, the strategic actions of elites under these contexts shape the
choice sets available to the electorate and its voting behavior at the micro level.
Electoral consequences, both expected and unexpected, in turn affect the
persistence of and change in the party system, institutional structure, and
eventually the sociopolitical cleavages. The principal idea behind this
comprehensive framework is simple: that is, social cleavage and constitutional
structures are part and parcel of studying the evolution and effects of electoral
systems.

Party Politics under the SNTV System in Taiwan

The evolution of the party system in Taiwan can be divided into three periods:
the dominance of the single-party system under the KMT2 before the Page 227

→ late 1980s; the gradual emergence of small parties in the early 1990s that
transitioned into a quite vigorous multiparty system after 2000 (Fell 2005); and
the reversal of the latter system (as a result of changes to the electoral system)
to a cleavage-based two-party system after 2008 (figure 10.1). In brief,
competitive party politics emerged only after 1986, when opposition forces
formed the DPP. The transition to full democracy was completed in the 1990s,
when the national legislature was subject to regular reelection beginning in
1992 followed by the first presidential election in 1996. The first power shift
occurred when the long-time opposition party, the DPP, won the presidency in
2000 followed by reelection in 2004. However, the KMT has retained
continuous control of Taiwan’s legislative branch, although the party went
from a single-party majority to a majority in coalition and then back to a single-
party majority. In the early part of the democratic era, the KMT retained a
degree of dominance. After 2001, splinter parties forced the KMT into
coalition arrangements, but the party returned to dominant status in the first
postreform election of 2008 (see, for example, Stockton 2010).

Taiwan has employed the simple plurality system for presidential elections
since 1996 (table 10.2), but the legislative electoral system is somewhat more
complicated. Before its 2005 electoral reform, Taiwan employed an SNTV
system for its national legislature (table 10.1).3

In the 1998, 2001, and 2004 legislative elections, for example, there were a
total of 225 seats. Of these, 168 representatives were elected from 29
geographically defined multimember districts, and another 8 members were
elected from 2 nationwide districts reserved for lowland and highland
aborigines. The average district magnitude was 5.79 seats per district. Several
districts had only 1 seat, while the largest district had 13 seats. In addition to
the 176 SNTV seats, there were also 49 seats elected by closed proportional
representation (PR) lists. The list designated for national party representatives
had 41 seats, whereas the list designated for overseas representatives had 8
seats. There was no separate party list ballot for the PR seats. Instead, all the
votes for the party nominees running in the SNTV districts were summed to
obtain each party’s national total. For all parties with at least 5 percent of the
national vote, these totals were used to apportion seats on the two lists using a
largest remainders formula (Farrell 2011; Wang 2012).

Under the SNTV system, Taiwan has developed from a single-dominant-party

system in the early 1990s to a period of multiparty politics after the 2000
presidential election. Scholars have cited one-party dominance as the reason for
the implementation of SNTV electoral rules. Under an SNTV system, political
parties must coordinate their supporters’ votes within Page 231 → constituencies

in order to more evenly distribute votes across candidates. Without successful

coordination, weaker candidates will get too few votes and stronger co-
partisans will receive too many votes. If effective, intraparty coordination can
result in political parties gaining an overrepresentation bonus. Governing
parties have a tremendous advantage because they can use the resources of the
state to overcome the coordination problems of nomination and division of
votes (Cox 1997; Cox and Niou 1994; Cox and Rosenbluth 1993, 1996;
Patterson and Stockton 2010; Rochon 1981). Meanwhile, small parties also face
much smaller coordination problems since they often only nominate one
candidate in any given district. With only one candidate, vote division is not a
challenge (Taagepera and Shugart 1989, 28). Besides, the vote share necessary
to win a seat decreases as the number of seats increases. Since candidates can
win with support from a small minority of voters, they can appeal to highly
personalized and niche voters (Flanagan et al. 1991). Under an SNTV system,
there is electoral space for small parties, either newly formed or splintered from
the existing big parties when antagonistic intraparty struggle cannot be
resolved.
Page 228 → Fig. 10.1. Taiwan’s political party system, 1986–2012. Source: Author. Note: 1. % in the
parentheses denotes percentage of seats in the Legislative Yuan; the broken line denotes an

alliance, loosely defined as some form of cooperation between parties. For example, many NP
candidates ran under the KMT’s umbrella in 2001 and 2004, and the PFP agreed not to run its

own party list in 2008 in exchange for KMT’s promise to nominate six former PFP legislators in
six districts and also allowed the PFP to share four seats on the KMT’s party list. Although

formally there was no NP legislator in 2008 and 2012, NP continues to be active in some local
elections.
Page 229 → Table 10.1 Features of the Legislative Electoral System in Taiwan since 1992

PR PR legal Average
Election Electoral Total ENEP ENEP ENEP
Term seats threshold district ENPP2
year system seats (SNTV) (PR) (SMD)
(%) % magnitude1

2nd 1992 SNTV 161 36(22.4) 5.0 4.41 2.46 2.52 — —

3rd 1995 SNTV 164 36(22.0) 5.0 4.52 2.63 2.90 — —

4th 1998 SNTV 225 49(21.8) 5.0 5.79 2.53 3.14 — —

5th 2001 SNTV 225 49(21.8) 5.0 5.79 3.24 4.14 — —

6th 2004 SNTV 225 49(21.8) 5.0 5.79 3.08 3.76 — —

7th 2008 MMM 113 34(30.1) 5.0 1.00 1.47 — 2.49 2.29

8th 2012 MMM 113 34(30.1) 5.0 1.00 1.97 — 3.03 2.32

Source: Huang (2008b, 132) and Central Election Commission, ROC “Election Statistics: Legislative
Elections.” http://db.cec.gov.tw/ (accessed September 23, 2013).
1. (total number of seats elected from districts)/(total number of districts). 2. Effective number of
parliamentary parties (ENPP) = , where Pi is the proportion of seats of the ith parties. Likewise, ENEP
=, where Qi is the vote shares of the ith parties.
Page 230 → Table 10.2 Presidential Election Results in Taiwan, 1996–2012

PFP (James
KMT DPP NP Others1
Electoral Soong)
Year
system Vote Vote Vote Vote
Vote Vote Vote Vote Vote % Vote
% % % %

C:
9.98
1,074,044
1996 plurality 5,813,699 54.00 2,274,586 21.13 — — — —
L:
14.90
1,603,790

2000 plurality 2,925,513 23.10 4,977,697 39.30 16,782 0.13 4,664,972 36.84 H: 79,429 0.63

2004 plurality 6,442,452 49.89 6,471,970 50.11 — — — — — —

2008 plurality 7,659,014 58.44 5,444,949 41.55 — — — — — —

2012 plurality 6,891,139 51.60 6,093,578 45.63 — — 369,588 2.76 — —

Source: Central Election Commission, ROC, “Election Statistics: Presidential Elections.”

http://db.cec.gov.tw/ (accessed September 23, 2013).
1. C refers to Chen Lu-an; L refers to Lin Yang-Kang; H refers to Hsu Hsin-liang.

Electoral Reform: Tangles of Two Archrivals

Taiwan has experienced numerous institutional challenges, such as fractional

politics, extremism, intraparty competition, money politics, party splintering,
and an inefficient parliament, and many studies have attributed these
disadvantages to the SNTV system (Cox and Rosenbluth 1993; Cox and Thies
1998; Richardson 1988; Wang 2012). Because of continual election scandals, the
call for legislative electoral reform has often enjoyed widespread attention.
However, it took the tension of stagnation and a stalemate caused by the
divided government after the first power shift in 2000 to shake up the half-
century-old SNTV system.

The hurdle of changing the legislative electoral system was unusually high in
Taiwan when the DPP government took over power in 2000. The Legislative
Yuan’s organization and election are specified in Article 4 of the Additional
Articles of Constitution of the Republic of China. Changing the electoral
system therefore requires amending the constitution, the procedure for which
is also specified by the constitution. As of June 2000,4 a constitutional
amendment required two stages: a proposal passed by the Legislative Yuan and
then ratification by the National Assembly.5 The proposal had to be initiated by
at least one-fourth of the total seats of the Legislative Yuan and passed by at
least three-fourths of the members present at a meeting attended Page 232 → by
at least three-fourths of the total members of the Legislative Yuan. Once
passed, the proposal had to be publicly announced for six months. Then 300
delegates had to be elected by proportional representation to the National
Assembly to deliberate and vote on the Legislative Yuan proposal. Obviously,
changing the legislative electoral rule was extremely difficult, not only because
a constitutional amendment was called for but also because it took the
incumbent Legislative Yuan, a beneficiary of the status quo, to initiate and pass
the constitutional amendment proposal before sending it to the ad hoc
National Assembly for ratification. Electoral engineering in Taiwan required
Herculean efforts of both intra- and interparty coordination to achieve. Yet it
did happen in 2005. Why?
 Based on the comprehensive framework laid out previously, I argue that it was
the introduction of popular presidential elections in 1996 and the subsequent
power shift in 2000 due to the KMT’s internal split that set the momentum of
legislative electoral reform on track in order to “bring back order and end the
parliamentary chaos.” The power of the president in Taiwan, under its
“president-parliamentary” semipresidential system (Shugart and Carey 1992,
24), makes the presidency a big enough prize to motivate political elites to
cooperate in the electoral process in order to win by a plurality. Before 2012,
Taiwan also had peculiar staggered electoral schedules for multiple political
offices at different levels and branches of government (Huang and Lin 2013), as
shown in figure 10.2. The temporal proximity of election schedules between the
presidential and legislative elections in turn make the interparty competition
for the former spill over into the latter like an endless election campaign.6
Anxious to appeal to the electorate, the major parties campaigned on a
reformist platform and each advanced its image as the true champion of
electoral reform. Repeated promises as part of the continuing election
campaigns not only suppressed opposing intraparty views but elevated the
clamor for reform to such a point that party leaders resorted to party discipline
to get legislators to pass the constitutional amendment act. The following
paragraphs chronologically trace the trajectory of Taiwan’s electoral reform so
as to highlight the sequential interactions among cleavages, institutions, and
agents that shaped reform politics in path-dependent ways.

In January 1994, the Japanese Diet passed electoral law reform bills that
abolished the old SNTV system and adopted a new MMM system (Curtis 1999;
Reed and Thies 2001). The 1994 electoral reform in Japan and its initial
implementation in the 1996 House of Representatives election sent a shock
wave through neighboring Taiwan, where the SNTV system had been blamed
for intraparty competition and factionalism, as well as for money Page 233

→ politics (see, for example, Cox 1996; Cox and Niou 1994). In late 1996, at the
National Development Meeting summoned by President Lee Teng-hui of the
KMT after his win in the March 1996 presidential election, a consensus was
reached to replace the SNTV system with a mixed-member system. However, a
proposed constitutional amendment about electoral reform failed in 1997
because the then-ruling KMT insisted on a Japanese-style MMM system, which
favors big parties, while the then-opposition DPP and the New Party
supported a more proportional, German-style mixed-member proportional
system. Given that no consensus was reached on that point, the electoral
reform proposal was dropped from the agenda. Instead, an amendment was
ratified to increase the total number of seats in the Legislative Yuan from 164
seats in 1995 to 225 seats in 1998 so as to accommodate members of the then
to-be-abolished Taiwan Provincial Council who might run for the legislature.
As a result, the average district magnitude increased from 4.52 to 5.79 seats per
district, and only 40,000 votes were usually sufficient to elect a district
candidate. This 37 percent jump in the number of seats did defuse some
resistance to the streamlining of the Taiwan Provincial Government, yet it
unexpectedly increased the electoral opportunities for small parties and seeded
the next-round initiatives of slimming the Legislative Yuan.
 Fig. 10.2. Timing of Taiwan’s major national and local elections. Source: Huang and Lin 2013.
Notes: 1. Circles with dashed lines denote concurrent elections with the same set of ballots. 2.
Dotted line in 2014 means special-city elections were held on the same day as other
cities/counties but with a different set of ballots.

Three years later, the DPP won the March 2000 presidential election, while
the Pan-Blue parties maintained control of the majority of seats in the
Legislative Yuan and dragged their feet over the new DPP administration. The
first power shift plus the first divided national government ever experienced in
Taiwan intensified the fraught relationship between the executive and
legislative branches and incited much mudslinging between the governing Page
234 → DPP and the opposition KMT and its allies. The opposition Pan-Blue
parties were often portrayed as the source of chaos in the parliament. Frequent
brawls that broke out on the floor further ruined the public’s image of the
Legislative Yuan as a whole.

On November 24, 2000, a year before the 2001 legislative election, a DPP
legislator placed a campaign advertisement in the newspapers calling for
halving the number of legislative seats.7 This first call for such a reduction
immediately caught the nation’s attention and became a campaign slogan. It
constituted a simple issue that could easily gain popular support and was
difficult for opponents to disagree with in public. But it also motivated
politicians to bundle their own political agenda with it. The effects went far
beyond the December 2001 legislative election and extended to the March 2004
presidential election and then to the December 2004 legislative election.
Among these three consecutive national elections within four years, the 2004
presidential election was the impetus for the final showdown on electoral
reform. To a large extent, what happened in the May 2005 National Assembly
election and with the assembly’s ratification of the constitutional amendment in
June was not so much a critical election, as Fell (2010) argued, but only the
consequence of the long, harsh four-year election campaigns. The presidential
hopefuls of the two major parties were betting that the new MMM electoral
system would not only be favorable to their own party but also reduce the
opportunities for the small parties. The two archrival parties, the DPP and the
KMT, both dreamed of first winning the presidency and then a majority of
seats and unwittingly appeared to act like silent partners in carrying out
electoral reform.

In mid-April 2001, the DPP began openly advocating an MMM system, a

reversal of the position it had held since 1997, when it was in opposition. In
late October 2001, President Chen Shui-bian pledged that if the DPP became
the largest party in the Legislative Yuan, he would advance bold and decisive
parliamentary reforms, including replacing the SNTV system with an MMM
system and downsizing the legislature. This was the first time that electoral
system reform had been explicitly linked with the popular issue of assembly
size. In mid-November, the KMT responded with a five-point parliamentary
reform proposal that also included an MMM system and a reduction in the
number of legislative seats. In late November 2001, President Chen Shui-bian
accused the KMT and the NP of not supporting seat reduction and reiterated
his pledge to reduce the total number of seats to 150 and replace the SNTV
system with an MMM system, and, in addition, to synchronize legislators’ three-
year term with the president’s four-year term. For fear of being labeled
antireform, the KMT and NP responded Page 235 → that they had
wholeheartedly supported the seat-reduction proposal from the very beginning.
Under pressure from public opinion, the party caucuses promised to work on
these proposals after the 2001 legislative election. The December 2001 election
was a debacle for the KMT, whose seats decreased from 54.7 percent to only
30.2 percent. Meanwhile, the ruling DPP substantially increased its seats from
31.7 percent to 38.7 percent and became the largest party in the legislature. The
election results definitely boosted the morale of the DPP.

Not long after the new legislature convened in 2002, the Taiwan Solidarity
Union, which was formed months before the 2001 election and was considered
to be an ally of the DPP, took the lead in proposing a constitutional
amendment to reduce the assembly size by half. Other parties responded by
proposing various versions of that proposal. In reality, however, the reform was
unpopular with the rank-and-file legislators in both major parties, since they
understandably were not happy about half of their seats in the legislature being
eliminated. The two major parties paid lip service to the proposal, but the
constitutional amendment committee did not even convene. The atmosphere
began to change, however, after the KMT and the People First Party leaders,
Lien Chan and James Soong, teamed up as running mates for the 2004
presidential election. While the KMT and PFP were still squabbling about their
common platform, especially the PFP’s objection to the MMM system, the
Government Renovation Committee of the Presidential Office announced in
early May a parliamentary reform plan that reiterated President Chen Shui-
bian’s campaign promises of the previous year. Soon the DPP party caucus and
the DPP Standing Committee reached agreement on the reform plan and
pledged to complete the constitutional amendment and put it into effect for
the December 2004 legislative election. In June 2003, the KMT-PFP party
caucuses dared the DPP to reduce the seats actually by half, that is, to 113 seats
instead of 150, but they still criticized the MMM proposal as a campaign
gimmick. In October, the KMT-PFP caucuses even began talking about
reducing the number of seats to 100 but remained vague about whether they
supported an MMM or mixed-member proportional system.

The tone of “reform bidding” turned acrimonious as the presidential election

drew near. On November 13, Premier You Xi-Kun opened another battlefront
by proposing to formulate a completely new Taiwan constitution via public
referendum. This move touched upon the fundamental issue of Taiwan
independence versus future unification with China (see Huang 2005, 2006).
Caught by surprise and again worried about being labeled antireform by the
DPP, the KMT chair, Lien Chan, countered by advocating Page 236 → a new
Republic of China constitution. Lien, cajoled by the acquiescence of the PFP,
vowed a sweeping reform based on 10 principles, the fifth principle of which
concerned parliamentary reform: the MMM system; seat reduction to 100, 113,
or 150; and synchronized terms. The content of this principle was suspiciously
similar to the DPP’s earlier proposal. Meanwhile, the KMT-PFP caucuses took
the lead in late November to initiate a constitutional amendment based on
Lien’s pledge but with an explicit seat number of 113. While this initiative was
still collecting the signatures of legislators, on December 30 the DPP’s Central
Standing Committee also reduced its original 150-seat proposal to 113 seats,
partly at the strong urging of the DPP’s former chair Lin Yi-Xiong. The two
arch rivals now engaged in strange tangles.

On February 25, 2004, the Constitutional Amendment Committee of the

Legislative Yuan convened and on March 10 passed a draft of the amendment
proposal. It seemed that the major parties intended to rush the process, and
they sent the bill to the floor right before the March 20 presidential election.
Yet on March 15 the party caucus of the Nonpartisan Solidarity Union, a long-
time opponent of the MMM system, motioned to send the bill for cross-party
consultation, which would take four months according to the legislative
procedure rule. This motion blocked the bill one step short of reaching the
floor.

After winning reelection in 2004 by a thin margin, President Chen Shui-bian

vigorously pushed for electoral reform. In his inauguration speech on May 20,
he vowed to fulfill his campaign promises by completing the constitutional
renovation by the end of his second term in 2008. As before, electoral reform
was unpopular with most rank-and-file legislators. Not surprisingly, the TSU
and NPSU continued to oppose the MMM system. But even the KMT-PFP
alliance began to weaken after losing the presidential election. The KMT chair,
Lien Chan, again worried about public support in the upcoming December
2004 legislative election, threw his support behind the reform. Some PFP
legislators, however, began to voice their objections to the MMM system for
fear of a bleak future. The PFP chair, James Soong, who was also worried
about being accused of breaking campaign promises, persuaded the PFP
legislators to support the electoral reform bill so as to maintain a proreform
image for the December legislative election. The pressure of the upcoming
election was so great that eventually not only the DPP and the KMT-PFP
caucuses went along with their party leaders but the TSU and NPSU legislators
all voted unanimously, though reluctantly, on August 23, 2004, to send the
constitutional amendment proposal to the National Assembly for ratification.
The December 2004 legislative election outcome turned out to be a great
disappointment to the PFP, for it lost a dozen seats (table 10.3).
Page 237 → Table 10.3 Legislative Yuan Election Results in Taiwan, 1992–2012

KMT DPP NP PFP TSU OTHER

Electoral
Year Seats
system
Vote
% % Vote Seat Vote Seat Vote Seat Vote Seat Vote Seat
% (%) % (%) % (%) % (%) % (%)
Seat ( )

51 15
1992 161 SNTV 52.7 95 (59.0) 31.4 — — — — — — 15.9
(31.7) (9.3)

54 21
1995 164 SNTV 46.1 85 (51.8) 33.2 13.0 — — — — 7.8 4 (2.4)
(32.9) (12.8)

70 11 21
1998 225 SNTV 46.4 123(54.7) 29.6 7.1 — — — — 16.9
(31.1) (4.9) (9.3)

87 46 13 10
2001 225 SNTV 28.3 68 (30.2) 33.8 2.7 1 (0.4) 18.4 7.8 9.1
(38.7) (20.4) (5.8) (4.4)

89 34 12 10
2004 225 SNTV 32.8 79 (35.1) 35.7 0.1 1 (0.4) 13.9 7.8 9.6
(39.6) (15.1) (5.3) (4.4)

13
79 SMD 53.5 61 (77.2) 38.2 — — 0.3 1 (1.3) 0.9 0 (0.0) 7.1 4 (5.1)
(16.5)

14
2008 34 PR 51.2 20 (58.8) 36.9 4.0 0 (0.0) — — 3.5 0 (0.0) 4.4 0 (0.0)
(41.2)

27
113 (Total) — 81 (71.7) — — 0 (0.0) — 1 (0.9) — 0 (0.0) — 4 (3.5)
(23.9)

27
79 SMD 48.2 48 (60.8) 43.8 0.1 0 (0.0) 1.3 1 (1.3) — — 6.6 3 (3.8)
(34.2)

13
2012 34 PR 44.6 16 (47.1) 34.6 1.5 0 (0.0) 5.5 2 (5.9) 9.0 3 (8.8) 4.9 0 (0.0)
(38.2)

40
113 (Total) — 64 (56.6) — — 0 (0.0) — 3 (2.7) — 3 (2.7) — 3 (2.7)
(35.4)
 Sources: Data from the Taiwan Provincial Election Commission of the Department of Civil Affairs and
the Central Election Commission of the Ministry of the Interior, Republic of China.
Note: 1. In 2008 and 2012, SMDs included 6 SNTV seats for aboriginals.

Page 238 → As the election ended and campaign pressure subsided, political
reality took over. In late March 2005, the PFP announced a change of position
on the constitutional amendment bill that it had voted for the previous August.
The TSU and NPSU soon followed suit. The election for the 300 National
Assembly delegates held on May 14, 2005, was the only election in Taiwan
based solely on the party list system. The record-low turnout rate of merely
23.4 percent signified not just the confusion about the PR rule but also the
public’s unfamiliarity with the debates on electoral rules (Huang, Wang, and Lin
2012, 2013; Huang and Yu 2011). The DPP earned 42.5 percent of the votes
and the KMT garnered 38.9 percent. Not surprisingly, the smaller parties,
including the TSU, PFP, NP, and NPSU, all opposed the bill, fearing a bleak
future. But the minor parties did not have the clout to block the amendments
since the two major parties, the DPP and the KMT, combined accounted for
244 (81.3%) of the 300 seats and exerted strict party discipline over their
members. Eventually, the ad hoc National Assembly ratified the amendment in
June 2005 with a vote of 249 for and 48 against with staunch support from the
two big parties, which expected gains from forming a unified government
under the new electoral rules.

