This is to certify that the
dissertation entitled

THE LINK BETWEEN DOMESTIC POLITICAL
INSTITUTIONS AND COMMITMENTS TO
INTERNATIONAL TRADE AGREEMENTS

presented by
Jung In Jo

has been accepted towards fulfillment
of the requirements for the

Ph.D. degree in Political Science

jwm

Major Pfofe'ssor’s Signature

9I3~sio~t

 

Date

MSU is an Affirmative Action/Equal Opportunity Institution

 

 

. LIBRARY
Michigan State
University

 

 

 

PLACE IN REI‘URN Box to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJCIRC/DateDue.p65-p.15

THE LINK BETWEEN DOMESTIC POLITICAL INSTITUTIONS AND
COMMITMENTS TO INTERNATIONAL TRADE AGREEMENTS

By

Jung In Jo

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Political Science

2004

ABSTRACT

THE LINK BETWEEN DOMESTIC POLITICAL INSTITUTIONS AND
COMNIITMENTS TO INTERNATIONAL TRADE AGREEMENTS

By

Jung In Jo

Since the early 19803 many states have liberalized their trade. To do this, many
countries have joined Preferential Trading Arrangements that reduce trade barriers.
Almost every country in the world is a member of a preferential trade arrangement.
However, the puzzle scholars face is why so many countries have rushed to enter a free
trade agreement by lowering trade barriers at this historical juncture. Although a recent
expansion of preferential trade agreements has stimulated many debates, we lack an
understanding as to what motivates states to make a commitment to international trade
agreements since the 1980s. Moreover, economists have not paid enough attention to the
changing political motivations of leaders to sign Preferential Trade Agreements (PTAs).
In this dissertation, I ask how and whether domestic institutions of a country make a
difference in committing to international trade agreements. The purpose of this
dissertation is to provide a more convincing theoretical framework and empirical support
incorporating an interaction of leaders’ preferences and institutional constraints in order
to explain why more countries pursue interstate commercial cooperation by entering
PT As.

In this dissertation, I argue that the rush to PTAs since the 19808 is linked to the
change of domestic constraints. A decision-maker’s preferences can be constrained

through political institutions in the dynamic process of trade policy. In this dissertation,

I aim to demonstrate how politicians may implement their particular trade policy
preferences in different regimes and how this is linked to a decision to sign tariff
reduction agreements. I also aim to demonstrate how the constraints of political
institutions may be linked to a decision to sign tariff reduction agreements. I claim that
the propensity of signing a trade agreement increases as the survival of political leaders
becomes more dependent on having a large size of winning coalition. Decision-makers
have greater political motivations to enter international trade agreements as the level of
democracy increases. In order to test the theories, I use a large N quantitative study. This
dissertation tests key propositions drawing on a comprehensive database of international
trade agreements from 1950 to 1992 with 150 countries. I find that leaders have greater
political incentives to conclude international trade agreements as the size of their winning
coalition grows. The results also suggest that the level of democracy is associated with a
leader’s political motivation to conclude international trade agreements. I also find that
electoral rule and the number of veto players have an impact on interstate commercial
cooperation. By highlighting this dynamic, this dissertation makes important
contributions to theories of international relations and international political economy
more generally. Domestic institutional constraints deserve more attention in international
political economy. Domestic constraints and international constraints are intertwined.
On the basis of the arguments and evidence provided in this dissertation, it is clear that
domestic institutional analysis facilitates our understanding of trade policy in

international politics.

This dissertation is dedicated to my parents, my most ardent supporters and most valuable
friends.

iv

ACKNOWLEDGMENTS

I thank my greatest supporter, Dr. Kyu Dok Hon g for giving me a good start in my
expedition. Your advice and encouragement is always greatly appreciated. The presence
of Dr. Woosang Kim from my undergraduate days helped steady my course. I also owe a
great deal to Dr. Kyungsook Lee, Dr. KyungOk Chun and Dr. Eugene Lee. I found gentle
support and a demand in them.

During the writing this dissertation, I benefited from the support and suggestions
of many advisors and colleagues. I thank my chair and advisor, Scott Gates. He offered
invaluable advice throughout. He read my repeated drafts and offered many invaluable
insights and suggestions. I am deeply grateful to him. Thank you. It has been an honor to
work with you. I learned much from you.

I owe a special debt of gratitude to my committee members. To Mark Jones I owe
the dissertation’s focus on veto players. I am also grateful to Nicholas Van de Walle and
Michael Schecter for their significant contributions. Beyond my committee, I also thank
Dr. Seokwoo Kim and Dr. Wook Heo for their helpful comments. Edward Mansfield
kindly sent me his data. Marc Busch gave me some invaluable suggestions when I
presented a draft of my dissertation at the Midwest Annual Meeting. Helen Milner helped
inspire the ideas here. I want to thank Wonjae Hwang and David Letkzian for their
suggestive advices. I also want to thank William Gordon Bewick and Scott Symendra for

their help. Thanks also go to Dennis Patterson, Dr. Giljin Lim, Dr. Kiwon Kang, Brian

Silver, William Reed, Eunmyo Cho, Jinhee Cho, Sang Eun Byun, Sayaka Hirayanagi,
Sooyoung Cho, Thomas Kang, Janice Lee, Wonbin Cho, Doorae Kim, and ZeeSun Yun.
I would like to thank my family who provided a mental and financial support over the
years of graduate school. My most affectionate note of acknowledgement must go to my
mother and father. They were very supportive. I especially wish to thank to God for his

lead.

vi

TABLE OF CONTENTS

LIST OF TABLES ........................................................................................................... x
LIST OF FIGURES ........................................................................................................ xi
Chapter One. Introduction... ....................................................................................... 1
1.1 Puzzling Motivation ................................................................................................... l
1.2 The Importance of Free Trade Agreements ............................................................... 3
1.3 Domestic Constraints of External Trade Policy: Argument ....................................... 4
1.4 Testing the Theory ..................................................................................................... 7
1.5 Contributions of the Dissertation ............................................................................... 7
1.6 Outline of the Dissertation ......................................................................................... 9
Chapter TWO. Why Preferential Trade Agreements? ............................................... 10
2.1 Difficulty of Analysis ............................................................................................... 10
2.2 History of Preferential Trading Agreements and the Resurgence

of Preferential Trade Agreements .................................................................................. 11
2.3 The Significance of Preferential Trade Agreements: Does It Real]

Matter? ........................................................................................................................... 14
2.3.1 WTO and Preferential Trade Agreements ............................................................. 14
2.3.2 Evidence of Preferential Trade Agreements ......................................................... 15
2.3.2.1The Recent Trend of Preferential Trade Agreements ......................................... 15
2.4 The Impact of Preferential Trade Agreements ......................................................... 21
2.5 Why do Countries Join Preferential Trade Agreements instead of

Choosing Unilateral or Multilateral approaches? .......................................................... 25
2.5.1 External and Economic Factors ............................................................................ 26
2.5.2 Internal Factors .................................................................................................... 28
2.5.3 Political Motivations ............................................................................................. 29
Chapter Three. Literature Review ............................................................................. 31
3.1 Theories of Domestic Societal Preferences ............................................................. 31

vii

3.2 Theories of Political institutions .............................................................................. 34

3.3 International Factors ................................................................................................ 37
3.3.1 Neo-realism and Hegemonic Stability Theory ..................................................... 38
3.3.2 Neo-liberalism ...................................................................................................... 38
3.3.3 Power Transition Theory ...................................................................................... 40

Chapter Four. Variations of Domestic Political Institutions and the Determination
to Enter International 'fi'ade Agreements ............................................................ 41

4.1 The Linkage between Variations of Domestic Political

Institutions and International Trade Policy Outcomes ................................................... 42
4.1.1 Bueno de Mesquita’s Theory and International Trade Policy .............................. 42
4.1.2 Veto players and International Trade Policy ......................................................... 47
4.1.3 Electoral system and International Trade Policy .................................................. 49
4.1.4 Structure of the Government and International Trade Policy ............................... 50
4.2 A Perspective on the Public Goods Aspects of Free Trade ...................................... 51
4.2.1 Free Trade and Preferential Trade Agreements .................................................... 55
4.2.2 Regime Type and Free trade ................................................................................. 56
Chapter Five. Research Design and Hypotheses... ................................................... 58
5.1 Introduction .............................................................................................................. 58
5.2 Assumptions ............................................................................................................. 59
5.3 Research Design ...................................................................................................... 62
5.3.1 Hypotheses: Domestic Institutional Constraints ................................................... 62
5.3.1.1 Regime type ....................................................................................................... 62
5.3.1.2 Veto players ........................................................................................................ 66
5.3.1.3 Electoral system ................................................................................................. 71
5.3.1.4 Presidentialism vs. Parliamentarism .................................................................. 73
5.4 Data and Measurement ............................................................................................ 75
5.4.1 Dependent Variable ............................................................................................... 75
5.4.1.1 The Decision to Sign Preferential Trade Agreements ........................................ 75
5.4.2 Proxy of winning coalition and selectorate size ................................................... 80

5.4.2.1 Comparison between Bueno de Mesquita’s measurement
and Vanhanen’s measurement on democracy ................................................................ 82
5.4.3 Electoral Rule and Veto Players ............................................................................ 84

viii

5.4.4 Control Variables .................................................................................................. 85

Chapter Six. Empirical Estimations and Results... .................................................. 89
6.1 Model Specifications and Theoretical Expectations ................................................ 90
6.1.1 The Effect of the Winning Coalition and Selectorate ........................................... 91
6.1.2 The Effect of Regime type .................................................................................... 91
6.1.3 The Effect of Veto Players .................................................................................... 91
6.1.4 The Effect of Electoral Rule ................................................................................. 92
6.1.5 The Effect of Presidentialism ............................................................................... 93
6.2 Methodology: GEE method and Weak Link Analysis ............................................. 93
6.3 Estimation Results ................................................................................................... 96
6.3.1 The Size of Winning coalition and Selectorate ..................................................... 99
6.3.2 Democracy .......................................................................................................... 100
6.3.3 Veto Players ........................................................................................................ 100
6.3.4 Electoral Rule ..................................................................................................... 101
6.3.5 Presidential Political System .............................................................................. 101
6.4 Variations among Democracies ............................................................................. 102
6.5 Comparison with Alternative Approaches ............................................................. 106
6.5.1 Neo-Realism and Hegemonic Stability Theory .................................................. 106
6.5.2 Neo-liberal Institutionalism ................................................................................ 107
6.5.3 Power Transition Arguments .............................................................................. 108
6.5.4 Interpretation of Alternative Explanations .......................................................... 109
6.5.4.1 Structural approach: Neo-Realism, Neo-liberal Institutionalism,

Hegemonic Stability theory, and Power transition Theory .......................................... 109
6.6 Interpretations and Implications ............................................................................ 111
Chapter Seven: Conclusion ....................................................................................... 113

7.1 Recapping the Motivating Questions and Contributions of the

Dissertation .................................................................................................................. 1 13
7.2 Limitations of the Argument .................................................................................. 116
7.3 Implications ........................................................................................................... 118
7.3.1 How Domestic Politics Influences International Relations ................................ 118
Bibliography ................................................................................................................ 120

ix

LIST OF TABLES

Table 2.1. Number of Regional Trade Agreements (RTAs) Notified to GATT ............ 17
Table 2.2. Intra-Regional Trade as a Percentage of Total Trade .............................. 19
Table 2.3. Intra-regional trade divided by total trade of each region ........................ 19
Table 2.4. Intra-Regional Trade Intensities ...................................................... 20
Table 5.1. Levels of 171’ As ......................................................................... 77

Table 6.1. Descriptive Summary Statistics on PTA Formation Database Variables ...... 98

Table 6.2. Political Institutions and the Leader’s Determination to Join PTAs (Dyadic
analysis) GEE Population-Averaged Model with Weak Link Principle ................... 104

Table 6.3. Political Institutions and the Leader’s Determination to Join PTAs (Dyadic
analysis) GEE Population-Averaged Model with Weak Link Principle ................... 105

LIST OF FIGURES

Figure 2.1. The Number of Preferential Trading Arrangements notified to the GATT

1948-94 ............................................................................................... 16
Figure 5.1. Sources of the Formation of Trade policy .......................................... 59
Figure 5.2. PT As in Force and Under Negotiations ............................................. 78

xi

Chapter One. Introduction

1.1. Puzzling Motivation

Since the early 1980s many states have liberalized their trade. To do this, many countries
have joined Preferential Trading Arrangementsl that reduce trade barriers. Almost every
country in the world is a member of a preferential trade arrangement.2 Sixty percent of
world trade occurs within preferential trading blocs. However, the puzzle scholars face is
why so many countries have rushed to enter a free trade agreement by lowering trade
barriers at this historical juncture.

Although a recent expansion of preferential trade agreements has stimulated many
debates, we lack an understanding as to what motivates states to make a commitment to
international trade agreements since the 19808.3 Moreover, economists have not paid
enough attention to the changing political motivations of leaders to sign Preferential Trade
Agreements (PT As).4 Economists focus on the consequences of PTAs, not the political

sources or causes of PTAs. This dissertation seeks to improve our understanding of why

 

' I use the term preferential trade agreement (PTA) to refer to international trade agreements including
regional trade agreement, free trade agreement, custom unions, and common markets.

2 Even South Korea’s protectionists could not resist any longer. “After three failed attempts, the National
Assembly finally ratified the country’s first-ever free-trade agreement (PTA) with Chile recently although
serious protests had delayed passage of its FI‘A with Chile for over a year.” The Economist: “Why Asian
countries are racing to sign bilateral trade deals with each other”, (March 8, 2004).

3 Out of the 194 RTAs notified to the General Agreement on Tariffs and Trade and the World Trade
Organization between 1948 and 1999, 54% were notified from 1989 to 1999.

‘ RTA and PTA can be used interchangeably. It may be misleading to understand Regional Trade
Arrangements as strictly regional. Regionalism refers to a political process where countries agree to reduce
the trade barriers between each other. In other words, it induces institutionalized cooperation between nations
through the elimination of barriers on trade and investment flows. Regionalism does not necessarily mean
that Regional Trade Agreements will be made in the same region. Free trade agreements are formed with
countries from different regions. The United States and Israel signed a free trade agreement in 1985.
Similarly, under the Lorne convention exports from developing countries of Africa, the Caribbean and Pacific
are given preferential access to European markets. Yet, the majority of the formation of free trade agreements
usually occurs between geographically adjacent countries. For example, the Andean Community consists of
Bolivia, Colombia, Ecuador, Peru and Venezuela.

countries rushed to join trade-liberalizing5 agreements by examining the impact of political
motivations of leaders.

A growing body of literature argues that domestic institutional constraints affect trade
policy patterns. That is, political institutions influence the propensities of “state leaders to
make credible commitments and the willingness of states to enter international agreements”
(Leeds 1999). In this dissertation, I ask how and whether domestic institutions of a country
make a difference in committing to international trade agreements. This dissertation is
organized around the following questions: are domestic preferences related to the recent
international change of economic interactions? What role do domestic political institutions
play in the decision to sign international trade arrangements? Current studies leave these
important questions ambiguous.

Scholars have argued that states choose free trade policies because of external factors
(Mansfield 1998, Gilpin 2001) and domestic factors such as preferences of political leaders
(Rogowski 1989, Frieden 1991), and preferences of pressure groups (Grossman and
Helpman 1995, Magee, Brock, and Young 1989), and political institutions (Milner 1999).
More recently, scholars have claimed that interplay between domestic institutions and

preferences should be incorporated within the literature regarding the recent change of

 

5 The question is whether a free trade agreement is a stepping-stone or a stumbling block towards multilateral
liberalization. This is an important matter as the external trade policy of countries becomes more liberal over
time. There is little consensus on this issue in part because free trading arrangements differ. The effects of
free trade agreements depend on the specific models and are sensitive to the characteristics of the member
countries. Many studies suggest that the answer may be positive or complementary (Lawrence 1996). For
instance, the expansions of the European Economic Community (EEC) contributed to pressures by GATT
members to initiate multilateral trade negotiations. It led to significant reductions in trade barriers. It has also
been argued that free trade agreements may facilitate liberalization in areas that are too complex to be
negotiated successfully in the WTO. This is an argument for the claim that free trade arrangements are
stepping-stones to multilateralism. However, despite some positive effects of free trade agreements, the
disagreements remain. Arriving at a satisfactory explanation of the consequences of the free trading bloc is
difficult because developments are still moving at a rapid pace. It is not my purpose to focus on consequences
of expanding free trade agreement, but I will address this question later. For a detailed analysis of the
consequences of regional trading blocs on the world economy, see Frankel 1997.

external trade policy across countries. However, with the exceptions of Milner, Mansfield
and Rosendorff (1997, 2002), the literature does not provide a well-developed model as to
why political leaders may change their trade policy as domestic political institutions
change. Therefore, establishing a more complete model of how leaders’ preferences and
domestic institutions interact to make a change of trade policy is an important focus of the
dissertation. Accordingly, the purpose of this dissertation is to provide a more convincing
theoretical framework and empirical support incorporating an interaction of leaders’
preferences and institutional constraints in order to explain why more countries pursue

interstate commercial cooperation by entering PTAs at this historical juncture.

1.2. The Importance of Free Trade Agreements
As noted above, there has been a surge of international trade agreements notified to the
GATT and subsequently to the WTO. Yet, the impact of PTAs is contested. There are
various discrepancies in terms of the size and importance of trade agreements. For this
reason, some say that except for institutionalized trading arrangements such as EU and
NAFTA, most smaller trading arrangements are not influential. In this regard, some
scholars underestimate the importance of the growth of PTAs. However, if PTAs are
illiberal and insignificant, how can we explain the rush to create them?

As many PTAs have been formed since World War II, the level of free trade has
increased. It is clear that “rising trends of PT As have become crucial elements of the

international political economy.”6 Even if we do not have a clear conclusion regarding “the

 

6 Mansfield and Reinhardt 2002.

welfare effects of Preferential Trade Agreements,7 it is difficult to underestimate the rising
importance of Preferential Trade Agreements and their attraction to many governments in
the multilateral trade system and in the WTO.”8

Second, the WTO recognizes that trading agreements benefit countries. Countries
turn to PTAs as a mechanism for economic development. The importance of international
trade agreements is being recognized even by small countries because they find that global
free trade is much harder to achieve. In this sense, PTAs are attracting more countries in
recent years and becoming a bigger part of international trade.

Third, the characteristics of PTAs have changed. “Many of the PTAs that formed in
the 19608 and 1970s were based on a model of import substituting development. The
majority of PT As implemented a high level of external trade barriers. However, recently
formed PTAs are more outward looking and more committed to boosting international
commerce.”9 Trading agreements have clearly accelerated the level of liberal trade and

globalization; accordingly, the rising importance of PTAs cannot be overlooked.

1.3. Domestic Constraints of External Trade Policy: Argument

A growing number of studies suggest that domestic institutional constraints affect divergent
trade policy patterns. For instance, Nlilner (1999) argues that a global rush toward tariff
reduction and the expansion of democracy is not coincidental. Other researchers also claim

that these two phenomena must be correlated.

 

7 Scholars suggest that the welfare implications of regionalism are varied across time and trade blocs as
mentioned earlier.

8 Kerremans and Switky, 2000.

9 Schiff and Winters, 1998.

In this dissertation, I argue that the rush to P’I’As since the 1980’s is linked to the
change of domestic constraints. The model I develop demonstrates that leaders have greater
political incentives to conclude international trade agreements as the size of their winning
coalition grows. That is, decision-makers have greater political motivations to enter
international trade agreements as the level of democracy increases. I claim that the
propensity of signing a trade agreement increases as the survival of political leaders
becomes more dependent on having a large size of winning coalition.

In reality, many countries which experience trade liberalization also go through the
process of democratization. In Mexico, the period of the growth of political competition
and the decline of the dominant one party system coincide with the liberalization of trade
policy. However, they do not necessarily go together all the time. The trade liberalization
of many developing countries occurred before they started democratizing their political
system. Chile, Turkey, Taiwan, and South Korea were successful cases of trade
liberalization. They all began their trade liberalization processes before their democratic
transitions. Furthermore, “fragile new democracies have difficulty carrying out economic
liberalization because its costs make them unpopular and hence it can be politically suicidal
for elected officials. Authoritarian governments should be more capable of initiating and
sustaining major economic reforms.”lo Therefore, it would appear dubious that
democracies induce more liberalizing trade policies. However, scholars argue that any
change in political regime may motivate trade reforms. According to Milner, “although
strong evidence has not yet been presented, changes in political regimes, and specifically

the spread of democracy, may be the institutional change that helps explain the rush to free

 

‘° Geddes r995.

trade” (Milner 1999). In this context, an important goal of this dissertation is to assess the
proposition whether domestic institutional change pressures interstate commercial
cooperation toward freer trade.

Other than regime type, there are other factors that need to be assessed. For instance,
party fragmentation, the number of veto player with different ideological backgrounds, and
the structure of government system may mediate and channel domestic preferences in
divergent trade policy. Under different circumstances, how can we account for the effects
of different domestic institutions in explaining the rush to free trade? There are few
systematic studies which analyze different domestic institution’s impact on the change of
commercial policy across states with large N empirical tests.

I claim that whether states choose to enter international trade agreements will be
influenced by the preferences of political leaders as well as constraints of domestic
institutions.” Existing studies leave important theoretical and empirical issues unresolved
regarding the interplay between preferences of political leaders and domestic institutions.
Whether states choose to enter international trade agreements will be primarily influenced
by the preferences of political leaders in the first place. Yet, the nature of domestic
institutions may constrain the responsiveness of leaders to the pressure of social demands
to change international trade policy. In this dissertation, I aim to demonstrate how
politicians may implement their particular trade policy preferences in different regimes and
how this is linked to a decision to sign tariff reduction agreements. I also aim to
demonstrate how the constraints of political institutions may be linked to a decision to sign

tariff reduction agreements.

 

” Mansfield and Milner 1999.

1.4. Testing the theory

As mentioned, despite a prevalent interest on the enormous expansion of PT As, very little
systematic empirical research has been employed to analyze the impact of domestic
constraints. Most work on interstate trade cooperation has been theoretical rather than
empirical with few exceptions. These works have used a small-n case study design due to
data constraints. Therefore, there have been limitations in deducing a generalization on
interstate tariff reduction cooperation. In order to test the theories, I use a large N
quantitative study. This dissertation tests key propositions drawing on a comprehensive
database of international trade agreements from 1950 to 1992 with 150 countries. The
quantitative analysis leads to a systematic evaluation of the role of domestic institutions,
controlling for political factors and economic factors. The empirical study aims to validate
the accuracy of the theoretical predictions made in the dissertation. In doing so, this
dissertation aims to contribute to our understanding of interstate trade cooperation focusing

on pressures of domestic political institutions.

1.5. Contributions of the Dissertation

The direction of the growth of tariff reduction agreements will play an important role in
determining the structure of the international economy in the present century. In this
context, understanding the recent global expansion of free trade agreement is an important
task. Furthermore, the increasing division of the world into competing trade blocs is an
exciting topic, and the research on political factors is inconclusive. As one leading scholar
observes, “None of the existing theories seems to do very well in explaining this important

change in trade policy since the end of World War II” (Milner 1999).

In this context, I identify the interplay between motivations of leaders and political
institutions and assess whether and why countries with certain political institutions are
more likely to commit to free trade agreements. I attempt to provide a solid theoretical
background and rigorous empirical research on how political leaders mediate their trade
preferences through political institutions and how these preferences are linked to a decision
to commit to free trade agreements. In doing so, this dissertation provides theoretical and
empirical contributions.

Most predominant theories of international political economy focus on international
constraints rather than domestic constraints. They assume that states are unitary and
rational actors. They leave domestic actors’ preferences unexplored. They also leave
significant factors such as domestic institutional constraints as ambiguous. Such theories
and models yield incomplete conclusions based on their assumptions. This dissertation
suggests that we need to examine more closely the process of government preference
manifestation by exploring how leaders and political institutions interact. This dynamic has
received little attention in the literature. By focusing on domestic political factors, the
following significant points will be demonstrated. First, domestic institutional constraints
deserve more attention in international political economy. Domestic constraints and
international constraints are intertwined. By underestimating one of those factors, we
develop a misleading understanding of the link between domestic politics and international
trade. This is important not only for the field, but because its findings have significant
practical implications for policymakers. This dissertation also performs a comprehensive
empirical test of the domestic institutional argument. The statistical part of the work

features a systematic dyadic test of institutional theories on the growth of PTAs in the

international political economy. Accordingly, it aims to provide a valid generalization of

domestic institutional theories related to international trade policy.

1.6. Outline of the Dissertation

The dissertation proceeds as follows. Chapter 2 asks why we need to evaluate the
expansion of PTAs and the widespread liberalization of trade policies that has taken place
since the early 19808. Chapter 3 reviews current theories on the recent change of trade
policy. This chapter analyses existing theories and explains their weaknesses. Chapter 4
provides a detailed explanation of domestic institutional theory, by highlighting the link
between domestic institutions and variations in terms of trade policy. It deduces key
propositions on the effect of domestic institutions to the expansion of international trade
agreements. Chapter 5 discusses the rationale for key hypotheses of this dissertation and
the research design of its empirical tests. Chapter 6 presents the methodology and
empirical tests of the hypotheses. Chapter 7 concludes the dissertation. It evaluates how the
models resolve the puzzle of the dissertation. It then discusses some weaknesses of the
dissertation. It also elaborates on some implications of the dissertation for our
understanding of the interplay between domestic institutions and preference formation in
international trade policy outcomes and for theories of international relations more

generally.