The new MMM electoral system, specified by the new Article 4 of the
amended Additional Articles of the ROC Constitution, has two tiers with a
total of 113 seats. The tiers are not linked, so seats are determined
independently in each tier. The nominal tier includes 73 seats (64.6%), which
are elected by plurality in SMDs, and 6 seats (5.3%) are in two national SNTV
districts for lowland and highland aboriginals. The remaining 34 seats (30.1%)
comprise a single national list tier and are apportioned by a largest remainder
rule. Note that the nominal tier seats far outnumber the list tier seats. Giving
such heavy weight to the nominal tier is disadvantageous to smaller parties,
since smaller parties often find winning a plurality in an SMD to be a daunting
challenge and rely heavily on seats from the list tier for survival. With fewer list
tier seats available, smaller parties find it harder to survive. Even worse, to win
any of these seats, parties must win at least 5 percent of the national list tier
vote. This 5 percent legal threshold also discriminates against smaller parties,
since, without a legal threshold, parties winning at least 2.14 percent of the
valid vote would be able to win a seat.8

Taiwan’s electoral reform has important implications. Literature on electoral

reform usually considers the costs and benefits to legislators. As might be
expected, legislators, especially from small parties, were generally against
reform. However, the critical players were actually the national party
leaders/presidential contenders who effectively sold out their legislators in
order to Page 239 → play to public opinion in hopes of advancing their own
careers. We cannot understand the electoral reform without reference to the
semipresidential system. Additionally, these key presidential contenders
represented the two camps of the identity cleavage. That is, the identity
cleavage was hidden at the very root of Taiwan’s electoral reform.

Effects of Electoral System Change on the Party System

Preludes to the Postreform Legislative Election

Most studies evaluating the impact of the new MMM system in Taiwan focus
on the January 2008 legislative election (Hsieh 2009; Jou 2009; O’Neill 2013;
Shyu 2011; Stockton 2010). My comprehensive framework clearly points out
that the new electoral system directly affects politicians’ strategic entry and exit
decisions as well as the formation and dissolution of intraparty and interparty
coalitions. Interactions among political elites, in turn, determine the voters’
choice sets on ballots in elections.

Indeed, the impact of the MMM system had been felt almost immediately after
the ratification of the constitutional amendment. The effects were particularly
acute for legislators of small parties. They were squeezed from both ends: by a
50 percent decrease in total seats, as well as by a plurality rule in the newly
drawn SMDs. Small-party incumbents intending to run for reelection were
fighting an uphill battle. But of course this had been part of the plan of the two
big parties pushing for the MMM system. As a result, the first and immediate
impact of MMM on the party system manifested itself among the political
elites, including large-scale party switching during 2006 and 2007 as well as
interparty negotiations inside the Pan-Blue and Pan-Green camps. In the two
years after the 2005 electoral reform, as many as 22 incumbent PFP legislators
switched to the KMT while 5 TSU legislators switched to the DPP. Therefore,
party realignment started long before the first postreform legislative election
held in January 2008. More important, the multiparty system began to converge
toward a two-party system split along the preexisting fundamental cleavage,
that is, independence versus unification.
 However, interparty coordination proves to be not as easy as party switching
on the part of individuals. Two years before the 2008 legislative election, the
KMT reached out to and eventually negotiated successfully with the PFP to
nominate six former PFP legislators in six districts and also allowed the PFP to
share four seats on the KMT’s party list. The Page 240 → KMT also made way to
NPSU candidates in three SMDs, and promised not to nominate a candidate in
the first district in Pingtung so as to allow an independent candidate, also
affiliated with the NPSU, to compete against the DPP candidate. Apparently,
the then-opposition KMT, after losing the 2000 and 2004 presidential elections,
was anxious to form a Pan-Blue coalition aimed at the presidential election in
March 2008 and thus was more willing to make compromises with its allies. In
contrast to the coordination in the Pan-Blue camp, the DPP and the TSU
squabbled with each other and eventually failed to reach any substantive
agreement. The DPP seemed to believe that the TSU would simply back down
and follow its lead. Yet the TSU eventually fielded 13 candidates in districts to
fight its battle for survival. The effects of such differences in the interparty
coordination within each camp were further magnified by the new electoral
system, since the mechanical effect of the SMD tier favors the party capturing
the majority of popular votes.

Effects at the National Level

As discussed in the previous section, the reform of the legislative electoral

system was initiated by the then-ruling DPP to win an absolute majority in the
Legislative Yuan and gain full control of both the executive and legislative
branch. However, the lame-duck president, Chen, was unable to reach an
agreement with the TSU. Even worse, a series of scandals involving the
president and his family broke out in 2006 and tarnished the clean image of the
DPP. As a result, the outcome of the 2008 legislative election starkly
demonstrated the new MMM system’s disproportionality effect on the losing
party. Actually, the DPP received 38.2 percent of total district votes but only 13
(16.5%) out of 79 SMD/SNTV seats. For the PR ballot, the DPP received 36.9
percent of the total votes and 14 (41.2%) of 34 party seats. On the other hand,
the KMT garnered 53.5 percent of the total district votes and 61 (77.2%) out of
the 79 SMD/SNTV seats, as well as 51.2 percent of the total at-large votes and
20 (58.8%) out of the 34 party seats (Huang and Hsiao 2009). If we count the 3
seats won by the NPSU, the Pan-Blue, indeed, secured an overwhelming victory
over the DPP in the 2008 legislative election, which also paved the way for the
landslide victory of the KMT presidential candidate, Ma Yin-jeou, in March
2008. In 2008, 12 parties filed lists on the PR ballot, but only the two big
parties, the KMT and the DPP, surpassed the 5 percent threshold. Although
the two small parties, the NP and TSU, did receive 4 percent and 3.5 percent
for the PR votes, respectively, Page 241 → neither of them managed to reach the
5 percent threshold required for parties to be allocated PR seats.

The presidential and legislative elections became synchronized on January 14,

2012, when Taiwan held its first-ever concurrent presidential and legislative
elections. The elections also served as a test of the tenacity of the new two-
party system under Taiwan’s MMM rules embedded within the semipresidential
system. The PFP chair, James Soong, lamenting being cheated by the KMT, ran
as an independent presidential candidate and also nominated 10 legislative-
district candidates in order to file the party list.9 The TSU, on the other hand,
ran on the party list only, without nominating any district candidates. So each
party essentially ran on its own in 2012, with the exception of a tacit alliance
between the KMT and NPSU. In this three-way presidential race, the
incumbent president, Ma Ying-jeou of the KMT, defeated his main challenger,
Tsai Ing-wen of the DPP, by a substantial margin. The third candidate, James
Soong, received only 2.8 percent of the popular vote (Huang and Wang 2014).
Meanwhile, the two major political parties took the lion’s share in the legislative
election (table 10.3). The ruling KMT received 48.2 percent of the total district
votes and 48 (60.8%) SMD/SNTV seats, as well as 44.6 percent of the total
party list votes and 16 (47.1%) party seats. The DPP, on the other hand,
performed much better than it did in 2008 by garnering 43.8 percent of the
total district votes and 27 (34.2%) SMD/SNTV seats. For the PR ballot, the
DPP received 34.6 percent of the total votes and 13 (38.2%) party seats. Three
PFP candidates were elected: one through the aboriginal SNTV district and
two through the list tier with 5.5 percent PR votes. The TSU also won three
seats from its 9.0 percent list votes. Besides PFP and TSU, none of other seven
parties that filed lists got as much as 2.9 percent. Notably, the outcome of the
2012 election indicated the consolidation of the two-party system without
coalitions.

The last four columns of table 10.1 present the Laakso-Taagepera effective
number of parliamentary parties (ENPP) and the effective number of electoral
parties (ENEP) in moving from an SNTV to an MMM system (Laakso and
Taagepera 1979; Taagepera and Shugart 1989). According to Duverger’s law,
the number of parties in SMDs would shift toward two, due to mechanical
effects and strategic voting owing to psychological effects. But Duverger’s
hypothesis predicts that multiple parties remain in the PR tier of the system
because voters have stronger incentives to vote sincerely. Figure 10.3 clearly
indicates that there was an immediate sharp drop in the ENPP from 3.08 to
1.47, which is almost one-party dominance in parliament. With a much
stronger performance by the DPP, and after the PFP split from the Pan-Blue
coalition and ran its own candidates in the 2012 Page 242 → legislative election,
the ENPP rose to 1.97, which is much closer to 2.00. ENEP indexes are not
directly comparable between single-ballot SNTV and two-ballot MMM
systems. Still it is not difficult to see that the growing trend of ENEP under
SNTV was interrupted after the change of electoral system. Furthermore, as
Duverger’s law and hypothesis expect, the ENEP in the SMD tier hovers
around 2.30 while its PR counterpart is slightly higher. Actually, the ENEP in
the PR tier rose from 2.49 in 2008 to 3.03 in 2012 because both the PFP and
TSU’s party votes exceed the 5 percent legal threshold.

Fig. 10.3. Effective numbers of parliamentary parties and presidential candidates. Source: Huang
2013.

Effects at the District Level

The national-level indexes were further checked and tested at the district level.
It can be shown that the changes at the district level are also congruent with
the electoral-system theories. Specifically, I exploit the SF ratio (the ratio of the
second to the first loser’s vote total) patterns over time and across two tiers.
Cox (1997, 85) used the SF ratio to test the M + 1 rule, a generalization of
Duverger’s law. The M + 1 rule holds that, at the district level, the effective
number of candidates will decline toward the district magnitude (M) plus one.
SF ratios demonstrate the extent to which the number of votes cast for less
competitive candidates trail behind M + 1. Cox reasoned Page 243 → that in
Duvergerian equilibria, when strategic voting by voters occurs, the (M + 2)th
candidate will be deserted and thus the SF ratio will be near zero. In non-
Duvergerian equilibria, however, the first and second losers receive nearly the
same number of votes, and thus the SF ratio will be near one. When M is small,
such as in SMDs, where M = 1, strategic voting is more likely to occur in
closely contested districts. As M grows larger, such as in an SNTV system with
multiple seats in each district, strategic voting becomes more difficult due to
the lack of clear information for voters. In other words, we should expect a
greater number of districts concentrated at the lower end (closer to zero) in the
nominal tier under an MMM system than in an SNTV system. If we plot the
histogram of SF ratios of all the districts, the M + 1 rule predicts a right-
skewed distribution with most districts having near-zero values concentrated
on the left side and only a small number of exceptions on the right.

The SF ratios of Taiwan’s seven legislative elections at the district level were
computed and then classified into 10 intervals, as shown in figures 10.4 and
10.5. An examination of these figures reveals a dramatic reversal of district
distributions after the electoral system switched from SNTV (figure 10.4) to
MMM (figure 10.5). During the SNTV period from 1992 to 2004, many of the
multimember districts fell at the higher end of SF ratio (i.e., closer to 1.00). In
the last two legislative elections of 2008 and 2012 under MMM, however, most
single-member districts’ SF ratios indeed had values close to 0.00 and thus
concentrated at the lower end. This means that two-party competition has
become the norm at the district as well as at the national level.

Effects at the Voter Level

As mentioned earlier, Duverger (1959, 205) asserted that single-member district

plurality would tend to generate two-party competition, and he also proposed
that PR systems would encourage multiparty competition. Besides the
mechanical factor of SMD that leads to the underrepresentation of the weaker
parties, the psychological factor that supporters of the third party tend to
“transfer their voter to the less evil of its two adversaries” (Duverger 1959,
226) also causes strategic voting. A mixed-member electoral system is
characterized by the hybrid of both SMD and PR tiers. If identifiers with small
parties indeed vote strategically on the SMD ballot and sincerely on the PR
ballot, this ticket-splitting pattern becomes a micro-level evidence of electoral
system effects.
 Fig. 10.4. SF ratio of Taiwan’s Legislative Yuan elections, 1992–2004. Source: Huang 2013.

Fig. 10.5. SF ratio of Taiwan’s Legislative Yuan elections, 2008–2012. Source: Huang 2013.
Page 244 → There is an abundant extant literature on voting choices and ticket-
splitting under the MMM system at the voter level based on survey data,
especially the TEDS 2008L and TEDS 2012-T surveys (Huang 2008b; Huang
and Chou 2013; Huang and Hsiao 2009; Huang and Lin 2009; Huang, Wang
and Kuo 2008; Wang, Lin and Hsiao 2016). It emphasizes that partisan voting
is dominant for either of the two major party identifiers on both SMD and PR
ballots. Yet certain supporters of the smaller Page 245 → parties indeed tend to
vote strategically for the major party candidates of their own “color” camp, but
vote sincerely for their most preferred party on the PR ballot. Besides
partisanship, ethnic identity and attitude toward Taiwan independence vs.
unification with China remain important explanatory variables of voting
choices. For example, compared to those who identify themselves as Taiwanese,
those who identify themselves as Chinese or as both Taiwanese/Chinese are
more likely to vote for KMT candidates in districts and vote for the Pan-Blue
camp on the party ballot.

In spite of the tendency of strategic voting from minor parties’ supporters,

however, the MMM electoral system still squeezes smaller parties from two
ends. That is, PR portion in Taiwan accounts for only 30.1 percent of the total
113 seats plus its relatively high 5 percent legal threshold causes the MMM
electoral system to exert heavy pressure on small parties and thus push down
the number of parties. Meanwhile, the force of gravity of the SMD portion in
Taiwan’s MMM electoral system tends to favor any major party that can garner
near 50 percent vote share in legislative elections.

Conclusions

This chapter has employed a comprehensive framework with which to study

the evolution and effects of electoral systems embedded within social and
institutional contexts. It has examined the change in legislative electoral system
in Taiwan and evaluated its consequences on political party systems from the
perspective of this broad framework. It is interesting to note that after
switching from an SNTV system to an MMM in 1994, Japan experienced only a
gradual evolution from a multiparty to a two-party system, taking five House of
Representatives election cycles, whereas the party system in Taiwan changed
immediately and dramatically in the first postreform election (Huang 2011;
Huang, Kuo, and Stockton 2016). Taiwan’s fast convergence toward two-party
competition can be partly attributed to institutional factors. That is, the
powerful presidency of Taiwan’s semipresidential system motivates party
leaders to synchronize the legislative electoral system to make it easier to form
a unified government.

However, institutions alone do not explain the direction in which multiple

parties converge. This chapter argued that the fundamental cleavage in Taiwan,
that is, independence versus unification, has played a hidden yet significant role
in three ways. First, it directed the centripetal force of the plurality system of
presidential election toward the two camps. Although the old SNTV system
left room for splinter parties within camps, the cleavage Page 246 → motivated

elites to push for a new electoral system that was more compatible with the
enduring social division. Second, it accelerated the speed of convergence
toward the two-party system along the borderline of the preexisting cleavage by
skipping time-consuming trial and error in coalition permutations. Last, but
certainly not least, it crystallized a relatively stable two-party competition with
regular power shifts under the semipresidential system.

Notes

1. In Chinese, the new MMM rule is literally called “single-member

district, two-ballot system.” This chapter adopts the classification and
terminologies of two subtypes of mixed-member systems used by Shugart
and Wattenberg (2001, 13–14). In the MMM systems, there is no link
between nominal and list tiers in the allocation of seats to parties. The
plurality formula used in the nominal tier can lead to significant
disproportionality, and the list tier merely mitigates rather than erases this
disproportionality. In contrast, the mixed-member proportional systems
prioritize the list tier, that is, the second ballot, and each party’s total seat
share is proportional to the list tier vote share.

2. The KMT, which was founded on the mainland, was the ruling party of
the Republic of China on the mainland. It moved to Taiwan in 1949 after
losing the civil war with the Chinese Communist Party.

3. As explained below, the prereform system was technically a mixed

system since it had a nominal tier, with voters choosing specific
candidates in the SNTV tier, as well as a list tier (see Farrell 2011).
However, for the purpose of clarity, this chapter will refer to the
prereform system as an SNTV system. In an SNTV system, each district
has one or more seats and each voter can cast only one ballot for one
specific candidate. There is no provision for preference rankings, so if a
voter supports a candidate who does not win, the vote cannot be
transferred to a second-favorite candidate.

4. Before 2000, the constitution could be amended by the National

Assembly alone upon the proposal of one-fifth of the total delegates and
by a resolution of three-fourths of the delegates present at a meeting
having a quorum of two-thirds of the entire Assembly. The KMT
government took advantage of its overwhelming majority in the National
Assembly to pass six constitutional amendments during the 1990s.
 5. The second stage was changed into a national referendum when the
2005 constitutional amendment abolished the National Assembly.

6. Both the learning effect of the 2000 presidential election and the
centripetal effect of the plurality rule pushed the effective number of
presidential candidates toward two in the ensuing presidential elections, as
shown in figure 10.3. This downward trend may well affect the effective
number of parties in legislative elections.

7. The DPP legislator Zhang Xue-shun used a sensational approach,

pointing to the majority KMT as the “source of chaos and disorder” in
the Legislative Yuan and argued that halving the seats would bring back
order. See Commercial Times, November 24, 2000.

Page 247 → 8. According to Lijphart (1997, 74), given the district magnitude
M, the effective threshold is 0.75/(M + 1).

9. Article 24 of the Election and Recall Act provides that unless a party
has attained 2.0% and more of the total valid votes in the recent
presidential election or has won at least 2.0% of the votes in the previous
three legislative elections or has five seats or more in the Legislative Yuan,
it has to nominate at least 10 candidates in SMDs or indigenous districts
in order to qualify for a party list.

References

Ahmed, Amel. 2013. Democracy and the Politics of Electoral System Choice:
Engineering Electoral Dominance. Cambridge: Cambridge University Press.
Batto, Nathan F., and Gary W. Cox. 2016. “Introduction: Legislature-Centric
and Executive-Centric Theories of Party Systems and Faction Systems.” In
Mixed-Member Electoral Systems in Constitutional Context: Taiwan, Japan, and
Beyond, ed. Nathan F. Batto, Chi Huang, Alexander C. Tan, and Gary W.
Cox. Ann Arbor: University of Michigan Press.

Benoit, Kenneth. 2007. “Electoral Laws as Political Consequences: Explaining

the Origins and Change of Electoral Institutions.” Annual Review of Political
Science 10: 363–90.

Chang, Alex Chuan-Hsien, and Yu-Tzung Chang. 2009. “Rational Choices and
Irrational Results: The DPP’s Institutional Choice in Taiwan’s Electoral
Reform.” Issues & Studies 45 (2): 23–60.

Clark, William Roberts, and Matt Golder. 2006. “Rehabilitating Duverger’s

Theory: Testing the Mechanical and Strategic Modifying Effects of Electoral
Laws.” Comparative Political Studies 39 (6): 679–708.

Colomer, Josep M. 2004. “The Strategy and History of Electoral System

Choice.” In Handbook of Electoral System Choice, ed. Josep M. Colomer, 3–78.
Hampshire, UK: Palgrave Macmillan.

Cox, Gary W. 1996. “Is the Single Nontransferable Vote Superproportional?

Evidence from Japan and Taiwan.” American Journal of Political Science
40(3): 740–55.

Cox, Gary W. 1997. Making Votes Count: Strategic Coordination in the World’s
Electoral Systems. Cambridge: Cambridge University Press.
Cox, Gary W., and Emerson Niou. 1994. “Seat Bonuses under the Single Non-
Transferable Vote System: Evidence from Japan and Taiwan.” Comparative
Politics 26 (2): 221–36.

Cox, Gary W., and Frances Rosenbluth. 1993. “The Electoral Fortunes of
Legislative Factions in Japan.” American Political Science Review 87 (3): 577–89.

Cox, Gary W., and Frances Rosenbluth. 1996. “Factional Competition for the
Party Endorsement.” British Journal of Political Science 26 (2): 259–69.

Cox, Gary W., and Michael F. Thies. 1998. “The Cost of Intraparty
Competition: The Single, Non-Transferable Vote and Money Politics in
Japan.” Comparative Political Studies 31 (3): 267–91.

Curini, Luigi, and Airo Hino. 2012. “Missing Links in Party-System

Polarization: How Institutions and Voters Matter.” Journal of Politics 74 (2):
460–73.

Page 248 → Curtis, Gerald L. 1999. The Logic of Japanese Politics. New York:
Columbia University Press.

Duverger, Maurice. 1959. Political Parties: Their Organization and Activity in the
Modern State. 2nd English ed. New York: John Wiley and Sons.

Farrell, David M. 2011. Electoral Systems: A Comparative Introduction. 2nd ed. New
York: Palgrave Macmillan.

Fell, Dafydd. 2005. Party Politics in Taiwan: Party Change and Democratic Evolution
of Taiwan, 1991–2004. London: Routledge.
Fell, Dafydd. 2010. “Was 2005 a Critical Election in Taiwan? Locating the Start
of a New Political Era.” Asian Survey 50 (5): 927–45.

Flanagan, Scott C., Shinsaku Kohei, Ichiro Miyake, Bradley M. Richardson, and
Joji Watanuki. 1991. The Japanese Voter. New Haven: Yale University Press.

Hicken, Allen. 2009. Building Party Systems in Developing Democracies. New York:
Cambridge University Press.

Hicken, Allen, and Heather Stoll. 2008. “Electoral Rules and the Size of the
Prize: How Political Institutions Shape Presidential Party Systems.” Journal
of Politics 70 (4): 1109–27.

Hsieh, John Fuh-sheng. 2009. “The Origins and Consequences of Electoral

Reform in Taiwan.” Issues & Studies 45 (2): 1–22.

Huang, Chi. 2005. “Dimensions of Taiwanese/Chinese Identity and National

Identity in Taiwan: A Latent Class Analysis.” Journal of Asian and African
Studies 40 (1–2): 51–70.

Huang, Chi. 2006. “The Evolution of Taiwanese Identity: A Pseudo Panel

Analysis.” Presented at the 102nd Annual Meeting of the American Political
Science Association, Philadelphia, August 31–September 3.

Huang, Chi. 2007. “Assessing the Impact of Mixed Electoral System in Taiwan:
Methodological Challenges of Testing Interaction Effects.” Presented at the
International Symposium on Mixed Electoral Systems in East Asia,
National Chengchi University, Taipei, May 26.
Huang, Chi. 2008a. “Contexts and Consequences of Electoral Systems: An
Introduction” [in Chinese]. In Ruhe Pinggu Xuanzhi Bianqian: Fangfalun de
Tantao [The consequences of electoral system change: Methodological
perspectives], ed. Chi Huang and Chin-hsing Yu, 1–17. Taipei: Wu-nan
Publications.

Huang, Chi. 2008b. “Voting Choices under the Mixed-Member Majoritarian

System: A Multilevel Mixed Logit Model” [in Chinese]. In Ruhe Pinggu
Xuanzhi Bianqian: Fangfalun de Tantao [The consequences of electoral system
change: Methodological perspectives], ed. Chi Huang and Chin-hsing Yu,
129–50. Taipei: Wu-nan Press.

Huang, Chi. 2011. “Political Consequences of the MMM Electoral Systems in

Taiwan and Japan.” Presented at the 2011 Annual Meeting of the American
Political Science Association, Seattle, September 1–4.

Huang, Chi. 2013. “Taiwan’s Political Geography Information System

(TPGIS).” http://tpgis.nccu.edu.tw/nccu/. Accessed September 30, 2013.

Huang, Chi. 2016. “Mixed-Member Systems Embedded within Constitutional

Systems.” In Mixed-Member Electoral Systems in Constitutional Context: Taiwan,
Japan, and Beyond, ed. Nathan F. Batto, Chi Huang, Alexander C. Tan, and
Gary W. Cox. Ann Arbor: University of Michigan Press.

Page 249 → Huang, Chi, and Chang-chi Lin. 2009. “Vote Choices under the
SNTV and MMM Electoral Systems: A Panel Study of Taiwan’s 2004 and
2008 Legislative Elections.” Presented at the 2009 Annual Meeting of the
American Political Science Association, Toronto, Canada, September 3–6.
Huang, Chi, and Chang-chi Lin. 2013. “The Effects of Concurrent Elections on
Voter Turnout: A Causal Analysis” [in Chinese]. In 2012 Nian Zongtong yu
Lifa Weiyuan Xuanju: Bianqian yu Yanxu [Continuity and change in Taiwan’s
2012 presidential and legislative elections], ed. Lu-huei Chen, 47–83. Taipei:
Wu-nan Publications.

Huang, Chi, and Ching-hsin Yu. 2011. “Political Cycle of Voters’

Understanding of the New Electoral System: The Case of Taiwan.” Japanese
Journal of Electoral Studies 27 (2): 60–76.

Huang, Chi, and T. Y. Wang. 2009. “The Effect of the 2008 Legislative Yuan
Election on the 2008 Presidential Election in Taiwan: Pendulum or
Bandwagoning?” [in Chinese]. In 2008 Nian Zongtong Xuanju: Lun Erci
Zhengdang Lunti zhi Guanjian Xuanju [The 2008 presidential election in
Taiwan], ed. Lu-huei Chen, Ching-hsin Yu, and Chi Huang, 31–56. Taipei:
Wu-nan Publications.

Huang, Chi, and T. Y. Wang. 2014. “Presidential Coattails in Taiwan: An

Analysis of Voter- and Candidate-Specific Data.” Electoral Studies 33:175–85.

Huang, Chi, and Yi-ching Hsiao. 2009. “Vote Choices under the Mixed-
Member Majoritarian System in Taiwan’s 2008 Legislative Election.”
Presented at the 2009 Asian Election Studies International Conference,
“Elections in Taiwan, Japan, and Korea under the Mixed-Member
Majoritarian Systems,” Election Study Center, National Chengchi
University, Taipei, Taiwan, May 24.
Huang, Chi, and Ying-Lung Chou. 2013. “Split-Ticketing in Taiwan’s 2012
Concurrent Presidential and Legislative Elections” [in Chinese]. In 2012
Nian Zongtong yu Lifa Weiyuan Xuanju: Bianqian yu Yanxu [Continuity and
change in Taiwan’s 2012 presidential and legislative elections], ed. Lu-huei
Chen, 85–124. Taipei: Wu-nan Publications.

Huang, Chi, Ding-ming Wang, and Ming-feng Kuo. 2008. “Straight- and Split-
Ticket Voting in a Mixed-Member Majoritarian System: An Analysis of the
1996 House Election in Japan” [in Chinese]. Xuanju Yanjiu [Journal of
Electoral Studies] 15 (2): 1–35.

Huang, Chi, Hung-chung Wang, and Chang-chih Lin. 2012. “Knowledge of the
Electoral System and Voter Turnout.” Taiwan Zhengzhi Xuekan [Taiwanese
Political Science Review] 16 (1): 239–79.

Huang, Chi, Hung-chung Wang, and Chang-chih Lin. 2013. “Knowledge of the
Electoral System and Voting: The Case of Taiwan’s 2012 Legislative
Election.” Issues & Studies 49 (4): 1–45.

Huang, Chi, Ming-feng Kuo, and Hans Stockton. 2016. “The Consequences of
MMM on Party Systems.” In Mixed-Member Electoral Systems in Constitutional
Context: Taiwan, Japan, and Beyond, ed. Nathan F. Batto, Chi Huang,
Alexander C. Tan, and Gary W. Cox. Ann Arbor: University of Michigan
Press.

Jou, Willy. 2009. “Electoral Reform and Party System Development in Japan
and Taiwan: A Comparative Study.” Asian Survey 49 (5): 759–85.
Laakso, Markku, and Rein Taagepera. 1979. “Effective Number of Parties: A
Measure with Application to West Europe.” Comparative Political Studies 12
(1): 3–27.

Page 250 → Lijphart, Arend. 1994. Electoral Systems and Party Systems: A Study of
Twenty-Seven Democracies 1945–1990. Oxford: Oxford University Press.

Lijphart, Arend. 1997. “The Difficult Science of Electoral Systems: A

Commentary on the Critique by Alberto Penadés.” Electoral Studies 16 (1):
73–77.

Lin, Jih-wen. 2016. “The Consequences of Constitutional Systems on Party

Systems.” In Mixed-Member Electoral Systems in Constitutional Context: Taiwan,
Japan, and Beyond, ed. Nathan F. Batto, Chi Huang, Alexander C. Tan, and
Gary W. Cox. Ann Arbor: University of Michigan Press.

Lin, Tsong-Jyi. 2011. “Institutionalization of Party System in Taiwan: An

Analysis on the Mass Political Attitudes” [in Chinese]. Taiwan Minzhu Jikan
[Taiwan Democracy Quarterly] 8 (4): 141–66.

Lipset, Seymour M., and Stein Rokkan. 1967. “Cleavage Structure, Party
Systems, and Voter Alignments: An Introduction.” In Party Systems and Voter
Alignments: Cross-National Perspectives. ed. Seymour M. Lipset and Stein
Rokkan. New York: Free Press.

Moser, Robert G., and Ethan Scheiner. 2012. Electoral Systems and Political
Context: How the Effects of Rules Vary Across New and Established Democracies.
Cambridge: Cambridge University Press.
Neto, Octavio Amorim, and Gary W. Cox. 1997. “Electoral Institutions,
Cleavage Structures, and the Number of Parties.” American Journal of Political
Science 41 (1): 149–74.

Norris, Pippa. 2004. Electoral Engineering: Voting Rules and Political Behavior.
Cambridge: Cambridge University Press.

O’Neill, Daniel C. 2013. “Electoral Rules and the Democratic Progressive

Party’s Performance in the 2004 and 2008 Legislative Elections in Taiwan.”
Journal of Asian and African Studies 48 (2): 161–79.

Ordeshook, Peter C., and Olga V. Shvetsova. 1994. “Ethnic Heterogeneity,

District Magnitude, and the Number of Parties.” American Journal of Political
Science 38 (1): 100–123.

Patterson, Dennis P., and Hans Stockton. 2010. “Strategies, Institutions, and
Outcomes under SNTV in Taiwan, 1992–2004.” Journal of East Asian Studies
10 (1): 31–59.

Powell, G. Bingham, Jr. 1982. Contemporary Democracies: Participation, Stability, and

Violence. Cambridge: Harvard University Press.

Reed, Steven R., and Michael F. Thies. 2001. “The Causes of Electoral Reform
in Japan.” In Mixed-Member Electoral Systems: The Best of Both Worlds?, ed.
Matthew Soberg Shugart and Martin P. Wattenberg, 152–72. Oxford:
Oxford University Press.

Renwick, Alan. 2010. The Politics of Electoral Reform: Changing the Rules of
Democracy. Cambridge: Cambridge University Press.
Richardson, Bradley M. 1988. “Constituency Candidates versus Parties in
Japanese Voting Behavior.” American Political Science Review 82 (3): 695–718.

Riker, William H. 1982. “The Two-Party System and Duverger’s Law: An Essay
on the History of Political Science.” American Political Science Review 76 (3):
753–66.

Rochon, Thomas. 1981. “Electoral Systems and the Basis for the Vote: The
Case of Japan.” In Parties, Candidates and Voters in Japan: Six Quantitative
Studies, ed. J. Page 251 → C. Campbell, 1–28. Michigan Papers in Japanese
Studies, No. 2. Ann Arbor: University of Michigan Center for Japanese
Studies.

Samuels, David J., and Matthew S. Shugart. 2010. Presidents, Parties, and Prime
Ministers: How the Separation of Powers Affects Party Organization and Behavior.
Cambridge: Cambridge University Press.

Sartori, Giovanni. 1994. Comparative Constitutional Engineering: An Inquiry into

Structures, Incentives and Outcomes. New York: New York University Press.

Shugart, Matthew S., and John S. Carey. 1992. Presidents and Assemblies:
Constitutional Design and Electoral Dynamics. New York: Cambridge University
Press.

Shugart, Matthew Søberg, and Martin P. Wattenberg. 2001. “Mixed-Member

Electoral Systems: A Definition and Typology.” In Mixed-Member Electoral
Systems: The Best of Both Worlds?, ed. Matthew Søberg Shugart and Martin P.
Wattenberg, 7–24. Oxford: Oxford University Press.
Shyu, Huo-yan. 2011. “Taiwan’s Democratization and the Freezing of the Party
System.” In Political Parties, Party Systems, and Democratization in East Asia, ed.
Liang Fook Lye and Wilhelm Hofmeister. Singapore: World Scientific.

Stockton, Hans. 2010. “How Rules Matter: Electoral Reform in Taiwan.” Social
Science Quarterly 91 (1): 21–41.

Taagepera, Rein, and Matthew Søberg Shugart. 1989. Seats and Votes: The Effects
and Determinants of Electoral Systems. New Haven: Yale University Press.

Wang, T. Y., Chang-chih Lin, and Yi-ching Hsiao. 2016. “Split-Ticket Voting
under MMM.” In Mixed-Member Electoral Systems in Constitutional Context:
Taiwan, Japan, and Beyond, ed. Nathan F. Batto, Chi Huang, Alexander C.
Tan, and Gary W. Cox. Ann Arbor: University of Michigan Press.

Wang, Yeh-Lih. 2012. Bijiao Xuanju Zhidu [Comparative electoral systems]. 6th
ed. Taipei: Wu-nan Publications.
Page 252 →
 Chapter 11

Political Participation in Taiwan

Chung-li Wu with Tzu-Ping Liu

Political participation by ordinary citizens is the essence of democracy. Verba

and Nie (1972, 3–4) stress that political participation is at the heart of
democratic theory and has “a particularly crucial relationship to all other social
and political goals.” Dahl (1971, 1) also posits a strong link between the two:
“A key characteristic of a democracy is the continuing responsiveness of the
government to the preferences of its citizens, considered as political equals.” To
this end, a democracy must provide its citizens with equal opportunities to
formulate preferences, to signify those preferences, and to have their
preferences influence the formation of the government. Therefore, higher levels
of participation by citizens in political activities can be viewed as a norm that
supports a democratic political regime.

Political participation comes in many forms, including contacting public

officials, participating in political demonstrations, and many others. To an
ordinary citizen, however, voting is the commonest, simplest, and least costly
form of participation in electoral politics, although it has profound
implications for the political system. Under a system of voluntary suffrage,
voter turnout not only indicates how much interest the electorate has in the
election but it also reveals the electorate’s degree of psychological attachment
to political affairs (Milbrath and Goel 1977, 46–47; Rosenstone and Hansen
2003, 245–48). Likewise, voting is the key mechanism for responsiveness in
democratic society. Citizens cast ballots to choose among candidates from
competing political parties, and their choices are important in the selection of
political leaders and public policies. The desire for office certainly makes
political elites modify or even totally change policies to meet the expectations
of voters (Almond and Powell 1988, 49; Nie and Verba 1975, 9–10).

Page 253 → Thus electoral participation in general, and voter turnout in

particular, are important elements in the maintenance of democracy. Both have
been widely studied in Western countries; however, they have received
comparatively little attention elsewhere. This chapter explores political
participation in Taiwan since the early 1990s, with a particular focus on
electoral participation and voter turnout from 2001 through 2012. Our main
purpose is, first, to describe trends in voter turnout since democratization, and
second, to analyze the personal attributes of Taiwan citizens that have led them
to become involved in politics over the past decade.

Voter turnout in Taiwan has been relatively high compared to other

democracies, with an average of about 70 percent; however, it has declined
substantially since 2000. To understand this pattern, we will use individual-level
survey data to explore the question of who votes in Taiwan, comparing people
of different ages and ethnicities. We will also examine party identification and
political knowledge to learn their effects on the decision to vote. Finally, we
will discuss how each of these demographic and cognitive factors influences
other forms of political participation. The survey data used originate from the
multiyear Taiwan Election and Democratization Study conducted by the
National Chengchi University Election Study Center.1

The Framework of This Chapter

The term political participation encompasses many different activities. This

chapter adopts Rosenstone and Hansen’s (2003, 4–5) definition, so our research
is focused on voting, persuading others, campaigning, giving money, contacting
others, attending meetings or rallies, and signing petitions. To explain these
activities, we focus first on two sociodemographic variables—age and ethnicity.

Demographic factors are important for understanding political participation

because individuals’ social backgrounds are central to the development of their
political attitudes and behaviors. It is a generally accepted proposition that age
is a predictive variable where political attitudes and participation are
concerned. According to the life-cycle effects theory, political information and
experience rise steadily with age, a phenomenon confirmed by previous
empirical studies in the United States (Campbell et al. 1960, 485–87; Conway
2000, 19–24; Milbrath and Goel 1977, 114–16). Political participation increases
through a person’s thirties, forties, and fifties, and is at its height in Taiwan
when individuals are in their late fifties and early sixties. Participation among
those older than sixty-five declines Page 254 → primarily because these people
are in poorer health and are less physically mobile. Thus a person’s ability to
participate in political activities will be closely related to age.
 Among the socioeconomic variables used to explain political participation,
race/ethnicity is probably the most thoroughly researched factor (Rosenstone
and Hansen 2003; Tate 1991; Verba and Nie 1972). Ethnicity has been a critical
issue in Taiwan’s political life, especially in relation to the ethnic differences
between Minnan and mainlanders (Wang 1994, 1998).2 In the light of Taiwan’s
unique historical background and sociopolitical environment, ethnicity and its
related issues (e.g., ethnic consciousness, national identity, and disputes over
unification with or independence from the mainland) may all be regarded as
social cleavages.3 Mainlanders have political attitudes and voting behavior that
are distinct from those of other ethnic groups, mainly Minnan and Hakka
(Hsiau 2000; Hughes 1997; Wu 2008; Wu and Hsiao 2006).

In addition to individual characteristics, we consider the effects of cognitive

variables as well. Previous studies have demonstrated that these subjective
psychological determinants could be more important than the
sociodemographic factors noted above. Party identification, which refers to
how closely a person identifies with one of the major political parties, is in
theory closely related to political involvement (Abramson, Aldrich, and Rohde
1995, 72–75; Campbell et al. 1960, 121–23; Miller and Shanks 1996, 154–56). It
is a key part of an individual’s belief system and is characterized by long-term
stability. Previous studies have demonstrated that party identifiers have a
greater degree of political interest than those without any party preference
(Conway 2000, 52–55; Milbrath and Goel 1977, 54). In Taiwan, many voters
think of themselves as Pan-Blue supporters, others consider themselves Pan-
Green.4 The rest—with the exception of the few who identify with a minor
party—are labeled independents.
 Research has shown that independents tend to be less concerned about
politics, have less political information, are less interested in political activities,
and tend to vote less often. Partisans, in contrast, are more involved and
informed and more likely to register and vote, to talk about politics, to evaluate
the outcomes of elections, to discuss candidates’ campaign promises, to try to
influence others, to engage in campaign activities, and so on (Wu and Hsu
2003; Wu and Huang 2007). Partisans are also treated as the object of
mobilization efforts by political parties, so they tend to participate more
actively in the political process.

Intimately connected with ethnicity and party identification is the issue of

national identity, or what is sometimes termed “ethnic consciousness,” which is
widely regarded as an important issue that attracts the most attention Page 255 →
in Taiwan’s politics, and is also considered to be a key variable in research on
political behavior (Fell 2005; Hsiau 2000; Hsieh 2005; Wachman 1994).
National identity comprises individuals’ attitudes and beliefs toward their own
nationality—Taiwanese, Chinese, or both Taiwanese and Chinese. Over the
decades when Taiwan was ruled by a Kuomintang-controlled, mainlander-
dominated authoritarian regime, a China-centered political ideology was the
mainstream value. However, since the beginning of the democratization
process in the mid-1980s, a Taiwan-centered consciousness has gradually risen
to prominence. This is confirmed by there being a greater number of
respondents who identified themselves as “Taiwanese only” compared to those
who chose “Chinese only.”
 It is necessary to explain that although ethnicity and national identity should
be closely related, the former is an objective characteristic, while the latter is
based on a subjective psychological sense of belonging. Regarding the direction
of political participation, it is hypothesized that an individual having a more
intense ideological identification (i.e., those who identified themselves as
“Taiwanese only” and “Chinese only”) is predisposed to participate in politics
more actively.

In addition to party identification and national identity, political knowledge is

another subjective cognitive factor related to political participation. In theory,
participation is strongly linked to information about government and politics.
The available empirical evidence on this point indicates that individuals who
have more information about what the government is doing tend to be more
active politically (Delli Carpini and Keeter 1996, 62–104; Stone and Schaffner
1988, 204–5). Actually, there is a positive-feedback relationship between
participation and information: as individual citizens gain more political
information, they participate more, thereby acquiring more experience and
skills, thus further increasing their political knowledge. Previous studies on
Taiwan politics have confirmed that a person’s political knowledge is directly
related to sociopolitical involvement (Liao 2006; Tsai 2001; Yang 2003). In the
data collected for this chapter, political knowledge is self-reported by the
respondents in surveys. In other words, it reflects the respondents’ confidence
about their own level of political knowledge, which may contain some degree
of bias.

The Historical Trend in Voter Turnout in Taiwan

In 1949 the KMT government retreated to Taiwan after the civil war in
mainland China. From the early 1950s through the mid-1980s, the KMT
regime, in view of its comprehensive domination over the government’s
ruling Page 256 → apparatus, bore the characteristics of an authoritarian one-
party state with elements of totalitarianism (Tien 1989; Winckler 1984). The
control seized by the KMT was comparable to that of the ruling party in a
Leninist-style state, with two exceptions. Those exceptions were the existence
of private ownership and, more significantly, the institutionalization of local
elections (Cheng 1989, 477–78).

The first local elections for executive posts were held in two stages in 1950
and 1951. Over the next five decades, voting was gradually expanded from local
to national elections. From 1950 to 1968, electoral competition was limited to
the chief executives and representative bodies at the city, subcounty, and county
levels and to the provincial assembly. In these elections, no organized political
opposition was permitted to compete with the governing KMT. In 1969, the
authorities initiated limited electoral competition for supplementary
representative seats at the national level. With the lifting of martial law and the
end of the Period of Mobilization for the Suppression of Communist
Rebellion, all members of the National Assembly and the Legislative Yuan were
subject to direct election in 1991 and 1992. The most important development
was the first popular presidential election, which took place in 1996. At
present, all representative bodies and major executive officials—except the
premier, who is appointed by the president—are elected by popular vote.
 After the lifting of martial law in 1987, voting participation in Taiwan’s
national elections, at an average of approximately 70 percent, was relatively
high in comparison with that of other democratic countries. Nevertheless,
Taiwan’s aggregate-level turnout rate for presidential and parliamentary
elections has exhibited a downward trend similar to that in many developed
countries (Powell 1986). Generally speaking, the rates of voter turnout in
presidential elections are somewhat higher than those in parliamentary
elections, as shown in table 11.1. For presidential races, turnout increased from
76 percent in 1996 to 83 percent in 2000, and then slipped to 80 percent in
2004, 76 percent in 2008, and 74 percent in 2012.

The turnout rates for legislative elections exhibit an obviously declining trend.
The turnout in the 1992 Legislative Yuan elections was about 72 percent, which
was high compared to the elections that followed. In 1995, 1998, and 2001, the
average was approximately 67 percent. The level of voter turnout then plunged
to 59 percent in the 2004 year-end legislative election. Save for 2012, which was
the first occasion upon which a presidential election coincided with a legislative
vote, and when, as might be expected, there was a higher turnout rate (Fornos,
Power, and Garand 2004; Nikolenyi 2010), the turnout rate for legislative
elections has consistently Page 257 → decreased, reaching its lowest point in 2008
at about 58 percent. Overall, Taiwan’s voter turnout has declined for both
presidential and legislative elections at the aggregate level. This drop in citizen
engagement raises worrisome questions for the health of Taiwan’s democracy
to which we will return at the end of this chapter.

Explaining Voter Turnout in Taiwan

To understand who votes in Taiwan, we examine the TEDS multiyear survey
data mentioned above. These data allow us to trace the pattern of changes in
the turnout rate in recent elections. In each election year, the survey data
constitute a nationally representative, multistage probability sample of adults
living throughout Taiwan.5

The respondents were asked whether they had cast ballots, but in Taiwan, as
in other countries, people sometimes report that they voted when in fact they
did not.6 The actual turnout of individual citizens is known to the Taiwan
Election Commission, but their records are secret. Unlike in some other
democracies, researchers are not given access to these records, even on a
confidential basis. Thus turnout reports in Taiwan cannot be “validated,” that
is, checked against official records. The result is that turnout in Taiwan as
reported in surveys is higher than the official records, as may be seen in Page 258

→ the charts below.7 Thus surveys are not very good at estimating absolute
levels of turnout. Fortunately, however, turnout comparisons across groups are
less affected by this problem, and that is our focus in this chapter. In order to
provide clearer ideas, we present the results in figures (11.1–11.5) and tables
(11.2–11.6).
Table 11.1. Voter Turnout in Taiwan’s National Elections, 1989–2012

Year Type of election %

Turnout ( )

1989 Supplementary representatives of the Legislative Yuan 75.2

1991 2nd session representatives of the National Assembly 68.3

1992 2nd session representatives of the Legislative Yuan 72.0

1995 3rd session representatives of the Legislative Yuan 67.7

1996 9th presidential election 76.0

3rd session representatives of the National Assembly 76.2

1998 4th session representatives of the Legislative Yuan 68.1

2000 10th presidential election 82.7

2001 5th session representatives of the Legislative Yuan 66.2

2004 March 11th presidential election 80.3

2004 December 6th session representatives of the Legislative Yuan 59.2

2008 January 7th session representatives of the Legislative Yuan 58.5

2008 March 12th presidential election 76.3

2012 January 13th presidential election 74.4

2012 January 8th session representatives of the Legislative Yuan 74.7

Source: Election Study Center, National Chengchi University, http://www.esc.nccu.edu.tw/.