Chapter 2. Why Preferential Trade Agreements?
Preferential Trade Agreements (PT As)12 have become increasingly prevalent during the
past 50 years.13 In the period 1948-1994, the GATT received 124 notifications of
Preferential Trade Agreements. Since the creation of the WTO in 1995, over 100 additional
arrangements covering trade in goods or services have been notified. As of March 2002,
250 P'I‘As had been notified to the GATT/WTO, of which 168 are currently in force.14
Compared to previous decades, the 1990s saw PTAs proliferate at an unprecedented rate.
The purpose of this chapter is to ask why we need to evaluate the expansion of PTAs
and widespread liberalization of trade policies that has taken place since the early 19805.
This chapter offers a general overview on the various aspects of PTAs. In this chapter, first
I address the difficulties in analyzing PT As. Second, I provide an overview of the relevant
history and the growth of PT As. Third, I address the significance of PTAs. Fourth, I
examine the static and dynamic impact of PTAs. Finally, I address why countries join
PTAs instead of focusing on unilateral or multilateral approaches. In doing so, I highlight
the rush to join PTAs as one of the most important puzzles in international political
economy.
2.1. Difficulty of Analysis
The tangible resurgence of PTAs has stimulated many debates among scholars, but few

have employed clear analytical models and empirical rigors to analyze this phenomenon.

 

'2 Preferential Trading Agreements refers to all trade agreements that reduce barriers between countries
without including the most~favored~nation rule. This includes partial and full trade areas, customs union, and
single markets (Mansfield and Milner 2000, Mansfield 1999). PTAs are recognized by the WTO in GATT
Article 24 and Article 5 of the General Agreement on Trade in Services (GATS).

'3 Mansfield 1998.

'4 Among the 125 PT As notified during the GATT years, only 51 remain in force.

10

The endogenous aspects of postwar regionalism and the lack of data make it hard to
analyze the systematic causes of the resurgence of trade agreements (WTO 1995).
Furthermore, trade agreements are various in their scopes and the extents of liberalization
actually achieved.

Free Trade Agreements are often considered by scholars to be irrelevant. They
highlight the fact that Free Trade Arrangements are discriminatory, and conclude that FT As
“rarely eliminate external trade barriers. Economists consider them inferior to arrangements
that liberalize trade worldwide” (Mansfield and Milner 1999). However, these arguments
do not explain why PTAs are popular. If P’TAs have little effect on liberalization of trade,
debating the significance of their expansion may not be especially relevant. If free trade
agreements do not make a difference in trade, how can we account for the growth of P’I‘As
and why are they popular? These puzzles are important because they have significant

practical implications15 for policymakers.

2.2. History of Preferential Trading Agreements and the Resurgence of Preferential
Trading Agreements

In this section, I examine the background of the growth of PTAs and clarify their relevance
in international political economy. PT As are not a new phenomenon in international trade.
Before the 19th century, there were several trade alliances associated with global empire
building, but its origins are traced to the 19th century, which saw “the establishment of the

German Zollverein (customs union) in 1833 and the 1860 Cobden-Chevalier Treaty

 

’5 This debate may help us in answering practical political economy issues, such as “whether regional trade
agreement stimulates investment or whether it leads to multilateral trade liberalization agreements” (Schiff
and Winters 1998).

ll

between Great Britain and France, which allowed for free trade. A more recent form of
regional trade agreement is the trade that took place among countries of the British
Commonwealth and Empire on a preferential basis” (Srinivasan 1998). The Lomé
convention and the General System of Preference (GSP)16 are further examples of trading
agreements. However, it was not until after World War H that the resurgence of PTAs
became a hot issue in world trade.

The history of PTA can be divided into two different timeframes “the pre-Cold War
phase and the post-Cold War phase” (Bhagabati 2001). The pre-Cold War phase continued
up to the 1960s and amongst developed countries it was mostly restricted to Western
Europe such as the European Economic Community and amongst developing countries to
Latin America. Negotiations began in 1958 among six European countries led by France
and Germany, to create a Common Market by removing trade barriers among members. In
North America, there was little regionalism because the US wanted to pursue the
multilateral approach. However, the US was a major “external federator to the process of
integration in Western Europe because of the compulsions of Cold War politics”

(Bhagabati 2001). The United States opposed regionalism, but US. saw a Western Europe

 

‘6 “The objectives of the generalized, non-reciprocal, non-discriminatory system of preferences in favor of the
developing countries, including special measures in favor of the least advanced among the developing
countries, should be: to increase their export earnings; to promote their industrialization; and to accelerate
their rates of economic growth. Under GSP of preference-giving counties, selected products in developing
countries are granted reduced or zero tariff rates over the MFN rates. The least developed countries (LDCs)
receive special and preferential treatment for a wider coverage of products and deeper tariff cuts. The idea of
granting developing countries preferential tariff rates in the markets of industrialized countries was originally
presented by Rat’rl Prebisch, the first Secretary-General of UNCT AD, at the first UNCT AD conference in
1964. The GSP was adopted at UNCT AD 11 in New Delhi in 1968. Later, the Contracting Parties decided to
adopt the 1979 Enabling Clause, titled “Differential and More Favorable Treatment, Reciprocity and Fuller
Participation of Developing Countries”, creating a permanent waiver to the most-favored-nation clause to
allow preference- giving countries to grant preferential tariff treatment under their respective GSP schemes.
The following countries grant GSP: Australia, Belarus, Bulgaria, Canada, the Czech Republic, the European
Community, Hungary, Japan, New Zealand, Norway, Poland, the Russian Federation, the Slovak Republic,
Switzerland, Turkey and the United States of America” (cited from www.unctad.org).

12

as an effective deterrent to the growing Soviet threat. Except in Europe and Latin America,
PT As diminished in the 19708 because the majority of liberalization efforts had failed.
“The primary motive behind the movement in the developing world was industrialization
through regional import substitution. It was thought that infant industries could first learn
to export within a protected regional market and then face world competition. But for the
same reasons that import substitution failed in countries, it also failed in the regions. By the
late 19708, outward-oriented policies had begun to capture policymakers. In the years that
followed, unilateral, nondiscriminatory trade liberalization became the order of the day,
and regionalism was pushed into the background.”17 After the 19708, unilateral trade
liberalization became dominant instead of regionalism. After the relative silence of the
19708, regionalism again revived in the late 19808, primarily in North America. The most
important reason was the change in the trade policy of the US towards regionalism. The
most noticeable event was the formation of NAFTA. The US-Israel Free Trade Area,
MERCOSUR, Australia-New Zealand Closer Economic Relations Agreement
(ANZCERTA), AFTA, APEC, SAFTA, the Free Trade Area of the Americas (FT AA) and
the European Union were also formed.

A number of factors motivated countries to pursue a bilateral or regional approach
rather than a multilateral approach. One important explanation for the increase in regional
trade agreements since the 19808 is that there will not be a final multilateral agreement at
the end of Uruguay Round negotiations. However, no consensus exists on the conditions
that motivated the resurgence of RTAs in the early 19808. Furthermore, there is no

consensus on the political conditions that accelerate the formation of PTAs.

 

17 de Melo and Panagariya 1992.

13

2. 3. The Significance of Preferential Trade Agreements: Does It Really Matter?
2.3.1. WTO and Preferential Trade Agreements

Forming a free trade area would “violate the WTO’s principle of equal treatment for all
trading partners (“most-favored-nation”). In this sense, trade agreements are regarded as an
exception to the GATT/WTO trading system, which is based on the non-discrimination
principle. The exceptional treatment is accorded with three conditions under Article XXIV
of the GATT/WTO; (1) Trade barriers on non-members shall not be more restrictive, (2)
Trade barriers shall be eliminated on “substantially all the trade”, (3) RTAs shall be
established within a reasonable amount of time. Although these conditions are explicitly
stated in the GATT/WTO rules, the imposition of these conditions on RTAs is not
straightforward. In this context, the Committee on Regional Trade Agreements (CRTA)18
was created to monitor developments concerning PTAs in the Uruguay Round. However,
there are many difficulties encountered by WTO members in assessing the consistency of
RTAs, despite the CRTA was created to ensure their consistency. For example, no rules as

to the definition of “substantially all the trade”19 have been established. (World Bank)

 

‘8 During the GATT years, the examination of P’I‘As was conducted in individual working parties. In order to
ensure consistency in their examination, the General Council established in February 1996 a single
Committee to examine all PTAs, the Committee on Regional Trade Agreements (CRTA). In addition to
examining individual agreements, another important task of the Committee is to consider the systemic
implications of the PT As for the multilateral trading system and the relationship between them. The
Committee also develops procedures to facilitate and improve the examination process and to ensure that the
reporting on the operation of the regional agreements is adequately carried out by the parties to the
agreements.

'9 “Article XXIV requires that “substantially all the trade” between the constituent members be covered by
the RTA, and the same condition is laid by GATS Article V which requires a “substantial sectoral coverage”
in services. But there is no agreement among members on the meaning of these wordings. In fact, many
agreements omit from their coverage large and sensitive areas such as agriculture and textiles. More broadly,
with the increasing question of the consistency of important RTAs exarrrined vis-a-vis WTO rules, the
relationship between regionalism and multilateralism has become a critical issue.” (WTO) This will be one of
the most important challenges that trade policymakers have to face in the near future.

14

2.3.2. Evidence of Preferential Trade Agreements

2.3.2.1.The Recent Trend of Preferential Trade Agreements

The recent proliferation of PT As has been one of major developments in international
relations. The emergence of the European Union, the North American Free Trade
Agreement (NAFTA), the common market of the South American Southern Cone
(MERCOSUR), the ASEAN Free Trade Area (AFTA), the Asia Pacific Economic
Cooperation (APEC), the South Asian Free Trade Agreement (SAFT A), led to major
changes in the trading arrangements of the world’s economy.

Over the past decade, trade liberalization and global free trade has been pursued
through both multilateral and regional (or bilateral) approaches. Historically, whenever
multilateral trade negotiations have failed, bilateral and regional free trade agreements have
replaced the void. Such trade arrangements enabled many states to move toward free trade
at their own pace and for their own benefit. The International Monetary Fund (IMF)
compiled a list of such agreements, numbering sixty-eight in 1994. It is estimated that
about 90 percent of all the contracting parties in the General Agreement on Tariffs and
Trade (GATT) are signatories to such arrangements (Lawrence 1996).20 Since January
1995, a further 125 Regional Trade Agreements have been notified, with an average of 15
notifications every year to the WTO, compared with an annual average of less than three
during past decades of the GATT. In part, the increase in notifications is a reflection of
increased WTO membership and new notification obligations (WTO). It is obvious that

the rate of growth of Regional Trade Agreements has continuously increased. It is

 

so “In mid-1998, 100 of the 132 or 76 per cent of all WTO Members were participants in one or more notified
regional trade agreements. If participation in both notified and non-notified agreements is taken into account,
then the share of participating WTO Members rises to 97 per cent.” Crawford and Laird, 2000.

15

estimated that “by 2007, a further 87 RTAs will be in force, if those being planned or under
negotiation are concluded.”21 Figure 2.1.and table 2.1. show the growth of notifications of
PTAs to the GATT. As shown, the number of PTAs notified to the GATT has expanded
enormously as an unprecedented number of regional trade agreements have been signed.

Figure 2.1. The Number of Preferential Trading Arrangements notified to the GATT
1948-94

I Number
of PTA:

 

 

1948-541955-591960-641965-691970-741975—791900—841985-891990-94

Source: Mansfield and Milner 1999

 

2' Boonekamp 2002.

11W ‘I“‘11I
' 'I,‘ 1*, I
1‘ “‘Idfl‘v :H , 111‘ 1,111
I‘I11'I1I1IQIIIJ.
r

1.11
I‘M} 11‘“ .II" ,Ilui’l‘

I . l . .III

"3:: U II": ,III‘V1‘1I"‘1IIUI,.III l11V“ 111i
1:11.11111ari111 1111 ,II'
‘II'

III! 1,1I
MI,11II"”",;‘I‘IM\I;I
111 ”I" I),-

I

. ‘IIU‘.‘IIIU

“II" 1'1I‘,
‘ I

‘I I
I

II!

3,1331“
.' II
I'»"
.HIIH‘

III
.11“,

II‘
IIIIII" "
‘IIUI “1“?"

11““, 1:,“ \II I
I1,,11,,II.

’1‘.I,'I':‘LI Him
.I‘ 111‘“ “HINT
Iii
ENE”. 11“”

II.

(It
> ,1 I IIIIIIIUII P”-M‘ “”1””..updnlth
1:1,“; IIIII
lr.“1:‘flllillI.Ilthi"iI W

I
1,1.IIIrIII'I""""-I
,1,I,.,I,,

 

I“

‘.I‘.‘.'
. ' I
t"

11111111IIII III
IIII'I‘

Itil'l’ '
“IIIHHH'I‘I”

:‘I'l’t't‘lllt‘llls‘ (RIASI Notilwtl It)

h

IIIII‘L
111,211“.

.III .,~
“III “II ,
11:ll‘:l‘1illH 1H 1

(,H‘l

I101“, I , '
1””. I“ ' .II
It" , ‘ .‘IM
1 , , I 111 “1111.

1:11“: “I“, I 11“
1:31:1111111111I‘1111

ii I
,1II‘;1,IIII“:,1IIIE
I‘,11‘L'I':'1'111Ii,

“I
"II”, N
I‘

.II
I“ II].
:1111IIU11III
II.‘II
. I
1.1I\:“11.II
11111. ‘11II'

,II,‘

11II ”MIN.“
II“I I. ,I
,,II'II

111”}

‘I

H,rl 11‘3II'11‘“ 1111i“
flu“

’IIII‘

,I ‘
II
JIM,“ ”Jill

,II “‘

,.-1
1,1111”

_I,'\t|

'I'
,1; 1:“

mull

II, I IIII WM" 1
II

II1II I .II‘ I"

U \,
tut-lulu“ ,1

”IIIL'III
I "1' ‘I,‘.'II|I'IIIIIII,.I,
.IIII .

{'I
I‘I'III'IIA‘IIII1 ,’
I11” III‘IIII‘W.
HII‘I'III'II‘HII‘M 1111

1,113 IIIJI'I'I
‘ ‘ ,1. mm .
,Hfiyntlvllww
I.III‘III-II‘:II‘»rlI 3‘4“.” "m
' v":21111'11‘t'l!|VLI\Ih1l‘Ii‘lt‘1.|
\“‘|l‘l‘.A:4:‘:A‘1‘\A“l“““‘iil“““ '
u \l' “.1
lull-“'1. IIIIIIII ‘IIIII‘ III “will." ““‘H ‘
W'I' ‘ “2:? “1,1,, II-,I III,'. II.I III, III IIIIIIIIMI.
‘IIII‘ '

IIIIIII' """'
1p .111‘1“|IIII

.llgl‘n “I,“

,III
"NH”.“hu,-1Iv..II)I1”1.11”,In,:III”1I,,A1‘~,IHI"II
IIII IIIllIII'I "HMHH‘UWUW ”1 W “WW“
I III ,

I'i'lll‘ II'I‘" II'H-III‘,‘ \III II I111IIII;'I1I,IIITI

Illhlutliiuk-nln

Source. erdTrade Organization
One way of measuring the significance of PTAs is to calculate the percentage of world

trade among countries that are participating in them. The WTO estimated that in 1999 57

17

percent of world trade was occurring within preferential trade blocs such as the EU and
NAFTA. PTAs resulted in an increase in trade among their members (Athanasios 1999,
Mansfield and Milner 2000).

Whether intra-regional trade has been affected by PT As is another important question.
Table 2.2. shows some data on trade flows with the three major regions of the world. It
indicates that regional integration has increased in recent years in Europe and Asia, while
remaining stable in North America. Similarly, table 2.3. shows more data on intra-regional
trade divided by total trade of each region. The evidence of regionalization over the period
is less clear. The number of countries in the region, as well as the relative openness of the

economies will affect the share of trade between members of the same region.

18

Table 2.2. Intra-Regional Trade as a Percentage of Total Trade

 

 

 

 

 

 

 

 

 

 

 

1965 1970 1975 1980 1985 1990 1993
East Asia 3 23.2 27.4 27.5 35.0 37.1 41.1 44.5
European 47.0 51.2 50.6 51.8 52.8 58.0 53.5
Communityb
North America ° 35.9 38.8 35.7 33.2 38.4 37.2 40.6

 

 

 

Source: International Monetary Fund, Direction of Trade Statistics.

3: East Asia includes Brunei, China, Hong Kong, Indonesia, Republic of Korea, Japan,
Malaysia, the Philippines, Singapore, Taiwan and Thailand.

b: The EC 12 pre-l995 enlargement, excluding the former German Democratic Republic.

c: Canada, Mexico and the United States of America.

Table 2.3. Intra-regional trade divided by total trade of each region

 

 

 

 

 

 

 

 

 

Region 1965 1970 1975 1980 1985 1990
East Asia 0.199 0.198 0.213 0.229 0.256 0.293
Western Hemis* 0.315 0.311 0.309 0.272 0.310 0.285
EC* 0.358 0.397 0.402 0.416 0.423 0.471
EFI‘A* 0.080 0.099 0.104 0.080 0.080 0.076
Mercosur 0.061 0.050 0.040 0.056 0.043 0.061
Andean Pact 0.008 0.012 0.020 0.023 0.034 0.026
NAFTA* 0.237 0.258 0.246 0.214 0.274 0.246

 

 

 

 

 

 

 

 

Source: Frankel, Stein, and Wei 1995

EFTA: European Free Trade Area

EC: European Community

NAFTA: North American Free Trade Agreement

19

An intra-regional trade intensity index indicates the share of one country’s exports to
another as a proportion of that country’s overall share in world markets. The advantage of
using trade intensities is that they correct for any bias which stems from differences in
growth rates or in the size or openness of members. The results are shown in Table 2.4.
Again, evidence for the growth of regionalization is less clear.

Table 2.4. Intra-Regional Trade Intensities

 

 

 

 

 

 

 

 

 

 

 

1958 1968 1973 1979 1983 1990
Western Europe a 1.38 1.51 1.54 1.57 1.72 1.60
Asia " 3.15 2.84 2.88 2.77 2.41 2.31
North America ° 2.72 2.90 3.22 3.09 2.98 3.21

 

 

Source: K. Anderson and H. Norheim,1993. “History, Geography and Regional Economic
Integration”, in K. Anderson & R. Blackhurst, eds., Regionalism and the Global Trading
System, Harvester Wheatsheaf, London.

a Includes Turkey and Yugoslavia.

b Includes Australia and New Zealand.

c Canada, Mexico and the United States of America.

Economists compare the growth in trade between member countries and with other regions
before and after the conclusion of PTAs. Soloaga and Winters22 argue that the new wave of
PT As did not boost intra-bloc trade significantly. Vamvakidis also notes that countries in

PTAs still keep the high level of trade barriers.23 There also have been concerns about the

negative effects of PTAs on the trade of signatory states. For instance, there is evidence of

 

22 They identify the separate effects of PTAs on intra-bloc trade. members’ total imports, and members’ total
exports by testing the significance of changes before and after the trade blocs’ formation.

23 Vamvakidis 1999.

20

trade diversion in MERCOSUR. Yet, “while protection of certain sectors such as
automobiles limits market opportunities in MERCOSUR, overall these countries are now
much more open than they were in the 1980s. Imports from third countries have also been
growing rapidly. Scholars claim that the context of the new RTAs and their product
coverage are different from the unsuccessful trade agreements of the 1950s and 1960s,
which were mostly designed to achieve trade diversion. Based on a simple comparative
analysis, nonmember countries may be negatively affected by trade diversion and a
reduction in their trade” (Crawford Jo-Ann and Sam Laird 2000). However, it is difficult to
arrive at clear conclusions based on a crude dynamic analysis, especially in the case of the
faster growing RTAs. It is still puzzling why so many countries rush to sign to P’I‘As. In the

next section, I elaborate the implications of P'I‘As.

2. 4. The Impact of Preferential Trade Agreements

As PT As have proliferated, it becomes more important to examine their impacts. Although
much of the literature emphasizes the welfare implication of Preferential Trade
Agreements, the effect of PTAs is a vigorous question. It is less clear that there is any
systematic economic impact of PTAs, whether static or dynamic.

To begin with the static impact, it is often argued that PTAs have a “two-sided quality,
liberalizing trade among members while discriminating against non-members” (Mansfield
and Milner 1999). PT As cause trade creation and trade diversion effects. The trade creation
effect is observed when trade among PTA members expands because of the elimination of
trade barriers. Since the creation of trade leads to better resource allocation, the trade

creation effect is beneficial to PTA members. On the other hand, we can observe trade

21

diversion effect when this trade replaces interaction with non-members. If the loss from
tariff elimination is greater than the benefit of lower priced imports, the impact of trade is
negative. The trade diversion effect has an undesirable impact on non-members, as PTAs
reduce their export opportunities. In addition, the negative impact may come from possible
retaliation from non-members with a disadvantageous situation.24 For instance, Bhagwati25
argues that regional trade agreements are “mostly welfare reducing since the partner
countries will end up with severe welfare losses due to trade diversion. Further expansion
and creation of free trade areas, instead of concentration on multilateralism at the WTO, is
a mistake.” Krueger (1999) also argues that, “regional liberalizations lead to the formation
of new interest groups who oppose multilateral free trade.” Even if a PTA may be trade
diverting, there may be welfare benefits gained by increasing productive efficiency.
Member countries can realize economies of scale in production because of the increased
market power through PT As. Also, domestically, companies may reduce the cost of
production by expanding the scale of production through expanded trade and greater
competition. This factor was one of reasons for the expansion of PTAs established by
developing countries during the 1960s (Mansfield and Milner 1999).

The impact of PT As has varied over time. “During the nineteenth century, it was more
trade creating. But between World Wars I and H, it tended to be more trade diverting.
Although scholars have been debating this issue over the past years, case specific

characteristics of empirical study and lack of reliable data prevented assessing the

comprehensive consensus about this issue” (Mansfield and Milner 1999). Overall, with

 

2.4 Bhagwati (1992) points out the negative effects of the growth of regionalism on global free trade. They
argue that RTAs divert attention from the multilateral trading system.

25 Bhagwati 1996.

22

regards to the welfare impacts of PT As, the net static impact of PT As is ambiguous,
because they have both positive and negative impacts.26 We cannot generalize the static
welfare effects of PT As.

It is even more difficult to evaluate the longer-term impact of P’I‘As. The more
“dynamic time-path question” (Bhagwati 1993) related to the welfare impact of P’I‘As is the
crucial question in determining whether or not these trade agreements promote global free
trade. Some argue that regional trade agreements may be positive in achieving multilateral
trade liberalization (Kemp and Wan 1976, Krugman 1993, Summers 1991, Perroni and
Whalley 1996, Lawrence 1996). On the other hand, some argue that PTAs inhibit global
free trade (Krugman 1991, Bhagwati 1993, Frankel, Stein, and Wei 1995). Opponents of
P'I‘As describe them as competing trade interests that hinder multilateral agreement.
Especially as PTAs transcend regional boundaries, political and economic tensions will
lead to increased hostility through anti-dumping charges and dispute settlement. In doing so,
PTAs complicate WTO efforts to unify dispute settlement procedures by inhibiting global
free trade. Others fear that PT As yield incomplete liberalization.” They worry that some
countries who benefit from regional trade agreements will be reluctant to open their
markets completely if they expect relatively insignificant benefits. In contrast, Kemp and
Wan (1976) show that trade blocs can be constructed in a way such that nonmembers’
welfare is not degraded. Some argue that PTAs allow countries to gradually liberalize by

slowly increasing the level of competition and allowing domestic industries time to adjust.

 

2“ Urata 2001.

27 Lawrence argues that the appropriate comparison is not between a preferential arrangement and complete
multilateral liberalization, but between second-best situations of multilateral liberalization that are only partial
with preferential trade liberalization.

23

Most member countries think that if they conclude PT As, these agreements can be a
complementary way to trade liberalization. In this context, some PT As are implemented for
“the purpose of economic reforms, which stimulate multilateral openness” (Mansfield and
Milner 1999). In addition, Summers (1991) argues that, “holding the degree of multilateral
progress constant, the world will be better off with more regional liberalization.”
Nordstrom (1995) finds that regional trade agreements can provide countries with stronger
motivation to pursue multilateral tariff reduction by allowing small countries to deal with
trade issues with large trading blocs more effectively.28 Moreover, Perroni and Whalley
(1996) claim that in recent regional trade agreements, member countries can choose their
own external tariff level. This new character of regionalism does not increase the monopoly
power of emerging trading blocs. In other words, PTAs do not necessarily imply higher
external trade barriers between the emerging trading blocs. They conclude that the
expansion of trade agreements is not a stumbling block to the multilateral trading system.29
However, despite some positive effects of trade agreements, the debate over the dynamic
time-path question on global free trade has not been concluded. Winters30 argues that,
based on various models, it is difficult to determine whether PTAs are building blocks31 or

stumbling blocks to globally freer trade.

 

28 Campa and Sorensen (1996) also suggest that if small countries with small economies form a trading bloc,
global free trade can be sustainable since the integration of small countries can undermine the market power
of the larger trading blocs.

29 NAFTA pressured countries to conclude more free trade agreements as a domino effect. This pressure
strengthens the power of exporters and pro free trade groups. Crawford and Laird, 2000.

3° Winters 1996.

3‘ Yet, there may be “little doubt that the main economic advantages to participants in regional trade
agreements would be even greater if the liberalization were carried out on a wider, multilateral scale. RTAs
may be a second-best solution.” Crawford Jo-Ann and Sam Laird, 2000. “Regional Trade Agreements and the
WTO”, Paper presented for a meeting of the North American Economic and Finance Association in Boston,
6-9 January. Kemp and Wan (1976) claim “there is a big incentive to form and enlarge a customs union until

24

2.5. Why do Countries Join Preferential Trade Agreements instead of Choosing
Unilateral or Multilateral Approaches?

While economists agree that free trade leads to positive welfare gains, they propose
different approaches to making a change from a protectionist policy to a free trade policy.
The three approaches to implement free trade reform are unilateral, multilateral, and
bilateral.32 Recently, unilateral trade liberalization has been an important way of
liberalizing trade. Britain gained substantial benefits from unilaterally lowering its tariff
barriers in the nineteenth century. Furthermore, its success caused other countries to lower
their barriers. In recent decades, Chile and South Korea implemented unilateral tariff
reductions. The argument for unilateral trade liberalization is that a country benefits from
increased trade by reducing its trade barriers. Countries that lower trade barriers
independently do not have to delay reform to persuade other nations in lowering their trade
barriers. Scholars argue that gains from unilateral trade liberalization are substantial.