Figure 11.1 shows the differences in voter turnout rates by age group. The
first thing to note is that the reported turnout rate in presidential elections is
higher than that in legislative elections, which is as expected. Moreover, in both
types of elections, the turnout rate increases with age. However, with the
exception of voters in their fifties and older, the reported turnout rate has
gradually decreased with time. Thus the problem of declining interest in voting
is concentrated among younger voters. Their turnout has declined faster than
the average seen in figure 1. This raises special concerns for Taiwan’s
democratic health as Taiwan’s older and more reliable voters inevitably reach
the age at which they can no longer participate actively.

Ethnicity has long been a politically sensitive issue in Taiwan, especially the
cleavage between the majority Taiwanese and the minority mainlanders (Moody
1992; Wachman 1994). Mainlanders are often thought to participate more.
Contrary to expectations, however, the results in figure 11.2 reveal that
mainlanders do not have a significantly higher turnout rate than Minnan and
Hakka. Thus subethnic differences in Taiwan have very little impact on
turnout. As we will see, however, differences reappear when we look at
electoral participation more broadly.

Next, we explore the relationship between an individual’s party identification

and turnout rate, as displayed in figure 11.3. Party identification, in theory, is
closely related to political involvement. As expected, an individual with a
preference for a specific political party, either a Pan-Blue or a Pan-Green
supporter, is more likely to vote in both presidential and legislative elections.
The differences are typically not large (5 to 10 percentage points), but they have
existed in all elections after the 2001 Legislative Yuan contest.

Contrary to expectations, the results in figure 11.4 indicate that there is no

specific relationship between voter turnout and national identity. Individuals
who have a clear national identity (either Taiwanese only or Chinese only) have
relatively high turnout rates; however, the differences are weak and
insignificant. More specifically, those who identify themselves as Chinese are
more likely to vote. Overall, the findings are consistent with the results
displayed in figures 11.2, 11.3, and 11.4—that individuals who have a distinct
idea of their ethnicity, party affiliation, and national identity are more likely to
vote, although these variables merely exert conditional effects on voting
participation.

Fig. 11.1. Voter turnout by age group. Note: From left to right in each year, bars represent voters
in one of five age groups: 20s, 30s, 40s, 50s, and 60s and above. Numbers on the bars represent
the percentage turnout for each age group. L: legislative; P: presidential.

Page 259 → Also as hypothesized, the turnout rate in presidential elections

increases with the level of (self-reported) political knowledge (see figure 11.5).
The turnout rate for individuals with a low level of political knowledge is
around 84–86 percent. The rate for those with either a moderate or high level
of political knowledge is generally higher than 90 percent. However, in the
elections to the Legislative Yuan, the relationship between turnout and
knowledge is more erratic. Not until the 2008 election does a legislative
election exhibit the expected pattern, with more knowledgeable individuals
voting at a higher rate. It may be that the parties formerly mobilized voters
differently in the two kinds of elections but no longer do so; this is a topic that
deserves further investigation.

In summary, an individual’s age (up to 65), party identification, national

identity, and level of political knowledge are all positively correlated with
turnout rates, just as they are in most democracies. Essentially, the more
experienced and more engaged citizens are more likely to vote, as one would
expect. However, we did not find strong differences between Taiwan’s
subethnic groups.
Fig. 11.2. Voter turnout by ethnicity. Note: From left to right in each year, bars represent Hakka,
Minnan, and Mainlanders. L: legislative; P: presidential.

Page 260 →

Explaining Other Kinds of Political Participation in Taiwan

Voter turnout is just one aspect of democratic participation. Citizens may

influence their representatives in many different ways, and all of those pathways
matter for democratic responsiveness. Thus we now proceed to examine other
types of participation, using the same explanatory factors we used to study
voter turnout. Due to the different types of participation being recorded in our
multiyear datasets, we divided respondents into two groups, participants and
nonparticipants, for analytical convenience. Participants are defined as those
who took part in at least one form of political activity in the surveyed time
period, while nonparticipants are those who indicated an abstention from all
such activities.8

Differences in electoral participation by age group, by an individual’s ethnicity,

by party identification, national identity, and by level of political knowledge are
displayed in figures 11.6–11.10.

Age, ethnicity, party identification, national identity, and political knowledge

are all more or less positively correlated with the level of an individual’s
electoral participation, as was the case with turnout rates. However, Page 261

→ all these explanatory factors make a bigger difference to electoral

participation broadly defined than they do to turnout. Age matters more;
mainlanders participate more, at least in recent elections; and partisanship,
political knowledge, and national identity matter much more significantly than
they did for turnout. Thus substantial inequality in political participation
persists in Taiwan, with older, more engaged mainlanders participating more.
 Fig. 11.3. Voter turnout by party identification. Note: From left to right in each year, bars
represent Pan-Blue identifiers, Pan-Green identifiers, and independents. L: legislative; P:

presidential.

If this trend of inequality in participation continues, it will have a negative

impact on the future development of democracy in Taiwan. If younger people
drop out of the politically active population, the nation’s political agenda will
be dominated by the interests of older, possibly more conservative, voters in
the future. Economically, this might result in more government resources spent
on welfare programs geared toward the elderly as well as protection of
inefficient traditional industries at the expense of financing future-oriented
policies, such as improving education and providing incentives for starting new
businesses. Politically, it might cause young people to become disillusioned
with the democratic establishment, pushing them toward ever more radical
methods of promoting their own interests. The Page 262 → recent Sunflower
Student Movement is an illustration of youth’s lack of confidence in a political
system led by an older generation with apparently little concern for the needs
of the young.

Fig. 11.4. Voter turnout by national identity. Note: From left to right in each year, bars represent
voters who identify themselves as: Taiwanese only, both Taiwanese and Chinese, and Chinese

only. L: legislative; P: presidential.

Conclusion

This chapter covers the development of voter turnout and electoral

participation in Taiwan. Both voting and participation in electoral activities are
essential elements in the formation and maintenance of democracy. Elections,
especially, are the most important way of promoting political participation
among ordinary citizens. Through elections, citizens select political leaders and
shape public affairs, while the government uses them to guarantee its legitimacy
(Jackman 1987, 405–6). According to Lipset (1981, 27), “democracy is a
complex society . . . which supplies regular constitutional Page 263

→ opportunities for changing the governing officials, and a social mechanism

which permits the largest possible part of the population to influence major
decisions by choosing among contenders for political office.” In a democracy,
political participation affects the distribution of social values, and one way of
judging its effectiveness is to see who plays an active role in the political
process, and how much they participate in it.
 Fig. 11.5. Voter turnout by self-assessed political knowledge. Note: From left to right in each
year, bars represent low, moderate, and high levels of political knowledge. Because different
survey questions on political knowledge were used, varying scoring scales were adopted. For

data from TEDS 2001, TEDS 2004L, TEDS 2004P, and TEDS 2008L, we recoded political
knowledge by level: low (score 0–1), moderate (2–3), and high (score 4–5); for data from TEDS
2008P, scoring was recoded to low (score 0–1), moderate (score 2), and high (score 3); for data

from TEDS 2012, scoring was recoded to low (score 0–2), moderate (score 3–4), and high (score
5–7). L: legislative; P: presidential.

Despite the importance of political participation to democracy, we find that

empirical research on this topic has been limited primarily to Western
countries, and that systematic analysis of electoral participation in developing
democracies such as Taiwan remains scarce. Research into the level of political
participation among Taiwan citizens is therefore likely to have implications for
government authorities and civic groups, as they seek to stabilize Page 264 → and
strengthen the island’s fledgling democratic institutions. The need to deduce
what drives political participation in Taiwan is even more pressing due to a
significant downward trend in turnout for both presidential and legislative
elections in recent years.

Fig. 11.6. Electoral participation by age group. Note: From left to right in each year, bars
represent the percentage of voters in one of five age groups: 20s, 30s, 40s, 50s, and 60s and
above. L: legislative; P: presidential.

We end this chapter as we began it, by highlighting that political participation

can be affected by sociodemographic characteristics and subjective cognition,
as well as people’s personal attitudes and life experiences. Political socialization
is a continuing process. Reviewing the similarities and differences displayed
above, we note in particular that voters in their fifties and sixties, those who
identify with a particular party, those who perceive themselves to be Chinese,
those who have high levels of political knowledge, and those who are
mainlanders are more likely to vote and engage in political activities. This more
engaged group can be seen as the established higher social class within Taiwan
society. They tend to be an economically well off sector whose interests are at
least in part opposed to those of younger voters.

In this sense, social and economic factors are fundamental to political

participation. The findings of this chapter confirm the proposition that the Page

265 → higher a person’s social status, the more likely they are to be an active
participant in political life. People in the upper and middle classes tend to be
more interested and involved in public affairs than lower class individuals. The
unequal distribution of political participation may have the effect of driving
Taiwan toward an “elite regime” in which the established upper class will be
able to use its financial resources and political knowledge to dictate the policy
agendas of elected politicians, gradually creating an environment more
favorable to corporate and business interests at the expense of labor. Such a
development would not only further strengthen the dichotomy between the
young and the old but also create an insurmountable gap in power between the
urban and the rural populations, the rich and the poor, as well as between the
politically connected and unconnected members of the public. Socioeconomic
disparity between groups and regions would be accompanied by political
unfairness, and the democratic system would no longer guarantee an equal
voice for all voters as it was originally meant to do.
Fig. 11.7. Electoral participation by ethnicity. Note: From left to right in each year, bars represent
Hakka, Minnan, and Mainlanders. L: legislative; P: presidential.

The potentially divisive nature of unequal political participation means that it

is necessary for Taiwan to reengage those groups of voters who are
underrepresented Page 266 → in the electorate. To be fundamentally and
sustainably effective, this process of reengagement must emphasize cooperation
between the higher levels of society and the grassroots. On the one hand,
populist civic groups must be established to educate people about the long-
term harm caused by political indifference. These groups should not only be
encouraged but also financed, if necessary, by the government. On the other
hand, the government must be tolerant of dissent. Negotiations with the
Sunflower Student Movement and even partial incorporation of their platform
into official policy, for instance, may trigger renewed interest among those
young people who have become disillusioned with politics.

Fig. 11.8. Electoral participation by party identification. Note: From left to right in each year,
bars represent Pan-Blue identifiers, Pan-Green identifiers, and independents. L: legislative; P:
presidential.

What is most important, however, is the need to continue monitoring the

levels of political participation by different groups, as has been done in the
research presented in this chapter. The ability to identify which groups have
become estranged from the mainstream political establishment allows measures
to be taken to rectify this state of affairs before the drifting away of certain
voters creates systemic problems for policymakers and for Taiwan society.
Fig. 11.9. Electoral participation by national identity. Note: From left to right in each year, bars
represent voters who identify themselves as: Taiwanese only, both Taiwanese and Chinese, and

Chinese only. L: legislative; P: presidential.

Page 267 →

Appendix 11.A1. Survey Questions and Coding of

Variables
Voter Turnout. “Did you vote for the Legislative Yuan Election?” (1 = yes; 2 = no) (TEDS 2001L)
(TEDS 2004L) (TEDS 2008L)
 “In this presidential election many people went to vote, while others, for various reasons, did not go to
vote. Did you vote?” (1 = yes; 2 = no) (TEDS 2004P) (TEDS 2008P) (TEDS 2012)
Electoral Participation. “Respondents’ total amount of political activities listed below: did volunteer
campaign work for either a candidate or a party; attended an election-related gathering or banquet;
joined a candidate’s support organization; reminded friends to watch candidate debates or campaign;
persuaded others to vote for a particular candidate or party; gave money to a political party or
candidate; purchased a candidate’s souvenirs; attended a candidate’s rally) (1 = none; 2 = at least one of
them) (TEDS 2001L) (TEDS 2004P) (TEDS 2004L) (TEDS 2008P) (TEDS 2012)

Fig. 11.10. Electoral participation by self-assessed political knowledge. Note: From left to right in
each year, bars represent low, moderate, and high level of political knowledge. L: legislative; P:
presidential.

Page 268 → Age. Respondent’s age measured in years. (1 = 20 to 29; 2 = 30 to 29; 3 = 40 to 49; 4 = 50 to
59; 5 = above 60)
Ethnicity. Ethnic background of respondent’s father. (1 = Hakka; 2 = Minnan; 3 = Mainlander;
aborigines coded as missing)
Party identification. “Among the main political parties in our country, including the KMT, DPP, PFP,
NP, and TSU, do you think of yourself as leaning toward any particular party?” “Which party is
that?”(1 = Pan-Blue supporter [KMT, NP, and PFP]; 2 = Pan-Green supporter [DPP and TSU])
 National identity. “In Taiwan, some people think they are Taiwanese. There are also some people who
think that they are Chinese. Do you consider yourself as Taiwanese, Chinese, or both?” (1 = Taiwanese;
2 = both; 3 = Chinese)
Political knowledge. Respondents’ total amount of correct answers to the following questions: “Who is
the current Vice President of our country?”; “Who is the President of the PRC?”; “Who is the current
President Page 269 → of the United States?”; “How many years are a legislator’s term?”; and, “Which
institution has the power to interpret the Constitution?” (1 = low [total correct answers are 0–1]; 2 =
moderate [total correct answers are 2–3]; 3 = high [total correct answers are 4–5]) (TEDS 2001L)
“Who is the President of the PRC?”; “Who is the current President of the United States?”; “How many
years are a legislator’s term?”; “Which institution has the power to interpret the Constitution?”; and,
“Who is the current Vice President of our country?” (1 = low [total correct answers are 0–1]; 2 =
moderate [total correct answers are 2–3]; 3 = high [total correct answers are 4–5]) (TEDS 2004P)
“Who is the current Vice President of our country?”; “Who is the President of the PRC?”; “Who is the
current President of the United States?”; “How many years are a legislator’s term?”; and, “Which
institution has the power to interpret the Constitution?” (1 = low [total correct answers are 0–1]; 2 =
moderate [total correct answers are 2–3]; 3 = high [total correct answers are 4–5]) (TEDS 2004L)
“Who is the current President of the United States?”; “Who is the current premier of our country?”;
“Which institution has the power to interpret the Constitution?” (1 = low [total correct answers are 0–
1]; 2 = moderate [total correct answers are 2]; 3 = high [total correct answers are 3]) (TEDS 2008P)
“Who is the current Vice President of our country?”; “Who is the current President of the United
States?”; “Who is the President of the PRC?”; “Which institution has the power to interpret the
Constitution?”; and, “How many years are a legislator’s term?” (1 = low [total correct answers are 0–
1]; 2 = moderate [total correct answers are 2–3]; 3 = high [total correct answers are 4–5]) (TEDS
2008L)
“Who is the current President of the United States?”; “Who is the current premier of our country?”;
“Which institution has the power to interpret the Constitution?”; “Which of these persons was the
finance minister before the recent election?”; “What was the current unemployment rate in Taiwan as
of the end of last year”; “Which party came in second in seats in the Legislative Yuan?”; and, “Who is
the current Secretary-General of the United Nations, Kofi Annan, Kurt Waldheim, Ban Ki-moon, or
Boutros-Ghali?” (1 = low [total correct answers are 0–2]; 2 = moderate [total correct answers are 3–4];
3 = high [total correct answers are 5–7]) (TEDS 2012)
Page 270 →
Notes

1. Data analyzed in this chapter were collected as part of the research

project entitled “Taiwan’s Election and Democratization Study, 2012:
Presidential and Legislative Elections” (TEDS 2012) (NSC 100–2420-
H002–030). The coordinator of the multiyear TEDS project is Chi Huang
of the Department of Political Science at National Chengchi University.
The principal investigator is Professor Yun-han Chu of the Institute of
Political Science at Academia Sinica. More information is on the
TEDS website (http://www.tedsnet.org). The following were responsible
for distributing the date: the Department of Political Science, National
Taiwan University; the Department of Political Science, Soochow
University; the Graduate Institute of Political Science, National Sun Yat-
Sen University; the Department of Political Science and Graduate
Institute of Political Economy, National Cheng Kung University; the
Department of Political Science, Tunghai University; and the Election
Study Center, National Chengchi University. The authors appreciate the
assistance of the institutes and individuals aforementioned in providing
data. This research is partially supported by National Chengchi
University’s Top University Project. The views expressed in this chapter
are those of the authors alone.

2. For a discussion of ethnic and subethnic differences in Taiwan, see

chapter 1.

3. Although Minnan and mainlanders have different attitudes and

characteristics, in reality politics is the main factor generating the ethnic
consciousness of both groups. See chapters 1 and 3.
4. As a reminder for readers, the Taiwan party system is generally divided
into the Pan-Blue camp, which espouses eventual political unification
with China, and the Pan-Green camp, which consists of supporters of
Taiwan independence. The major parties of the Pan-Blue camp are the
Kuomintang, the People First Party, and the New Party, while the Pan-
Green parties are the Democratic Progressive Party, the Taiwan Solidarity
Union, and the Green Party.

5. The TEDS data are weighted by gender, age, and education to achieve
national representativeness. The population statistics are based on census
data reported in the official documents, Taiwan-Fukien Demographic
Statistics, Republic of China, released by the Ministry of the Interior,
Republic of China.

6. The topic of vote misreporting is important both theoretically and

practically (Wu 2006, 224). Research on both electoral turnout and vote
choice depends heavily on self-reported behavior, but it is generally found
that a number of respondents do not accurately report their electoral
behavior. One cause of errors in survey research is that more respondents
claim to have voted in postelection interviews than have actually cast
ballots. A possible consequence is that misreporting does indeed produce
some misleading conclusions, since much of the scholarly work tests
models of electoral behavior based on survey measurements containing a
relatively large amount of error.

7. In figure 11.2, for example, all but the small, youngest group reported
turnout rates exceeding 80% for the 2001 Legislative Yuan election, but as
figure 11.1 shows, the actual turnout rate in that election was only 67%.
 8. The 13 electoral activities in the TEDS are as follows: read the official
election notice; read candidates’ leaflets, newsletter, or newspaper ads;
watched candidate Page 271 → debates or campaign speeches on TV; did
volunteer campaign work for either a candidate or a party; attended an
election-related gathering or banquet; joined a candidate’s support
organization; reminded friends to watch candidate debates or campaign;
persuaded others to vote for a particular candidate or party; gave money
to a political party or candidate; purchased a candidate’s souvenirs; invited
to participate in a candidate’s rally; attended a candidate’s rally; and visited
a candidate’s website.

References

Abramson, Paul R., John H. Aldrich, and David W. Rohde. 1995. Change and
Continuity in the 1992 Elections. Washington, DC: Congressional Quarterly
Press.

Almond, Gabriel A., and G. Bingham Powell, Jr. 1988. “Political Socialization
and Political Culture.” In Comparative Politics Today: A World View, 4th ed.,
ed. Gabriel A. Almond and G. Bingham Powell, Jr., 34–48. Glenview, IL:
Scott, Foresman and Company.

Campbell, Angus, Philip E. Converse, Warren E. Miller, and Donald E. Stokes.

1960. The American Voter. New York: John Wiley and Sons.

Cheng, Tun-jen. 1989. “Democratizing the Quasi-Leninist Regime in Taiwan.”

World Politics 41 (4): 471–99.
Conway, M. Margaret. 2000. Political Participation in the United States. 3rd ed.
Washington, DC: Congressional Quarterly Press.

Dahl, Robert A. 1971. Polyarchy: Participation and Opposition. New Haven: Yale
University Press.

Delli Carpini, Michael X., and Scott Keeter. 1996. What Americans Know about
Politics and Why It Matters. New Haven: Yale University Press.

Fell, Dafydd. 2005. Party Politics in Taiwan. London: Routledge.

Fornos, Carolina A., Timothy J. Power, and James C. Garand. 2004.

“Explaining Voter Turnout in Latin America, 1980 to 2000.” Comparative
Political Studies 37 (8): 909–40.

Hsiau, A-chin. 2000. Contemporary Taiwanese Cultural Nationalism. London:

Routledge.

Hsieh, John F. 2005. “Ethnicity, National Identity, and Domestic Politics in

Taiwan.” Journal of Asian and African Studies 40 (1–2): 13–28.

Hughes, Christopher. 1997. Taiwan and Chinese Nationalism. London: Routledge.

Jackman, Robert W. 1987. “Political Institutions and Voter Turnout in the

Industrial Democracies.” American Political Science Review 81 (2): 405–24.

Liao, Yih-Hsing. 2006. “The Research of Voter Turnout: Case Study of

Taiwan.” Journal of Chinese Public Administration 3: 185–202.
Lipset, Seymour M. 1981. Political Man: The Social Bases of Politics. Baltimore:
Johns Hopkins University Press.

Milbrath, Lester, and M. L. Goel. 1977. Political Participation. Chicago: Rand

McNally.

Miller, Warren E., and J. Merrill Shanks. 1996. The New American Voter.
Cambridge: Harvard University Press.

Moody, Peter R., Jr. 1992. Political Change in Taiwan. New York: Praeger.

Page 272 → Nie, Norman H., and Sidney Verba. 1975. “Political Participation.”
In Handbook of Political Science: Volume 4, Nongovernmental Politics, ed. Fred I.
Greenstein and Nelson W. Polsby, 1–74. Reading, MA: Addison-Wesley.

Nikolenyi, Csaba. 2010. “Concurrent Elections and Voter Turnout: The Effect
of the De-Linking of State Elections on Electoral Participation in India’s
Parliamentary Polls, 1971–2004.” Political Studies 58 (1): 214–33.

Powell, G. Bingham, Jr. 1986. “American Voter Turnout in Comparative

Perspective.” American Political Science Review 80 (1): 17–43.

Rosenstone, Steven J., and John Mark Hansen. 2003. Mobilization, Participation,
and Democracy in America. New York: Longman.

Stone, William F., and Paul E. Schaffner. 1988. The Psychology of Politics. 2nd ed.
New York: Springer-Verlag.