Since the 1980s and 1990S, many countries have signed on to PTAs. Some countries
liberalize unilaterally and join PTAs at the same time. This happened mostly in EU
countries. However, recent political economy history shows that more countries “enter into
Regional Trade Agreements instead of implementing unilateral liberalization” (Schiff and
Winters 1998). If the positive welfare gain of unilateral free trade is so substantial, why do

so many countries pursue Regional Trading Agreements instead?” Addressing the

 

the world is one big customs union, that is, until free trade prevails.” (Cited from Crawford Jo-Ann and Sam
Laird)

32 Multilateral and bilateral approaches (or regional approach) include diminishing trade barriers by
cooperating with other countries.

33 Leaders enter into a trade agreement not solely for economic reasons. Leaders may care more about the

25

motivations of countries to conclude PTAs is important for resolving this question. There

are various relevant external and internal factors that are political as well as economic.

2.5.1. External and Economic Factors

Because of the difficulty in achieving trade liberalization on a global level, many countries
decide to turn to bilateral or regional free trade agreements. Many countries realize the
benefits of free trade for the promotion of economic growth.34 Despite the desire of many
policy makers for the promotion of trade liberalization, under the GATT/WTO it has
become difficult to pursue on a global level. As the number of GATT/WTO members
increased, their views on the pace and the extent of trade liberalization have diverged. The
necessity for consensus among GATT’s participants limits the range of trade reform.
Indeed, it took the GATT members eight years, twice as long as originally planned, to
conclude the Uruguay Round. This indicates the difficulty in reaching a global consensus
on trade liberalization. The difficulty in reaching a consensus led to a failure by the WTO
members to start new trade negotiations in Seattle in 1999. Given these barriers, countries
often pursue PTAs as an alternative means of trade liberalization.

In addition, PTAs can be valuable in dealing with THE intractable trade problems that
arise in areas such as agricultural subsidies and trade in services. While GATT relatively
successfully reduced tariffs on industrial goods, it was less successful in liberalizing trade

in agriculture, services, and other areas of international trade. Moreover, the relative

 

political incentives. In this sense, “unilateral trade barrier reductions are less credible to voters and more
easily repealed than are mutually agreed international trade agreements.” Leaders always consider the utility
of their actions. In this context, leaders assess the utility of trade agreement formation. They may use the
signaling function of trade agreement for the promotion of their political survival and reelection.” For more
detailed discussion on this point see Mansfield, Milner, Rosendorff 2002.

34 Many studies demonstrate, “income grows more rapidly in countries open to international trade than in
those more closed to trade” (World Bank 2003).

26

recession of the world’s economies in the seventies and eighties increased protectionist
pressures among countries. These pressures caused a proliferation of new trade barriers
such as voluntary limits on exports of steel and cars to the United States. Recent
negotiations including the Uruguay Round aimed to extend GATT rules to new areas of
trade. These negotiations, however, faced several problems and the possibility of their
success is far from certain. Consequently, many countries “turn their attention from GATT
to bilateral or regional trade agreements” (Irwin 2003).

External factors also include the economic gains for domestic companies from the
reduction of trade barriers. The production from the growth of exports enables companies
to take the benefits of economies of scale. By broadening markets, mutual liberalization of
trade increases competition among countries by enhancing efficiency and consumer
incomes. As a result of the expansion of P’TAs, the companies in non-member countries are
discriminated against in many markets. To overcome this problem and compete against
companies from other countries under the same conditions, companies want to form PTAs.
For instance, it was very important for companies in Canada and Mexico to gain access to
the US market as part of NAFTA. Companies in Eastern and Central Europe are in a
similar position today with regards to the EU.35 The importance of securing markets by
entering into PT As is increasingly important because of the recent expansion in free trade
agreements. Countries and their companies do not want to be excluded from these access

opportunities.36 Therefore, strategic incentives may emerge for outside countries to form a

 

’5 Urata 2002.

36 In this way, preferential trade arrangements create incentives for nonmembers to join. Even traditionally
protectionist countries may be tempted. Winters (1997) argues that, “RTAs are like street gangs. You may not
like them, but if they are in your neighborhood, it is safer to be in one.”

27

free trade agreement to make up for the loss of welfare37 due to the formation of the

original trade bloc.

2.5.2. Internal Factors

To make a policy change to join a PTA, countries make a political commitment. When
countries make a political commitment to a bilateral or multilateral trade agreement to
lower trade barriers, it may reduce political opposition to free trade. Consequently,
preferential trade agreements can be a potential alternative for liberalizing trade (Irwin
2003). Domestically, the larger the veto group38 is, the more difficult it is to reach any
agreement because the cost of negotiation is high. Diverse views of veto players will be a
great barrier in reaching any agreement. Trade reform is difficult to achieve through
domestic measures. Therefore, some governments use the external pressure of a trade
agreement to push for economic reform. By making domestic policies a part of
international agreements, they can increase the credibility of reform. In this context,
scholars argue that governments join free trade agreements in order to make their
commitments credible.39 For example, Fratianni and von Hagen (1992) claim that by

joining the European Monetary System (EMS), political leaders made a costly signal in

 

37 PTAs may “benefit the member countries but worsens the welfare of the rest of the world even if member
countries do not raise the external tariff rates” (Goto and Hamada 1998).

38 A veto player on a given issue dimension may be the person, political party, or faction of a political party,
that exercises a veto. There are two types of veto players. Partisan veto players are individuals or parties who
wield vetoes, while institutional veto players are those entities with veto power, such as presidents and
legislatures. Tsebelis argues that if there are more ideologically opposed or polarized veto players on the
same policy dimension, the political system may have a greater difficulty in order to agree on policy change. I
elaborate the impact of veto player over trade policy in detail in chapter 4. See Tsebelis 1999.

39 A country’s policy choice can reveal information about a country’s preference through repeated interaction.
The decision of whether or not to cooperate with a trade agreement can reveal a country’s ability to commit to
trade liberalization. While one means of signaling is to reveal one’s intention through multilateral trade
agreement, another way is to sign and join a trade liberalization agreement through a free trade agreement.

28

terms of the credibility of their commitments to low-inflation policies. Likewise, Haggard
(1995) argues that developing countries’ governments may join international trade
agreements to demonstrate their commitment to policy reforms and promoting

investment. Commitment arguments emphasize governments’ efforts to demonstrate their

credibility to domestic actors and international regimes.”0

2.5.3. Political Motivations
Political goals can be the main motivation for joining a PTA. First, some trade agreements
reduce the probability that “mi grants’ bloodshed will spread across international borders.
Second, trade agreements respond to outside threats by binding relations between members.
Third, hostile nations can potentially reduce tensions” (Shiff and Winters 1998).41 The
typical example of this is European Union.42

Some countries also try to increase their bargaining powers by entering into a PTA.
One of the reasons behind regional integration in Europe after World War II was for the
European countries to increase their influence against the US and to strengthen their
political power against the Soviet Union and Eastern Europe. The increased size of the

unified European market had substantial impacts on other regions in the world. Part of the

 

‘0 However, this approach is inconclusive in answering some questions. First, it does not explain why
governments decide to reveal their commitment through free trade agreements rather than through some other
means. Second, this approach is silent about the policy preferences to which governments choose to

commit. It must be “incorporated with more specific theories of economic preference formation and the
political process” (Kono et al. 2000).

4' However, political cooperation is possible without the linkage with trade. Free trade does not necessarily
guarantee peace.

‘2 Political and security considerations were important in the decision to form a number of RTAs, especially
in Europe. These RTAs have been established even if strict economic criteria were not met (Crawford and
Laird 2000).

29

reason for the United States was eager to succeed with NAFTA was because of the need to
make a balancing power in North America against the EU.

Developing countries may benefit from joining a PTA if accession requires political
or social reforms. Often it is not easy to adopt reforms without external conditionality.
Because of the expected benefits, interest groups opposing political or social reform may
permit it. For example, some cases of external conditionality include the adoption of
democracy, as in the EU and Mercosur. In reality, this rule had an impact in the case of
Greece, Portugal, Spain and Paraguay.

PTA is one of the most debatable issues43 in international trading systems. In this
chapter, I highlighted their significance by noting the expansion of PT As in international
political economy. I clarified why we need to evaluate the expansion of PT As and their role
in the widespread liberalization of trade policies. I addressed the difficulty of the analysis
of P'TAs. Next, I provided an overview of the relevant history and the growth of PTAs and
the impact they have had. Finally, I presented the main reasons why countries join PT As
instead of focusing on unilateral or multilateral approaches. In doing so, I addressed
important puzzles in the international political economy literature about the rush to join

PTAs.

 

‘3 The new wave of PT As has posed major challenges in international economy such as “do PTAs stimulate
growth and investments, induce political stability and cooperation in the regions or do they divert trade in
negative directions? Most importantly, do they contribute to the multilateral free trade?” The answer depends
on different approaches and particular circumstances of each PTA (Bhagabati 2001).

30

Chapter 3. Literature Review

The purpose of this chapter is to present the major theoretical approaches relevant to trade
policy. I explore existing theories and examine weaknesses of the arguments. In the first
section, I present domestic preference formation models. In the next section, I present the
approach which analyzes the role of domestic political institutions. In the final section, I
address the structural theories.

Domestic preference formation models include factoral and sectoral models. However,
this approach does not explain how social actors’ preferences are translated into political
outcomes. The second approach addresses the role of domestic political institutions and
how they influence preference formation and aggregation. One of the important puzzles in
international political economy is the link between domestic institutions and the policy
decisions to join Preferential Trade Arrangements. Few systematic studies answer “how
domestic institutional pressures of countries induce trade policy change” (Milner 1999).
The third approach includes structural theories. Systemic theories include neo-realism,
neo-liberal institutionalism, and hegemonic stability theory. Structural perspectives focus
on international conditions that motivate trade policy change. However, it lacks the ability
to account for “preference formation explaining who wants free trade and why” (Kono et al.

2000).

3.1. Theories of Domestic Societal Preferences
There are clear interactions between internal and external politics (Putnam 1988). In many
analyses, domestic politics alone can be what determines international trade policy. This

literature focusing on domestic politics emphasizes the distributional consequences of trade

31

policies for domestic groups. That is, whether a state decides to liberalize trade depends on
how much influence different interest groups have in terms of distributional consequences
of trade policies (Grossman and Helpman 1995).

In this approach, the influence of pressure groups is assumed to be the important
factor in determining trade policy. In general, scholars posit that the losers will oppose the
change of trade policy, while winners will support the change of policy (Rogowski 1989).
Trade policies often reflect “the preferences of the more powerful and best organized
interest groups” (Milner 1997, Rogowski 1989). More specifically, scholars provide several
perspectives on how trade liberalization will affect the interests and the preferences of
various groups. The Heckscher-Ohlin and Stolper—Samuelson theorems demonstrate that
interests and preferences of various groups reflect factoral preference (Stolper and
Samuelson 1941). Free trade will benefit the owners of abundant factors and harm those
holding scarce factors. On the other hand, trade protection will hurt abundant-factor
owners and benefit scarce- factor owners. Factoral models provide an explanation that
“trade-policy preferences of different groups will be determined by their countries’ relative
factor endowments” (Milner 1997).

Another main approach to explaining trade policy preferences stresses that interests
and preferences of various groups reflect sectoral interests. Proponents of sectoral models
(Frieden 1991, Gilligan 1997, Grossman and Helpman 1994, Magee, Brock and Young
1989, Milner 1988) point out weaknesses such as factoral models’ assumption of factor
mobility between different sectors. However, if factors do not have mobility, then their
interests are directly connected to the sector. In this model, scholars argue that export

industries go for free trade and import-competing industries lobby for protection. For

32

instance, the government is willing to sign a PTA when it makes reasonable gains for
export groups to the partner country given that there are the costs of import competing
industries and voters (Grossman and Helpman 1998). In addition, when leaders exclude
some sensitive industries, which can impose influential “political costs”, the possibility of
signing a PTA rises (Helpman 1998).

In general, the empirical evidence on the sectoral-factoral debate is not conclusive.44
Magee, Brock and Young (1989) find that US. firms tend to lobby along sectoral rather
than factoral lines. However, there are also empirical findings which oppose this theory
(Rogowski 1987). The level of factor mobility depends on technological and institutional
factors. Moreover, other factors such as domestic political institutions may impose other
kinds of constraints. Furthermore, there is little evidence as to why particular domestic
groups prefer free trade agreements to multilateral trade agreements. In addition, evidence
is lacking regarding whose interests groups serve. It is not clear that “export groups prefer
regional trade agreements to multilateral trade agreements” (Mansfield and Milner 1999).
This approach remains incomplete if it is silent “how an exogenous force shifted political
influence away from protectionists and in favor of those preferring free trade” (Milner
1999). The influence of societal groups cannot be the entire story, even domestically.
Politicians have their own preferences and interests, and their desire to keep office is a
crucial factor in trade policy-making. Moreover, competition among interest groups to

influence governments’ policies is usually not fixed or structured. Domestic political

 

‘4 There are also arguments related to the logic of collective action (Olson 1965). Political action for the
pursuit of free trade or protectionism requires that political actors overcome collective-action problems. This
is easier for small groups. Since sectors are smaller than factors, collective action should be easier for sectors
than for factors. Thus, trade policies should reflect sectoral rather than factoral preferences. However, the
collective-action problems related to factors and sectors are not “fixed”. It may be “exogenous” (Kono et al.
2003).

33

institutions may influence the patterns of interactions between domestic groups. Also, they
may structure which groups’ preference will be translated to governments’ policies (Garrett
and Lange 1995, Putnam 1988). In this context, the preferences of political leadership in
trade policy choice become a key factor that shapes economic policy. Thus, it is necessary
to explore such aspects as the ideological orientations of the top political leaders and those
of their political support bases. However, there are few systematic theories on such
preferences of policy makers. For instance, some scholars claim that leaders conclude free
trade agreements to stimulate economic and political reform. However, the effect of
commitment of free trade agreements for the pursuit of political reform and liberal
economic reform is not yet clear. “There are few theories which specify the conditions
under which policy makers will adopt liberal trade policy preference” (Milner 1999).
Although the preference approach nicely addresses the importance of domestic
groups’ preferences, the explanations hardly examine the dynamics of domestic political
institutions and changes in international factors. Theories that combine domestic political
institutions and domestic groups’ preference aggregation mechanism are necessary to
determine which groups reflect specific preference and whose interests will be predominant.
Consequently, explanations should combine the consideration of domestic preferences,

domestic institutional factors, and international (or structural) factors.

3.2. Theories of Political Institutions
Domestic pressure groups’ access to trade policy choice and their interest formation are
influenced by the constraints of domestic political institutions. Therefore, it is important to

incorporate domestic institutional factors into the study of trade policy decisions. Domestic

34

political institutions are significant in explaining trade policy in that “they shape the
patterns of interactions between domestic groups and whose interests will be represented in
governments’ policies” (Putnam 1988).

I begin with the link between international trade and democracy. In terms of regime
type, there are debates over the monadic impact of regime type on the propensity of trade
policy. Scholars have found a positive link between international trade and democracy
(Gaubatz 1996, Verdier 1998, Remmer 1998). Gaubatz argues that democracies are more
likely to cooperate for freer trade policy due to the system’s transparency. Transparency
tends to decrease the suspicions of defection. Barbieri and Schneider, however, draw a
different conclusion about the link between democracy and trade (Barbieri and Schneider
1999). Penubarti and Ward (2000) maintain that previous studies rely on a misspecification
of the trade model and fail to account for the spatial context of international trade flows.
Therefore, debates over the monadic impact of regime type on the propensity of free trade
policy are still incomplete (Reinhardt 2000).

Some political scientists discuss the dyadic effect of democracy on trade cooperation,
arguing that “liberal democracies are more likely to commit to rules that reflect their
institutional biases” (Simmons 2000). For instance, Mansfield, Milner and Rosendorff
(2001) show that “democratic dyads are more likely to be proponents of cooperation to
lower trade barriers than are autocratic ones due to their need for domestic ratification of
agreements” (Mansfield, Milner and Rosendorff 1999, 2000, Bliss and Russett 1998,
Morrow, Siverson and Tabares 1998). Dixon and Moon suggest that any “two nations with
similar regime types are more likely to trade more freely with each other, because

similarity decreases political conflict and increases each other’s business exchange”

35

(Reinhardt 2000). On the other hand, Verdier (1998) argues that democracies tend to
choose protection due to the political clash caused by trade. Remmer and Mansfield and
Bronson (1997) also argue that democratic dyads are not more likely to cooperate in terms
of treaty counts and trade flows.

One of the important aspects of democracy is “its relative sensitivity to mobilized
political pressure” (Reindhart 2000). In terms of trade policy-making, top decision-makers
should care about different preferences of their domestic constituents because voters
influence the survival of decision-makers. Differences in producer groups’ and consumers’
trade policy preferences may explain divisions. However, it is difficult to make some
demands at the aggregate level for consumers due to collective action problems. Moreover,
general voters and consumers do not consider trade policy as their critical determinants of
the choice of party or candidate although we cannot draw a general conclusion about this
issue. On the other hand, producers are more effective in aggregating their own interests
and they are more active as an interest group.45 Producer groups will have access to
decision-makers through lobbying and campaign finance whereas general voters as
consumers are relatively ineffective in demanding specific trade policies.“6 Verdier argues
that democracies are less sensitive to pressure groups because a winning coalition must be
larger than the selectorate size in democratic countries. In general, democracies “vary

widely in the extent to which they are open to interest group pressure” and autocratic

 

‘5 This group has a strong incentive to press for protection, and is in a better position to overcome the free-
rider problems are very important in determining who will be more effective in exerting their political
influence. As mentioned, political action for the pursuit of free trade or protectionism requires that political
actors overcome collective-action problems (Olson 1965). This is easier for small groups. Since producer
groups are smaller than consumers, collective action should be easier for producers than for general
consumers.

‘6 Grossman and Helpman show that the government decides to join trade-liberalizing agreements when it

leads to welfare gains for the majority of voter and “adversely affected interest groups fail to make
coordination in their efforts to block the accord” (Helpman 1998).

36

countries are more sensitive to small groups of their elites (Reindhart 2000). In this context,
incorporating the more specific features of domestic institution types such as the electoral
system, the number of veto players and so forth may be necessary in order to explore
variations of international trade policy. I explore this issue in the next chapter in detail
because more attention is necessary to stress the relationships between different domestic
institutions and trade policy. Little theoretical or empirical research has been conducted to
“sort out the complex relationship between the different components of domestic
institutions and trade policy” (Milner 2000).

Change in political regimes may be “the institutional change that helps explain recent
global free trade” (Milner 2001). However, strong evidence has not yet been presented
about the effect of democracy on trade cooperation. Simmons suggests that the presence of
a democratic regime has no independent effect on the propensity to commit to openness
(Simmons 2000). Moreover, Dai suggests, “domestic political institutions alone are not
sufficient to predict a higher level of cooperation among democracies regardless of the
preferences of the decision maker” (Dai 2002). Consequently, theories that incorporate
both preferences and institutions are necessary, because “both are jointly determined”

(Milner 1999).

3.3. International Factors
International factors may play a major role in the change of trade policy. The approaches
that consider them include neo-realism, hegemonic stability theory, neo-functional

explanations (Haas 1958) and theories of complex interdependence (Keohane and Nye

1977).

37

3.3.1. Neo-realism and Hegemonic Stability Theory

Neo-realism stresses that the distribution of power influences international foreign policy
outcomes. Although relative gains from trade tend to hinder international cooperation
(Waltz 1979, Grieco 1988), neo-realists do not exclude the possibility of trade agreements.
For example, some scholars (Gowa 1994, Gowa and Mansfield 1993) argue that trade
cooperation is more likely among states that are close in terms of political and military
relationship. Hegemonic stability theory emphasizes that powerful states committed to
promoting global trade liberalization are the key to successful economic cooperation
(Gilpin 2001). That is, the presence of a hegemonic state is a necessary condition to sustain
liberal international commerce. Therefore, the erosion of hegemony tends to cause

protectionism.

3.3.2. Neo-liberalism

States join trade agreements because there are “increasing technological, economic, and
other complexities that countries can no longer effectively solve at the nation-state level”
(Haas 1964). In addition, governments enter cooperative arrangements in order to improve
economic welfare. If leaders think that economic cooperation fits with their interest, they
pursue economic cooperation with others, which means one issue-area spills over into other
issue-areas. However, “functional needs do not determine change that countries choose to
pursue all the time” (Pentland 1973). For instance, East Asia is the region where
international interactions are high, but also the region where we find few preferential trade

agreements between countries.

38

Neo—liberalism argues that international institutions promote cooperation by
minimizing collective action problems. By enlarging the shadow of the future, international
institutions induce international commitments and make them more credible (Axelrod and
Keohane 1986, Martin 1992, Simmons 2000). Keohane (1986), for example, claims that
sustaining international regimes is to reduce market failures and transaction costs. More
specifically, for the advanced countries, the GATT/WTO allowed countries to foster broad
liberalization, as did the EU. Mansfield and Reinhardt (2002) argue that, “developments

within the GATT/WTO have influenced when member-states have sought the insurance

I

and bargaining power offered by PTAs.” For the developing countries, the IMF and World
Bank promoted trade liberalization although economic crises motivated countries to
implement economic reforms (Milner 1999, Haggard 1997).47 For the purpose of
explaining commercial cooperation, however, this approach has some weaknesses. This
view can provide a partial explanation about the recent trend of trade liberalization.
However, this approach is not appropriate for understanding non-European commercial
cooperation. Moreover, the different variation and extent of institutionalization itself needs
explanation. In general, this approach has not developed any general theories of preference
formation that explain how changes in the level of transactions affect the preferences of
political actors. Also, it is silent as to why and when political leaders are more or less

responsive to interstate commercial c00peration initiatives.

 

‘7 According to Haggard, the difference between the developments of regionalism of East Asian and Latin
American countries can be explained by the difference in the convergence or divergence of preferences of
countries about the direction and extent of economic integration. In Latin America, severe economic crises of
the 1980s promoted preferences in various countries for the expansion of economic cooperation, whereas the
lack of a similar crisis in Asia (before 1998) has made preferences of Asian nations relatively different
(Haggard 1997).

39

3.3.3 Power Transition Theory

Some scholars (Efird and Genna 2002) show that the level of trade agreements improves
when the power of a subordinate state exceeds the power of the preponderant state if they
are jointly satisfied with their relationship. Yet, this perspective is silent about the sources
of states’ preferences regarding the expansion of trade arrangements, reflecting its tendency
to underestimate the issue of preferences. It cannot tell us why governments want to
conclude free trade agreements in the first place. It only features systemic factors, and the
sources of variation are “underspecified” (Kono et al. 2000). Domestic preferences and
institutions are expected to contribute to economic policy formulation. We should identify
which preferences matter and how they interact with domestic institutions in determining
the propensity for interstate commercial cooperation. In sum, changing preferences among
political leaders and interest groups, institutional changes, and the increased influence of
international institutions must all be accounted for to explain the recent global expansion of
trade liberalization. Research on this puzzle is not complete.

In the next chapter, I show how interplay between the preferences of decision-makers
and nature of domestic institutions help us to resolve the questions associated with the
expansion of international trade agreements. I demonstrate that the interplay between the
preferences of decision-makers and nature of domestic institutions will provide us with key
insights into the puzzle being addressed in this dissertation. Recent researches leaves
various questions unresolved, especially regarding how and why domestic political
institution pressures affect the recent expansion of trade agreements (Mansfield and Milner

1999). In this dissertation, I attempt to establish a solid theoretical background and

40

rigorous empirical research to explain how political leaders form their trade preferences and

how domestic institutions condition these preferences.

Chapter 4. Variations of Domestic Political Institutions and the Determination to

Enter Preferential Trade Agreements

While many theories address the causes of international trade agreement formations, they
provide little convincing answers as to why the formation of PTAs varies over time across
countries. In large part, this is because they do not consider institutional factors. There are
only “little studies, which systematically explore the domestic institutional structures of
government and how these domestic political institutions impact the preferences of
decision-makers over trade policy” (Ehrlich 2003). Therefore, it is hard to sort out the
dynamic relationships between domestic political institutions and international commercial
policy. In this chapter, I examine how institutional theories explain recent trade
liberalization, especially focusing on the underexplored linkage between political
institutions and the liberalization of trade policies. I highlight how political institutions
condition the motivations of policymakers and their decisions (Boix and Adsera 2003).

In the previous chapter, I examined why regime type is one of the most important
variables in explaining how political institutional structures influence changes in
international trade policy. Furthermore, even within regime types, there are many
variations in terms of international commercial policy. When the regime type variable is
disaggregated into sub-level regime type variables such as veto players, we can more
clearly understand the political influence of domestic institutions on disparate commercial

policies across countries. This chapter offers a detailed theoretical framework by

41

incorporating institutional structures into models of trade liberalization by linking
government preferences and domestic political institutions. In the first section, I introduce
Bueno de Mesquita’s theory, which explains how regime types interact with the
motivations of political decision-makers in terms of supplying trade policy. Second, I
provide an overview of different explanations how the preferences of actors are translated

and aggregated into trade policy.

4.1. The Linkage between Variations of Domestic Political Institutions and

International Trade Policy Outcomes

Only recently have researchers started to pay attention to how domestic institutions affect
trade policy. Government decision-makers respond to the incentives created by political
and legal institutions. The incentives that institutions channel for decision-makers will
influence how the government adopts trade policy. In other words, institutions aggregate
domestic preferences. Thus, different institutions may result in divergent trade policy
outcomes. In this context, understanding the linkage between institutions and trade policy
outcomes is significant for grasping the dynamic interplay between the supply side and the

demand side of trade policy (Milner 1999, Nelson 1988).

4.1.1. Bueno de Mesquita’s Institutional Theory and International Trade Policy

In this dissertation, I claim that different combinations of institutions affect the incentives
of politicians to adopt a particular trade policy. Institutional difference in international trade
policy outcome rests on assumptions that institutions affect how the divergent preferences

of policymakers translate into choosing a particular trade policy outcome. In other words,

42

“institutions constrain the choice set by defining what preferences are expressed and then
aggregated to determine a policy” (Boix and Adsera 2003). Although preferences and
institutions are analytically distinct variables, which can both influence the process and
outcome of the change of trade policy, it is necessary to combine both of these distinct
categories to make a complete model of trade policy formation and explain its more
dynamic aspects. In this section, I attempt to make a dynamic model of trade policy
formation by understanding the ways preferences are aggregated into particular policy
decisions. It highlights how incentives and motivations of leaders toward international
trade policy vary across institutional settings.

Political leadership in trade policy-making is an important factor that structures the
trade policy preferences of countries (Milner 1999). We need well-developed theories that
show why and how leaders come to prefer to liberalize trade. Therefore, in this section, I
focus on why political leaders commit to a particular trade policy by entering free trade
agreements. In this context, I highlight differences in the incentives of autocratic leaders
and democratic ones.

Mansfield, Milner and Rosendorff (1999, 2000) state, “democratic audiences exert a
positive effect on free trade agreements.” By showing a repeated Prisoner’s Dilemma
game, they show that “two democracies are more likely to enter into trade agreements than
mixed countries since leaders in democracies derive more gains from trade liberalizing
agreements” (Mansfield, Milner and Rosendorff 1998). Other scholars also argue that the
domestic regime type is important to understanding international law compliance
(Simmons 2000). Related to the compliance issue, Mansfield, Milner and Rosendorff argue

that cheating on the rules of trade agreements are more likely to lead to the removal of

43

” 48 in democracies. If so, we should see a higher rate

leaders by generating “audience costs
of compliance with trade agreements in democracies (Mansfield, Milner and Rosendorff
2000).49 However, the major weakness of their study is that it does not help us understand
what positive political gains motivate democratic leaders to cooperate in trade. Also, it is
not clear why democratic political leaders manifest a preference for entering free trade
agreements in the first place.

Deciding whether or not to join trade agreements is a political decision. “The shape of
free trade agreements will be a rational response to political leaders’ domestic political
situation” (Mansfield and Milner 1997). In general, leader’s decisions related to trade
policy have domestic distributional and electoral implications. In this section, I examine the
dynamic link between institutions and policy choices in the context of trade policy.

A leaders’ political survival depends on successful performance, which is defined by
his ability to please the segments of his electoral coalitions. Leaders do not need to satisfy
everyone to hold power; in fact, attempting to do is inefficient in a strategic sense. Leaders
are more likely to do what is necessary to satisfy their major constituents’ desire in order to
retain office. Building from this principle, Bueno de Mesquita (2000) provides an
interesting perspective for understanding a country’s external trade policy. As the winning

coalition grows, he argues, “the prospects of political survival increasingly hinge on

successful policy performance” (Bueno de Mesquita 2000). Depending on the size of the

 

’8 Scholars argue that domestic “audience costs” exist in democracies. Audience costs in democracies make
the statements of democratic leaders more credible. Domestic audience costs involve punishments imposed
by voters when leadersmake not credible threats. If leaders fail to act, they are more likely to be punished. As
a result, the governments of democratic leaders are less likely to bluff and more likely to be very careful in
their selection of foreign policy tools. More detailed explication for audience cost, See Fearon 1994 and
Schultz 2001.

49 However, the proposition has not been put to a rigorous test in their study. Furthermore, the argument
seems questionable and their empirical evidence needs more cautious interpretation.

44

winning coalition and selectorate, some political systems give leaders the incentive to rob
their own people. Such political arrangements make leaders indifferent to whether they
produce prosperity for their countries. Other domestic political arrangements encourage
leaders to be attentive to peace and prosperity. In general, the strategic allocation of
resources depends on the size of the winning coalition. Leaders have an incentive to limit
the distribution of goods to members of a winning coalition. For a given budget, the
quantity of goods available to individuals shrinks as the size of the winning coalition
grows.

A winning coalition is defined as “those members of the selectorate whose support is
essential to keep the incumbent leadership in office” (Bueno de Mesquita 1999). Winning
coalition members share any of the private benefits that the leadership distributes. The
selectorate is defined as “all people in a country who have a granted right that gives them a
say in choosing the government” (Bueno de Mesquita 1999). Systems with a large winning
coalition and a large selectorate are democracies, and those with small coalitions and large-
selectorate polities are autocracies for the most part. Systems with a small coalition and a
small selectorate tend to be monarchies or military juntas.

Leaders should maintain the loyalty of enough people to keep a winning coalition to
stay in office. A small winning coalition implies that a leader needs only the support of a
few individuals. Even if the leader’s economic policies fail, the essential supporters of an
incumbent can be expected to remain loyal, as long as they are happy with their private
benefits. Since political leaders in this situation are able to keep their positions by
allocating more political resources into private goods, their policy performances are less

likely to influence their political survival. That is, leaders do not significantly increase their

45

prospects of remaining in office by spending resources for good public policies. In contrast,
leaders in democracies may prefer different strategies since they know that private goods
are distributed very widely. Therefore, it is difficult for leaders to provide supporters with
enough private goods to keep them loyal. It is obvious that coalition members are more
likely to defect if the incumbents policies are seen to have failed.50 How do democratic
leaders stay in office? In this situation, policy performances are more likely to impact
political leaders’ survival. Leaders have the stronger incentive to provide good policy for
their citizens (Bueno de Mesquita et al. 1999). That is, if the size of the winning coalition
is larger, more resources will be allocated to produce public goods and therefore better
policies such as those which promote free trade will be expected.51 On the other hand, if the
size of a winning coalition is small, poor policy performances are generally expected,
because leaders have to use resources to distribute private benefits. Therefore, democratic
leaders have a strong incentive to perform well by producing good public policy which
autocratic leaders lack. Although this is a simplified explanation of the interactive role of
motivations and institutions on trade policy choices, the logic of public and private benefits
in terms of the strategy of leaders provides a reasonable explanation for the different
incentives of leaders regarding a foreign economic policy in different institutional systems.
Why authoritarian leaders prefer protectionism as a means of satisfying their winning
coalition is a more complicated matter. Autocratic leaders may prefer closed trade policies
because they face lower costs in closing the economy to international trade. That is, lower

political costs increase the likelihood that autocracies will choose protectionism. Free trade

 

so There is supportive empirical evidence that poor policy performance increases leaders’ risk of being
removed, creating an incentive to work hard to prevent policy failure (Bueno de Mesquita et al. 1999).

5‘ More detailed explication for this statement, see chapter 5.

46

policies hurt inefficient domestic businesses. Workers in inefficient industries and cronies
of leaders suffer from greater losses if they compete under free trade. Thus, for autocracies,
there is not much pressure for policy change.52

In the next chapter, I elaborate on how this theory can be used to account for the
decision to join PT As. I clarify how the notions of winning coalition and the size of the
selectorate provide special insights for differentiating the decision to join PT As from
unilateral or multilateral decisions to open trade, and conclude that regime type can
influence the probability of signing a trade agreement. Yet, Haggard and Kaufmann (1995)
argue that differences within regime types are likely to have more effect on the capacity to
initiate trade policy reform than the regime type itself. Thus, based on the need to
investigate the disaggregated impact of political institutions on the decision to liberalize
trade, I examine how changes of international trade policy are influenced by the variations

of domestic political institutions.

4.1.2. Veto players and International Trade Policy

Veto players in political systems are an important institutional factor in shaping trade
policy. Political institutions “structure the advantages and impediments accruing to
different groups” in the political process (Bennett and Duchesne 2000). In other words,
political institutions influence how political pressure may be exerted over policy outcomes

among different interests in society.

 

52 On the other hand, in democracies, leaders cannot afford much protection. Democratic leaders face several
problems if they pursue protectionist policies. Democratic leaders need a broad base of support. This implies
that protection will get substantial resistance. Benefits generated by high protection may not make people
happy, since the average consumer sees no more than small gains from each protectionist trade policy.

47

Joining a trade agreement is a change of commercial policy. In democratic systems,
the process of policy decision involves a variety of institutional and partisan actors. Under
these conditions, certain political institutions make it difficult for the government to adopt
policy changes. These players represent veto players, who have the power to block policy
change in democratic countries. For example, Tsebelis emphasizes “the capacity for policy
change” by differentiating political systems on the basis of “veto players”. Tsebelis defines
a veto player as “the number of actors whose agreement is required for a change of the
status quo” (Tsebelis 1995). He argues that an increase in the number of veto players
increases “policy stability” by impeding significant departures from the status quo
(Tsebelis 1995). In doing so, Tsebelis claims that countries with many veto players such as
coalition governments, bicameral political systems, and presidents with veto power will
have a difficulty altering the current policy, resulting in only incremental policy change
(Tsebelis 2000). In other words, we should expect countries with many veto players to
display a lower likelihood of joining a trade agreement. There may be greater institutional
resistance to trade policy changes as the number of veto players increases. In the same
context, if countries have many fragmented party systems and if the dominant parties are
separated by large ideological differences, these countries may have a hard time initiating
changes in commercial policy. That is, countries with fragmented or polarized party
systems with large ideological differences are less likely to conclude trade agreements.
There is a different expectation related to the argument about the number of veto players.
Changing status quo policy to comply with PTA conditions requires the approval of other
domestic actors with veto power. In this context, a change-oriented leader makes domestic

rejection of these policy changes more costly by tying their commercial policy to an

48

international trade agreement. In some contexts, policy change inflexibility induces a
stronger tendency to join a PTA since decision-makers may strategically use a PTA by
making government commitments more costly. In this situation, the number of veto
players may not systemically influence the propensity of leaders to join international trade

agreements.

4.1.3. Electoral System and International Trade Policy

McGillvray (1997) shows that trade policies, such as the size of tariffs, may vary according
to the types of electoral systems across democracies. Specifically, decision-makers can be
more or less insulated from domestic pressures. Some institutions may give interest groups
greater access to lobby decision makers. On the other hand, other institutions tend to
insulate policy makers from the pressures of interests groups. The electoral system
influences the extent of the motivation of decision-makers to respond to pressures from
societal groups.

Alt and Gilligan (1994) consider the impact of majoritarian systems and non-
majoritarian systems such as an insulated bureaucracy or a legislature elected on the basis
of geographic constituencies in determining trade policy. Majoritarian institutions tend to
reflect the preferences of the median voter and facilitate the formation of broad-based
coalitions. On the other hand, non-majoritarian institutions are likely to stimulate the
growth of specific interests. In this context, bureaucrats are easier to lobby than politicians
accountable to a large electorate. Similarly, if legislatures are composed of representatives
elected in small districts, sector-based lobbies may have strong access to the policy—making

process.

49

In majoritarian voting systems, tariffs are more likely to be concentrated in districts.
In these systems, voters’ support is very crucial to the incumbent government. However, in
systems with proportional representation (PR), tariffs are not concentrated (Bueno de
Mesquita 2001). Under this condition, policy makers are more likely to be isolated from
societal pressures for protectionism. In large electoral districts and proportional
representation (PR) systems, “pressure groups are restrained where campaign resources or
the legal control of nominations are centralized in the hands of party leaders, such control
is achieved quite effectively in rigid list system” (Rogowski 1997). Thus, there may be a
positive relationship between free trade and PR systems. Conversely, Mansfield and Busch
claim that larger districts and a PR system result in more protection because of
“undermined insulation and autonomy of decision-makers” (Mansfield and Busch 1995)
from pressure groups. Therefore, it is less clear that greater insulation of policy makers
from societal groups will always produce policies that promote free trade. The preferences

of policy makers also become an important factor.

4.1.4. Structure of the Government and International Trade Policy

The structure of the government is an important institutional factor influencing trade
policy. Party systems also interact with the structure of the government (Milner 1999).
Scholars have demonstrated that divided government structures in presidential systems are
associated with a higher level of protectionism (Lohmann and O’Hallonran 1994). Milner

(1997) 53 also claims that international cooperation is more difficult among countries with

 

53 Milner posits that the state is composed of three important political actors in terms of international
bargaining: the executive, the legislature, and interest groups. These actors have different preferences in
terms of bargaining outcomes, and each plays a role in the bargaining process. The executive bargains with

50

divided governments. This can be an important variable for understanding decisions to join
trade agreements, but there is a debate on this issue among scholars. Divided governments
are more likely to have a lower level of trade barriers (Milner and Rosendorff 1996), but
there are no persuasive theories to explain why “divided government is conducive to
international cooperation” (Cowhey 1993).

Looking at the American case, some scholars argue that the congressional delegation
of trade policymaking authority to the President allows the President to negotiate reciprocal
trade agreements more effectively (Gilligan 1997, Lohmann and O’Halloran 1994). They
assert that the delegation of trade policymaking authority to the president “eliminated
protectionists’ logrolling when passing trade legislation and made more salient the costs of
tariffs to consumers”. In the same context, Haggard and Kaufman claim that the
concentration of trade-policy making capabilities in the hands of the executive branch may
motivate strong trade liberalization tendencies across countries (Haggard and Kaufman
1995). These models provide important implications for understanding the role of
constraints in international commercial cooperation. It cannot explain governments’
incentives to support or oppose trade agreement formation. Political institutions do not
entirely determine every aspect of disparate trade policy outcomes.54 Political institutions

shape which preferences become dominant in policy making (Milner 1999).

 

other states to reach some agreement along a policy space. The legislature ratifies or vetoes these agreements.
Interest groups shape the policy preferences of political actors.

5‘ The trade policy preferences of the essential players such as interest groups, parties, and decision-makers

are important for the outcome. The scholars who focus on this do not say enough about how interest group or
politicians’ preferences are actually formed.

51

4.2. A Perspective on the Public Goods Aspects of Free Trade

In this dissertation, I assume that free trade is a public good but the public goods aspects of
free trade need to be clarified.55 The implications of free trade liberalization as public
goods are not sufficiently studied in the literature.56 In this section, I clarify whether a
preferential tariff reduction policy is a public good at least in regional terms.

Public goods, which are often provided by governments, are usually underprovided
because of free-rider problem. Public goods benefit all of society whereas private goods
only benefit some members. Public goods in policy include national defense, income
redistribution, welfare benefits, free trade and so forth. Private goods in policy include
spoils of war, tariff policies that benefit key supporters, and so forth. Trade restrictions
belong to private goods because they create private benefits to only a few people. It is
known that one of important collective action problemss7 in international economics is free
trade. Free trade goods are generally considered as public goods. Public goods have two
features, non-rivalry and non-excludability. Scholars often claim that free trade is a good
public policy for general voters and consumers because it generates lower prices, more
quality products, and greater product variety. In this way, free trade policies benefit voters

and consumers across the entire nation, while the costs of free trade policies are often

 

55 On the basis of the institutional theory of Bueno de Mesquita, in this dissertation I argue that with a large
winning coalition, more benefits will be focused on public goods and free trade will be enhanced. That is, to
secure better public policy performance, policy makers will promote the level of free trade. “The consensus
among main-stream economists on the desirability of free trade remains almost universal.” In other words,
countries are better off with free trade than with policies restricting trade. The numerous theories show that
economies that are more open are likely to grow faster. If the faster growth is sustainable, we will see
substantial increases in terms of economic well-being. The findings of empirical studies reinforce this theory
by showing through numerous models that the growth of a nation is improved by free trade policies.

56 The implications of trade liberalization and bilateral or regional trade agreements as public good provision
have received little attention from political scientists.

57 It refers to situations in which individual incentives to be a free rider lead to inefficient collective
outcomes. This is also called a market failure. Collective action problems arise over public goods.

52

concentrated in particular industries and in particular locations. Free trade brings other
benefits. Economic research supports the notion that free trade increases national
economic welfare. According to one study, “those countries that favor free trade policies
tend to promote relatively greater income equality and better returns to labor. They also
attract more investment and more capital. Those that eschew free trade and protect special
domestic interests foster inequity, poor returns to labor, and low investment in the society’s
future growth in general. Thus, those governments that rarely interfere in foreign trade
produce the best outcome for most of their citizens.”58
Free trade agreements (FTAs) allow more reciprocal market access and involve the
harmonization of trading rules and the reduction of trade barriers among states. At least at
the regional level,59 they reduce the probability of costly trade disputes through mutually
binding rules in international trade. Free trade agreements provide benefits across nations
by diminishing inefficient protectionism and fostering the distribution of goods and
services.60 In doing so, more consumers and producers can benefit from expanded trade at

least at the regional level.61 In this sense, free trade agreements are desirable since most

countries do not want to be excluded from the benefit of FTAs.

 

58 Mendoza and Chandrika 2002.

59 Public goods can be discussed at many levels, such as the national, global, as well as regional level.
Regional public goods are different from global public goods because their benefits are mostly to countries in
a specific region. It should reduce the problems of mobilizing collective action among actors with diverse
interests. This is partly because cooperation among neighboring countries often leads to less information
problems. Mendoza and Chandrika 2002.

6° By facilitating efficient trade, free trade agreements can provide public goods to any member that chooses
to trade. As to non-excludability, exclusion from increased levels of free trade is not desirable because it
diminishes increased gains.

6’ It is often difficult to straightforwardly discuss free trade as a public good. There are certainly positive
gains from free trade, yet the effects of free trade are diverse in different contexts. We should consider how
the gains from free trade are distributed and have evolved over time. Sometimes, benefits from increased free
trade will be concentrated within exporters and consumers. At the global level, the benefits from increased

53

Although free trade generates good public policy effects in general, the adoption of
free trade will depend on the beliefs of political decision makers. If they believe that free
trade is socially and economically desirable, they will more likely to be motivated to make
a policy change. In general, executive leaders should be more focused on the pursuit of
free trade as good public policy than legislators. Leaders pursue policy reform toward free
trade to “appeal to broad voters” (Nielson 2003). However, leaders are motivated to do so
only when they think that free trade generates public goods and it will benefit their political
survival. Otherwise, leaders are not going to be very determined to adopt free trade policy
reform. On some occasions, political decision makers will be influenced by negative
impacts of trade policy reform. Free trade is generally beneficial for the general consumer
but at the same time it may harm particular industries or those who own particular factors
of production. A key part of free trade is that it deprives inefficient domestic businesses of
protection from a strong foreign competition. This is good for the economy, but bad for
businesses that were being protected (Bueno de Mesquita 2003). Accordingly, policy

change involves winners and losers in different sectors."2 Who will be banned depends on

 

trade will often be too concentrated within the developed countries because they already enjoy dominance in
terms of competition in international markets. Abe shows that tariff reductions may lead to lower welfare in a
small open economy. Gray (1990) suggests that greater economic integration limits the scope of policy
makers to provide national collective goods in accordance with a divergent preference across nations. “Policy
harmonization, as far as public goods are concerned, is undesirable when country’s policies are aimed at
providing public goods at a nationally desirable level. Higher levels of integration involve labor mobility,
foreign direct investment and policy harmonization. The reason for that speculation is that the scope for
provision of public goods, such as income distribution and even the goals of stabilization policy, is reduced
by economic integration beyond some point. If governments do not act in the general interests of their citizens
then public-good provision may be sub-optimal” (Feehan 1996).

62 For instance, despite economic benefits, free trade policies are opposed by a large percentage of the US.
public. The opposition may come from labor unions, environmentalists, and human rights activists. The
implementation of trade policies creates winners and losers, and the potential losers oppose free trade
policies. Others oppose free trade because of concerns that free trade contributes to the abuse of workers, as
well as to environmental degradation. However, as mentioned, economists agree that a nation’s economic
well-being is promoted by free trade (Coughlin 2002).

54

the mobility of labor and capital and on the ability of specific groups to overcome free-rider
problems by pressuring the government for protection.63 Furthermore, there are short-term
side effects of trade policy reform, such as employment failures and economic downturns.
If these kinds of negative effects influence the political ambition of leaders sufficiently,
then leaders will not adopt free trade. Although diverse issues complicate the discussion of
the public goods aspects of free trade, few reject that free trade is a desired public policy
(Gowa 1988, Nielson 2003). In conclusion, despite some potential negative effects of free
trade, free trade policies generate many advantages in economic sectors and should be

considered a good public policy.

4.2.1. Free Trade and Preferential Trade Agreements
Another important question that should be clarified is whether a Preferential Trade
Agreement is a stepping-stone or a stumbling block to free trade. Many studies suggest
that the answer may be “positive or complementary” (Lawrence 1996).64

First, trade will be expanded because of the elimination of trade barriers. For instance,

the expansions of the European Economic Community (EEC) pressured GATT members to

 

”3 Who wins and who loses is not systematically related to whether they are wealthy or poor. It is more likely
to be influenced by the fact that which group has a strong incentive to press for protection and which group is
in a better position to overcome the free-rider problems by using their political influence to get protection.

6‘ It is difficult to say precisely what the consequences of the growth of PTAs are. However, the expansion of
RTAs includes two other recent phenomena. “First, the 1990s were a period of rapid growth of accessions to
the GATT and the WTO, from some 80 GATT Contracting Parties in 1990 to over 130 WTO Members
today. In the accession process, new GATT/WTO Members committed themselves to reduced protection and
the implementation of WTO rules, which include the notification of RTAs to which they are party. Second,
this was a period of unilateral liberalization, particularly among developing countries and economies in
transition, and this liberalization was largely consolidated in the Uruguay Round. There have been a decline
in the use of non-tariff measures as well as considerable rationalization of tariff structures, tariff reductions to
moderate average levels and a major expansion in binding coverage. This background of unilateral and
multilateral liberalization considerably reduces the scope for trade diversion. Baldwin claims that almost all
empirical studies of European and North American arrangements find positive impacts on member’s living
standards” (Crawford Jo-Ann and Sam Laird 2000).

55

initiate multilateral trade negotiations leading to significant reductions in trade barriers.
Second, the creation of trade leads to a better resource allocation. There are welfare
benefits gained by increasing productive efficiency. Member countries may also realize
economies of scale in production because of the increased market power through PT As.
However, despite some of these positive effects of PTAs can cause toward free trade, no
arguments about their effects on productivity and overall trade liberalization have been
conclusive.“ Arriving at a complete explanation of consequences of the preferential trading
blocs is difficult because developments are still moving at a rapid pace. Furthermore, the
debate about the impact of PTAs on the dynamic time-path question in global free trade has
not been resolved. Although there is no explicit conclusion on this question, many scholars
tend to agree that the growth of P'TAs is a positive step towards free trade liberalization
(Lawrence 1996, Kemp and Wan 1976, Krugman 1993, Summers 1991, Nordstrom 1995,

Perroni and Whalley 1996, Campa and Sorensen 1996).66

4.2.2. Regime Type and Free trade

Although the adoption of free trade or protectionism will depend on the beliefs of political
decision makers, political institutions constrain how preferences of leaders aggregate to
become trade policy. Democratic leaders tend to be dependent on a large winning coalition,
whereas autocratic leaders are dependent on small winning coalitions. These institutional

differences provide leaders with incentives to use different trade policy strategies.

 

”5 PTAs have a “two-sided quality, liberalizing trade among members while discriminating against non-
members” (Mansfield and Milner 1999). Bhagwati (1995) discusses the trade diversion effect, and claims that
regional trade agreements are “mostly welfare reducing since the partner countries will end up with severe
welfare losses due to trade diversion.” Additionally, PTAs may inhibit global free trade (Krugman 1991,
Bhagwati 1993, Bond and Syropoulos 1996 and Frankel, Stein, and Wei 1995).

6" For more detailed discussions and explications on this question, refer to the section of the impact of
Preferential Trade Agreements in chapter 1.

56

So how do democracies strike a balance between groups pressuring leaders to use
trade barriers and the other groups pulling leaders to adopt free trade? The answer is not
straightforward. We should keep in mind that, not all trade policies will affect a democratic
leader’s major supporters equally. Democracies are responsive to a larger constituency.
Leaders in democracies need to enact successful trade policy in order to help maintain their
position. Therefore, we see more liberal trade policies in democracies because their leaders
are more concerned about good public policy performance, which is measured in large part
by economic growth.67 Autocratic governments are less supportive of free trade than are
democracies. Because they mostly rely on small winning coalitions, they use private
benefits to reward supporters more than do democratic systems. Therefore trade
protectionism is more prevalent in autocracies than in democracies. In sum, free trade
produces diffuse benefits, whereas protectionism produces concentrated benefits in specific
business sectors and districts. In an authoritarian system, the primary political
consideration of leaders is to keep their cronies happy. If this need is satisfied, leaders in
autocracies are less likely to be concerned about producing good public policies. Also
relevant, it is the difficulty faced by democracies in adopting protectionist policies.68 In a
democracy leaders need a broad base of support. We should expect that they will try to
promote more effective trade policies since this appeals to a broader group of their

electorate.69

 

”7 This is an example of successful provisions of public goods.

68 If protectionism is not good for an economy as a whole, why does it persist? Decision-makers may pursue
protection based on narrow political considerations. Asymmetry between general public benefits and
concentrated costs from trade is an important factor that determines the level of trade protectionism. If
interest groups or other coalitions who benefit from trade restriction are powerful elements of a leader’s
election coalition, we can expect to see more protectionism.