Tate, Katherine. 1991. “Black Political Participation in the 1984 and 1988
Presidential Elections.” American Political Science Review 85 (4): 1159–76.
Tien, Hung-mao. 1989. The Great Transition: Political and Social Change in the
Republic of China. Stanford, CA: Hoover Institution Press.

Tsai, Chia-Hung. 2001. “Why Do Taiwanese Vote?” Journal of Electoral Studies 8

(2): 125–58.

Verba, Sidney, and Norman H. Nie. 1972. Participation in America: Political

Democracy and Social Equality. New York: Harper and Row.

Wachman, Alan M. 1994. Taiwan: National Identity and Democratization. Armonk,

NY: M. E. Sharpe.

Wang, Fu-chang. 1994. “Ethnic Assimilation and Mobilization: An Analysis of

Party Support in Taiwan” [in Chinese]. Bulletin of the Institute of Ethnology
Academia Sinica 77 (Spring): 1–34.

Wang, Fu-chang. 1998. “Ethnic Consciousness, Nationalism, and Party

Support: Taiwan’s Ethnic Politics in the 1990s” [in Chinese]. Studies on
Taiwan’s Sociology 2 (July): 1–45.

Winckler, Edwin A. 1984. “Institutionalization and Participation on Taiwan.”

China Quarterly 99: 481–99.

Wu, Chung-li. 2006. “Vote Misreporting and Survey Context: The Taiwan
Case.” Issues & Studies 42 (4): 223–39.

Wu, Chung-li. 2008. “Ethnicity, Empowerment, and Political Trust: The 2005
Local Elections in Taiwan.” Issues & Studies 44 (1): 105–32.
Wu, Chung-li, and Cheng-tai Hsiao. 2006. “Empowerment Theory and Ethnic
Politics in Taiwan.” Issues & Studies 42 (1): 103–36.

Wu, Chung-li, and Wen-pin Hsu. 2003. “Who Are Partisans and Independents?
Determinants of Party Identification of Taiwan’s Voters in 2001” [in
Chinese]. Political Science Review 18: 101–40.

Wu, Chung-li, and Chi Huang. 2007. “Divided Government in Taiwan’s Local
Politics: Public Evaluations of City/County Government Performance.”
Party Politics 13 (6): 741–60.

Yang, Meng-Li. 2003. “Voter Turnout and Economic Adversity in Taiwan.”

Journal of Electoral Studies 10 (2): 159–91.
Page 273 →
 Chapter 12

Conclusion

The Power of Identity in Taiwan

Christopher H. Achen and T. Y. Wang

Most countries are politically divided along lines that reflect their internal
cleavages. Those cleavages may be religious, linguistic, ethnic/racial, regional,
or class-based. Most countries contain several politically relevant divisions. In
multiparty systems, particularly those with proportional representation and a
low threshold for gaining parliamentary seats, some minor cleavages (farmers,
small ethnic or linguistic minorities) may have their own party in the
legislature. Larger parties usually represent coalitions. In two-party systems,
both parties are large, and in consequence each party will represent a broad
combination of groups. Large parties frequently reflect many entirely distinct
and unrelated cleavages in the society, sometimes in ways that make common
sense, and sometimes not. In the United States, for example, liquor distributors
and conservative evangelical Christians, once bitter enemies in the Prohibition
era, now find themselves side by side in the Republican Party.

Thus, in most countries the major parties embody many different identities.
The voters, too, are often a jumble of identities, some more strongly felt than
others, but all of them subject to activation and mobilization under the right
circumstances. Studying the role of social identities in politics is typically quite
difficult. Different voters will identify with a particular party for very different
reasons. Many voters themselves will have more than one identity that drives
them toward a particular party, and perhaps some other identities that are in
conflict with that party.

As the parties adopt new positions, some voters will feel conflicted. Most Page

274 → will stay with their partisanship, simply living with the tensions or tuning
them out. Others will move toward political independence, perhaps eventually
switching to another party with which, again, they are in imperfect agreement.
But the result is that at any given time, there is no simple relationship between
identity, partisanship, and issue positions. All three are tangled up in not wholly
consistent ways.

In consequence, the study of how identities relate to partisanship and political

attitudes is very complex in most countries. A small proportion of voters, often
discriminated-against minorities, may have one main identity driving their
political stances. But for most voters, too many things are affecting too many
other things. Voters have too many identities. Even when identities are the
main factor driving party choices and issue positions, the catch-all nature of the
parties and the complexity of the voters’ own political lives create a vortex of
causal arrows, making it extremely difficult to discern why the voters are
thinking and choosing as they do. Any attractive theoretical account comes up
against mixed empirical support and plausible counterarguments. Put another
way, most countries are not very good places to study how social identities
connect to political identities, issue preferences, and vote choices.
 Taiwan, however, is a happy exception, as this book has demonstrated. Social
cleavages are few. Apart from a tiny minority, Taiwan voters do not differ
racially. The great majority adhere to a low-intensity, syncretist religious
tradition with a mixture of Buddhist, Taoist, and Confucian elements, which
has never been a source of political divisions. In the absence of an exploitative
industrial revolution and without the associated development of strong
employee unions, social class has not been central to Taiwan politics. (Compare
the heavy emphasis on class in treatments of European politics—a recent
example is Evans and De Graaf 2013.) In Taiwan, the principal division is
linguistic and ethnic—whether the language one speaks at home is Mandarin,
Hakka, Hokkien (Fujianese), or one of the aboriginal languages. In turn, this
cleavage relates to the historical time of arrival on Taiwan from elsewhere, and
the associated history and culture of each group.

As we discussed in chapter 1, events of the past 70 years have reinforced those

linguistic and cultural divisions and made them politically salient. Other
divisions are much weaker. That means that Taiwan has just one central
cleavage—a gift to scholars trying to understand how identity operates in
politics. Of course, as we have seen, that cleavage has evolved. Once tied more
strictly to ethnicity, with a Mandarin-speaking ruling elite enforcing their
culture on everyone else, the issue is now more closely related to differing
conceptions of national identity.

Page 275 → Ethnicity still matters: only a minority of mainlanders identify as

strictly Taiwanese, and extremely few Minnan identify as purely Chinese. But
with a more open society, intermarriage, and the passage of time, the categories
have blurred. Some Hakka and aboriginal citizens, with their own historical
grievances against the Minnan, feel free to line up politically with mainlanders,
for example. What matters less and less are the divisions of the 1950s. What
matters more and more, regardless of ethnicity, is where one stands on the
status of Taiwan. Is China the ancestral homeland of which Taiwan is an
integral part, even if currently administered separately? Or is China a different
country from Taiwan, home to a fundamentally different people? Debates of
this kind are familiar from the history of many countries.1

In this book, we have addressed the question of how Taiwan voters make
their decisions when they go to the polls. We have found that the central
political cleavage and its associated social and political identities are central to
voters’ thinking. Candidates’ personal traits, the domestic issues of the day,
cross-Strait relations, and Taiwan’s institutional arrangements all play a modest
role as well. But what shapes politics on the island much more than anything
else is “the China factor,” the central dispute over Taiwan’s national identity.
Over and over again in this book, we have found that it dominates voters’
decisions. And because left-right language is used to describe that division by
almost no one in Taiwan, the conventional Western view that “left” and “right”
apply everywhere in one form or another among knowledgeable citizens simply
collapses when applied to Taiwan, as chapter 9 showed.

Because the two main parties are perceived to take opposite sides on the
fundamental cleavage, partisanship embodies the same electoral division. Thus,
whether the national identity issue directly shapes some policy dispute, or
whether partisanship structures it instead, the result is the same: the dispute
will be molded by the underlying cleavage over Taiwan’s national identity.
Nothing else matters to the same degree, and certainly not the conventional
left-right dimension that gives form to politics in most Western countries.

Repeatedly, therefore, the analyses in the various chapters have identified

partisan identification as the most important factor in Taiwan voters’ electoral
calculus.2 That is, the island citizens’ self-declared partisan affiliations with the
Pan-Blue Alliance and the Pan-Green Alliance exert the most significant effects
on how and why they support specific candidates. As figure 12.1 shows,
partisanship has an extremely powerful effect in Taiwan. 3 In the 2012
presidential election, knowing the voter’s partisanship was tantamount to
knowing how he or she voted in the vast majority of the cases. Only
independents fell toward the middle: everyone else was polarized. In
contemporary Page 276 → Taiwan, there simply is not much about the vote left
to explain once partisanship is accounted for.

Fig. 12.1. Pan-Blue vote in 2012 Taiwan presidential election

 The finding that partisan identification plays an important role in
understanding citizens’ political behavior in democratic polities is not new, of
course. More than half a century ago, the authors of The American Voter
convincingly demonstrated “the role of enduring partisan commitments in
shaping attitudes toward political objects” (Campbell et al. 1960, 135). They
concluded that “the strength and direction of party identification are facts of
central importance in accounting for attitude and behavior” (121).

What Taiwan voters add to this familiar story is that their partisan
identifications largely embody the single most important political cleavage on
the island—Taiwan’s future relationship with China, generally characterized as
the issue of unification vs. independence. Because this key political cleavage is
also intertwined with the island citizens’ psychological attachment to China (or
detachment from it), partisan identification mirrors Taiwan voters’ conception
of Taiwan’s national identity. The Pan-Green Alliance sees Taiwan’s de jure
independence and its permanent separation from China as the ultimate
objective, whereas the Pan-Blue Alliance does not preclude the island’s eventual
unification as a possible outcome.

In the public’s view, the Pan-Green Alliance is pro-independence while Page

277 → the Pan-Blue Alliance is pro-unification. As these are two diametrically

different positions, the Taiwan voters’ decision to adopt a partisan
identification is easier than in many other countries—the choice is clear and
unmistakable. Cross-Strait relations and Taiwanese/Chinese identity have
always been key campaign issues in presidential elections. All other issues are
secondary, as Sheng and Liao’s chapter demonstrates. Subjects related to
Taiwan’s relationship with China have been main instruments employed by the
two political alliances to energize their supporters. Also, because Taiwan’s party
structure reflects the key political cleavage in the society, all other issues, even
essentially administrative policy issues such as absentee voting or the adoption
of an English translation system, can be formulated as aspects of the main
cleavage and debated in partisan terms, as we saw in chapter 1. Of course, the
impact of partisanship varies across countries, as hundreds of studies have
shown. (The case of Latin America, for example, is discussed in Lupu 2015.)
The clarity of choice and the polarization provided by the party system are
often thought to enhance the development of partisanship. (The European case
is treated in Berglund et al. 2005 and van der Eijk, Schmitt, and Binder 2005,
177–80; see Zechmeister 2015 for the Latin American counterpart.) Both clarity
and polarization apply to Taiwan, and the strength of partisanship there
confirms the usual comparative logic of how citizens become partisans. These
Taiwan findings handsomely coincide both with Western studies and with
Bartels’s conclusion in the American political setting that partisanship is “a
pervasive dynamic force shaping citizens’ perceptions of, and reactions to, the
political world” (Bartels 2002, 138).

Thus, partisan identification plays a central role in the Taiwan voter’s electoral
calculus, as it does elsewhere. Yet other countries’ voters often give weight to
the character of the candidates and the state of economy in making their vote
choices. Is that true in Taiwan? The chapter by Wang and Chen (chapter 8)
clearly shows that the notion of candidate issue ownership is applicable in
Taiwan. That is, voters’ perceptions of candidate traits are closely connected
with party labels. In general, the Pan-Blue candidates, or more precisely the
KMT candidates, are perceived as more capable of handling issues related to
economic development and cross-Strait relations, whereas the Pan-Green, or
the DPP, candidates are associated with eliminating corruption and initiating
political reforms. Yet the perception of personal traits does not provide
substantial advantages or disadvantages to candidates during elections.
Citizens’ electoral decisions continue to be conditioned by their partisan
affiliations.

As Lewis-Beck (1988) and Paldam (1991) demonstrated, and as many

subsequent studies have confirmed, the strength of economic voting varies Page

278 → dramatically from one country to the next. Powell and Whitten (1993)
and Anderson (2007) each argued that context matters: if clarity of
responsibility is unclear, for example, or if no credible alternative government
exists, economic voting may diminish or disappear. (Gelineau and Singer 2015
review the literature and discuss the Latin American case.) In the case of
Taiwan’s essentially two-party system, the president’s responsibility is clear and
the opposition is credible. Economic voting should occur. Indeed, the Taiwan
public’s assessment of the economy does appear to have an effect on their
voting decisions. However, Taiwan voters appear to be rather tolerant of
incumbents’ past economic performance. Rather than punishing sitting
presidents for a bad economy, they apparently have a tendency to reward
candidates that offer a promising economic future. That is, prospective
economic voting is more prevalent in Taiwan than retrospective voting, as both
the postelection survey data and the aggregate cross-county economic evidence
suggest. Nevertheless, as Tsai concludes in chapter 8, Taiwan voters’
“partisanship is a better predictor” of electoral outcomes than their assessment
of the state of the economy.
 Comparative Strength of the Factors Influencing Vote
Choice

In this book, we have reviewed a variety of factors that influence vote choice—
ethnicity, national identity, issues, and economic evaluations—and we have
compared each with partisanship. Repeatedly, we have found that partisanship
was the controlling factor, with an impact much stronger than any of the other
single forces at work. However, we have not yet assessed these competing
factors jointly. We have seen only that partisanship is dominant, but not which
of the remaining influences on the vote are also somewhat consequential. It is
to that task that we now turn.

We begin by using all available explanatory variables to account for the

presidential vote in 2012. The small vote (2.8%) for James Soong is grouped
with that for Ma Ying-jeou to form the Pan-Blue vote; the vote for Tsai Ing-
wen is the Pan-Green vote. Logit analysis is employed to model this
dichotomous variable. In addition to party identification, the explanatory
variables include the three principal ethnicities (with Hakka as the excluded
category), and the three types of national identification (“Taiwanese,”
“Chinese,” and “Dual”), with “Dual” (i.e., “both Chinese and Taiwanese”) as
the excluded identity. We also create a category (“No ethnic identity”) for the
small but distinctive group who replied “other,” “don’t know,” or who refused
to answer the question. This group is more senior (a large majority older than
49) and less educated (a majority with junior high school or less). Page 279

→ More than 70 percent have no partisanship, but they often retain the one-
party KMT voting loyalties of their youth.
 Other variables included were measures of opinion on
unification/independence, on social welfare expenditures, and on the cross-
Strait economic agreement (ECFA). Economic evaluations were also included
—first, the county-level change in disposable income per capita, and then also
the respondent’s prospective and retrospective evaluations of the national
economy. (See chapter 6 for the definitions of these variables.) Preliminary
exploration of the data indicated that the only category of occupation with a
possible substantial effect on vote choice was being a postsecondary student.
Similarly, having a junior high school education or less was the only possibly
influential category of education. Dummy variables for each were included in
the explanatory model, along with two variables for demographic factors—
gender and age. All variables were scaled to 0–1 for easy comparison.4

The first column of table 12.1 displays the result of this initial exploratory
model.5 Partisanship is by far the most powerful factor, as expected. The other
variables generally take on their expected sign, although not all are statistically
significant. Retrospective evaluations in particular fail to be influential, just as
one would expect from the results in chapter 8. Age is also neither powerful
nor significant, and in most alternate specifications it looked even less
influential than it does here. And of course, that is to be expected, especially
because the age distribution of ethnicities and identities does not differ in
Taiwan: there are young and old in all groups.

The second column of table 12.1 therefore drops retrospections and age, and
reestimates the model. As expected, very little changes. Thus, if retrospections
or age somehow have small effects, we can be comfortable that their exclusion
is not distorting the impact of other factors.
 The issue that arises next concerns the prospective evaluations. The TEDS
survey is conducted after the election. Thus, when voters are asked whether
“the state of the economy of Taiwan will get better, stay about the same, or get
worse,” they already know who won the election. We have known since The
American Voter (Campbell et al. 1960, 397–400) that economic outlooks are
powerfully influenced by partisanship: winners are optimistic, losers are
pessimistic. Thus, as chapter 8 discussed, some scholars have expressed the
concern that the TEDS prospective economic evaluations appear influential
only because they are caused by the dependent variable, and not because they
have genuine explanatory power on their own. Thus, the impact of prospective
evaluations may be substantially inflated in column 2. There is no way to know
for certain with the data available.
Page 280 → Table 12.1: The Pan-Blue Presidential Vote in 2012

(1) (2) (3) (4)

9.72*** 9.71*** 9.68*** 9.59***

Party identification
(.702) (.687) (.626) (.620)

−.464 −.492 −.150 −.138

Minnan
(.326) (.324) (.287) (.286)

1.68*** 1.65*** 2.19*** 2.16***

Mainlander
(.603) (.604) (.571) (.567)

-.414* -.392 -.424 -.472**

Taiwanese ID only
(.252) (.250) (.228) (.226)

.084 .045 -.128 -.166

Chinese ID only
(.772) (.777) (.651) (.648)

1.23 1.39 .200 .080

No ethnic ID
(1.11) (1.06) (.598) (.589)

1.54** 1.54** 1.79*** 1.80***

Unification/independence
(.705) (.701) (.596) (.593)

.603 .634* .410

Social welfare
(.379) (.379) (.322)

1.85*** 1.79*** 1.95*** 1.89***

ECFA
(.409) (.392) (.356) (.354)

.773* .808* .407

2011 change in county disposable income
(.459) (.457) (.400)

1.11*** 1.08***
Prospective evaluation
(.388) (.355)

−.222
Retrospective evaluation
(.395)
Postsecondary education only −.926 −.840 −.511

(.616) (.592) (.535)

.935** .839*** .913*** .987***

Primary education only
(.394) (.312) (.248) (.244)

.326 .327 .436** .435**

Female
(.229) (.227) (.201) (.200)

−.484
Age
(.797)

−6.61 −6.78 −6.69 −6.42

Intercept
(.733) (.705) (.622) (.587)

Pseudo-R2 .67 .67 .64 .64

N 1,190 1,203 1,393 1,393

Note: * significant at .10. ** significant at .05. *** significant at .01. Dependent variable is 1 if the
respondent voted for pan-Blue (Ma or Soong); 0 if for DPP (Tsai). Logit parameter estimates (with
standard errors in parentheses).

Page 281 → To allow for this possibility, column 3 of table 12.1 drops
prospective evaluations from the statistical model. Partisanship remains as
powerful as ever, as do most of the other variables, but the impact of social
welfare attitudes drops, and the estimated effects of changes in county
disposable income and postsecondary education are cut in half. None of the
latter three coefficients is large and none is even close to statistical significance:
they may have a modest effect, but there is not enough evidence to support
keeping them as explanatory factors.6

Column 4 of table 12.1 then drops these three variables with unprovable
effects. The remaining coefficients change little, letting us assess the ethnicity
and national identification effects. Column 4 shows that being a mainlander has
an important positive impact on voting pan-blue, as expected, but being
Minnan has only a small and statistically insignificant effect. That is, the
Minnan are not distinguishable from the excluded category (Hakka). And
second, being a Taiwanese identifier has a substantial and significant negative
effect on voting pan-Blue, again as expected, but the other categories of ethnic
identity (“Chinese” and none) have small and statistically insignificant effects;
that is, they are indistinguishable from the excluded category (“dual identity”).
Thus, within the limits of the available data, there are actually just two
dichotomies where ethnicity and national identification are concerned:
mainlanders vs. everyone else, and “Taiwanese” identification vs. everything
else.7

Table 12.2 uses these two dichotomies to replace the more elaborate coding of
table 12.1. All the other remaining variables are retained, generating our final,
preferred model. Every variable is now comfortably statistically significant and
most coefficients are quite large. Partisanship retains its overwhelming effect,
and apart from the demographic factors of being female or having little
education, all the other explanatory factors are closely related to the central
political cleavage on the island. Being a mainlander and identifying as
“Taiwanese” make a difference even after partisanship is controlled, as do
opinions on unification/independence and on the cross-Strait economic
agreement, ECFA.

Thus again we find that partisanship, ethnicity, national identity, and the main
political dimension are more tightly linked in Taiwan than elsewhere. Ethnicity
is fixed, but the other three can be chosen. We find that they are causally joined
with each other, so that for most politically knowledgeable Taiwan citizens,
conceptions of national identity, partisan identification, and position on issues
related to China are strongly connected. Other political issues tend to be swept
up into this complex of interrelationships: the clarity of that relationship on the
island suggests strongly that politics elsewhere Page 282 → works the same way,
even if the researcher cannot otherwise penetrate the murkiness of most
countries’ multiple identities, issues, and parties. In that sense, the bright clarity
of Taiwan’s political life helps us see into the shrouded complexity of other
countries’ polities. That is what we have tried to help the reader accomplish in
this volume.

A New Cleavage?

Recent political developments in Taiwan have led some scholars to speculate

that “class politics based on wealth gap has become a new driving force of
Taiwan’s party politics” (Wu 2014, 1). That is, the dominant political cleavage
has shifted away from the unification-independence issue to the widening gap
of economic distribution. Indeed, beginning in the early 2000s, Taiwan’s
economy experienced a gradual slowdown, economic inequality worsened, and
the rate of unemployment rose as well.
Table 12.2: The Pan-Blue Presidential Vote in 2012

(1)

9.46***
Party identification
(.600)

2.24***
Mainlander
(.506)

−.470**
Taiwanese ID only
(.210)

1.71***
Unification/independence
(.580)

1.92***
ECFA
(.349)

.423**
Female
(.195)

1.00***
Primary education only
(.238)

−6.42
Intercept
(.516)

Pseudo-R2 .63

N 1,428

Note: * significant at .10. ** significant at .05. *** significant at .01. Dependent variable is 1 if the
respondent voted for Pan-Blue (Ma or Soong); 0 if for DPP (Tsai). Logit parameter estimates (with
standard errors in parentheses).