”9 I elaborate on this point in detail in chapters 4 and 5 (Bueno de Mesquita, Bruce. 2000)

57

 

Chapter 5. Research Design and Hypotheses

5.1. Introduction

Chapter 4 presented different paths to the growth and expansion of reciprocal trade
agreements. The interplay between preferences70 and domestic institutions along with
systemic international factors" can help explain trade policy, as shown in figure 5.1. An
understanding of domestic institutional structures is crucial to explaining trade policy
outcomes. As Hankla (2002) notes, “Arguing that variations of trade policy are mostly the
result of domestic political processes, some scholars have relaxed the unitary actor
assumption. They examine the policy preferences of actors within states. Preferences
oriented theories are necessary to understand the trade policy puzzle because it is difficult
to understand the policy actions of a state without analyzing the push and pull of different
interests on decision-makers. Yet, societal explanations of trade policy are incomplete
because they do not explicitly explain how the preferences of competing domestic actors
are aggregated into policy.” Without an understanding of how domestic institutions affect
the way domestic preferences are aggregated and transformed into trade policy, our
explanation is incomplete. Therefore, explaining how domestic institutions constrain the

preferences of leaders into policy outcomes is crucial, and is not done sufficiently in the

 

7° Some scholars indicate that preference oriented approach can be called as endogenous pressure group
approach. However, since the preferences of domestic actors are not easily observable, it is difficult to
incorporate the preferences into an empirical model of the expansion of Preferential Trade Agreements
although theories can always incorporate non-measurable factors. There may be some indicators that could
serve as a proxy for preferences. For more detailed discussion, see Mansfield and Busch, 1995 and Mansfield
and Milner, 1997.

7' Systemic international factors are one of significant factors in explaining global trends in terms of trade

policy. Nevertheless, this approach is less useful in explaining the cross-national variation in levels of the
conclusion of trade agreements.

58

current literature. This dissertation focuses on developing a better domestic institutional
theory of trade policy.

This chapter elaborates on the basic assumptions of the domestic institutional theory.
I clarify key hypotheses about the growth of PTAs. I elaborate how political institution
variables such as the size of winning coalition provide special insights into the decision to
sign international trade agreements as opposed to protectionist policies to open trade. I also
clarify how disaggregated domestic institutions may influence leaders in deciding whether
to sign PTAs. This chapter finally presents a research design that includes data and

measurement for analysis.

Figure 5.1. Sources of the Formation of Trade policy

Leader’s
willingness to sign
a Preferential Trade
Agreement

 

 

5.2. Assumptions

In order to grasp the models that stress political institutions adequately, understanding
assumptions, which reflect the sources of the variations of domestic preferences filtered by
institutions in the trade policy formulation dynamic, is crucial. In this dissertation, the

main assumption is that domestic endogenous preferences will be aggregated to a specific

59

trade policy outcome through the filter of regime type and other domestic institutional
constraints.

Political institutions “create rules for decision making, help to structure agendas, and
offer advantages to certain groups while disadvantaging others. Over time, institutions may
shape actors’ policy preferences. The institutional context conditions the incentives of
interest groups and politicians. For instance, different regime types may refract incentives
of politicians in a divergent way. Non-democratic systems should respond differently than
democratic ones in terms of trade policy-making” (Milner and Keohane 1996). Political
institutions also “shape the patterns of interactions between domestic groups and whose
interests will be represented (O’Halloran 1994)” in terms of trade policies.72 In other words,
how divergent preferences of domestic groups are unified into a policy outcome depends
on “how insulated political leaders are from public and interest group pressure, and on
which political leaders and how many have power and influence over policy-making”
(Hankla 2002). Democratic systems may be divergent73 in the way in which they are
“responsive to domestic group pressure” (Reinhardt 2000). Regime types also constrain the

motivations of political leaders to the pursuit of particular trade policies.

 

72 For more detailed review see, Putnam, Robert. 1988. “Diplomacy and Domestic Politics: The Logic of
Two—Level Games.” International Organization. 42. Rogowski, Ronald. 1987. "Trade and the Variety of
Democratic Institutions.” International Organization 41(2): 203-22. O’Halloran, Sharyn. 1994. Politics,
Process, and American Trade Policy. Ann Arbor: University of Michigan Press. Garrett, Geoffrey and Peter
Lange. 1995. “Internationalization, Institutions, and Political Change.” International Organization, 49: 627.
Politics, Process, and American Trade Policy. Ann Arbor: University of Michigan Press. Cox, Gary and
Mathew McCubbins. 2001. “The Institutional Determinants of Economic Policy Outcomes.” In Presidents,
Parliaments and Policy, ed. Stephan Haggard and Mathew McCubbins. New York: Cambridge University
Press. Nielson, Daniel. 2003. “Supplying Trade Reform: Political Institutions and Liberalization in Middle-
Income Presidential Democracies”, American Journal of Political Science, Vol. 47, No. 3, July: 470-491.

73 Even among democracies, there may be more variations in the reaction of leaders or legislators to domestic
pressures in terms of electoral rules. Electoral rule as a preference aggregation method may refract policy
preferences of political leaders by providing different constraints to interest groups and societal groups
(Milner and Keohane 1996). In doing so, political institutions “channel societal demands” for trade policy-
making (Nielson 2003).

60

Few models explicitly test how leader’s preferences aggregate to the particular
outcome in different regimes and different institutional contexts. To fill this gap, chapter 4
explained how leaders’ preferences and the distinguishing features of regime types such as
democracy and autocracy are connected to the formulation of international trade
agreements. The theory is derived from the institutional theory of Bueno de Mesquita and
was applied in particular to external trade policy. With the application of this theoretical
perspective, this dissertation attempts to redress the current limitations in the study of
political institutions and trade policy. 74 I present how domestic institutional constraints
such as the number of veto players and electoral rule may play a role in mediating social
demands in terms of trade policy formulation. I control for systemic factors such as the
erosion of hegemony, the existence of the GATT/WTO and other economic conditions.75
The dissertation aims to contribute to a systematic analysis of the role of domestic

institutions in the expansion of free trade agreements.

 

7’ Mansfield and Busch highlight “the interplay between factors that cause demands for protection and those
that control the provision of protection by policymakers has not been treated adequately in the field of
international commercial policy. Pressure groups approach and domestic institutions approach should be
combined more appropriately.” They conclude that the interplay of these two approaches is necessary to study
determinants of trade policy (Mansfield and Busch 1995).

75 The analysis allows us to do “simultaneous tests of both domestic and international factors,” which
influence the decision of the formation of a Preferential Trade Agreement. Without considering the
interaction between domestic and international constraints, the explanation related to economic cooperation
remains incomplete (Reinhardt 1996).

61

5.3. Research Design
5.3.1. Hypotheses: Domestic Institutional Constraints

5.3.1.1. Regime type

The likelihood of a leader choosing to join international trade agreements differs across
levels of regime types. Depending on regime type, a leader’s policy goals and motivations
become different with regards to the strategic distribution of available resources. Some
leaders distribute their resources toward the pursuit of public policies that benefit all
citizens. Other leaders focus on the pursuit of the distribution of private goods that benefit
their key supporters. Different regime types give different motivations to leaders in terms
of their need to pursue policies that benefit their economies.

In this dissertation, I extend Bueno de Mesquita’s institutional theory to explain the
factors which determine the likelihood of the formation of PTAs. An increase in the size of
the selectorate and the winning coalition will change a leader’s trade policy. Democracies
with large winning coalitions induce the leader to commit to policies benefit the populace
in general.76 As a consequence, with a large winning coalition, more effort will be focused
on policies that bring public goods such as free trade. On the other hand, small winning

coalitions associated with large selectorates induce leaders to provide their key supporters

 

7‘ Because resources are limited, leaders want to limit the distribution of private goods to members of the
winning coalition. Why? Because the members of this coalition are the ones that keep the leader in power,
and so they are the individuals that the leader cares about. Wasting private goods on individuals outside
winning coalition group is a waste of resources. Public policies benefit all of society, whereas private goods
only benefit some members. If the selectorate and winning coalition are large, then it is more effective for a
leader to focus on providing public goods, rather than private goods. If the winning coalition is small and the
selectorate is large, then the government will focus on providing private goods to its supporters. These
members will be very loyal because they would likely be shut out of alternative winning coalitions. For
democracies the implications are the opposite. Democracy encourages a norm of disloyalty when a leader
fails to provide effective public policies. That is, in democracies, the contest is over who can provide the best
bundle of public goods. In short, democracies focus on public policy, while autocracies do not.

62

with private goods, not public policies. These leaders can remain in office even by bringing
corruption and economic uncertainty.77 Compared to democratic leaders, autocrats have
“little incentive to concern themselves with the welfare of their society” (Bueno de
Mesquita, Morrow, Siverson and Smith 2000). To test the theory, I develop a hypothesis as
to whether countries with a larger winning coalition are more likely to sign a free trade
agreement.

Democratic leaders care about their reputation as effective policy makers, and often
desire to be seen as strong proponents of free trade. Aside from the actual benefits of a
trade agreement, successfully concluding international agreements is valuable in itself.
When democratic leaders conclude a trade agreement, they tend to be more credible with
regard to free trade and in foreign policy generally. By concluding a Preferential Trade
Agreement, governments can make a costly signal to their audiences by revealing their
willingness in lowering trade barriers. Mansfield et al. (2002) demonstrate that democracies
do not enter into a PTA solely because they want to lower the level of trade barriers.
Entering into international trade agreement has purely political implications aside from
economic considerations. “International trade agreements can solve an informational
problem that limits the head of state’s credibility. From the point of view of the voters,
entering into a PTA strengthens the credibility of offers made by the executive to reduce
trade barriers. In this process, leaders can enhance the prospect of their political survival. In

this sense, both voters and leaders can be made better off with an international trade

 

77 Consider Saddam Hussein, who had a small winning coalition. Even economic sanction cannot hurt him.
Sometimes, economic sanctions may create black market opportunities. Saddam allowed his supporters to
exploit for personal gain. He provided private goods to his supporters, and they were very loyal. Another
example would be Mobutu Sese Seko, who was of Zaire. Seko succeeded in plundering the country,
accumulating a personal fortune over six billion dollars. He successfully maintained his office by satisfying
his small group of supporters (Bueno de Mesquita 2003).

63

 

agreement as a country becomes more democratic. Little would be sacrificed in the trade
negotiations, but much would be gained in terms of the domestic polity” (Mansfield,
Milner, Rosendorff, 2002). This is another crucial factor that makes democracies more
likely to join international trade agreements.

Disaggregated aspects of domestic political institutions may have different impacts in
influencing the propensity of leaders to join a PTA. It is not satisfactory to highlight only
the aspect of regime type in considering leader’s determination in concluding international
trade agreements. Even among democracies, different institutional structures motivate
different interests in the policy-making process. Thus, disaggregated aspects of political
institutions such as the role of veto players, the legislative-executive relationship, and the
kind of electoral system may also influence the likelihood that political leaders decide to
sign a PTA (Milner and Kubota. 2001).

Executives play a crucial role in terms of a country’s incentive and ability to initiate
the implementation of trade policy change. National leaders of executives tend to focus on
overall economic policy performance in factors such as growth. It is argued by some that
political leaders in all countries prefer free trade to protectionism. This is a simplified
generalization, but there is supporting evidence. Executives generally have national
constituencies, and so they should be less vulnerable to interest group pressure. For
instance, American presidents, regardless of their party identification, have tended to
support free trade and trade liberalization, despite strong protectionist lobby groups. They
are also capable of strengthening free trade even in protectionist times by encouraging the
participation of free trade interests (Hankla 2002). However, if that is the case, why is there

so much variance in trade policy preferences across countries and over time? In reality, the

process of trade policy formulation and outcome is much more complex in different
institutional contexts. For instance, when it comes to trade policy formulation, we must
integrate the role of legislatures rather than exclusively focusing on executives.
Legislatures do not take an initiative on the implementation of particular trade policy, but
they are often required to ratify and certify policy proposals.78

It is argued that, “the legislature is often regarded as a passive organization,
uninformed and with intentions that are transparent to other players” (Reinhardt 1999).
However, the structure of domestic legislatures79 plays an important role in terms of trade
policy formulation. The actors in legislatures also play a significant role in the specific
institutional setting because of their impact in the aggregation of the supply of trade policy
(McGillivray 1997). Legislators may have an interest in facilitating national policy
performance, but compared to national leaders, they are more likely to care about narrow
constituencies. They are often more concerned with satisfying specific interests such as
labor, business, export groups, import groups. Without the strong support of parties to
achieve their political goals, legislators are more likely to be vulnerable to pork barrel
politics. Thus, the institutional structures of legislatures interact with leaders’ policy
preference in the policymaking process. In this context, underestimating the role of the
institutional structure of legislatures in determining trade outcomes leads to a failure in

accounting for the likelihood of successful agreements (Reinhardt 2000).

 

78 In general, leaders of executives initiate policy and implement it. They also conduct the negotiations with
other governments. On the other hand, legislators often approve, amend, and veto a president’s policy

proposal.
79 There may be variations. In some democracies, the legislature has enormous power over law and policy. In

other democratic systems, agenda setting power is almost entirely in the hands of the executive. When this is
the case, the legislature is often forced to support the cabinet (Hankla 2002).

65

Understanding the uncertainties stemming from domestic institutional structures in the
formation of trade policies involves gauging the impact of the legislature, veto players, the
legislative-executive relationship, the electoral system etc. In the next section, I elaborate
how disaggregated domestic institutions may motivate leaders to sign a PTA. By relating
the impact of disaggregated domestic institutions to foreign economic policy, I argue that
understanding the impact of domestic institutional structures across states in determining
trade outcomes is crucial. Researchers have argued that political leaders’ preferences, as
constrained by domestic institutional constraints, shape trade policy outcomes. Such
dynamics may include the number of veto players, electoral systems and parliamentary-
executive relations.80 I show how disaggregated aspects of domestic institutions may

influence a decision of leader to sign PTAs.

5.3.1.2. Veto players 8'

Tsebelis compares different political systems in terms of the stability of their policy
outputs, by specifying the ability of political and institutional actors to veto policy change.
Tsebelis claims, “the number of veto players82 in a political system, and the conflict of
interest among those veto players, affects the legislative productivity. He shows that an

increase in the number of veto players or the diversity of veto players will reduce the

 

8° The number of political parties, the presence of a strong chief executive and so forth can be also discussed.

3' Some scholars use the notion of divided government in order to explore the impact of political institutions
in trade policy making. However, “divided government indicates a simple dichotomy. A more sophisticated
distinction is the number of separate veto players. A veto player on a given issue dimension may be the
person, political party, or faction of a political party, that exercises a veto on that issue by itself” (Cox and
McCubbins 2001).

82 For example, he argues, “oversized coalitions will be locked into the previous policy pattern whatever that
pattern happens to be” (Tsebelis 1999).

66

likelihood of new laws passing” (Haggard and McCubbins 2001). Cox and McCubbins
(2001) also claim that a “govemment’s ability to change or to commit to policy depends
mostly on the effective number83 of vetoes in political decision-making. As the effective
number of vetoes increases, there is an increase in the transactions costs that must be
overcome in order to change policy. As more actors must be taken into account in a policy
logroll, it will become increasingly difficult to structure negotiations. As more interests are
provided with vetoes, it becomes increasingly difficult to ensure that every party to the
negotiations receives sufficient value to accept the deal. Hence changing policy becomes
increasingly costly as the number of parties to a negotiation, or as the diversity of their
preferences, increases. Costs hinder policy change. As the effective number of vetoes
increases, the polity becomes more resolute, and less decisive.”

When a political decision-maker enters a PTA, the policy change requires the approval
of other political actors with veto power. For example, when an executive of a country
enters into a free trade arrangement, the agreement will set certain regulations. These
conditions entail specific trade policy changes such as tariff reductions, the decision-maker
may have to overcome the domestic barrier of veto players because changing trade policy
require their approval.

Consider a country interested in making trade policy reforms. The decision-maker of
this country may want to conclude a PTA in order to overcome the opposition of veto
players because they can lock them in through this external mechanism. This tendency is

more likely to be seen when various segments of society oppose trade policy reforms and if

 

83 The issue how to count veto players is not easy. It becomes more problematic if one wants to examine the
preferences of all subunits. In this sense, Cox et al. use some shorthand rules. They count a veto actor that
controls a veto, ignoring the issue of whether their preferences are distinctive, or if their preferences are not
completely distinct from the remainder of the party (Cox and McCubbins 2001).

67

 

domestic institutions make leaders vulnerable to public pressures. Under these conditions,
leaders must have the institutional means to overcome domestic opposition in order to enter
such free trade agreements. The costs of violating a PTA must be high enough to ensure
that reforms will be hard to reverse.84 In this vein, governments may enter into PTAs to
push through unpopular policies of economic reform because rejection of these policies by
opponents becomes more costly when the policies are tied to international trade agreements.
By tying policy to international trade treaties, a reform-oriented executive makes rejection
of these policy changes more costly and difficult to implement in the future.85 According
to Mansfield and Milner (1999), “PTAs have been created by policymakers who preferred
to liberalize trade but faced domestic obstacles to doing so unilaterally. Columbia and
Venezuela decided in November 1991 to turn the previously moribund Andean Pact into
what is now one of the world’s most successful FTAs. Policymakers in these countries
explain their decision as a politically easy way to dismantle protectionist barriers to an
extent that their domestic legislatures would never have allowed had the policy not been
pursued in a regional context. Even if dominant domestic actors oppose commercial
liberalization altogether, institutional factors sometimes create opportunities for
policymakers to sidestep such opposition by relying on regional or bilateral trade

strategies.” Mansfield and Milner (1999) also point out “the situation Napoleon III faced on

 

8’ Consider Mexico in NAFTA, where the Mexican government locked in domestic policy reform through
NAFTA. For more detailed discussion, see Whalley 1998.

85 This is called “the tied hands” hypothesis. It claims, “leaders may undertake a PTA to achieve economic
reforms through free trade agreement means if it may not be politically popular back home. The strategy of a
national decision-maker is to commit his country to an international institution. It may be helpful in imposing
the politically opposed commercial policy. The decision-maker can claim that his hands are tied. Also, free
trading arrangements provide external credibility for unpopular domestic programs. This is important because
it suggests that political systems that permit effective domestic opposition will be more likely to seek external
credibility than authoritarian or single party governments. In doing so, decision-makers may avoid domestic
criticism for policies that they actually prefer. In other words, international agencies help top decision-makers
against the negative effects resulting from unpopular measures” (Kerremans and Switky 2000).

68

the eve of the Anglo-French commercial arrangement. Anxious to liberalize trade with
Great Britain, he encountered a French legislature and various salient domestic groups that
were highly protectionist. But although the legislature had considerable control over
unilateral trade policy, the constitution of 1851 permitted the emperor to sign international
treaties without this body’s approval. Napoleon was able to skirt well-organized
protectionist interests much more easily by concluding a bilateral commercial agreement
that would have been impossible had he relied solely on unilateral instruments.”

Even when governments choose to conclude PT As to promote domestic reforms,86 veto
players can create substantial institutional resistance to policy change, especially when
there are strong veto players. Veto players are able to block the policy reform of economic
liberalization or tariff reduction. Consequently, many veto players, even those separated by
large ideological differences, may influence the likelihood of trade policy change. There are
two types of veto players. Partisan veto players are individuals or groups (parties) who
wield vetoes, while institutional veto players are those entities with veto power, such as
presidents and legislatures. As Tsebelis summarizes, “the number of institutional veto

players is specified by the Constitution, while the number of partisan veto players will be

 

8" Based on this logic, we can make an interesting counter-intuitive argument with regard to the impact of
veto player over trade policy. When dominant domestic actors oppose trade reform or liberalization,
institutional factors sometimes create opportunities for leaders to sidestep such opposition by relying on
regional or bilateral trade strategies as mentioned above. That is, political opposition groups with veto power
may accept more economic reform than they would like because they are concerned with the costs of
rejecting the international trade agreement. By rejecting the international trade agreement, they may send a
negative signal to other countries in the international trade market. In this case, domestic political institutions
make executives more likely to rely on the strategy of concluding an international free trade agreement to
push through reform. While executives in different types of regimes may use that kind of a strategy, it is most
likely to be pursued when there is greater institutional resistance to policy change. That is, executives facing
more veto players are more likely to turn to a preferential trade agreement. An international free trade
agreement may serve as a potential ally in the face of veto player opposition to economic reform.

69

decided endogenously by the party system and the government coalitions of each specific
country” (Tsebelis 1995).87

Tsebelis argues that if there are more ideologically opposed or polarized veto players on
the same policy dimension, the political system may have a greater difficulty in creating the
agreement necessary for change. For instance, if parties, presidents, and legislatures are
opposed over an issue, they will have a more difficulty in agreeing on a policy than in those
areas with less explicit disagreements. Tsebelis also argues that policy change will be more
difficult if veto players are more cohesive, because such players will more strongly demand
their group’s preferences. For example, party defections rarely occur if party discipline is
high. To sum up, Tsebelis argues that increasing the number of actors, their polarization,
and their cohesion (party discipline) will decrease the government’s ability to pass
significant laws such as the conclusion of a PTA.

Trade issues are more vulnerable to veto pressures because they have distributive
implications (Milner 1997). In this context, international cooperation on preferential trade
agreements is more difficult to achieve as the number of veto player rises in political
systems. Thus, in this dissertation, I attempt to investigate how the number and preferences
of veto-wielding political actors influence international commercial cooperation,
specifically in terms of the conclusion of PTAs. Controlling for economic and political
factors, more veto powers tends to decrease the government’s ability to pass significant
laws such as the conclusion of a Preferential Trade Agreement. Using cross-national

datasets, I test this proposition.

 

87 Sometimes, the number of partisan and institutional veto players overlaps; for example, when a veto-
wielding president is of the same party as the majority party in one or both chambers of a bicameral
legislature.

7O

5.3.1.3. Electoral system

Leaders in democracies develop their preferences based on the preferences of their
constituencies as well as their own beliefs and the preferences of their political parties.
How these divergent preferences are aggregated into a policy outcome mostly depends on
how insulated leaders are from public pressure. Determining the degree to which leaders
are insulated from public pressure is a growing topic in the literature. The impact of
constituency size and electoral system on trade policy are important. That is, external trade
policy of countries can be influenced by electoral rule88 (Rogowski 1997,. McGillvray 1997,
Mansfield and Busch 1995, Nielson 2003). The electoral rule and party strength affect how
legislators’ preferences can be aggregated into policy. As explained in chapter 4, it is
generally argued that large electoral districts are associated with less influential interest
groups. On the other hand, small electoral districts are associated with more influential
interest groups because of their concentration. That is, small districts tend to increase the
power of large interests, while larger districts may diminish the influence of single groups.
As mentioned earlier, the interests of protectionist groups tend to be more concentrated in
particular geographical locations. As a result, political representatives with smaller
constituencies are more vulnerable to protectionist interest groups than leaders chosen from

a broader electorate. For this reason, small electoral districts tend to foster patronage and

 

8” See Nielson, Daniel. 2003. “Supplying Trade Reform: Political Institutions and Liberalization in Middle-
Income Presidential Democracies”, American Journal of Political Science, Vol. 47 , No. 3, July: 470-491.
McGillivray, Fiona. 1997. “Party Discipline as a Determinant of the Endogenous Formation of Tariffs.”
American Journal of Political Science. 41(2), April. Mansfield, Edward and Marc Busch, 1995. “The
Political Economy of non-tariff barriers: a cross-national analysis”, International Organization 49,4. Autumn.
723-49.

71

pork-barrel politics (Rogowski 1987).89 The number of a country’s legislative districts may
influence trade policy outcomes.

It is also argued that the electoral system influences trade policy outcomes90 in terms
of its institutional influences on the preferences of legislators. Majoritarian electoral
systems with a plurality rule are more likely to be associated with more protection.
Meanwhile, proportional systems tend to be associated with freer trade. That is, legislators
elected from many single-member districts are more vulnerable to protectionist lobbying
than those elected from larger proportional districts (Rogowski 1987, Mansfield and Busch
1995). Legislators from single member districts need to maintain a plurality of votes in
their districts to stay in office. As a result, they should keep complex and fragile voting
coalitions. This situation creates influential narrow interests within a district, which are
often protectionist. It is agued, for example, that “United States senators from
heterogeneous states are freer to pursue policies in keeping with their party ideologies than
senators from homogenous states, who must appease specific constituent interests in order
to be reelected. Senators from heterogeneous states are able to ensure their reelection with
the help of a variety of possible voting coalitions, thus making them less reliant on specific,
and usually protectionist, interests” (Hankla 2002).

Another reason proportional representation (PR) systems display a tendency towards
more liberal trade is that they tend to produce more cohesive and disciplined parties than in
single-member district systems. When there is PR, the power of the party leadership is very

influential since the “Party leader generally decides who will appear on election lists and

 

8” For more detailed explanation on this point, refer to chapter 4.

9° In this dissertation, I focus on the impact of electoral rule to the trade policy outcomes.

72

ballots, legislators must follow the national party line in general. This party cohesiveness
becomes stronger by the tendency of PR systems to be parliamentary, creating pressure for
majority party (or coalition) legislators to support the government in all its initiatives.”
Finally, because each district elects more than one representative in a proportional
representation system, legislators tend to have larger constituencies. In this case, specific
interests are not dominant. In sum, plurality systems are more likely to “undermine the
insulation and autonomy of decision-makers” (Mansfield and Busch 1995) from the
demand of pressure groups. Mansfield and Busch argue that a greater insulation of policy
makers from societal groups tends to promote free trade. Accordingly, plurality systems are
more likely to lead to protection since policymakers are more likely to be constrained to
societal pressures. On the other hand, legislators representing large electoral districts under
a PR system will be more free trade oriented. However, the argument is far from
determined. Proportional representation systems (PR) may lead to more protection.
Consequently, this proposition regarding the impact of electoral rule needs to be assessed
more systematically with alternative explanations, especially those regarding the

motivation of leaders to sign a PTA.

5.3.1.4. Presidentialism vs. Parliamentarism

Many comparative studies suggest that there are qualitative differences between
parliamentary and presidential systems91 in terms of the process and formulation of their
policy making (Lijphart 1992, Weaver and Rockman 1993). “In a parliamentary system,

the executive and the legislature are often linked. Sometimes the legislature has almost no

 

9’ Shugart and Carey, 1992.

73

independent power. It makes the executive leaders the sole relevant policymakers. On the
other hand, in a presidential system, both the president and the legislature retain
independent power. Relevant policy makers can be found in both branches, suggesting that
presidentialism should lead to higher tariffs” (Elrich 2003). However, if the legislature
delegates trade-policy authority to the president, this may change the situation. It is argued
that “delegation of trade policy decision authority to presidents and party leaders is
significantly related to trade liberalization. Further, presidents are more insulated from
interest group pressure than prime ministers. Therefore, with presidential insulation, we
expect that presidential system with strong delegating power should reduce protectionists
bias and lead to lower tariff as a result” (Nielson 2003). Therefore, empirical models

92 in order to test if

include the distinction between presidentialism and parliamentarism
presidentialism is more prone to international commercial cooperation such as the
conclusion of PT As.

In addition, the number of veto players may function differently under presidential
and parliamentary systems. Policy change may require the approval of veto players,
including the signing of free trade agreement. Under a presidential system, the legislature is

likely the prominent veto player. The legislature is not a player who can directly influence

the origin and survival of the executive. Under a parliamentary system, however, the veto

 

92 This distinction is a crude measure of executive and legislative relations because there are many
institutional variations even among presidential systems and among parliamentary systems. Even among
presidential systems, presidents have different levels of power over policy making. For example, the president
may possess reactive powers, such as the authority to veto or delay legislation. The president may possess a
package or line-item veto. Presidents have different levels of proactive power, up to the authority to take
unilateral action that does not require legislative approval. These different factors will lead to different effects
over international commercial cooperation. “Although the relative merits and consequences of presidential
and parliamentary systems have received substantial attention, this distinction between macro institutions is
not enough. Explaining political outcomes often requires greater focus on the details of institutional structure
given different institutional arrangements have systematic effects on policy making” (Haggard and
McCubbins 2001).

74

player is likely to be a coalition partner that can influence the survival of the executive. If
the coalition partners in the government are strongly opposed to the policy change, they
may cause the downfall of the government. In this sense, the effect of the number of veto
players should be different under presidential systems and parliamentary systems.93 I test
this hypothesis by investigating how different government systems influence international

commercial cooperation.

5.4. Data and Measurement
5.4.1.Dependent Variable

5.4.1.1.The Decision to Sign International Trade Agreements

I evaluate whether factors such as the size of the winning coalition and the selectorate
affects the probability that a leader chooses to sign a free trade agreement. I use the

decision of a state to enter a free trade agreement as the dependent variable.94

 

93 Scholars highlight a different process of legislative ratification of trade policy in parliamentary and
presidential systems. “The chief executive such as prime minister or president in a democracy needs the
approval of a legislative majority to implement her preferred trade policies. Legislative ratification of trade
policy occurs in a different way in parliamentary and presidential systems. In parliamentary systems,
ratification is often ex ante. The potential prime minister must negotiate an acceptable trade policy with her
legislative majority prior to taking office. This internal bargain is enforced afterward because the majority can
call for a vote of (no) confidence if the prime minister strays from the accepted position in international
economic negotiations. In presidential systems, major changes in trade policy usually require legislative
ratification ex post” (Mansfield, Milner, Rosendorff 1997).

94 As mentioned earlier chapter, there are many factors that influence the decision of leaders to join
Preferential Trade Agreements. Countries join a PTA for economic and political reasons. On the economic
side, the expansion of the domestic market can create economies of scale, while competition from partners
can make domestic manufacturers produce goods more efficiently. A PTA can create broader liberalization
among their members than may be possible multilaterally. Countries also can have political benefits from
joining a PTA. Studies show that countries that are trade partners are less likely to go to war. Leaders may use
preferential trade agreement membership as a means to consolidate domestic economic reforms and provide
political stability. Some PTAs require a commitment to democracy as a condition of membership. By joining
a regional group, smaller countries may increase their bargaining power. Countries may also decide to join a
PTA to achieve their objectives regionally, if such objectives are difficult to achieve at the multilateral level.
Countries may also conclude PTAs to avoid political or economic isolation. Plus, the more countries that are
parties to PT As, the greater the incentive for others to conclude their own PTAs (WTO 2002).

75

Commitment to a PTA can be seen as a costly signal by governments to convey their

95 Internationally, leaders commit to

future willingness in eliminating trade barriers.
lower trade barriers among members by signing PTAs since they diminish trade barriers
among members and restrict a states’ capability to raise trade barriers in the future
(Mansfield 1997).96 Once they commit to lower trade barriers, it is difficult for a leader to
“refrain this commitment by simply altering policy in the future.”97 Trade liberalizing
agreement commitment is one way for governments to secure their credibility if
international markets doubt their willingness to maintain and enhance trade liberalization
policy. In this vein, employing the conclusion of a PTA as a dependent variable is a
reasonable way to measure the willingness of a country to cooperate with other countries in
liberalizing commerce. Among trade agreements, I focus on Preferential Trade Agreements
in the dissertation. Preferential Trade Agreements (PTAs) are “actions by governments to
liberalize or facilitate trade (WTO 2002).” As indicated earlier, free trade agreements are
increasingly concluded among geographically non-contiguous countries. They include
“cross-regional preferential agreements linking countries around the globe” (WTO 2002).
Trade arrangements take various forms as shown in table 5.1. Free trade agreements
eliminate internal tariff and non-tariff trade barriers but do not harmonize external barriers.

Customs unions remove internal barriers and establish a common external tariff. The

establishment of a common external tariff involves a number of stages to be completed

 

95 Rejecting and abrogating international trade commitments is costly to the governments because it sends
negative signals to other members. By harming their credibility in international markets regarding their
willingness to cooperate, the entire nation may suffer from economic isolation in the future.

96 See for more detail, Femanadez and Portes 1998, Mansfield 1997, Whalley 1998, Yarbrough and
Yarbrough 1992.

97 See Simmons 2000.

76

over time. In some cases, these stages are either not clearly defined or often extended.
Common markets are customs unions in which barriers to the mobility of labor and capital
are eliminated. Countries can also cooperate to establish a common currency and common

98 I take preferential trade

economic policies through an economic and monetary union.
agreements, free trade agreements, customs union, common markets, and monetary

(economic) union as dependent variables.

 

 

 

 

 

Table 5.1. Levels of PTAs

Types Examples

Free Trade Areas Members eliminate tariffs NAFTA

(FT As) among themselves but keep EU-CEE agreement
their original levels with Proposed FTAA
the rest of the world

Customs Unions Customs Union members MERCOSUR

(CUs) eliminate tariffs among Andean Pact
themselves and adopt a Central America
common tariff (CET) Common Market
against the rest of the world

Common Markets Members eliminate tariffs European Economic
among themselves, adopt a Community
common external tariff, and
remove impediments to
movements of factors of
production across member
countries.

Economic Unions Members move beyond the European Union
common markets to unify
their fiscal and monetary
pohcy

 

 

 

 

 

 

98 See Lawrence 1996.

77

 

Figure 5.2. PTAs in force and under negotiation as of July
2000, by type of PTA

 

FT As in force
I CU s in force
I FTAs under

negotiation
El CUs under negotiation

 

 

 

 

 

 

Cited from WTO, 2003

Free trade agreements are more prevalent than any other forms of PTAs.99 They account for
almost 90 per cent of all PTAs identified. FTAs do not provide for the complete
elimination of all tariffs and non-tariff measures between their parties. Most provide for
the elimination or reduction of existing tariffs, with varying degrees of exclusions.

PT As differ considerably in scope, and their configurations are diverse. The simplest
one is a bilateral agreement formed between two parties, but some agreements unite three
or more. More complex are agreements in which one (or more) of the parties to an
agreement is a PTA itself. This can occur either in the case of a customs union (e.g. the
agreements signed between the EC and the Euro-Mediterranean countries) or a free-trade
area (e. g. the similar web of agreements signed by the European Free Trade Association
(EFTA) (WTO 2003).

In this dissertation, PTA data is derived from the dataset of Mansfield, Milner and

Rosendorff (2002). To list PTA formations, Mansfield, Milner and Rosendorff use the list

 

99 Free-trade areas are more prevalent partly because of its flexibility. That is, members of FTAs eliminate
tariffs on trade with each other but retain autonomy in determining their tariffs with nonmembers.

78

of PT As notified to the GATT [WT 0 under Article XXIV and the Enabling Clause100 from
the World Trade Organization’s data from “Regionalism and the World Trading System
(1995).”101

The dependent variable in this dissertation is the occurrence that states form or join a
PTA102 where we observe 1 if this occurs and 0 otherwise for dyad i during year t. That is,
it is the binary indicator of whether a dyad forms a PTA in a given year. States are
considered to have joined a PTA if they form a new PTA, or if another state joins a PTA in
which they are already a member (Mansfield et al. 2002). Thus, in this dissertation, the
dependent variable is measured in terms of a policy change, not trade flow outcomes.

The decision to join a PTA is not limited to domestic political factors. As Leeds
suggests, “making conclusive statements regarding international interaction based on the
domestic politics of a single state is not complete. International cooperation is better
understood as an interaction among leaders with both domestic and international

constraints.”103 That is, decisions to cooperate by entering a free trade agreement are

 

'00 Formally, the “Decision on Differential and More Favorable Treatment, Reciprocity, and Fuller
Participation of Developing Countries” was negotiated during the Tokyo Round as Part I of a new
“Framework Agreement” on International Trade. The “Enabling Clause” legalized the extension by
developed Contracting Parties of GATT of preferences to developing countries, notwithstanding the most-
favored-nation treatment required under GATT Article 1.

'0' Because the WTO lists only PTAs formed by its members and those of the GATT, Mansfield et al.
supplement data by including other PTAs listed in other studies. They include the Council on Mutual
Economic Assistance (CMEA), the Southern African Customs Union, the Economic Community of West
African States, the Southern African Development Coordination Conference, and the Eastern and Southern
African Preferential Trade Area, since various studies have identified each of these groupings as PTAs. Each
PTA is coded as starting in the year it enters into force. Obviously, the decision to join a PTA is different
from the decision to stay in a PTA that changes status, say from a free trade area to a common market. I do
not consider agreements strengthening or superceding an existing PTA to be new preferential trade
agreements. Also, considering disaggregated types of PTAs rather than treating them identically is beyond
the scope of this study.

102 Notifications may also refer to the accession of new parties to an agreement that already exists, e. g. the
notification of the accession of Bulgaria to the CEFTA Agreement.

"’3 See Leeds 1999.

79

dependent “not only on one’s own preferences, but on one’s expectations regarding the
behavior of others” (Leeds 1999). In order to observe the effect of domestic institutional
type in the context of interactive behavior in the international system, I use a dyadic
observation as the unit of analysis (Smith 1996, Leeds and Davis 1997), as opposed to
simply focusing on the determinants of one state’s trade policy. I test a hypothesis whether
a probability of trade cooperation rises as domestic institutions change in dyads. In other
words, I am more interested showing not that democracies tend to cooperate more

monadically, but rather that they do so only with other democracies dyadically.104

5.4.2. Proxy of Winning Coalition and Selectorate Size
The measures of domestic institutions are important in the dissertation for testing
hypotheses. In order to measure the size of winning coalition and the selectorate, I draw on

the Vanhanen Polyarchy datasetws

Although Polity data is used widely in the international
relations field, its emphasis on the institutional side of democracy neglects certain aspects
of political reality such as the extent of actual participation. It also tends to take coded

features of the institutional democracies without being able to assess their actual

performance (Gates et al. 2002). Vanhanen’s (2000) dataset106 has two indicators:

 

104 For empirical studies about dyadic hypotheses, see Mansfield, Edward, Milner Helen, Peter Rosendorff.
1997. “Free to trade: democracies and international trade negotiations ”. Presented at Annual Meeting
American Political Science Association, Washington, DC, September. Dixon William and Bruce Moon,
1993.” Political Similarity and American Foreign Trade Patterns,” Political Research Quarterly, 46. Bliss,
Harry and Bruce Russet. 1998. “Democratic Trading Partners: The Liberal Connection, 1962-1989.” The
Journal of Politics 60 (4): 1126-47. Morrow James, Randolph Siverson, Taressa Taberes. 1998. “T he
Political Determinants of International Trade. American Political Science Review 92(3).

'05 Vanhanen 2000.

'06 There are also some limitations with the measurement of democratic participation. Political systems with
many parties are considered to be more democratic than systems with few parties. In other words, “the share
of the votes of the largest party in parliament may produce certain distortions favouring highly fragmented

party systems with many small parties which may be a result of the respective electoral systems, e. g. highly

80

“Participation” and “Competition”. Two theoretical dimensions of democracy - public
contestation and the right to participate - examined by Dahl (1971) seem to correspond to
those two indicators in the data set that differentiate more democratic systems from less
democratic ones. Legal competition means that individuals and groups are free to organize
themselves and to oppose the government. It also indicates the degree to which the freedom
of different groups is equal in their competition for power (Vanhanen 1992). The level of
participation is also very crucial in democracy as indicated by the relative number of
people taking part in politics. The higher the degrees of competition and participation are,
the higher the level of democracy is. I use Vanhanen’s two measurements as proxies of
winning coalition size and selectorate size. The participation variable is a useful measure
for the selectorate size, and is measured by the percentage of the population that actually
voted in these elections. This variable is prone to large variations even in the same
regime.107
The measure of winning coalition is the actual size of members of the selectorate
whose support is essential to keep the incumbent leadership in office. However, the direct
measurement of winning coalition size is more complicated. I use the competition
(Vanhanen 2000) and democratization indicator as a proxy for the winning coalition size.

Competition measures the electoral success of the smaller parties. It measures the

proportion of the votes won by those parties in parliamentary or presidential elections, to

 

proportional ones” (Gates et al. 2002). Vanhanen also does not control for systems that have no choice, but
have elections (e. g. Iraq under Hussein).

'07 It should be noted that this percentage is calculated from the total population, not from the adult
population or from the enfranchised population (Vanhanen 1992). To take the share of voters of the total
population, instead of the adult population only, causes a distortion of this index disadvantaging young
nations with a high population growth, but the author defends this problem on practical grounds based on
better availability of total population data rather than more differentiated demographic census results.

81

indicate the degree of competition in a political system. The competition measurement is
useful to distinguish political systems that have had very high participation in elections
where in fact there were really no alternatives, as in the former communist states (Gates et
al. 2001). This figure is calculated by subtracting the percentage of the votes won by the
largest party from 100 percent. If both parliamentary and presidential elections are taken
into account, the arithmetic mean of the two percentages is used to represent the smaller
parties’ share of the vote cast. Because these two variables - participation and competition -
are assumed to represent different dimensions of democratization, it is plausible to assume
that a combination of the two would be a more realistic indicator of democratization than
either of them alone. Vanhanen weights them equally in the construction of an index of
democratization. Multiplying the variables and dividing the outcome by 100 forms the
democracy index. This means that a low value of either of the two variables indicates a low

level of democratization.

5.4.2.1. Comparison between Bueno de Mesquita’s Measurement and Vanhanen’s
Measurement on Democracy

In this section, I provide reasons why Bueno de Mesquita’s measurement is not satisfactory
in order to measure winning coalition size and selectorate size. Then, I explain why the
measurement of Vanhanen outperforms indicators of the Polity data in terms of the
measurement of the winning coalition size and the selectorate. Bueno de Mesquita,
Morrow, Siverson, and Smith (1999) use the Polity data, which includes a number of
institutional variables in order to test institutional theories on external trade policy. Polity
data includes five factors that capture the institutional differences between democracies and

autocracies: 1) the competitiveness of the process for selecting a country’s chief executive,

82

 

2) the openness of this process, 3) the extent to which institutional constraints limit a chief
executive’s decision-making authority, 4) the competitiveness of political participation
within a country, and 5) the degree to which binding rules govern political participation
within it (Gurr et al. 1989, J aggers and Gurr 1995). Each of these five measures is directly
related to BDM’s measurement on political competition and the size of the selectorate.

Bueno de Mesquita uses “legislative selection” to measure the size of the
selectorate. The legislative selection variable “measures the breadth of the selectiveness of
the members of each country’s legislature”. In Polity 111, this variable is coded as a 0,
meaning that there is no legislature. 1 means that the legislature is chosen by heredity,
ascription, or is chosen by the effective executive. 2 indicate that members of the
legislature are directly or indirectly selected by popular election. Bueno de Mesquita
interprets that the larger the value of legislative selection, the more likely it is that the
selectorate is large. Bueno de Mesquita et al. add 1 to this indicator, so that it varies
between 1 and 3. They refer to this variable as a Selectorate size.

To estimate the size of the winning coalition, Bueno de Mesquita et al. construct a
composite index based on the variables Regtype, chomp, Xropen, and Parcomp in Polity
1]]. When the regime type was not military or military/civilian, they give one point to
Winning coalition size. chomp indicates the competitiveness of executive recruitment. If
it is larger than or equal to 2, then another point is assigned to winning coalition size.
chomp represents the method of executive selection. When it is coded 1, it means that the
chief executive was selected by heredity or in rigged, unopposed elections. Code values of
2 and 3 indicate greater accountability to the populace, and is interpreted it as a larger

winning coalition. Xropen is the openness of executive recruitment. They add an additional

83

 

point to winning coalition size if the executive is recruited in a more open setting than
heredity. Finally, one more point is added to the index of winning coalition size if
Parcomp, the competitiveness of participation is coded as 4. It means that there are stable
political groups, which regularly compete for political influence. Thus the winning
coalition size indicator varies between 1 and 5 (Bueno de Mesquita et al. 2000). However,
there are weaknesses in their measurement.108 One limitation is that this measure may
explain only differences between regime-types such as democracy and autocracy.
Therefore, it may not explain variations within same regime type such as US. and Britain.
Further, by using the regime type variable as one of the measures of the size of the winning
coalition, Bueno de Mesquita et al. (1999) mistakenly concludes that there are differences
in terms of policy performance level such as economic growth and political survival at the
regime level. This drawback stems from inappropriate measurement of the size of the

selectorate and the winning coalitions.

5.4.3. Electoral Rule and Veto Players

I draw on Beck, Clark, Groff, Keefer, and Walsh’s political institution database109 for the
measurement of electoral system, the number of veto player and the distinction between
presidentialism and parliamentarism. Based on this data, I test how electoral rule, the
number of veto players, and presidentialism affect the decision of a country to enter a PTA.
The number of veto players is calculated with the following criteria drawn from the

Database of Political Institutions: 1 for the executive, 1 if multiple parties are legal and

 

"’8 For instance, in majoritarian systems of democracies, the size of the winning coalition is mostly related to
the size of the selectorate due to electoral rules.

‘°° Beck, Clark, Groff, Keefer. Walsh, 2001.

84

compete in executive elections, 1 for each legislative chamber, and 1 for every veto player
whose political orientation is closer to that of the first opposition party than to the party of
the executive. For presidential systems, the veto player is the first government party. For

parliamentary systems, the veto players are the 3 largest government parties.

5.4.4. Control Variables

It is important to include other confounding variables that may be responsible for the
relationship between domestic institutions and a commitment to preferential trade
agreements. Studies110 find that political factors such as alliance positively influence the
likelihood of leaders to make a commitment to a PTA. This research suggests that alliances
may account for any effect of domestic institutional types on the propensity of signing a
tariff reduction agreement.111 That is, “alliances are likely to promote free trade because
countries may be more likely to enter PTAs with their allies when they trust the credibility
of their mutual political relationship. When a leader is concerned about the possibility that
his ally will cooperate with other opponents, he is less likely to be willing to enter a
preferential trade agreement with that ally.”l ’2 Previous studies also found that states are
likelier to enter a free trade agreement if they are members of the General Agreement on

Tariffs and Trade (GATT) or WTOl '3 and if they are not involved in a war (Dispute)“"

 

”0 Mansfield, Milner, Rosendorff, 2002 and Mansfield 1998.
1” Mansfield 1998.
“2 Morrow and Siverson, 1998.

"3 “The increase on the occurrence of countries entering PT As has been more precipitous within the GATT
than for the world as a whole” (Mansfield 1998).

85

1

(Gowa 1994, Mansfield 1998). Studies also show the effects of hegemony on the expansion
of PTAs (Gilpin 1975, Mansfield 1998). The growth in Preferential Trade Agreements
results from the decline in US. leadership and power. That is, states join PTAs more
extensively as hegemony declines. In this context, I control for the effect of hegemony
(Hegemony) (Mansfield 1998). Hegemony can be calculated as “the percentage of total
global trade (the sum of imports and exports) conducted by the state engaging in the most
commerce in year t - 1” (Mansfield 1998). The geographic distance between countries
(Distance) is also controlled for because this factor stimulates the formation of PTAs
(Mansfield 1998).

Controls for other economic conditions are also included to observe the effects of
institutional variables on trade relations. Domestic economic factors can influence a
leader’s decision to join a preferential trade agreement. For example, states with larger
home markets are less likely to be motivated to enter free trade agreements compared to
states with smaller home marketsl ’5 (Milner and Mansfield 1997, Mansfield, Milner and
Rosendorff 2002). Thus, the model includes the size of their economy (GDP). I use the
lagged per capita real GDP as a control in models. Empirical studies of trade policy also
emphasize the level of macroeconomic conditions as the cause for increasing levels of
protection. Trade is a field where decision-makers tend to face strong domestic pressures
for import protection frequently. When economic conditions are bad, domestic pressures

for protection may suddenly occur. In other words, economic conditions influence the level

 

”4 Related to war, conflict has some effects on trade relations, since leaders assess the conflict as a disruption

to enter PTAs. The more likely the conflict, the less attractive it is for leaders enter international trade
agreements (Mansfield 1998).

“5 “If the home market is large, the pursuit of scale economies through regional agreements will be less
important for firms since they may be close to efficient scale already” (Milner and Mansfield 1997).

86

I

of pressures for protection. “Recessions may lead states to join a PTA comprising countries
whose principal industries do not rival each other. Doing so is one way a government can
address domestic pressures for protection by restricting imports” (Mansfield 1998). If there
are recessions, forming a PTA rather than imposing unilateral trade barriers gives some
benefits to member countries of the GAIT/WTO. Pomfret (cited from Mansfield 1998)
argues, “if a country intends to raise Most Favored Nation tariffs or introduce import
quotas, it is both contravening GATT and risking retaliation. A bilaterally negotiated
discriminatory trade barrier is a way to sidestep GATT obligations without an open
breach.” In this sense, increasing levels of unemployment and real exchange rate influence
the level of trade policy. These two factors contribute to the satisfaction level of voters.
That is, political decision-makers should meet the demands for protection from voters
based on unemployment level and exchange rate level (Rosendorff and Milner 2000,
Mansfield and Busch 1995). Appreciation of the exchange rate may increase the demands
for protection because it increases imports and decreases exports by influencing the balance
of domestic trade preferences.”6

Despite the general trend toward global trade openness, an understanding of the
significant cross-national variations in trade policy is necessary for a complete
understanding of this important policy area. Current analyses are not effective in resolving
the puzzle as to how domestic institutions affect the way domestic preferences are

aggregated into interstate commercial cooperation. To fill this gap, this chapter presented

key domestic institutional constraints hypotheses. I elaborated on how the notions of the

 

”6 Much debate continues over the impact of the appreciation of the exchange rate. Some studies suggest that
the exchange rate does not systematically impact protectionist pressures. I omit some economic condition
variables because of data constraints.

87

size of winning coalition and selectorate provide special insights for the decisions to join
PT As. This institutional explanation provides a sound framework for understanding much
of what has previously been regarded as the puzzling behavior of nations in the rush to join
PTAs, and the theory leads to dyadic hypotheses. I also clarified how disaggregated
domestic institutions influence the motivation of leaders to join international trade
agreements. This chapter dealt with the data and measurement issues related to the analysis

in chapter 6.

88

Chapter 6. Empirical Estimations and Results

This chapter mainly provides the quantitative analyses of the dissertation’s arguments
about the impact of domestic institutional constraints on cooperative trade policy, measured
in the conclusion of international trade agreements. The key hypotheses to be tested were
presented in the preceding chapter. To recap, I apply Bueno de Mesquita et al.’s (1999)
institutional explanation to the realm of commercial policy. That is, the inherent differences
between democracies and autocraciesll7 should cause systematic differences in the trade
policy strategies of their leaders. Building on this theory, I argue that the varying sizes of
political institutions, the winning coalition and selectorate, explain the variation in
concluding international trade agreements. I also discuss the impact of other institutional
constraints on the conclusion of trade agreements, highlighting those which deserve more
attention. I present estimation results using comprehensive data from the period 1950-1992.
The last section interprets the empirical results and observations. By doing so, the chapter
shows how domestic institutions influence the decision of politicians to sign international
trade agreements. The model I develop demonstrates that leaders have greater political
incentives to conclude international trade agreements as the size of winning coalition
grows. Decision-makers also have greater political motivations to enter international trade

agreements as other indicators of the level of democracy increases.

 

”7 Most studies of trade policy focusing on domestic institutional factors “focus only on variations in policy
among democracies, and avoid comparing democracies with other types of regimes.” This dissertation’s focus
is different from other studies that emphasize how domestic institutional variations among countries with the
same regime type affect trade policy (Mansfield, Milner and Rosendorff 2000). In this sense, this dissertation
provides new insight into the impact of political institutions on trade policy by comparing different regime

types.