Page 283 → As a remedy for Taiwan’s economic misfortunes, leaders of the Pan-
Blue Alliance see the Chinese mainland as an economic opportunity, and they
argue that an economic liberalization policy is the means to the island’s
economic revival. Access to China’s vast market, however, is contingent on an
improved cross-Strait political relationship. Thus, after taking office in 2008
the Ma administration of the KMT implemented a policy of rapprochement
toward Beijing. As was explained in chapter 1, cross-Strait tension attenuated
after a series of accords were signed between the two governments, including
the landmark trade deal known as the Economic Cooperation Framework
Agreement. Unfortunately, the expansion of cross-Strait economic exchanges
did not improve Taiwan economic fortunes in the short term, and the island’s
unemployment rate and economic inequality continued to worsen. Leaders of
the Pan-Green Alliance therefore criticized Ma’s liberalization policy as only
allowing the import of cheap Chinese goods, thereby hurting the island’s own
industries, widening the gap between haves and have-nots, and worsening the
job prospects of the younger generation. Pan-Green leaders also argued that
Ma’s rapprochement policy of advancing cross-Strait relationship on the basis
of the “1992 Consensus”—or “one China with different interpretations”—is
tantamount to surrendering Taiwan’s sovereignty to the Beijing government.8
These concerns were a prominent feature of the 2012 presidential election, and
they constituted the underlying appeal of the 2014 protest known as the
Sunflower Movement, led by a group of young people.9

But do the recent economic concerns and the disputes over ECFA herald the
arrival of a new central dimension in Taiwan political life (Wu 2014)? Or is
ECFA just another issue being swept up into the usual partisan cleavage? We
have argued throughout this book that on Taiwan, issues are primarily
consequences of partisanship and national identity, not causes. Disagreements
over ECFA certainly involve differences over trade liberalization vs. autarky,
and thus they might represent purely economic disagreements—a new
dimension, perhaps even a partisan realignment. But ECFA also requires closer
integration with China, and that aspect of the policy might be dominant in
people’s thinking in the way that we have seen repeatedly in this volume.

Fortunately, the 2012 TEDS survey permits a test of these alternate

explanations. The survey contains a question about conventional left-right
attitudes, which we have called the “social welfare” issue:

Regarding the question of social welfare, some people believe that the government should merely
maintain the current system in order Page 284 → not to increase people’s tax. Other people believe that
the government should promote social welfare, even though it will lead to a tax increase.

As Wu (2014, 16) notes, this question is not directly about redistribution.

However, it is certainly related to it: maintaining the status quo is precisely the
means to avoid tax increases of all kinds, including those for reducing
inequality through redistribution. Laissez-faire domestic attitudes should
predict laissez-faire international trade policy attitudes, at least to some degree,
if a new economic dimension is really emerging. Thus, if arguments about
ECFA are truly economic arguments, answers to this social welfare question
will inevitably be related to attitudes toward ECFA. On the other hand, if
ECFA is just another aspect of the main cleavage, then partisanship and
attitudes toward unification/independence should be far better predictors of
what people think about ECFA.

Table 12.3 shows the result of an ordered logit analysis of ECFA opinions on
partisanship, unification/independence, and the social welfare question. As the
table shows, the first two factors are powerful and statistically significant
predictors of ECFA attitudes. The coefficient on the social welfare item, on the
other hand, is not only small and highly statistically insignificant, but Page 285
→ it actually has the wrong sign. Net of partisanship and unification attitudes,
laissez-faire attitudes toward social goals have zero effect on what people think
about the laissez-faire policies embodied in ECFA. The obvious conclusion is
that in the voters’ minds, ECFA was about China, not about economics. Like
postal voting, ECFA is yet another policy pulled into the main Taiwan
cleavage. Indeed, a survey conducted after the Sunflower Movement confirms
the above observation (Yen, Kay, and Chen 2015). Their results indicate that
the perceived economic interests of Taiwan citizens do affect their positions on
trade policies, but that Taiwanese nationalism and security concerns play a
more paramount role in their positions.
Table 12.3. Attitudes toward ECFA in 2012

(1)

3.50***
Party identification
(.201)

1.57***
Unification/independence
(.283)

−.129
Social welfare
(.152)

.089
cut 1
(.150)

3.25
cut 2
(.174)

Pseudo-R2 .14

N 1,826

Note: * significant at .10. ** significant at .05. *** significant at .01. Dependent variable is 1 if the
respondent believes that ECFA will make his family better off; 0 if worse off, and .5 if the same or not
sure. Ordered logit parameter estimates (with standard errors in parentheses).

If Taiwan’s democracy survives its external threats long enough, then some
day, undoubtedly, Taiwan’s internal political divisions will represent something
other than disputes over national identity. That day may arrive quickly, or it
may not. But as of 2012, the date of the most recent presidential survey data
available, such additional considerations had at best only a small impact on the
presidential election. We find no credible evidence that the beginnings of a new
dimension had emerged in voters’ minds.

Looking to the Future

Now if partisanship in Taiwan is strong and other issues matter relatively little,
how stable is Taiwan voters’ electoral behavior? What are the implications of
our findings for Taiwan’s future elections? Taiwan’s party system has
consolidated into what is effectively a two-party system, as chapter 10 explains.
The resulting stability corroborates Lipset and Rokkan’s (1967) “freezing
hypothesis”: a party system reflecting key political cleavages in the society is
expected to be fairly stable. In fact, a 2012 TEDS panel study reinterviewed the
2008 respondents, finding that among those who voted KMT in the 2008
presidential election, 89 percent voted for the KMT again in 2012. The
corresponding number for the DPP was precisely the same—89 percent. These
stability rates are similar to those in other two-party democracies, such as the
United States, and they indicate that Taiwan’s democracy is consolidating its
party system and the loyalty of partisans. Whether this stability is due to
American-style psychological identification with the parties or to the steady
force of other attitudes and identities makes no difference for our purposes.

The analysis in chapter 2 has shown that the Pan-Blue Alliance has enjoyed on
average a 5–10 percentage point electoral advantage at the national level since
2000, and yet the DPP has won three of the six presidential Page 286 → elections

since democratization. How is that possible? First of all, as past experience has
shown, the Pan-Blue Alliance is notoriously susceptible to internal division. As
was discussed in chapter 2, the KMT was severely divided in the 1996, 2000,
and 2012 presidential elections. When a number of party members failed to
secure the party nomination for the presidency, they ran as independents or
became party switchers. Indeed, it was Pan-Blue candidates, Soong and Lien,
who divided nearly 60 percent of the electoral support and thereby delivered
the victory to the DPP-affiliated Chen in the 2000 presidential election, who
won the election with only 39.3 percent of the votes. Even in the absence of
Pan-Blue divisions, a properly engineered campaign strategy by the Pan-Green
Alliance may secure a win, especially when economic prospects look weak
under Pan-Blue control.

Furthermore, even if Taiwan voters’ electoral behavior is largely determined

by their partisan affiliations, the vision for the country that a candidate presents
continues to play an important role in citizens’ electoral calculus. As chapter 3
shows, the majority of the island citizens consider Taiwan an independent state
separate from China, but they are risk-averse and pragmatic on cross-Strait
relations. The majority of them are not willing to sacrifice their hard-won
democratic way of life and economic prosperity for such radical political
changes as declaring de jure independence or unification. Precisely because the
island’s future relation with China is the key political cleavage of the society,
any presidential hopeful will need to present a realistic and workable vision to
the electorate or suffer the consequence of losing the election.

The 2012 presidential election best illustrates this logic.10 Indeed, observers
generally believe that the loss of Tsai Ing-wen, the DPP nominee, in Taiwan’s
2012 presidential election was related to her ambiguous stand on cross-Strait
policy. Tsai’s campaign strategy was to deemphasize cross-Strait issues, a strong
suit for the KMT but a major DPP weakness, and to focus more on social
issues such as economic growth and inequality. While Ma’s policy had a proven
record of engaging China and had been praised by Washington, Tsai’s “Taiwan
Consensus” was necessarily short on specifics due to the internal politics of the
DPP. On the one hand, the DPP needed to court the backing of its core
supporters, who were generally in favor of Taiwan’s de jure independence from
China. On the other hand, the DPP also needed to win the support of
moderate partisan identifiers from both alliances who did not support
unification but wanted to avoid inflammatory rhetoric and policies that could
lead to cross-Strait tension. Tsai was thus trapped in the dilemma of needing to
mobilize the party’s core supporters by appealing to their political identity,
while not alienating the more moderate Page 287 → centrists. The ambiguity and
lack of specificity provided the KMT with ammunition for an attack. It also
raised uncertainty about cross-Strait relations should she win the election, with
a potential to destabilize Taiwan’s uneasy but carefully managed relationship
with the mainland. Ultimately, the election came down to a choice between
“1992 Consensus/stability” vs. “Taiwan Consensus/potential instability,”
which was a manifestation of Taiwan’s key political cleavage. The incumbent
Ma was a vulnerable candidate with many weaknesses, but he managed to win a
second term.

Interestingly, during the 2016 presidential election held on the island, the
cross-Strait relationship appeared to be a mute issue. If the China factor heats
up Taiwan’s dominant political cleavage, as we have argued, why were there no
sparks thrown off during the campaign? In our view, this is due to a divided
KMT and a moderate stance adopted by the DPP presidential nominee, Tsai
Ing-wen. As we noted earlier, the Pan-Blue Alliance is notoriously susceptible
to internal division. The 2016 election was no exception. Immediately after
Hung Hsiu-chu became the party’s presidential nominee, many KMT elites and
local leaders contemplated replacing her with someone else (Peng and Chin
2015). They eventually succeeded, making Eric Chu the KMT nominee. James
Soong, formerly of the KMT, also ran as the candidate of the PFP. Meanwhile,
realizing that cross-Strait policy is her Achilles’ heel, Tsai moderated her stance
by claiming that she would maintain the cross-Strait status quo if elected.
Although Tsai’s vague China policy presented the KMT with an opportunity
for attack, as it did in the 2012 election, the divided KMT could not launch an
effective offensive against Tsai’s stance on cross-Strait relations (Lowther 2015).
The unpopularity of the Ma administration and the relatively stagnant economy
probably also helped Tsai. In the end, she won easily. Thus the China factor
may have played a less explicit role in the 2016 campaign than in past years. But
does that mean that the traditional Taiwan political cleavage will be any less
influential in voters’ decisions or in the near term?

As the old Danish proverb has it (sometimes attributed to the Danish

physicist Niels Bohr or to the American baseball player Yogi Berra),
“Prediction is very difficult, especially about the future.” We know no reliable
fortune teller who could tell us how Taiwan’s politics and international status
will evolve. For the medium term, however, we see no prospect that a de facto
independent Taiwan will develop another cleavage to replace the powerful role
of “the China factor.” China is too near, too antagonistic, and too strong. Each
of the two parties may trim, modify, and deemphasize, but inevitably one will
represent faster progress toward independence and the other will want to go
slow, engage with China, and avoid irreversible declarations Page 288 → that
would likely lead to war. Those differences, tied to deep social cleavages, will
not evaporate any time soon.

Chinese-American relations may have much to do with the outcome that

prevails. Taiwan is a small country, and its fate is not entirely in its hands. But
so long as it persists as a democracy, it can talk about the future it wants for
itself, and it can use elections to put alternate teams in place, teams that differ
on the right national identity for the island country. As we have shown in this
book, that is what politics in Taiwan is fundamentally about.

There are many lessons to be taken from the study of Taiwan, and we
conclude with just one that seems to us the most important. Taiwan illustrates
in great detail the power of identity in politics. In every country that we know,
the choice of partisanship is not primarily about issues; it is about identity.
Election campaigns are not centrally about proper positioning in some
ideological space; they are primarily about mobilizing identity groups.
Candidate personalities and the state of the economy matter at the margin, but
the major effects shaping democratic elections are determined well in advance
by the balance of partisanship in the electorate. And that balance is determined
by the complex, interconnected histories of the relevant identity groups and
their resulting affinities and antagonisms.

Yet there is dispute about these claims among scholars. Some believe that an
understanding of politics has to begin from preferences, and that preferences
are about self-interest, often material self-interest and social class. Indeed, in
most countries, it can be hard to tell the difference between self-interest and
identity explanations in the welter of competing identities and interests. But
Taiwan is a place where one can see those forces in undiluted form and without
the multiple cleavages and countervailing effects present in larger countries.
Because identity effects are not being obscured in the cross-section by equally
large and opposite identity effects, as they often are elsewhere, Taiwan lets us
see just how strong those identity effects are. And the answer from Taiwan is:
very strong indeed. Just as the Galapagos Islands showed Darwin how
evolution worked everywhere, but in a clear and indisputable way, so also the
island of Taiwan demonstrates in a clear and indisputable way that identities
are where to start in thinking about electoral politics.

In turn, that suggests that the first questions to ask about another country’s
party system are not: What are the main political issues? Or: How are the
parties positioned from left to right? Both questions may matter for
governmental policy but have little resonance in the public mind. For the
average citizen, perceptions of issues and ideologies are primarily derivative
rather than causal, as this book has repeatedly shown and as other scholars Page

289 → have demonstrated for other countries (Campbell et al. 1960, chaps. 6, 7;
Lenz 2012; Achen and Bartels 2016, chaps. 9, 10).

Thus an implication of this book is that the key questions for understanding
voting are not issues and party positions, as so many comparative studies
assume. Instead, scholars should ask: What are the principal identity groups?
And how are they connected to the political parties? Those are questions that
need to be posed everywhere. But we hope that we have convinced the reader
that nowhere is their value and power more obvious scientifically than in
Taiwan. For that reason, the study of Taiwan, valuable for its own sake, is even
more valuable for what it has to teach us about how elections should be
understood around the world.

Notes
1. Citizens of Britain and the United States may recall a similar dispute
from the 1770s.

2. In Taiwan’s TEDS survey, party identification is measured in the usual

international manner, as “leaning to” a political party. That language is
quite different from the original American survey item, and it also differs
from the wordings in use in some other democracies. Question wording
matters substantially where party identification is concerned (for example,
Sinnott 1998), and a better understanding of what wording is best for
Taiwan now that the party system has consolidated is an important topic
for future research. For our purposes, however, it makes no difference
whether Taiwan’s partisanship is a meaningful psychological identity or
simply a habituated behavioral partisanship. The point is that, either way,
it organizes attitudes and votes.

3. Throughout this chapter, partisanship is coded on a seven-point scale,

ranging from “very strongly” lean to the DPP to “very strongly” lean to
the KMT. The middle category represents those who do not lean to either
party.

4. The opinion items and the prospective and retrospective economic

evaluations all have three response categories, and they are coded 0, .5,
and 1, with upper values indicating more pan-Blue/proincumbent
attitudes. Age is coded so that age 20 = 0 and age 100 = 1. Percentage
disposable income changes are divided by 20, so that they range
approximately from -.5 to +.5. All other explanatory factors are dummy
variables coded either 0 or 1. We have not included candidate traits
because the list of such factors is very long, and because chapter 8
demonstrated that such evaluations are driven primarily by partisanship.
We also excluded left-right orientation because, as chapter 9 showed, that
variable is meaningless in Taiwan, and thus the variable was dropped from
the 2012 TEDS presidential study. Previous work on the 2008 presidential
election showed that left-right positions had only a small, statistically
insignificant effect on the vote (Hsiao and Lin, 2013).

5. We found repeatedly that the survey weights made no meaningful

difference, and so we have chosen to present unweighted results
throughout this chapter.

6. The county-level disposable income figures are quite variable from one
year to Page 290 → the next, especially for the smaller counties, which are
difficult to survey adequately in every country. Thus the available
measures may represent actual disposable income changes with substantial
error. In addition, there have been too few presidential elections to run
regressions with national-level data. Our provisional conclusion in this
book is that retrospective economic voting is not very consequential in
Taiwan, but the topic cries out for additional research.

7. We also tested the coefficients for Chinese identity and no identity to

see whether they were jointly significant. However, the Wald test in each
of the four columns of table 12.1 was very far from statistical significance,
meaning that deleting both variables was justified.

8. The “1992 Consensus” maintains that the notion of “one China”

should serve as the basis for cross-Strait interactions. However, the two
governments had different interpretations of what “one China” was. This
 is the tacit understanding presumably reached by Beijing and Taipei in
November 1992. See Su and Cheng (2002).

9. The “Sunflower Movement” was a protest against a proposed cross-

Strait trade-in-service agreement. It lasted more than 20 days between
March 18 and April 10, 2014, during which time student demonstrators
occupied the Legislative Yuan and damaged the main government
buildings of the Executive Yuan. The movement reflects the public’s
concern about Taiwan’s increasingly close economic ties with China. It
also led to a massive demonstration against the Ma administration’s cross-
Strait policies in front of the presidential office on March 30 (J. R. 2014).

10. For further discussion of the 2012 presidential election, see Romberg
(2011) and Paal (2012).

References

Achen, Christopher H., and Larry M. Bartels. 2016. Democracy for Realists.
Princeton: Princeton University Press.

Anderson, Christopher J. 2007. “The End of Economic Voting? Contingency

Dilemmas and the Limits of Democratic Accountability.” Annual Review of
Political Science 10: 271–96.

Bartels, Larry M. 2002. “Beyond the Running Tally: Partisan Bias in Political
Perceptions.” Political Behavior 24 (2): 117–50.
Berglund, Frode, Soren Holmberg, Hermann Schmitt, and Jacques Thomassen.
2005. “Party Identification and Party Choice.” In The European Voter, ed.
Jacques Thomassen, 106–24. Oxford: Oxford University Press.

Campbell, Angus, Philip E. Converse, Warren E. Miller, and Donald E. Stokes.

1960. The American Voter. New York: John Wiley and Sons.

Evans, Geoffrey, and Nan Dirk De Graaf, eds. 2013. Political Choice Matters.
Oxford: Oxford University Press.

Gelineau, Francois, and Matthew M. Singer. 2015. “The Economy and

Incumbent Support in Latin America.” In The Latin American Voter, ed.
Ryan E. Carlin, Matthew M. Singer, and Elizabeth J. Zechmeister, 281–99.
Ann Arbor: University of Michigan Press.

Hsiao, Yi-ching, and Tsong-Jyi Lin. 2013. “Exploration of Political

Polarization inTaiwan: Page 291 → Measurement and Analysis” (in Chinese).
In Chi Huang, ed. Methodology of Taiwan’s Election and Democratization Study
(TEDS): Retrospection and Prospection. Taipei: Wu-nan.

J. R. 2014. “Politics in Taiwan: Sunflower Sutra.” Economist, April 8.

http://www.economist.com/blogs/banyan/2014/04/politics-taiwan.
Accessed January 20, 2015.

Lenz, Gabriel S. 2012. Follow the Leader? How Voters Respond to Politicians’ Policies
and Performance. Chicago: University of Chicago Press.

Lewis-Beck, Michael. 1988. Economics and Elections. Ann Arbor: University of

Michigan Press.
Lipset, Seymour M., and Stein Rokkan. 1967. “Cleavage Structure, Party
Systems, and Voter Alignments: An Introduction.” In Party Systems and Voter
Alignments: Cross-National Perspectives, ed. Seymour M. Lipset and Stein
Rokkan, 1–64. New York: Free Press.

Lowther, William. 2015. “Tsai Vows ‘Consistent’ Cross-Strait Ties.” Taipei

Times, June 5.
http://www.taipeitimes.com/News/front/archives/2015/06/05/200361995
1. Accessed October 10, 2015.

Lupu, Noam. 2015. “Partisanship in Latin America.” In The Latin American

Voter, ed. Ryan Carlin, Matthew M. Singer, and Elizabeth J. Zechmeister,
226–45. Ann Arbor: University of Michigan Press.

Paal, Douglas. 2012. “Taiwan Election Has the United States and China on
Edge.” Carnegie Endowment for International Peace, January 11.
carnegieendowment.org/publications/index.cfm?fa=view&id=46441.
Accessed October 10, 2015.

Paldam, Martin. 1991. “How Robust Is the Vote Function?” In Economics and
Politics, ed. Helmut Norpoth, Michael S. Lewis-Beck, and Jean-Dominique
Lafay, 9–31. Ann Arbor: University of Michigan Press.

Peng, Hsien-chun, and Jonathan Chin. 2015. “KMT Politician Demands

Review of Hung’s Selection.” Taipei Times, October 4.
http://www.taipeitimes.com/News/front/archives/2015/10/04/200362921
6. Accessed January 12, 2012.
Powell, G. Bingham, Jr., and Guy Whitten. 1993. “A Cross-National Analysis of
Economic Voting: Taking Account of the Political Context.” American
Journal of Political Science 37: 391–414.

Romberg, Alan D. 2011. “The 2012 Taiwan Election: Off and Running.” China
Leadership Monitor, no. 35 (Summer).
http://media.hoover.org/sites/default/files/documents/CLM35AR.pdf.
Accessed December 31, 2011.

Su, Chi and Cheng An-guo (eds.). 2002. Yige Zhongguo, gezi biaoshu’ gongshi
de shishi (One China, Different Interpretations: An Account of the
Consensus). Taipei: Guojia zhengce yanjiu jijinhui.

Sinnott, Richard. 1998. “Party Attachment in Europe: Methodological Critique

and Substantive Implications.” British Journal of Political Science 28 (4): 627–50.

Van der Eijk, Cees, Hermann Schmitt, and Tanja Binder. 2005. “Left-Right
Orientations and Party Choice.” In The European Voter, ed. Jacques
Thomassen, 167–91. Oxford: Oxford University Press.

Wu, Yu-Shan. 2014. “Paradigm Shift in Taiwan’s Politics: From Identity to

Distribution.” Presented at the Conference on Democratic Governance,
Cross-Strait Security and Prosperity, Taiwan and Asia Program, College of
William and Mary, October 13.

Page 292 → Yen, Wei-ting, Kristine Kay and Fang-Yu Chen. 2015. “Unpacking
Support for Free Trade: Experimental Evidence from Taiwan.” Paper
presented at the 2015 Midwest Political Science annual meeting, Chicago,
April 16-19, 2015.
Zechmeister, Elizabeth J. 2015. “Left-Right Identifications and the Latin
American Voter.” In The Latin American Voter, ed. Ryan Carlin, Matthew M.
Singer, and Elizabeth J. Zechmeister, 195–225. Ann Arbor: University of
Michigan Press.
Page 293 →

Contributors

Christopher H. Achen is a professor in the Politics Department at Princeton

University, where he holds the Roger Williams Straus Chair of Social Sciences.
His primary research interests are public opinion, elections, and the realities of
democratic politics, along with the statistical challenges that arise from those
fields. He is the author or coauthor of five books, including Democracy for
Realists (with Larry Bartels), published by Princeton University Press in 2016.
He has also published many articles. He has been a member of the American
Academy of Arts and Sciences since 1995, and has received fellowships from
the Center for Advanced Study in the Behavioral Sciences, the National
Science Foundation, and Princeton’s Center for the Study of Democratic
Politics. He was the founding president of the Political Methodology Society,
and he received the first career achievement award from the Political
Methodology Section of the American Political Science Association in 2007.
He has served on the top social science board at the National Science
Foundation, and he was the chair of the national Council for the Inter-
University Consortium for Political and Social Research (ICPSR) from 2013 to
2015. He is also the recipient of awards from the University of Michigan for
lifetime achievement in training graduate students and from Princeton
University for graduate student mentoring.
Lu-huei Chen is Distinguished Research Fellow at the Election Study Center
and is a Professor in the Political Science Department, National Chengchi
University in Taiwan. He holds a PhD degree from Michigan State University.
His research focuses on Taiwanese political behavior, political socialization,
research methods, and cross-Strait relations. He has published articles in such
scholarly journals as Issues and Studies, the Journal of Electoral Studies (in Chinese),
Social Science Quarterly, and Taiwan Political Science Review (in Chinese). His edited
volumes include books on the 2012 presidential election in Taiwan (in Chinese)
and public opinion (in Chinese).