89

6.1. Model Specifications and Theoretical Expectations
I empirically assess the impact of domestic political institutional factors on the willingness
of leaders to commit free trade agreements across states over time. To test these

hypotheses, I construct following dyadic empirical models:

Model 1: Commitment to free trade agreements i,j = Bo + BlProxy of Winning
coalitionsize + [32 Selectorate size+ B3lagged GDP+ BaAlliance iJ +BSGATT +136
Distance i,j +137 Hegemony +B7Dispute ij + 8 1,1

Model 2: Commitment to free trade agreements i,j = Bo + BlDemocracy 1+ [3; lagged GDP+
B3Alliance ij +BaGATT +65 Distance 1,1 +86 Hegemony +B7Dispute i,j + 81d

Model 3: Commitment to free trade agreements 1; = Bo + BlProxy of Winning
coalitionsize + B; Selectorate size +133 PR+ B4Veto player+ BsPresidentialism +B6lagged
GDP+ B7Alliance 1,1 +BgGA'IT +BgDistance i,j +510 Hegemony +BnDispute ij + an

The main explanatory variables of the empirical models are the size of the winning
coalition and selectorate. These variables are mainly designed to assess the dissertation’s
basic argument. In the last empirical model, I also include domestic political constraints
such as electoral rule, the number of veto players and the structure of the political system in
order to investigate how disaggregated domestic institutional constraints channel domestic

demands and pressures in influencing interstate commercial cooperation.118 The models

also include a set of political and economic variables developed to control for mostly

 

”8 Some of institutional data relating channeling the sources of domestic preferences are not available for all
countries and times. Therefore, the empirical models incorporate some but not all of domestic constraints
variables. The size of electoral district, the number of party system (party fragmentation), party polarization,
divided government, delegation of power to presidents may also be domestic constraints in influencing the
level of formation of trade agreements. The absence of these data is regrettable given potential importance of
these measures in affecting trade policy.

90

confounding factors. Most of the explanatory variables are time varying. I discuss the

expected effects of each variable in what follows.

6.1.1.The Effect of the Winning Coalition and Selectorate

I evaluate a hypothesis as to whether a dyadic regime type with a larger size of the winning
coalition is more likely to sign a PTA. That is, whether a dyad with a larger size of the
winning coalition tends to conclude a higher level of PTAs. I expect that the propensity of
signing an agreement increase as a survival of political leaders becomes more dependent on

larger size of winning coalition and selectorate.

6.1.2. The Effect of Regime type
I test the hypothesis that there is a positive association between joint democracy and the

propensity that political leaders conclude a PTA.

6.1.3. The Effect of Veto Players119
As the number of veto players with ideological difference120 grows in a given dyad, policy
coordination of trade policy should be more difficult. As a result, the policy change

required for the formation of PT As is less likely. Yet, veto players do not play the trade

 

”9 Thanks to Prof. Mark Jones for bringing the importance of veto players to my attention.

120 Some scholars use the indicator of party fragmentation using the effective number of party in a given
country in order to investigate the effect of uncertainty or policy change inflexibility from the structure of
party system in affecting trade policy provision. For example, Neilson (2003) argues that, “stable party

system may lead to the enhanced level of trade reform by reducing the level of uncertainty.” Yet, illustrations
of successful pursuits toward free trade of the small European countries with many small parties show that the
effective number of party system is not an effective determining factor on the trade reform issue. In this sense,
I adopt a veto player as potentially more important explanatory variable.

91

policy game alone. Even a small number of veto players may be insufficient to bring about

policy change. Leaders should also be motivated to conclude a PTA.

6.1.4. The Effect of Electoral Rule121

It is often argued that tariff reduction orientation is more predominant in PR systems.
Proportional representation is more likely to reflect the interests of larger blocs of voters
and this leads to an enhancement of free trade policy. On the other hand, disproportional
representation-oriented systems such as those based on plurality rule are more likely to lead
to trade policies which favor small blocs of voters. Therefore, it is likely to cause the rise of
protectionism. I evaluate the hypothesis that the formation of PTA is more likely in PR
(Proportional Representation) systems compared to other forms of electoral rule such as
plurality rule. That is, the propensity to commit to PT As increases in countries with the
electoral rule of PR. However, the literature has not conclusively shown that PR systems
are more effective in cooperating with other countries in liberalizing trade. Though
electoral rule such as disproportionality may allow “protectionists forces to operate, it
neither determines their existence nor their success. In fact, we see countries with highly
disproportional electoral systems such as the US. and Britain liberalized their trade early in

their democratic histories. Furthermore, they implemented liberal trade policy even during

 

121 How insulated political decision-makers are from domestic and societal pressure makes a critical
difference in terms of affecting the responsiveness of leaders to social demands on trade policy choice. In this
regard, the impact of electoral system (or constituency size) on trade policy is one of important topics in the
trade policy literature. Electoral system is considered as an important intervening variable that shape the
interests and preferences of leaders to the demands of domestic pressure groups. It may affect under which
conditions different sectors have more opportunities to implement their preferences in the process of trade
policy making. In this context, electoral rule can be called as a “method of preference aggregation” (Garrett,
and Lange 1995).

92

economic hard times when protectionists pressures were very high. This result shows that

other factors are in play beyond electoral disproportionality” (Nielson 2003).

6.1.5. The Effect of Presidentialism

I assess whether there is a positive association between presidential system and a
commitment to sign international trade agreements. In other words, I test the hypothesis
that the formation of PTAs is more likely in presidential political systems.122 Again, the
literature on this has not shown that presidential systems are more effective in terms of
commercial cooperation with other countries. It is also possible that presidents are more

likely to be biased to protectionist trade policy.

6.2. Methodology: GEE method and Weak Link Analysis

The most common techniques for analyzing statistical models with dichotomous dependent
variables are logit and probit. However, logit and probit techniques require assumptions of
independence among cases that are inappropriate in a time-series cross-section context.
Given that the dependent variable is comprised of the decision to join PT As over time,

(1.123

there must be autocorrelations from year to year per dya To account for this

characteristic of the data, I rely on a generalized estimating equation (GEE) to analyze the

 

122 Yet, the theoretical basis of the argument is weak. No determinate direction can be made in comparing
presidentialism and parliamentarism in influencing the level of the formation of Preferential Trade Agreement.
Therefore, arguments about association between presidential political system and free trade policy tendency
should be made with caution.

123 It is notable that many of PTAs are interrelated. “The countries of the European Union and Mexico belong
to more than 10 agreements. Brazil, Colombia, Venezuela, Chile and some Central American countries
belong to between 5 and 10 agreements. Most other WTO members belong to at least one agreement. The
main exceptions to this pattern are Hong Kong, China, Japan, Macau and Mongolia” (Crawford and Laird
2000).

93

data. Liang and Zeger formalized a Generalized Estimating Equations (GEES) approach to
deal with correlated data involving repeated observations over time. According to Zorn,
the GEE technique is appropriate to use “when the standard assumption that observations in
the data are conditionally independent is called into question?”24 GEE analysis uses quasi-
maximum likelihood estimation techniques to control the effect of time dependence on
observations. To correct the temporal grouping of the dependent variable, cubic splines can
be included (Beck, Katz, and Tucker 1998).”5 As a direct extension of the Generalized
Linear Model (GLM), GEE is a quasi-likelihood method that provides information about
the relationship between the expected value of the dependent variable and the covariates in
the context of time-series cross-section data. The GEE method deals with panel data with a
binary or more complicated dependent variable. GEE considers the dependence of
observations as a nuisance, which has its primary impact on estimated standard errors. GEE
models are a quick way to deal with many problems posed by longitudinal data if we do not
care about the error structure. There are other advantages of the GEE approach for
estimating models with correlated data. These models incorporate the consideration of
within-unit interdependencies through specification of the working correlation matrix.
However, GEE can provide asymptotically unbiased estimates of the parameters of major
interest even when we do not know the exact nature of the intra-cluster dependence. The

parameter estimates are robust to misspecification of those correlations (Zorn 2001).

 

m Zorn 2001.

'25 The splines are designed as a corrective for assuming that observations are time-dependent in the time-
series cross-sectional framework. However, in this study, I use GEE method instead of using cubic splines
approach. It should be noted that using cubic splines as a cure for time dependence effects, does not make a
difference in terms of estimation results.

94

Also, given that my primary interest is in understanding how variance in domestic political
institutions affects the average probability of PTA formation across a population, I use the
GE method, because it is a population-averaged approach. It is particularly well-suited to
examining differences among groups identified by particular characteristics. In this study, I
attempt to estimate the general tendency of countries according to their domestic
institutional types to conclude PTAs. In this context, a marginal approach such as GEE is
more appropriate to the nature of the main question. If one is more interested in the effect
of domestic institutional constraints for particular states to join PTAs, then the conditional

approach is more appropriate (Zorn 2001).

Weak Link Analysis

I run the analysis with an indicator for each variable, ascribing the weak link principle. In a
dyadic analysis, it is difficult to measure major concepts and variables. For example, joint
democracy is a difficult concept to measure. Some studies have used a dummy variable,
which takes on a value of 1 when both states’ democracy scores are above some threshold.
While others have chosen the lower of the two democracy scores as sufficiently reflecting
the level of democracy needed for the dyad. I select the “weak link” procedure of relying
on the lower of the two democracy scores. Recently, it has become much more popular to
use the separate variables such as high democracy and low democracy that identify the
higher and lower of the two democracy scores in a dyad. This approach was first used by
Dixon based on the weakest link assumption that suggests that the democracy level of the
lower democracy state would drive the regime type relationship in any dyad. There are

variations on the inclusion of variables based on the weakest link argument. Some studies

95

include both high democracy and low democracy (Gartzke 2000, Mousseau 1998, Oneal
and Russett 1997), while others use low democracy alone (Beck, Katz, and Tucker 1998,
Mousseau 2000, Oneal and Russett 1999).

Following Oneal and Russett’s logic of the “weakest link in the chain,” I identify the
lower value of winning coalition size in the two countries as the level of winning coalition
size. That is, I identify the minimum constraint that the size of winning coalition places on
the policy coordination with regards to PTAs within the dyad. I also identify the lower of
the two democracy scores as the level of democracy in the dyad. For the size of veto player,
I choose the lower of the veto player numbers in the dyad. For the dummy variables, such
as proportional representation system and presidential system, I choose the lower value in
the dyad. That is, I choose one PR system instead of choosing two PR systems in the dyad.

I use the same method to identify the lower value of the presidential systems.

6.3. Estimation Results

In this section, I present the estimation results of the models by testing whether there is
systematic empirical evidence to support the dissertation’s arguments drawn from domestic
institutional theories. Analysis is divided into two sections. The first section focuses on the
impact of regime type constraints such as winning coalition size and selectorate to
interstate economic cooperation. It shows whether leaders display greater political
incentives to conclude trade agreements as the level of democracy grows. The impact of
disaggregated institutional constraints will be also analyzed. Such dynamics include the
number of veto powers, electoral system, and parliamentarism vs. presidentialism. In the

next section, I focus on variations among democracies. I use the core dataset of Mansfield,

96

Milner and Rosendorff’26 for dependent variable and control variables. I evaluate
hypotheses using pooled cross-sectional time-series data with a unit of analysis of non-
directed dyad-years from 1950 to 1992. Each dyad year is a single unit of observation.

Models are estimated by the GEE method as described above.

Descriptive Summary Statistics on PTA Formation Database
Table 6.1. highlights the relevant characteristics of the variables. It gives a general grasp of

the database.

 

'26 I am very thankful to Edward Mansfield for making their data publicly available. Mansfield, Milner,
Rosendorff, 2002.

97

Table 6.1. Descriptive Summary Statistics on PTA Formation Database Variables

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Variable Obs Mean Std. Dev. Min Max
Statea 254345 295 .7437 210.6671 2 987
Stateb 254345 594.968 218.9418 20 990

II) 254345 8161.483 4985.115 1 17194
Year 254345 1976.468 1 1.5287 1950 1992
Alliance 254345 . 1030136 .3039773 0 1
Distance 250626 8.237138 .8114752 1.871744 9.422935
HegemorLL 254345 . 1397812 .0096814 . 1227 . 1669
Colonial 254345 .00395 13 .0627354 0 1

History

Trade 231693 1754677 1994991 0 l.48e+08
PTA 240223 .0046374 .0679402 0 l

GATT 254345 .27 36087 .4458 1 l3 0 1
Regimea 131690 12.84156 7.782859 1 21
Reiimeb 131690 12.84214 7.789742 1 21
GDPa 131690 l.35e+08 4.02e+08 172360 4.58e+09
GDPb 131690 l.36e+08 4.05e+08 169328 4.58e+09
AGDPa 129100 4605731 1.66e+07 -2.02e+08 2.62e+08
AGDPb 129092 4617543 1.66e+07 -2.02e+08 2.62e+08
Dispute 254345 .0049657 .0702927 0 1
Selectoratea 171640 39.22086 16.88284 .1 74.6
Selectorateb 152883 35.94135 18.31392 .1 74.6
Compa (Wa) 171640 37.72855 23.48412 0 70
Compb (Wb) 152787 30. 10218 25.40498 0 70
Democracya 171640 15.10977 12.36155 0 47.11
Demorcrayb 152787 11.58085 12.60063 0 47.11
PRa 254345 .3987537 .4896429 0 l

PRb 254345 .2294325 .4204687 0 1
Presidentialisma 2465 19 .5 108044 .4998843 0 l
Presidentialismb 254345 .1398691 .3468518 0 1

Vetoa 88760 2.340412 1.355107 1 13
Vetob 75851 2.175581 1.516856 1 l3
Vetol 87102 3.047002 1.859145 1 16
Vet02 74416 2.551212 1.884678 1 16

 

The results of estimation of models are shown in Table 6.2. Results from the model

show that, except for a few variables, the estimation results are consistent with theoretical

expectations.

98

 

*** Table 6.2. About Here ***

6.3.1. The Size of Winning Coalition and Selectorate

l128 of the effects of

Table 6.3. presents the results127 of the population-averaged mode
domestic competition level (the proxy of winning coalition) on the determination of leaders
to commit to policy change for the formation of international trade agreements. The lower
winning coalition size has a strong positive effect on the commitment to tariff reduction
agreements. Winning coalition size in the model is statistically significant at the 0.001
level. In other words, the propensity of the determination of leaders to join Preferential
Trade Agreements increases as the level of winning coalition size increases. On the other
hand, the size of the selectorate shows a negative sign in a dyad on the policy change
regarding PTAs. It is not statistically significant. The theoretical implications of selectorate
size are less clear according to Bueno de Mesquita (2000).129 Accordingly, the results
show that trade policy performances with regards to PT As are associated with the lower

value of winning coalition size (measured as political competition level). These results are

consistent and robust across models with different specifications. To sum up, the level of

 

127 I use the command as xtgee ................... Family (bin), link(logit), robust

’28 Random-effects estimators (or other cluster-specific estimators) estimate the model: P(Y_ij=l | X_ij, u_i)
= F(X_ij b + u_i) whereas population-average estimators estimate the model: P(Y_ij=1 |X_ij) = G(X_ij b*).
Here, b and b* are different population parameters although in practice, b and b* are often close. Note, “the
population-averaged model does not fully specify the distribution of the population. The cluster-specific
model does specify the distribution (u_i is either given a distribution -- i.e., a random-effects model -- or is
considered fixed like X_ij -- i.e., a fixed-effects model). The population-averaged model only specifies a
marginal distribution. Hence, the term marginal is often used for GEE estimates. The crucial point is that
when using the GEE estimator, we should think about whether the averaging procedure makes sense for the
type of inference that we want to make. For instance, if we want to estimate how marriage makes a person get
their act together and get a job, then we want to go after b. If we want to look at employment for the average
married person compared to the average unmarried person, then we choose population average estimator”
(cited from stata.com).

'29 The effect of selectorate size in this sense is not as clear and more difficult to interpret.

99

competition (the size of winning coalition) positively influences the commitment of the

leader to the pursuit of international economic cooperation.

6.3.2. Democracy

A lower democracy score increases the likelihood of commitment to PT As. A lower
democracy score continues to have a strongly positive effect on the propensity of
commitment to PTAs. Table 6.3. confirms that there is a positive effect of low democratic
regime type, which is also statistically significant and robust across models. That is, the
propensity of commitment to sign PTAs increases between democracies.

There should be variations on the leaders’ decision on the formation of PTA
depending on other institutional constraints.130 That is, the likelihood of a leader choosing
to sign a PTA should differ even in the same regime type. In the preceding chapters, 1 show
that domestic institutional constraints such as the number of veto players, electoral system
and the structure of government play a role in mediating preferences of leaders in terms of
international economic cooperation. I present estimation results of the impact of
disaggregated political institutions on interstate economic cooperation using the weak link

function.

6.3.3. Veto Players
The number of veto players is negatively associated with the formation of PTAs across

models. The likelihood of signing PTAs decreases as the number of veto player increases.

 

13° Variation in autocracy should be considered. However, there will not be much difference in autocratic
countries in terms of electoral rules and other disaggregated aspects of political institutions compared to
democracy. Even if there may be an effect, it is less clear to discuss the effect of variation of institutional rule
in autocratic system. Because we know that this institutional difference may be spurious in the first place.

100

In other words, international commercial cooperation on preferential trade agreements is
more difficult to achieve as the number of veto player rises in political systems. The
number of veto player is an important determinant of Preferential Trade Agreement

formation.

6.3.4. Electoral Rule

The models with weak function application suggest that electoral rules do not influence the
likelihood of joint trade agreement cooperation. The estimation results indicate that there is
a positive association between PR system and trade policy change for tariff reduction
agreements across models. Put another way, the possibility of commitment to Preferential
Trade Agreements increases, as the level of the PR system increases. However, as
mentioned earlier, electoral rule such as PR neither determines the existence nor success of
interstate commercial cooperation. Electoral rule such as PR may allow the higher level of
insulation of leaders from the pressures of interest groups and may force to pro-trade
tendencies to predominate. Countries such as US. and Britain with plurality electoral
systems liberalize their trade with international trade agreements more extensively. This
result shows that other factors are more prominent in influencing interstate commercial

cooperation beyond electoral rule.

6.3.5. Presidential Political System
The models suggest that presidential a political system significantly influences the
propensity for joint PTA formation. The estimation results show that the direction of

estimates of this variable in the dyad is negative. It means that the possibility of

101

commitment to join Preferential Trade Agreements decreases, as the level of presidential
systems increase.

*** Table 6.3. About Here ***
Control variables
The control variables behave as expected and are unaffected by introducing weak function.
The results indicate that the economic size of countries affect the likelihood that leaders
sign PTAs. The estimates are negative and statistically significant. It means that as the
economic size of country increases, the propensity of leaders to join PTAs decreases.
Economic trade level and dispute are not statistically significant. Yet, geographically close
states are more likely to enter into PTAs. Geographic proximity is statistically significant
across models. Colonial history has a positive effect on the commitment to join PTAs and
statistically significant. Hegemony is negative and uniformly statistically significant across
models. Alliance has a positive effect on the commitment to join a PTA. The GATT
variable shows a positive sign. Countries that are parties to the GATT are more likely to

sign a PTA, but this variable is not uniformly statistically significant.

6.4. Variations among Democracies

In fact, variations of domestic political institutions in autocratic political systems may not
appropriately reflect the reality of their political environment. Even if they establish the
rules and laws, it is less clear that these rules function in an appropriate way. In addition, it
is not explicit that autocratic systems and dictatorships have much variation in terms of
electoral rule, the number of veto player, the structure of political system and so forth. In

contrast, within democracies, institutional structures channel and mediate divergent

102

interests to the policy-making process more explicitly than do autocratic systems and
dictatorships. In this regard, the findings regarding the impact of disaggregated political

131 which contain the mixture of democracy

institutions on trade policy including all dyads
and autocracy, may be less clear. In order to resolve this potential limitation, I estimate
models only focusing on OECD member countries. Results are almost consistent with the
previous estimations results. Winning coalition size in the model is statistically significant.
The propensity of the determination of leaders to conclude trade agreements increases as
the level of winning coalition size increases. '32 I also find that the propensity of
commitment to trade agreements increases as democracy increases. Regarding
disaggregated political institutions, the commitment to conclude PTAs increases as the
level of the PR system increases. As the number of veto player increases between states,
the propensity of leaders’ motivation to form international trade agreements decreases. The

possibility of commitment to join PT As decrease as the level of presidential system

increases.

 

131 The size of data is decreased because of limited available political institutions data even before 1 drop all
dyads except OECD member dyads.

'32 This finding highlights the limitation of Bueno de Mesquita’s measurement of winning coalition size. As
mentioned before, their measure may explain only differences between regime-types such as democracy and
autocracy. It does not explain variations within same regime type such as US. and Britain. However, the
finding of this dissertation shows that domestic institutions such as the size of winning coalition explain the
variations of trade policy even only within democracies.

103

Table 6.2. Political Institutions and the Leader’s Determination to Join P'I‘As (Dyadic

analysis) GEE Population-Averaged Model with Weak Link Principle

 

 

 

Dependent Variable: PTAs Model
Independent Variables
Model 1 Model 2 Model 3
Low Winning Coalition Size 0100425" .0068019***
(.0017992) (.0018708)
Low Selectorate -.0025331
(.002338)
Low Democracy .0170693**
(.0040256)
Low PR .0868132
(.0759712)
Low Presidentialism -.7453207***
(.0872658)
Low GDP -4.65e-09** -4.78e-09** -5.04e-09***
(l.30e-O) (l.30e-09) (1356-09)
Trade -7.60e—08 -7.08e—08 -8.44e-08
(7.03e-0) (6.65e-08) (7.52e-08)
GATT 5604896“ .5703853’” .4612153**"‘
(.082915) (.0830114) (.0849171)
Alliance 5444613“ 5437023" .5998672***
(.084974) (.0841387) (.0861073)
Distance -.7262246** -.7474049** —.7122335***
(.0389114) (.0403613) (.0442127)
Colonial history 1.121388M 1.103073" .9722216***
(.2462911) (.2533122) (.24175)
Hegemony -35.60616** -36.78196** -38.04256***
(3.980427) (4.000233) (4.055161)
Dispute -.5675965 -.6165609 -.4183728
(4448603) (.4478423) (.4315654)
Constant 5.19899*** 5.506632*** 5.707431***
(.6768774) (.6931587) (.716907)
P 0.0000 0.0000 0.0000
Wald test chi2 869.41 861.42 873.05
= 128199 128199 128199

 

 

*p < .05; **p < .01; ***p <.001
Huber-White robust standard errors allowed for clustering on each dyad
Robust Stande Errors for Coefficients in Parentheses

104

Table 6.3. Political Institutions and the Leader’s Determination to Join PTAs (Dyadic
analysis) GEE Population-Averaged Model with Weak Link Principle

 

Dependent Variable: PTAs

 

Independent Variables Model

Low Winning Coalition Size .024606l***
(.003838)

Low Selectorate

Low Democracy

Low Veto Player -.1306064*
(.0624248)

Low PR

Low Presidentialism

Low GDP -1.95e-09
(1.32e-09)

Trade -4.24e-08
(2.46e-08)

GATT .1916965
(.1397152)

Alliance .0179184
(. 1595204)

Distance -.768601***
(.06295 18)

Colonial history 1.121388
(.2462911)

Hegemony -24.4000***
(8.054606)

Dispute -.4878087
(.7200016)

Constant 3.496831* **
(1. 196047)

P 0.0000

Wald test chi2 277.52

N = 65921

 

 

*p < .05 ; **p < .01

Huber-White robust standard errors allowed for clustering on each dyad
Robust Standard Errors for Coefficients in Parentheses

105

6.5. Comparison with Alternative Approaches

As an important alternative to the domestic institutional arguments above, I explore
alternative explanations, which stress other systemic factors that are conducive to the
formation of PT As. These structuralist analyses include the realists approach, hegemonic
stability approach, neo-liberal approach and power transition theory. The goal of this
comparative analysis of alternative explanations is to compare those which focus on
systemic international factors with those under the domestic institutional approach. It is
also designed to reveal the independent effect of varying degrees of domestic political
institutional constraints on commercial policy cooperation. I briefly sum up each argument

and then present the estimation results in comparison.

6.5.1. Neo-Realism and Hegemonic Stability Theory

The neo-realist approach highlights systemic characteristics such as the distribution of
power among states as a key factor influencing international outcomes. The main argument
is that the relative gains from trade prohibit international cooperation (Waltz 1979, Grieco
1988). They argue that the relative disparity in the capabilities of states is likely to hinder
the development of international trade agreements, while the relative equality of
capabilities may foster the development of such trade arrangements. That is, the level of
international trade agreements is higher between countries when their economic sizes are
relatively similar. Also, Gowa and Mansfield (1993), and Gilpin (2001) argue that
interstate commercial cooperation is more likely among states that are political and military
allies than among enemy states. Alternatively, hegemonic stability theory emphasizes the

importance of hegemonic power. Proponents of hegemonic stability theory argue that a

106

hegemonic power can promote liberal international trade by overcoming the collective
action problems. Gilpin (2001) argues that the existence of hegemonic states committed to
economic integration is the key to successful economic cooperation. In sum, the existence
of hegemony can promote the reduction of trade barriers in general, while the erosion of
hegemony may cause a rise in protectionism. In contrast, some studies suggest that the
decline of hegemony may stimulate the expansion of PTAs (Mansfield 1999).

HI .° International tariff reduction agreement is more likely among states that are military
allies than among enemy states.

H2: The erosion of hegemony may increase the likelihood of the formation of PTA.

6.5.2. Nee-liberal Institutionalism

The major argument of neo-liberal institutionalism is that international institutions can
promote international cooperation by resolving collective action problems. By enlarging
the shadow of the future and increasing enforcement of cooperation, international
institutions facilitate international commercial commitments by making them more
credible.133 Keohane (1986) also claims that international regimes may solve problems that
are difficult to manage at the national level. International institutions may increase the
possibility that states enter into international free trade agreements.

H3: The GATT positively afiects the likelihood of formation of Preferential Trade

Agreement by fostering information transparency and commitment.

 

'33 See Axelrod and Keohane. 1986. Martin 2000. Simmons 2000.