Page 294 → Su-feng Cheng is a Research Fellow at the Election Study Center,
National Chengchi University, in Taiwan. Her research interests include voting
behavior, public opinion, and survey methods. She has published articles in
many journals, including Electoral Studies, Issues and Studies, Journal of East Asian
Studies, Journal of Electoral Studies, Taiwanese Political Science Review, and Taiwan
Democracy Quarterly. Her edited volumes include books on electoral studies and
survey methodology.

Karl Ho is Clinical Associate Professor of Public Policy, Political Economy,

and Political Science and Director of Academic Computing at the School of
Economic, Political and Policy Sciences, University of Texas at Dallas. His
research focuses on political behavior in new democracies and state behaviors
in response to challenges to democratic development. He is also interested in
public policy and the political economy of Taiwan, Hong Kong, and China. His
recent research examines civic engagement, political attitudes, and electoral
behavior in Taiwan and Hong Kong. He has published on elections in East
Asia, global women’s rights, and political research and technological advances.
He serves as the co-investigator of the Hong Kong Election Study and a
member of the Electoral Studies editorial board. His works have appeared in
Electoral Studies, Human Rights Quarterly, Journal of African and Asian Studies, and
the Journal of Information Technology and Politics.

Yi-ching Hsiao is an Associate Professor in the Department of Public

Administration at Tamkang University, Taipei, Taiwan. He received his PhD in
political science from National Chengchi University in 2009. He is interested in
voting behavior, electoral systems, public opinion surveys, and congressional
politics. His articles have appeared in such journals as the Asian Journal for
Public Opinion Research, Election Studies (Korea), Issues & Studies: A Journal of
Asian Pacific Studies (in Japanese), and the following journals in Chinese:
Taiwanese Political Science Review, Journal of Electoral Studies, Taiwan Democracy
Quarterly, Political Science Review, Soochow Journal of Political Science, and Chinese
Political Science Review.

Chi Huang is a University Chair Professor of Political Science and Research

Fellow of the Election Study Center at National Chengchi University, Taiwan.
His research interests focus on survey research, quantitative methodology,
electoral systems, and voting behavior. He has coauthored or coedited seven
books and published many articles in leading international and Taiwanese
journals, including the American Political Science Review, American Journal of
Political Science, Journal of Politics, Comparative Political Studies, Page 295 → Electoral

Studies, Party Politics, Political Research Quarterly, Japanese Journal of Electoral Studies,
Issues & Studies, and Taiwanese Political Science Review. He served as president of
the Taiwanese Political Science Association (2003–05), founded Taiwan’s
Election and Democratization Study (TEDS) interuniversity survey project in
2000 and has been the Coordinator since its inception.

Hsiao-chuan (Mandy) Liao is an Assistant Professor in the Department of

Political Science at National Taiwan University, Taiwan. She received her PhD
in the Department of Political Science from University of South Carolina. Her
current teaching field is international relations and China’s external relations in
particular. Her research interests include international conflict, international
relations theory, foreign policy and decision making, and Northeast Asian
studies. Her research has appeared in such refereed journals as Asian Politics and
Policy, Taiwan Political Science Review, and the Journal of International and Public
Affairs.

Tzu-Ping Liu is a PhD student in the Department of Political Science, the

University of California at Davis. His articles have been published in the
Japanese Journal of Political Science, Journal of Electoral Studies, and East Asia.

Shing-Yuan Sheng is a professor in the Department of Political Science at

National Chengchi University, Taiwan. She also serves as a member of the
Planning Committee of the Taiwan’s Election and Democratization Study
(TEDS). Her research interests include legislatures and legislative behavior,
election and voting behavior, as well as party and democracy. Her current
research includes “Party Institutionalization of the Legislative Yuan” and
“How the Taiwan Public Perceives Party in Legislature.” Her scholarly work
has appeared in the Taiwanese Political Science Review, Soochow Journal of Political
Science, and the Journal of Electoral Studies. She holds degrees from National
Chengchi University (BA and MA) and the University of Michigan (PhD).

Alexander C. Tan is Professor of Political Science at the University of

Canterbury in Christchurch, New Zealand, and University Chair Professor of
Political Science at National Chengchi University in Taipei, Taiwan. He has
published works in the areas of East Asian politics, comparative political
economy, and political parties, and his articles have appeared in such journals
as the Journal of Politics, Political Research Quarterly, Comparative Political Studies,
Comparative Politics, European Journal of Political Research, Electoral Page 296 →
Studies, Party Politics, and the Journal of Asian Security and International Affairs. His
most recent books include a coedited volume, Mixed-Member Electoral Systems in
Constitutional Context: Taiwan, Japan, and Beyond (University of Michigan Press,
2016), and a coauthored book, Taiwan’s Political Economy: Meeting Challenges,
Pursuing Progress (Lynne Rienner, 2012).

Chia-hung Tsai is Director and Research Fellow at the Election Study

Center, National Chengchi University, Taiwan, and holds a joint appointment
with the Graduate Institute of East Asian Studies. Tsai received his PhD. from
the Ohio State University in 2003. His fields of specialization include public
opinion, methodology, and comparative politics. He was a visiting scholar at
the Department of Political Science, MIT. He has published research in such
journals as Party Politics, International Political Science Review, Journal of Asian and
African Studies, and PLOS ONE.
Hung-chung Wang is a research fellow of the Taiwan Foundation for
Democracy (TFD). He had served as associate managing editor of the Taiwan
Journal of Democracy and the Taiwan Democracy Quarterly. Wang received his PhD
from the Department of Political Science at the University of New Orleans in
2010 and was a postdoctoral researcher at the Election Study Center (ESC),
National Chengchi University, from 2011 to 2013. His primary research
interests include political behavior, public opinion, legislative behavior, local
politics, and American politics. Wang has published articles in Issues & Studies,
Taiwan Democracy Quarterly, Journal of Electoral Studies, and the Taiwanese Political
Science Review.

T. Y. Wang is Professor of Politics and Government at Illinois State University

and currently serves as the coeditor of the Journal of Asian and African Studies.
He was the Coordinator of the Conference Group of Taiwan Studies
(CGOTS) of the American Political Science Association. His primary research
interests are national identity, East Asian politics, electoral behavior, U.S. policy
toward China and Taiwan, and research methodology. He has
authored/coauthored and edited/coedited seven books and special issues of
journals and published more than 40 articles/book chapters in such scholarly
journals as the American Political Science Review, Asian Survey, International Studies
Quarterly, Issues and Studies, Journal of Peace Research, Political Research Quarterly, and
Social Science Quarterly.

Chung-li Wu, PhD in Political Science, University of New Orleans (1997), is

Research Fellow of the Institute of Political Science at Academia Sinica Page 297

→ in Taipei, Taiwan. His research interests are American politics (political

institutions), comparative politics, urban and minority politics, comparative
politics, and electoral studies. He is the author of articles published in Party
Politics, China Quarterly, Parliamentary Affairs, Journal of Black Studies, Asian Survey,
International Relations of the Asia-Pacific, Japanese Journal of Political Science, and
several others. Among the academic awards he has received are Outstanding
Researcher Award (Ministry of Science and Technology and National Science
Council) in 2016 and 2009, the Investigator Award (Academia Sinica) in 2014,
Outstanding Research Award (College of Social Sciences, National Chung
Cheng University) in 2005, Outstanding Research Award (National Chung
Cheng University) in 2003, and Young Scholar Award (National Chung Cheng
University) in 2002.

Ching-hsin Yu is a Research Fellow at the Election Study Center at National

Chengchi University, Taiwan. He earned a PhD in political science from
Pennsylvania State University and specializes in election and party competition
in Taiwan. He has served on the editorial board and as a reviewer for academic
journals. He has also published widely in refereed journals on such topics as
identity, political knowledge, political accountability, and party system change.
He is currently studying the development of the new party system in Taiwan.
Page 298 → Page 299 →

Index

Page numbers in italics indicate tables and figures.

aboriginal people
defined, 21n8
KMT support, 41n3
legislative seats reserved for, 223, 238
absentee voting, lack of, 22n9
age
in multifactorial model, 279
national identity changes, 54–55, 55, 56
political generations definition, 67n6
political participation and, 253–54, 258, 259, 259, 260–62, 264
unification/independence issue, 62–64, 63
voting behavior connection, 35–36, 37
Aldrich, John, 167
The American Voter (Campbell et al.), 2–3, 170, 172, 276
Anderson, Christopher, 3, 278
assessment of candidates. See candidate evaluation

Bartels, Larry, 152, 277

Brody, R., 120
Bush, George H. W., 172
Butler, David, 3
Campaign for Rectifying the Name of Taiwan, 126
Campbell, Angus, 2–3, 72, 170
Canada, confusion with left-right labels, 215, 219n15
candidate evaluation, 170–97
data for, 173–80, 176, 178, 180, 190–91
formation of, 171–73
importance in electoral decision-making, 2, 12, 18–19, 277
overview, 170–71
partisan identification vs., 174, 180–88, 183–87, 191
role in pre-1986 elections, 76
survey questions, 192–94
vote choice vs., 174, 188–89, 189, 190, 191–92
Carlin, Ryan, 3
Carmines, Edward G., 99
characteristics of voters, 26–44
age, 35–36, 37
consolidation of two-party system, 15–16, 27–31, 28, 29, 39
education, 33–34, 35
ethnic identity, 31–33, 32, 36–39, 38
income, 33, 34
independent voters, 73
Page 300 → characteristics of voters (continued)
occupation, 34–35, 36
partisan identification, 36–40, 38
recent presidential election results, 26–27
wording and coding of demographic variables, 40
 See also specific characteristics
Chen, Wen-chun, 202
Chen Chih-ping, 134n9
Chen Lu-an, 27, 119, 175
Chen Shui-bian
assassination attempt, 27, 41n2, 178
candidate evaluation of, 173, 177–79, 183–87, 190–91
corruption charges, 104–5, 133n3, 195n8, 240
de-Sinicization measures, 10, 52, 65
economic policies, 108, 201
election of, 10, 286
electoral reform pledge, 234, 235, 236
evolution of unification/independence position, 115–16, 117, 126
Hakka appointments, 32
national referendum proposal, 155n3
New Central Way policy, 115, 121
nuclear power plant compromise, 112
partisan identification of supporters, 37, 39
presidential campaign, 27, 30
renaming of state agencies, 67n5
support for independence, 6, 52
trade restrictions, 7
voter perception of animosity from China, 60
Chiang Ching-kuo, 174, 175, 182
Chiang Kai-shek
crackdown in Taiwan, 6
mainlanders support for, 31
 re-Sinicization measures, 5, 45, 65
retreat to Taiwan, 58
China factor. See China-Taiwan relationship; unification/independence issue
China-Taiwan relationship, 158–169
in candidate evaluation, 185, 186, 186, 188–89, 189, 190t
Chinese vs. Taiwanese consciousness, 51
complex dynamics of, in vote choice, 18, 162–68, 163–66
concerns for peace, 14
cultural heritage, 4–5
diplomatic isolation of Taiwan, 6, 9–10, 59–60, 64, 65, 162
economic ties, 7–8, 142, 158–162, 160, 161, 166–67
importance in future elections, 285–89
issue ownership theory applied to, 182
military coercion, 59–60, 65
by national identity, 60–62, 60, 61
overview, 1–2
public perceptions of, 161–64, 163, 167–68
security threat aspects, 5–7, 63–64, 162
Chinese identity. See also dual Chinese-Taiwanese identity
cultural vs. political connotations, 56, 58
decrease in, 10, 64
by ethnicity, 53–54, 53
fostering of, 49, 51
one China beliefs, 8
Pan-Blue Alliance support, 11
by political generation, 54–55, 55
political participation and, 255, 258, 262
 private poll results, 67n7
Chinese Nationalist Party. See Kuomintang (KMT) Party
Chinese/Taiwanese consciousness boundary, 51–58, 53, 55, 57, 64
Chinese Youth Party, 77, 95n2
Chu, Eric, 287
Chu, Yun-Han, 100
citizen evaluation of candidates. See candidate evaluation
citizen identity. See voter identity
Page 301 → Clark, Cal, 168
class concerns
emergence of, 106–7, 129
occupation, 34–35, 36
unequal political participation, 264–66
See also education; income
cleavage theory, 224–26
Clinton, Bill, 172, 173
collectivity of common fate (sheng-ming-gong-tong-ti), 50
communism vs. democracy, left-right labels, 203, 205
Communist Party (China), 5, 7, 58, 203
Comparative Study of Electoral Systems (CSES), 198–99, 201–2, 218n2
competence
in candidate evaluation, 172, 190–91
data for, 174–80, 176, 178, 180
relationship with partisanship, 180–88, 183–87
survey questions, 192, 193, 194, 194n1
conservatism, as the “right,” 218n3
constructionism framework
instrumentalism vs., 67n2
overview, 46–48, 64
relational aspects of boundaries, 45–46, 65
socialization role in, 47, 65
Converse, Philip E., 121, 198, 202
Cooley, Charles, 46
corruption, 185–86, 187
Council for Hakka Affairs, 32
Council of Grand Justices, 112
Cox, Gary, 242–43
cross-Strait relations. See China-Taiwan relationship
CSES (Comparative Study of Electoral Systems), 198–99, 201–2, 218n2

Dahl, Robert, 252

Dalton, Russell, 3, 202, 214, 219n14
Dang-wai, 75, 76–77, 102, 103, 127
decision-making, electoral. See vote choice
De Graaf, Nan Dirk, 3, 200
democracy
political participation importance, 252–53, 262–63
political parties importance, 92
support for, 85–86, 86, 87
Democratic Progressive Party (DPP). See also Pan-Green Alliance
absentee voting opposition, 22n9
campaign strategies, 117
candidate evaluation, 19
 Chen Shui-bian leadership of, 177
Dang-wai involvement, 77
de-Sinicization measures, 52, 65
economic policies, 7–8, 111–14, 128, 142, 162
electoral system reform, 223–24, 234–36, 240–42
environmental policies, 111–14, 128
ethnic identity connection, 36–40, 38
formation of, 6, 75–76, 78, 91
framing of issues, 98, 102–3, 127
increasing power of, 10
irrelevance of left-right labels, 201, 209–10, 209, 211
issue ownership theory applied to, 181–88
national referendum proposal, 149, 155n3
NP alliance, 104
as opposition party, 227
Pan-Green Alliance leadership, 11, 15, 39
Peng Ming-min nomination, 175
post-1990s partisanship, 79–81, 81
public perceptions of, 104, 105
reform/stability issue, 103–5, 126, 127
social welfare policies, 106, 107–9, 109, 110, 129, 130
Taiwanese consciousness of, 58
unification/independence issue, 52, 114–22, 126, 128, 286–87
Democratic Socialist Party, 77, 95n2
Page 302 → democratization of Taiwan
discord with diplomatic isolation, 9–10, 59
 effects on ethnic divide, 49–50
first popular election of president, 27
overview, 1
pace of reform, 6
party system changes, 71, 227
reform/stability issue prominence, 100, 103
support for democracy, 85–86, 86, 87
support for independence, 52
demographics. See characteristics of voters
diplomatic isolation of Taiwan, 6, 9–10, 59–60, 64, 65, 162
Directorate General of Budget, Accounting and Statistics (DGBAS), 150,
155n4
disposable income per capita (DIPC)
effect on economic voting, 149–54, 151, 152, 153, 154
hypothesis for, 143
in multifactorial model, 281
district level elections, postreform changes, 242–43, 244
district magnitude (M), 224, 227, 233
Downs, Anthony, 140
DPP. See Democratic Progressive Party
dual Chinese-Taiwanese identity
cultural vs. political connotations, 58
by ethnicity, 53–54, 53
increase in, 10, 58, 64
by political generation, 54–55, 55
political participation and, 255, 258, 262
Duch, Raymond, 141
DuPont, 111
Duverger, Maurice, 224, 243
Duverger’s hypothesis, 224, 225, 241, 242
Duverger’s law, 224–25, 241, 242

Economic Cooperation Framework Agreement (ECFA)

as emerging issue, 283–85, 284
implications of, 160–61, 162
KMT vs. DPP views, 129
signing of, 7, 143
voter evaluation of, 164–66, 164–68
economic issues
in candidate evaluation, 185, 186, 186, 187
discord with diplomatic isolation, 9–10, 59
evolution of issue, 106–7, 128, 282–85, 284
importance in electoral decision-making, 17–18, 141–43
issue ownership theory applied to, 181–82
left-right political spectrum, 199–200
overview, 1
ties with China, 7–8, 142, 158–62, 160, 161
as valence issue, 123–24
economic voting, 139–57
bivariate relationship between economic evaluation and vote choice,
145–46, 145, 146
evaluations of national economic conditions, 144–45, 145
logistic regression models, 147–49, 148
mixed evidence for, 18, 139, 141–43, 154–55, 277–78
 overview, 18, 139
personal income factors, 149–154, 151, 152, 153, 154
survey questions and coding scheme, 143–44, 144
theoretical framework, 140–41
education
of independent voters, 79, 81–82, 82
in multifactorial model, 281
political interest among independents and, 83–84, 84
role in collective memories, 52
status quo support among independents and, 85, 85
support for democracy among independents and, 86, 87
Page 303 → vote choice among independents and, 90, 90
voting behavior connection, 33–34, 35
effective number of electoral parties (ENEP), 241–42, 242
effective number of parliamentary parties (ENPP), 241–42, 242
effective threshold, 247n8
elderly
political participation, 260–62
social welfare programs for, 107, 108, 109–10, 134n9
Election Study Center, National Chengchi University, 102, 104, 120, 150, 253
electoral decision-making. See vote choice
electoral participation, non-voting, 260–63, 264–68, 270n8
See also political participation
electoral system change, 223–51
convergence to two-party system, 245–46
district level effects, 242–43, 244
 first postreform legislative election, 239–40, 245
national level effects, 240–42, 242
overview, 223–24
reform, 231–39, 233, 237
SNTV system, 226–231, 228, 229, 230, 245
theoretical frameworks, 224–26
voter level effects, 243–45
ENEP (effective number of electoral parties), 241–42, 242
ENPP (effective number of parliamentary parties), 241–42, 242
environmental protection/economy issue
evolution of issue, 101, 111–14, 113, 128
importance in electoral decision-making, 17
irrelevance of left-right labels, 210, 212–13, 212, 213
questionnaire wording and results, 131, 132
status as most important problem facing Taiwan, 124, 125, 126–27
survey question wording, 217
equilibrium. tendency to, 22n10
Erikson, Robert S., 140, 149
ethnic identity
changing boundaries of, 48–51, 50, 64, 274–75
effects on unification/independence issue, 13
importance in electoral decision-making, 31–33, 32
in multifactorial model, 281
partisan identification connection, 36–40, 38
political participation and, 253, 254, 258, 259, 260–62, 260, 265
role in political cleavage, 16
wealth distribution connection, 106
evaluation of candidates. See candidate evaluation
Evans, Geoffrey, 3, 200
Executive Yuan, 112

factional politics, prevalence in Taiwan, 21n2

Fa-tong, 74
February 28 Incident (1947), 6, 9, 48, 73
Feldman, Stanley, 140
Fell, Dafydd, 122, 201, 234
Fiorina, Morris P., 140
Formosa Incident (1980), 177, 195n6
Four Noes plus One Without pledge, 116, 134n15
freezing hypothesis, 11–12, 224, 285
Fukushima nuclear incident (2011), 113, 128, 134n13
funnel of causality model, 170, 191

Geertz, Clifford, 47
Gellner, Ernest, 47
Gomez, Brad, 141
Page 304 → group consciousness
Chinese vs. Taiwanese consciousness, 52, 67n4
effects on constructed identities, 47

Hakka ethnic group

defined, 21n8
national identity changes, 53–54, 53
political participation, 254, 258, 260
pre-1986 lack of representation, 74
 support for KMT, 74, 75, 76
support for Lee Teng-hui, 175
views on ethnic tension, 50, 50
voting behavior, 31–33, 32
Hakka TV, 32
Hand-in-Hand Rally (2004), 118
Hansen, John Mark, 253
Hau Pei-tsun, 195n4
Hayes, Danny, 181
Hibb, Douglas A., Jr., 149
Hotelling, Harold, 198
Hsiao, Yi-ching, 142
Hsieh, John Fuh-sheng, 100, 142
Hsieh Chang-ting (Frank)
environmental and economic policies, 112
presidential campaign, 105, 179
unification/independence issue, 118, 119, 121
Hsu, Yung-ming, 155n3
Hsu Hsin-liang, 177
Huang, Chi, 225
Huang, Hsiu-tin, 155n3
Hung Hsiu-chu, 287
Hunt, W. Ben, 199–200

identity. See voter identity

income
classification of groups, 40–41
 effect on economic voting, 149–54, 151, 152, 153, 154
hypothesis for, 143
in multifactorial model, 281
voting behavior connection, 33, 34
incumbent reward-punishment voting. See economic voting
independence/unification issue. See unification/independence issue
independent voters
candidate evaluation, 188
demographics of, 73
education levels, 79, 81–82, 82, 92
literature review, 72–73
national identity connection, 57, 58
as non-KMT supporters, 78
overview, 16–17
pivotal role of, 73, 91, 92
political interest, 83–84, 84
political participation, 254
post-1990s partisanship, 79–81, 81
role in developing party system, 91
support for democracy, 86, 87
Taiwan Provincial Assembly elections, 77–78, 77
unification/independence issue, 85, 85
vote choice, 89–90, 89, 90
voting percentages, 87, 88, 89
industrial restructuring, 1, 159, 160
inequality. See wealth inequality
Inglehart, Ronald, 100, 114, 200
instrumentalism, 67n2
investment in China, by Taiwan
benefits of, 166–67
change over time, 159–62, 161
Ireland, confusion with left-right labels, 215, 219n15
issue orientation, 98–138
environmental protection/economy issue overview, 101, 111–14, 113,
128
importance in electoral decision-making, 2, 17–18, 98–102
irrelevance of left-right labels, 210, 212–13, 212, 213
partisan elites framing of, 102–3
public perceptions of most important problems, 123–27, 123, 124, 125
questionnaire wording and results, 130–31, 132
reform/stability issue overview, 98, 100, 101, 103–5, 105, 127
Page 305 → unification/independence issue overview, 98, 100–101, 114–
22, 116, 122, 128, 129–30
wealth distribution issue overview, 100, 101, 106–11, 109, 129
See also specific issues
issue ownership theory, 181, 188

Jansen, Guido, 200

Japan
adoption of mixed-member system, 19, 232, 245
colonial rule of Taiwan, 4–5, 9, 48
Jones, Calvin C., 121
Jou, Willy, 202, 215
Key, V. O., Jr., 140
Kiewiet, D. Roderick, 140
Kinder, R. Donald, 140, 172, 173, 174
KMT Party. See Kuomintang (KMT) Party
Kramer, Gerald, 140
Kuomintang (KMT) Party. See also Pan-Blue Alliance
authoritarian regime, 5–6, 10, 73–76, 255–56
candidate evaluation, 19, 175
Chinese consciousness of, 58
coalition nature of, 41n3
constitutional amendments, 246n4
control of legislative branch, 227
development of offshoot parties, 78, 116, 239
dominance under single-party system, 226–27
economic policies, 111–14, 128, 142, 162, 283
electoral reform advocacy, 223–24, 234–36, 240–42
environmental policies, 111–14, 128
ethnic identity connection, 36–40, 38
grassroots support for, 75
irrelevance of left-right labels, 201, 209–10, 209, 211
issue ownership theory applied to, 181–88
monopolization of political power, 16–17, 91
number of registered members, 95n3
Pan-Blue Alliance leadership, 11, 15, 39
post-1990s partisanship, 79–81, 81
postcolonial rule of Taiwan, 48–49
presidential campaigns, 27, 88, 90
 public perceptions of, 104, 105
reform/stability issue, 126
re-Sinicization measures, 45, 49, 51, 65
retreat to Taiwan, 5, 16, 31, 48, 58, 73
six-three-three slogan, 147, 155n2
social welfare policies, 106, 107–9, 109, 110, 129, 130, 134n9
suppression of left views, 203
unification/independence issue, 115, 116, 118, 119, 120–22, 128, 287

Laakso-Taagepera numbers, 241–42, 242

Lacy, Dean, 142
Lai, William, 21n7
Lanoue, David, 141
Laver, Michael, 199–200
Lazarsfeld, Paul, 2
Lee Teng-hui
candidate evaluation of, 173, 174–77, 182–83, 186, 188, 190
economic conditions during presidency, 182
evolution of unification/independence position, 51–52, 119, 120, 126
independent voter support of, 88
indirect election of, 49, 195n3
Japanese citizenship of, 21n4
Minnan support for, 31
missile crisis, 163
National Development Meeting, 233
popularity of, 33
reelection of, 6, 30
 social welfare policies, 134n9
left, cultural connotations of term, 203–4
Page 306 → left-right political spectrum, 198–222
citizen perception of, 204–7
concept of, 199–201, 218n3
conventional view of, 214
cultural connotations of, 203–4, 219nn6–7
lack of meaning in Taiwan, 19, 134n10, 201–2, 213–16, 218n4, 219n5
overview, 198–99
survey evidence, 207–13, 208–9, 211–13
survey questions, 216–18
legislative elections
electoral system change effects, 242–45, 242, 244
history of, 226–27, 228, 229, 231, 237
interaction with presidential elections, 232, 233, 241, 246n6
post-1990s partisanship, 80
preludes to postreform election, 239–40
unification/independence issue importance, 125–26
voter turnout, 256–57, 257, 258, 259
Legislative Yuan
change to direct election, 256
constitutional amendment of, 231–32
electoral system change ratification, 223
events leading to electoral reform, 231–39, 233, 237
nuclear power plant compromise, 112
Lewis-Beck, Michael, 140, 277
Lien Chan
candidate evaluation of, 183, 185–87, 190–91
electoral reform advocacy, 235–36
presidential campaigns, 27, 30, 115, 177, 178–79, 235
running mate candidacy, 174, 286
social welfare policies, 134n9
Lijphart, Arend, 247n8
Lin His-shan, 134n9
Lin Yang-kang, 27, 119, 175, 183, 195n4
Lin Yi-tze, 142, 146
Lin Yi-Xiong, 236
Lipset, Seymour M.
freezing hypothesis, 11, 224, 285
political cleavages, 100, 224
political participation, 262–63
list tier seats, 238
local elections, history of, 256
Lockerbie, Brad, 141

M + 1 rule, 224, 242–43, 244

MacKuen, Michael B., 140
mainlanders
defined, 21n8
monopolization of political power, 48–49
national identity changes, 50, 50, 53–54, 53, 58
Pan-Blue Alliance identification, 16, 51
political participation, 254, 258, 260
 support for KMT, 74, 76
support for Lee Teng-hui, 175
Taiwanese identity among, 56
voting behavior, 31–33, 32, 281
Malanchuk, Oksana, 172
Mandarin dialect, left-right terms, 203, 204, 214
manufacturing
move from Taiwan to China, 159
small factory participation, 133nn6–7
Markus, Gregory B., 121, 140
Ma Ying-jeou
candidate evaluation of, 173, 179–80, 186
economic policies, 112, 143, 160, 283
election of, 7, 10, 27, 31, 53, 240
engagement policies, 18
environmental policies, 112
mayoral election victory, 177, 179
Minnan support for, 31
peace dividend, 149
pinyin system use, 22n9
Page 307 → reelection of, 287
reform/stability issue, 105
Sunflower Student Movement protest, 290n9
support from business, 129
unification/independence issue, 118, 120
memory-based model of candidate evaluation, 171
Miller, Arthur, 172
Miller, Warren, 172
Minnan ethnic group
defined, 21n8
national identity changes, 9, 53–54, 53
Pan-Green Alliance identification, 16, 51
political participation, 254, 258, 260
pre-1986 lack of representation, 74
support for KMT, 74, 75, 76
support for Lee Teng-hui, 175
views on ethnic tension, 50, 50
voting behavior, 31–33, 32
missile buildup and tests, 7, 10, 14, 59–60, 163
mixed-member majoritarian system (MMM)
adoption of, 12, 19–20, 223, 238
impact of, 239–45, 242, 244
Japan adoption of, 19, 232, 245
SF ratio, 243, 244
subtypes of, 246n1
ticket-splitting, 243–45
Moser, Robert, 226
most important problem, as issue, 123, 123–27, 124, 125, 131

Nadeau, Richard, 140

National Assembly of Taiwan
change to direct election, 41n1, 195n3, 256
constitutional amendments, 246n4
 electoral system change ratification, 223, 231, 232, 236, 238
National Development Meeting (1996), 233
national health insurance programs, 107, 108
national identity
changing boundaries of, 58–62, 60, 61, 64, 65, 274–75
confusion with left-right labels, 215
by ethnicity, 53–54, 53
importance in electoral decision-making, 2, 8
in multifactorial model, 281
overlap with partisan identification, 275–77, 276
perception of animosity from China, 60–61, 61
by political generation, 54–55, 55
political participation and, 254–55, 258, 259, 260–62, 262, 267
private poll results, 67n7
unification/independence issue and, 10, 13, 16, 20–21, 61–62, 61
See also Chinese identity; dual Chinese-Taiwanese identity; Taiwanese
identity
Nationalist Party. See Kuomintang (KMT) Party
national referendum proposal, 149, 155n3
national security, as valence issue, 123–24
National Unification Council, 52, 119
National Unification Guidelines, 52, 119
New Central Way policy, 115, 121
New KMT Alliance, 119
New Party (NP)
DPP alliance, 104
formation of, 27, 78, 91, 119, 175
 Pan-Blue Alliance membership, 11
support for unification, 13, 117
new Taiwanese (xin-Taiwanren), 50, 52
Nie, Norman, 252
1992 consensus, 119, 134n16, 283, 287, 290n8
Niou, Emerson M. S., 100, 142
nominal tier seats, 238, 243, 246n1, 246n3
Page 308 → Nonpartisan Solidarity Union (NPSU)
cooperation with KMT, 240
opposition to MMM system, 236, 238
Pan-Blue Alliance membership, 11
nonpartisan voters. See independent voters
nonresponse rates to left-right question, 207–8, 208
Norris, Pippa, 108, 201
north-south divide. See regional divide
No Unification, No Independence, and No Use of Force pledge, 120
NP. See New Party
NPSU. See Nonpartisan Solidarity Union
nuclear power plant protests, 111, 112–14, 128

occupation, and voting behavior, 34–35, 36

One China principle
insistence on, 58–59
international effects, 9
KMT support for, 51–52, 74
1992 consensus, 119, 134n16, 283, 287, 290n8
online model of candidate evaluation, 171
Page, Benjamin, 120, 121
Paldam, Martin, 277
Pan-Blue Alliance
age demographics, 35–36, 37
consolidation of two-party system, 27–31, 28, 29
coordination in, 240
education demographics, 33–34, 35
electoral advantage of, 28, 31, 39, 285–86
ethnic identity connection, 31–33, 32, 36–40, 38
formation of, 11, 17, 27, 78–79, 270n4
income demographics, 33, 34
mainlanders identification with, 16, 51
national identity connection, 57, 58
occupation demographics, 34–35, 36
post-1990s partisanship, 79–81, 81
regional divide effects, 15–16, 30–31
wealth distribution issue, 129
See also Kuomintang (KMT) Party
Pan-Green Alliance
age demographics, 35–36, 37
consolidation of two-party system, 27–31, 28
economic policies, 283
ethnic identity connection, 33–34, 35, 36–40, 38
ethnic identity demographics, 33
formation of, 11, 17, 27, 78–79, 270n4
income demographics, 33, 34
increasing competitiveness of, 15
 Minnan identification with, 16, 51
national identity connection, 57, 58
occupation demographics, 34–35, 36
post-1990s partisanship, 79–81, 81
regional divide effects, 15–16, 30–31
wealth distribution issue, 129
See also Democratic Progressive Party
parliament. See Legislative Yuan
parliamentary elections. See legislative elections
partisan identification, 71–97
candidate evaluation connection, 19, 174, 180–89, 183–87, 191
categories of, 79, 94–95n1
democratization effects on, 90–91
ECFA evaluation connection, 164–67, 164, 166
economic voting connection, 143, 149, 153, 155
ethnic identity connection, 16, 36–40, 38, 57, 58
factors in, 273–74, 288
freezing hypothesis effects, 11–12
historical origins of, 73–79, 77
impact on voting, 87, 88, 89
Page 309 → importance in electoral decision-making, 2, 8, 16–17
independent and weakly partisan voters, 81–82, 82, 91–92
irrelevance of left-right labels, 213, 214, 219n12
issue position connection, 121–22, 122, 135n18
literature review, 72–73
measurement in TEDS survey, 289n2
 in multifactorial model, 278–82
overlap with national identity, 275–77, 276
overview, 71–72
political interest effects, 81–83, 82, 83, 84
political participation and, 254, 258, 259, 260–62, 261, 266
post-1990s partisanship, 79–81, 81
support for democracy and, 85–86, 86, 87
survey question wording, 218
unification/independence issue effects on, 11–12, 84–85, 84, 85
vote choice effects, 88–90, 89, 90
party list system, 223, 238, 241, 247n9
party polarization, confusion with left-right labels, 214–15
party switching, 239
party system
history of, 228
importance to democracy, 92
power of identity in, 288–89
pre-1986 elections, 73–79, 77
theoretical frameworks, 224–26
See also two-party system; specific parties
Peng Ming-min, 88, 115, 175, 188
People First Party (PFP)
cooperation with KMT, 116, 239
electoral reform advocacy, 235, 236, 238
formation of, 27, 78, 91
Pan-Blue Alliance membership, 11
postreform election, 241
 social welfare policies, 109
personal character
candidate evaluation of, 172–73, 174–80, 176, 178, 180, 190–91
relationship with partisanship, 180–88, 183–87
survey questions, 192, 193, 194, 194n1
personal income. See disposable income per capita
Petrocik, John, 181
PFP. See People First Party
Pierce, Roy, 198, 202
pinyin system, 22n9
Plebiscite Law, 118
Political Change in Britain (Butler and Stokes), 3
political generations, 67n6
political identity. See voter identity
political interest, partisanship and, 81–83, 82, 83, 84
political knowledge, political participation and, 255, 259, 260–62, 263, 268
political participation, 252–72
concerns with unequal participation, 261–62, 265–66
defined, 253
electoral participation, 260–63, 264–68, 270n8
factors in, 20, 253–55, 264–65
need for further research, 263–64
overview, 252–53
restriction of, during authoritarian period, 49
survey questions and coding scheme, 267–69
voter turnout, 255–59, 257, 259–63, 262–63
position issues, 101
 See also specific issues
postal voting, lack of, 22n9
Powell, G. Bingham, Jr., 225, 278
presidential elections
candidate evaluation in, 19
comparison of factors influencing vote choice, 278–82, 280, 282
Page 310 → presidential elections (continued)
ECFA evaluation and vote choice, 166, 166
economic voting in, 18, 142–43, 154
electoral system change effects, 242, 242
former indirect election practice, 41n1, 195n3
historical trends in voter turnout, 256–57, 257
independent voter influence on, 90
interaction with legislative elections, 226, 232, 233, 241, 246n6
partisanship changes over time, 80, 88–89
plurality system, 227, 230
recent results of, 26–27
unification/independence issue importance, 125–26, 286–87
voter turnout by age, 258, 259
primordialism framework
overview, 46–48, 64
relational aspects of boundaries, 45–46, 65
views on Taiwanese identity, 54, 56
proportional representation
Duverger’s hypothesis, 224, 225, 241, 242
electoral system change effects, 242, 242
 in postreform legislature, 240–41
in prereform legislature, 227
sincere voting, 243, 245
prospective economic voting
hypothesis for, 143
logistic regression models, 147–49, 148
mixed evidence for, 142–43, 154, 278
in multifactorial model, 279, 281
results of survey, 144–46, 145, 146
theoretical framework, 141
public perceptions
environmental protection/economy issue, 111–12, 113, 113, 114
most important problem facing Taiwan, 123–27, 123, 124, 125
response to elite framing of issues, 99
unification/independence issue, 115–22, 116, 122
wealth distribution issue, 108–9, 109

Qing dynasty, 4, 9, 48

Rectification Movement (zheng-ming-yung-dong), 52

reform/stability issue
diminishing importance of, 17
framing of, 103–5, 105
importance during democratization, 100, 101
irrelevance of left-right labels, 210, 212–13, 212, 213
issue ownership theory applied to, 182, 186, 187
questionnaire wording and results, 130, 132
 role in party competition, 98
status as most important problem facing Taiwan, 124, 125, 126
survey question wording, 217
regional divide, 15, 28, 30–31
Regulations for Managing the Relations between Citizens of the PRC and
Taiwan, 118
Republic of Taiwan, as term, 9, 115
Resolution on Taiwan’s Future, 115
retrospective economic voting
hypothesis for, 143
logistic regression models, 147–49, 148
mixed evidence for, 142–43, 154, 278
in multifactorial model, 279, 290n6
results of survey, 144–46, 145
theoretical framework, 141
right, cultural connotations of term, 203–4
right-left political spectrum. See left-right political spectrum
rising new nation (xin-xing-min-zu), 50
Rokkan, Stein, 11, 100, 224, 285
Rosenstone, Steven, 253

Page 311 → Scheiner, Ethan, 226

semipresidential system of Taiwan
interaction with legislature, 225–26, 232
role in electoral reform, 239, 245
seniors. See elderly
SF ratio, 242–43, 244
Shanks, J. Merrill, 172
Sheng, Shin-yuan, 142
sheng-ming-gong-tong-ti (collectivity of common fate), 50
Shih Ming-te, 133n3
Shils, Edward, 46
Shugart, Matthew, 246n1
Sigelman, Lee, 198
Singer, Matthew, 3
single-member districts (SMDs)
creation of, 223
Duverger’s law, 224–25, 241, 242
electoral system change effects, 242, 242
strategic voting in, 243, 245
single non-transferable vote (SNTV) electoral rules
disadvantages of, 231
history of, 226–31, 228, 229, 230, 246n2
reform of, 12, 223
SF ratio, 243, 244
six-three-three slogan, 147, 155n2
small business ownership, 107, 134n8
small parties
MMM system disadvantages, 238, 239
SNTV system advantages, 231
SNTV. See single non-transferable vote (SNTV) electoral rules
socialization, effect on constructed identities, 47, 65
social order, as valence issue, 123–24
social welfare policies
 as emerging issue, 283–85, 284
evolution of, 106–11, 109
irrelevance of left-right labels, 199–200, 210, 212–13, 212, 213
in multifactorial model, 281
survey question wording, 217
Soong, James
electoral reform advocacy, 236
Lien Chan alliance, 115, 177, 178
move to PFP, 27, 116, 179
presidential campaigns, 30, 235, 241, 286, 287
south-north divide. See regional divide
spatial model of left-right voting, 198
Stevenson, Randolph, 141
Stimson, James A., 99, 140
Stokes, Donald, 3
strategic voting, 241, 243–45
Sunflower Student Movement, 262, 266, 283, 290n9

Taiwan citizen, defined, 21n3

Taiwan Consensus, 119, 286, 287
Taiwan Election and Democratization Studies. See TEDS (Taiwan Election
and Democratization Studies)
Taiwan Election Commission, 257
Taiwanese, as term, 21n3
Taiwanese dialect, left-right terms, 203, 214
Taiwanese identity
by ethnicity, 53–54, 53
 increasing population with, 58, 64
in multifactorial model, 281
Pan-Green Alliance support, 11
perception of animosity from China, 60–61, 61
by political generation, 54–55, 55
political participation and, 255, 258, 262
primordial conceptualization of, 54, 56
reinforcement with independence views, 10
Taiwanese consciousness vs., 67n4
See also dual Chinese-Taiwanese identity
Taiwan Independence Clause, 114, 119
Taiwan National Security Survey, 4
Taiwan Provincial Assembly elections, 77–78, 77
Page 312 → Taiwan Provincial Council, 233
Taiwan Solidarity Union (TSU)
formation of, 78, 91
opposition to MMM system, 235, 238
Pan-Green Alliance membership, 11
party switching to DPP, 239
postreform election, 241
unification/independence issue, 13, 116–18, 126
Tan, Alexander, 168
Tanaka, Aiji, 214, 219n14
TEDS (Taiwan Election and Democratization Studies)
left-right political spectrum, 202, 204
partisan identification measurement, 289n2
 political participation data, 253, 257
as source, 4
ticket-splitting evidence, 244
Thomassen, Jacques, 3
Three Direct Links, 162
ticket-splitting, 243–45
Tien, Hung-mao, 142
tongdu issue. See unification/independence issue
trade between Taiwan and China
benefits of, 166–67
change over time, 159–62, 160
trait ownership theory, 181, 191
trust, in candidate evaluation, 172
Tsai Ing-wen
candidate evaluation of, 179, 180, 186
loss to Ma Ying-jiou, 241
support for cross-Strait status quo, 130, 287
Taiwan consensus, 119, 286
TSU. See Taiwan Solidarity Union
Tuan Yi-kang, 119
Tufte, Edward R., 140, 149
Tverdova, Yulia, 173
2-28 Incident (1947), 6, 9, 48, 73
two-party system
consolidation into, 15–16, 27–31, 28, 29, 39, 91, 227, 245–46
Dang-wai situation, 77
district level effects, 243
 institutional structures effects on, 19–20
largeness of parties in, 273
stability of, 285

unification/independence issue
changing boundaries of, 62–64, 63
Chinese vs. Taiwanese consciousness and, 51
dominance in Taiwanese politics, 8–12
economic voting connection, 143, 164–67, 164, 165
evolution of issue, 114–22, 116, 122, 128
importance in future elections, 285–89
irrelevance of left-right labels, 203, 206, 210, 212–13, 212, 213, 215
issue ownership theory applied to, 182
relationship with partisanship, 84–85, 84, 85, 276–77
relative importance of, 6, 14–15
role in electoral reform, 245–46
role in party competition, 98, 100–101
shifting public opinion of, 167–68
status as most important problem facing Taiwan, 124–26, 124, 125
survey questions, 131, 132, 133, 216
wealth distribution issue absorbed into, 129–30
United Nations, Taiwan application for membership, 118, 121

valence issues, 101, 124

See also specific issues
Verba, Sidney, 252
vote choice
 comparative strength of factors influencing, 2, 8, 13–15, 101, 277–82,
280, 282
funnel of causality model, 170, 191
Page 313 → power of identity in, 289
relationship with candidate evaluation, 174, 188–89, 189, 190, 191–92
relationship with partisanship, 88–90, 89, 90
See also economic voting; specific issues
voter identity, 45–70, 273–290
Chinese/Taiwanese consciousness boundary, 51–58, 53, 55, 57, 64–65
comparative strength of factors influencing vote choice, 278–82, 280,
282, 289n4
complexity of, 273–74
economic voting and, 277–78
ethnicity boundary, 48–51, 50, 64
evolution from ethnic to national, 274–75
implications for future elections, 285–89
importance in electoral decision-making, 2
national identity boundary, 58–62, 60, 61, 65
overlap of national identity and partisan identification, 275–77, 276
overview, 45–46
role of ECFA attitudes in central cleavage, 282–85
theoretical frameworks, 46–48, 64
unification/independence issue effects on, 62–64, 63, 65
See also ethnic identity; national identity
voter level changes, postreform, 243–45
voter turnout
 explanations of, 257–59, 259–63
historical trends in, 255–57, 257
importance to democracy, 252–53
misreporting of, 257, 270nn6–7
relationship with partisanship, 87, 88, 89

Wattenberg, Martin, 172, 246n1

wealth distribution
evolution of issue, 100, 101, 106–11, 109, 129–130, 282–85, 284
importance in electoral decision-making, 17
questionnaire wording and results, 130, 132
status as most important problem facing Taiwan, 124, 125, 126
wealth inequality
economic development effects, 155n1
increase in, 107, 110–11, 129, 282–85, 284
unequal political participation and, 265–66
Western democracies, left-right political spectrum, 198, 199–201, 215
Whitten, Guy, 278
Wilson, J. Matthew, 141
Wong Chung-chun, 134n9
Wu, Chin-en, 142, 146
Wu, Yu-Shan, 284

xin-Taiwanren (new Taiwanese), 50, 52

xin-xing-min-zu (rising new nation), 50

Yough, Syng Nam, 198

You Xi-Kun, 235
Zaller, John, 152
Zechmeister, Elizabeth, 3, 200–201
Zhang Xue-shun, 246n7
zheng-ming-yung-dong (Rectification Movement), 52