107

6.5.3. Power Transition Arguments

Like neo-realist theories, power transition theory134 is a power-based theory. It focuses on
power structure of states in the international system. Power transition scholars also stress
the satisfaction with the status quo between countries. Power transition occurs when a
challenger approximates and exceeds a predominant power in capabilities (Organski 1968,
Organski and Kugler 1980). Power transition theory suggests that the risk of war increases
when there is a power overtaking between challenger and dominant powers and if a
challenger is dissatisfied with the status quo. Yet, a peaceful transition can also occur if
both powers are jointly satisfied with the status quo. By using this theory to analyze
international commercial cooperation, scholars!” test if the level of free trade agreements
improves when the power of a challenging state exceeds the power of the preponderant
state and they are jointly satisfied with their relationship. They argue that the development
of international tariff reduction agreements improves between two satisfied powers.
Therefore, as the distribution of economic powers between countries comes closer, states
are more likely to enter negotiations on economic cooperation if both countries jointly find
it in their interest to do so. This may occur because states with comparable power may
pursue a strategy of economic cooperation in order to increase their competitive powers
outside the region. It can also occur because it is in the interest of the declining economic
power to constrain the future behavior of the rising challenger through economic
cooperation. Yet, if countries do not have similar interests and institutions, it is hard to sign

a free trade agreement (Efird and Genna 2002). In this sense, scholars argue that economic

 

’34 See Organski and Kugler. 1980.

’35 For detailed review, see Efird and Gaspare 2002.

108

cooperation is more likely among countries that share similar foreign policy preferences.
By measuring countries’ foreign relations preference profiles by alliance portfolios,
scholars argue that the level of free trade agreement is higher between countries as their
preferences converge.

H4: International tariff reduction agreement is more likely to occur when economic powers

of countries become approximate and they are satisfied with the status quo. ’36

6.5.4. Interpretation of Alternative Explanations

This section will evaluate these different arguments based on the findings of the study.

6.5.4.1. Structural approach: Neo-Realism, Neo-liberal Institutionalism, Hegemonic
Stability theory, and Power transition theory

The findings support the arguments of realist theory. Alliance has a positive effect on the
commitment to sign PTAs. Estimation results show that it is statistically significant. Also,
the decline of a hegemonic state will increase positive incentives for tariff reduction
cooperation. That is, the erosion of hegemony has a positive effect on the commitment to
join PTAs which is statistically significant. Meanwhile, the dyadic distribution of power
does not have a significant impact on trade policy change outcomes. Liberal
institutionalism predicts that international institutions positively affect cooperation on tariff
reduction agreements. The results support this hypothesis. Members of the GATT are more

likely to join a PTA even if it is not uniformly statistically significant. GATT is shown to

 

136I omit this hypothesis regarding the effect of similar preference in explaining the expansion of PTAs due to
data limitations.

109

be an effective tool in promoting trade cooperation.137 Structural international constraints
matter in international outcomes. In this regard, the structural approach, which stresses
external factors of decision-making, should not be downplayed. Yet, in reality, states often
choose trade policies based on domestic reasons. The structural approach downplays the
importance of the interplay between domestic constraints and international constraints.
Domestic leaders attempt to maximize the chances they have of holding on to political
power, and this is often at best only tangentially related to international factors. In addition,
domestic institutions inevitably constrain the responsiveness of decision-makers. Therefore,
leaders’ motivations and capacity to form Preferential Trade Agreements will be influenced
by domestic political pressures and institutional constraints. Accordingly, considering a
dynamic interplay between external factor and domestic factor is important.

To sum up, this section shows that the dissertation offers a better explanation by
providing a bigger picture of the trade policy decision-making process than to the structural
approach, which also includes neo-realist and neo-institutional explanations. Structural
perspectives such as neo-realist theories are silent about domestic sources of states’
preferences and institutional constraints regarding the recent expansion of tariff reduction
agreements. Given that each state’s domestic constraints are expected to play a crucial role
in the commercial policy formulation process, we should address how preferences and
domestic institutions matter in influencing trade policy. Exploring the interplay between
domestic politics and international relations is now being developed in various ways in the
international relations field. Indeed, this topic should be an important focus of international

political economy in the near future. Yet, we currently lack a specific theoretical

 

137 While liberalism identifies the systemic conditions under which trade liberalization can occur, it implies
that the incentive of defection prevents states from cooperating in the short term.

110

mechanism to account properly for the dynamic process of trade policy development. By
grasping how domestic politics and international outcomes are intertwined using a
quantitative analysis with large N dataset, this dissertation demonstrates that separating
internal factors and external constraints is not sufficient to provide an accurate picture of

the changes in trade policy.

6.6. Interpretations and Implications
The domestic institutional dynamic of the expansion of PTAs has received little attention in
the political science literature. In this context, providing a theoretical mechanism to explain
how decision-maker’s preferences can be constrained through political institutions in the
dynamic process of trade policy has significant implications. I attempted to delineate the
institutional contexts in which political leaders will be motivated to pursue international
trade agreements in the empirical section. Key hypotheses were formulated to investigate
the interplay between domestic preferences and domestic institutions to explain the
variations of preferential trade agreements. Among others, I focus on the effects of
institutions such as the size of winning coalition and the selectorate on the likelihood of
political leaders concluding a PTA. I also examine how other domestic institutions mediate
the policy preferences of domestic actors, since we must assuming that political decision-
makers respond to constraints. By grasping how disaggregated institutions may constrain
uncertain political competition between leaders, veto players and pressure groups, I aim to
provide a better understanding of interstate commercial cooperation.

I find that winning coalition size is significantly associated with a leader’s motivation

to form international trade agreements. That is, leaders have greater political incentives to

111

conclude trade agreements as the size of winning coalition grows. The probability of
signing an agreement increases when the survival of political leaders is dependent on the
support of a larger winning coalition. The results also suggest that democratic leaders are
more likely to form trade agreements with each other. I also find that government structure
and the number of veto players have an impact on interstate commercial cooperation.
Based on these empirical results, this dissertation demonstrates that domestic institutional
constraints provide an important mediating motivation for leaders to pursue PT As. That is,
domestic constraints can clearly influence policy action at the international level. The
dissertation suggests that not only does the domestic regime type influence the likelihood
of leaders to conclude PT As, but also that further disaggregated domestic institutions affect
this possibility. In doing so, the dissertation highlights the fact that institutional analysis
makes a significant contribution towards our understanding of international political
economy, especially in explaining how and which political institutions are most conducive
to a leader’s decision to commit to international trade agreements. Although many
economists claim that politics are not relevant in international economy, I have shown why
this is not the case. The formulation and implementation of trade policy involves a
“political decision” (McGillivray 1997), and the interplay between domestic institutions

and leaders’ preferences cannot be disregarded.138

 

’38 The political process on the recent expansion of PTAs should be explored furthermore with more
sophisticated political institutions measures. 1 hope that I could provide a more coherent picture of the impact
of a wide variety of political institutions on the formation of trade agreements.

112

Chapter 7. Conclusion

This chapter concludes the dissertation. The first section assesses what the dissertation has
accomplished by recapping the motivating questions. After reviewing the empirical tests, I
address some limitations of the dissertation. I reassess some implications of the argument
regarding the role of domestic politics in international trade and theories of international

relations more generally.

7 .1. Recapping the Motivating Questions and Contributions of the Dissertation
Since the 1980, there has been a marked revival of Preferential Trade Agreements and the
establishment of new ones. However, despite the popular resurgence of PTAs in the
international trading system, the recent rush to PTAs has received scant attention in the
literature. The global emergence of the “rush to free trade” (Rodrik 1994) is an interesting
puzzle. As trading arrangements have spread and enlarged over the last decade, intellectual
challenges have been posed. This dissertation began with motivating puzzles: what
explains the dramatic growth of PT As? Why have many countries determined to form
PTAs and why have they done so at this historical juncture? Why have they preferred PTAs
to unilateral liberalization and multilateral liberalization? Most importantly, in this
dissertation, I ask whether domestic institutional constraints across countries make a
systematic and measurable difference in the number of PTAs. I started from the
observation of scholars that the expansion of democracy and the recent rush to free trade is
not coincidental.

To address these puzzles, this dissertation uses an institutional perspective on the

dynamic linkage between domestic politics and international trade policy. The dissertation

113

makes a number of contributions to our understanding of the recent rush to PT As. First,
few researchers have examined whether countries’ political institutions influence
international trade relations (Mansfield et al. 2002). This dissertation provides a compelling
answer by showing how domestic politics affects the supply side of trade policy. In order
to explore the recent revival of PTAs, many scholars have focused on the demand side of
trade policy based on preference-oriented theories. With this approach, theories stress the
choice of trade policies to the distribution of individual preferences in the economy.
Therefore, theories emphasizing how domestic political institutions may constrain the
incentives and preferences of domestic political actors and policy-makers in shaping policy
outcomes are required, and as of yet underdeveloped. In this dissertation, I emphasize that
it is necessary to combine both preferences and institutions more appropriately into trade
policy models by explaining the dynamic aspects of the conclusion of PT As. Based on the
assumption, which shows “institutions constrain the choice set by defining what
preferences are expressed and then aggregated to determine a policy,”139 I show a dynamic
process of trade policy by demonstrating how preferences are translated into outcomes
through political institutions. 1 demonstrate that domestic institutions constrain the
incentives of politicians to pursue particular trade policies. Additionally, most studies of
trade policy focus on variations of trade policy among democracies. They do not compare
democracies with other types of regimes such as autocratic systems (Mansfield et al. 2002).
In this context, this dissertation is distinct from other theories of trade policy. In terms of
the recent interest about the effects of regime type on various aspects of foreign policy and
commercial policy (Geddes 1998, Leeds and Davis 1999, Mansfield et al. 2002), systematic

analysis of the links between regime type and trade policy is a challenging task. This

 

'39 Boix and Adsera, 2003.

114

dissertation identifies how politicians may manifest their particular preference in different
regime type and how this is linked to a decision to sign tariff reduction agreements as
domestic institutions change. Few studies directly assess the effect of disaggregated
domestic institutions on trade policy change. I have argued that political decision-makers,
as constrained by institutional arrangements, have varying levels of responsiveness in
shaping their incentives in terms of trade policy. For example, the dissertation shows that

40 may lead to a misleading conclusion

ignoring the role of veto players such as legislaturesl
in exploring the domestic institutional constraints of trade policy.

This dissertation is also distinct from other theories of trade policy in choosing a dyadic
analysis. In terms of trade policies, states do not act unilaterally. Instead, decision-makers
strategically interact with other leaders. In order to understand mutual cooperation in terms
of economic trade policy, it is necessary to observe the effect of strategic interaction'“
among leaders (Mansfield 1998, Bueno de Mesquita 2000, Leeds 1999). That is, leaders’
policy decisions regarding tariff reduction agreements are influenced by leaders’
expectations about other actors as well as the leaders’ own preferences (Leeds 1999). I
highlight the fact that any PTA formation involves a joint decision. Most studies of
endogenous trade policy underestimate the interaction between states. In particular, the
influences of domestic political characteristics of nations on these interactions, like the

level of democracy, are often underestimated. In this sense, the dissertation argues that

analyses focusing on single states may be potentially misleading in the study of trade

 

140 Proposed agreements on the conclusion of PTA often need to be approved by domestic political players
such as legislatures or other kinds of potential veto players. Yet, “it is not always clear which proposals will
be ratified at home and which will be rejected. Uncertainty often prevails”. For a more detailed review of this
point, see Reinhardt 1996.

141 For an excellent review on strategic interaction of PTAs, see Mansfield 1998.

115

policy. In this context, choosing a dyadic analysis is a unique feature of the dissertation
which accounts for the importance of strategic interaction between states in trade policy.
Considerable variation across countries regarding the conclusion of PTAs may be of
greater interest to politicians in terms of policy implications. Yet, systematic cross-national
comparisons on this topic have been rare. Cross-country empirical work has been
insufficient because of a lack of data on the political and economic characteristics of
countries over time. Instead, case studies with small samples are prevalent, but this limits
researchers who wish to generalizing their results. This dissertation performs a large-n
empirical test of the argument. With this advantage, the dissertation provides a general
picture of joint countries’ trade policy decisions over time. The statistical results
significantly support the dissertation’s argument about the effects of domestic institutions

on interstate commercial cooperation.

7.2. Limitations of the Argument

1 have attempted to identify how political leaders may manifest their particular preferences
in different regime types and how that is linked to their decisions to join tariff reduction
agreements. I have attempted to fill the gap in the literature by conducting more systematic
empirical analyses of the relationship between domestic institutional constraints and the
recent explosion of PTAs. I have demonstrated that joint democracy is more likely to lead
to the signing of a PTA. I also find that there is a significant impact at the disaggregated
level of institutions on the propensity of leaders’ decision to sign a PTA. Yet, although this
dissertation attempts to contribute to our understanding how political institutions can

structure the supply side of trade policy performances related to PTA formation, there are

116

potential limitations. First, empirical tests of the effects of domestic political institutions on
the change of trade policy confront some difficult problems. It is still questionable why
democracy as a regime type itself fosters tariff reduction cooperation among countries. We
need to explore further the effect of political institutions within democracies and
autocracies on interstate economic cooperation. Institutions are complex. Moreover, the
impact of a particular set of institutions can often be varied depending on the context. Also,
legislative systems and party systems with varying degrees of power vis-a-vis other
political actors may influence trade policy in a divergent way. For example, party
polarization may lead to different results in evaluating the effects of political institutions on
the change of trade policy. Even in presidential systems, the legislative powers of the
president (especially in the degree of delegated authority) in trade policymaking may make
a difference in terms of trade policy outcomes. The outcome of trade policy can be
different depending on the level of control legislatures have over trade policy.142 When
these variables are more clearly measured, we can understand the political influence of
domestic institutions on trade policy-making. Therefore, it will be interesting to explore the
effect of more various disaggregated political institutions in the future study.I43 Second,
this dissertation shows how to incorporate the ways which preferences are translated into
policies through mediating political institutions. It explores why political leaders may
change their trade policy preferences as domestic political institutions change.144 Yet, the

theory says little about preferences of voters and specific domestic business sectors even if

 

“2 Nielson 2003.

”3 I have not obtained all the data in order to explore other important political institutions. These under-
explored data sets will offer greater opportunities for further testing of a broader process of trade policy
making.

“4 Milner and Kubota 2001.

117

they may influence trade policymaking. Therefore, the theory is relatively ambiguous on
the preferences of general domestic audiences and interest groups including business
sectors. Third, each type of free trade agreement is a different type of PTA depending on
different levels of economic cooperation. Considering the different types of PTAs rather
than aggregating them together and treating them identically may be another important
unexplored area of study. Also, the decision to join a PTA is not the same as a decision to
stay in a PTA that changes status from a free trade area to a common market. In this
dissertation, I do not explore this issue. That is a subject for another project.

Exploring the impact of different political institutions on interstate commercial
cooperation is an important and underexplored area. Greater attention should be paid to the
impact of institutional pressures on the explosion of PTAs. Resurgence of PT As will be
more properly assessed with the effort toward the more dynamic understanding of

international trade policymaking process.

7.3. Implications
This dissertation makes important contributions to theories of international relations and
international political economy more generally. This section elaborates implications of the

dissertation for those areas.

7.3.] How Domestic Politics Influences International Relations
This dissertation provides some insights into the crucial role of domestic politics in
international relations. The dissertation’s empirical tests demonstrate that structural factors

such as the role of international institutions or the distribution of economic power are not

118

sufficient to explaining the outcomes of international trade agreements. Based on a large-n
empirical test of the argument, the dissertation challenges the empirical validity of the
theories put forward by the structural approaches. Even if the evidence here supports some
arguments put forward by structural theorists, the findings also highlight some of the
theory’s limitations. By showing how insights from the external factor oriented approach
and internal constraints oriented approach can contribute to an explanation of international
trade policy outcomes, the dissertation shows the need to include these distinct approaches
in any comprehensive account of this subject. Theories of international trade policy can
give a much more accurate picture by incorporating domestic politics. Second, this
dissertation suggests that disaggregated domestic constraints deserve more attention.
Theories of international trade policy will be limited if they do not incorporate
disaggregated domestic institutions. Most studies often fail to provide a useful answer to
the question how variations of domestic institutions within democracies (as opposed to the
larger distinction of regime type between democracy and autocracy), such as the role of
veto player, electoral rules, parliamentarism or presidentialism, the extent of divided
government, and so forth might constrain the incentives and abilities of politicians to
pursue different trade policies. Such institutions may structure the responsiveness of
political leaders to domestic pressures and condition trade policy making. By
underestimating the disaggregated effects of political institutions, studies deduce
incomplete conclusions about the relationship between institutions and trade policy
making. On the basis of the arguments and evidence provided in this dissertation, it is clear
that domestic institutional analysis facilitates our understanding of trade policy in

international politics.

119

References

Axelrod, Robert and Robert Keohane. 1986. “Achieving Cooperation under Anarchy,” in
Kenneth Oye (ed), Cooperation under Anarchy. Princeton: Princeton University Press.

Baldwin, Richard. 1993. “A Domino Theory of Regional Integration”. National Bureau of
Economic Research Working Paper 4465, Cambridge, MA.

Baldwin, Richard. 1997, “The Causes of Regionalism”, The World Economy, 20, 7, 865-
888.

Bhagwati, J agdish. 1992, “Regionalism versus Multilateralism”, The World Economy, 15,
5, 535-555.

Bhagwati, J agdish. 1996. “The Theory of preferential trade agreements: Historical
evolution and current trends.” The American Economic Review, Nashville, May Vol. 86:

(2)

Barbieri Katherine and Gerald Schneider. 1999. “Globalization and Peace: Assessing New
Directions in the Study of Trade and Conflict.” Journal of Peace Research 36(4): 387-404.

Bailey, Michael, Judith Goldstein and Barry Weingast. 1997. “The Institutional Roots of
American Trade Policy,” World Politics 49(3):309-38.

Beck, Nathaniel and Jonathan Katz. 1995. “What to do (and not to do) with Time Series
Cross-Section Data.” American Political Science Review 89: 634.

Beck Nathaniel, Jonathan Katz, and Richard Tucker. 1998. “Taking Time Seriously: Time-
Series-Cross-Section Analysis with a Binary Dependent Variable.” American Journal of
Political Science 42 (October): 1260-88.

Beck Nathaniel, 2001. “Modeling Dynamics in the study of Conflict: A Comment on Oneal
and Russett”, Mimeo.

Beck, Thorsten, George Clark, Alberto Groff, Philip Keefer, Patrik Walsh, 2001, New
Tools and New Tests in Comparative Political Economy, Policy Research Working Paper,
World Bank.

Bennett, Scott and Allan Stam. 2000. “E UGene: A Conceptual Manual.” International
Interactions 26: 179-204.

Bliss, Harry and Bruce Russet. 1998. “Democratic Trading Partners: The Liberal
Connection, 1962-1989.” The Journal of Politics 60 (4): 1126-47.

Boix Carkes and Alicia Adsera, 2003. “The Political Economy of Trade and Economic
Integration: A Review Essay, Unpublished manuscript.

120

Bruce Bueno de Mesquita, James Morrow, Randolph Siverson, Alastair Smith 1999. “An
Institutional explanation of democratic peace.” The American Political Science Review:
791-807.

Bueno de Mesquita, Bruce, James Morrow, Randolph Siverson and Alastair Smith. 2000.
“Political Institutions, Political Survival, and Policy success.” In Governing for Prosperity,
ed. Bruce Bueno de Mesquita and Hilton L. Root. New Haven, CT. Yale University
Pressz59-84.

Bueno de Mesquita Bruce. 2000. Principles of International Politics. Washington DC: CQ
Press.

Clemens Boonekamp. 2002. “WTO: Regional trade integration under transformation”,
Trade Policies Review Division, WTO website.

Cowhey, Peter. 1993. Domestic institutions and the credibility of international
commitments. International Organization 47: 299-326.

Conybeare John A. 1991. “Voting for Protection: An Electoral Model of Tariff Policy”,
International Organization, Vol. 45, No. 1. Winter, pp. 57-81.

Coughlin Cletus. 2002. “The Controversy Over Free Trade: The Gap Between Economists
and the General Public”, Review. J an/ Feb.

Cox, Gary and Mathew McCubbins. 2001. “The Institutional Determinants of Economic
Policy Outcomes.” In Presidents, Parliaments and Policy, ed. Stephan Haggard and
Mathew McCubbins. New York: Cambridge University Press

Crawford J o-Ann and Sam Laird, 2000. “Regional Trade Agreements and the WTO”,
Paper presented for a meeting of the North American Economic and Finance Association in

Boston, 6-9 January.

Dai Xinyuan. 2002, “Political Regimes and International Trade: The Democratic
Difference Revisited”, American Political Science Review 96 (1): 159-65

de Melo, Jaime and Arvind Panagari ya. 1993. New Dimensions in Regional Integration.
Cambridge University Press.

Dixon William and Bruce Moon, 1993.” Political Similarity and American Foreign Trade
Patterns,” Political Research Quarterly, 46.

Dur Andreas, 2003. “Theorizing the Contagious Effects of Regionalism: European
Integration and Transatlantic Trade Relations, 1957-1963, EUI Working Paper.

121

Efird, Brian and Gaspare M. Genna. 2002. “Structural Conditions and the Propensity for
Regional Integration.” European Union Politics. 3.

Ehrlich Sean. 2003, “Access to Protection Domestic Institutions and Trade Policy in
Democracies,” Paper prepared for 2003 Midwest Political Science Association Meeting.

Fearon James. 1994. “Domestic political audiences and the escalation of international
disputes.” American Political Science Review 881577—92

Feehan James, 1996. “Trade Liberalization, Economic Integration and National
Public Goods”, International Economic Journal. 51(10): 2, summer.

Fishlow, Albert and Stephan Haggard. 1992. The United States and the Regionalization of
the World Economy. Paris, OECD.

Frankel, Jeffrey A. 1997. Regional Trading Blocs in the World Economic System.
Washington, DC: Institute for International Econorrrics.

Garrett Geoffrey. 1995. “Capital mobility, trade, and the domestic politics of economic
policy.” International Organization. 492657—87

Garrett, Geoffrey and Peter Lange. 1995. “Internationalization, Institutions, and Political
Change.” International Organization, 49: 627.

Gaubatz, Kurt Taylor. 1996. “Democratic States and Commitment in International
Relations”. International Organization 50 (1): 109-39.

Geddes, Barbara. 1995. “Challenging the Conventional Wisdom.” In Larry Diamond and
Marc Plattner. Eds. Economic Reform and Democracy. Baltimore: Johns Hopkins.

Geddes, Barbara. 1995. “The Politics of Economic Liberalization,” Latin American
Research Review 30:2 (spring), pp: 195-214.

Gilligan, Michael. 1997. Empowering Exporters: Reciprocity, Delegation, and Collective
Action in American Trade Policy. Ann Arbor: University of Michigan Press.

Gowa, Joanne. 1988. “Public Goods and Political Institutions: Trade and Monetary Policy
Processes in the United States.” International Organization 42, Winter. pp: 15-31.

Gowa, Joanne. 1994. Allies, Adversaries, and International Trade. Princeton: Princeton
University Press.

122

Gowa, Joanne and Edward D. Mansfield. 1993. “Power Politics and International Trade.”
American Political Science Review. 87: 408.

Goto, Junichi, Hamada, Koichi, 1998, “Economic Integration and the Welfare of Those
Who Are Left Behind: An Incentive-Theoretic Approach,” Journal of the Japanese and
International Economies, 12, 25-48.

Grieco, Joseph. 1997. “Systemic Sources of Variation in Regional Institutionalization in
Western Europe, East Asia, and the Americas,” in Edward D. Mansfield and Helen Milner
(eds), The Political Economy of Regionalism. New York: Columbia University Press.

Grossman, Gene and Elhanan Helpman, 1994. “Protection for Sale.” American Economic
Review 84(4): 833-50.

Grossman, Gene and Elhanan Helpman. 1993. The Politics of Free Trade Agreements.
Working Paper 4597. Cambridge, Mass: National Bureau of Economic Research
(December).

Haas, Ernst. 1961. “International Integration: The European and the Universal Process.”
International Organization. 15: 366.

Haggard, Stephan. 1995. Developing Nations and the Politics of Regional Integration.
Washington, DC: The Brookings Institution.

Haggard, Stephan. 1997. “Regionalism in Asia and the Americas,” in The Political
Economy of Regionalism (eds) Edward D. Mansfield and Helen V. Milner. New York:
Columbia University Press.

Haggard Stephan and Mathew D. McCubbins. 2001. “Introduction: Political Institutions
and the Determinants of Public Policy.” In Presidents, Parliaments and Policy, ed. Stephan
Haggard and Mathew McCubbins. New York: Cambridge University Press.

Hankla Charles, 2002. “Legislative-Executive Relations and International Trade Policy: A
Cross-National Study”, Presented at International Studies Association Conference.

Hoekman Bernard, Maurice Schiff and Alan Winters, 2002. “Regionalism and
Development: Main Messages From Recent World Bank Research”, Development
Research Group, World Bank

Hufbauer, Gary and Schott, Jeffery J. 1994. Western Hemisphere Economic Integration.
Washington, DC: Institute for International Economics.

Hudgins Edward L.,‘ 1996. “Regional and Multilateral Trade Agreements: Complementary
Means to Open Markets”, Cato Journal, Vol. 15, Nos. 2-3 (Fall/Winter).

123

International Monetary Fund (IMF). Various years. Exchange Arrangements and
Restrictions. Washington, DC: International Monetary Fund.

J aggers, Keith and Ted Robert Gurr. 1995. “Tracking Democracy’s Third Wave with the
Polity IH Data.” Journal of Peace Research 32:469-482.

Jaggers, Keith and Ted Robert Gurr. 1996. Polity III: Regime Change and Political
Authority, 1800-1994. ICPSR version. Boulder, CO: Keith

Jaime de Melo and Arvind Panagari ya, 1992. The New Regionalism in Trade Policy, World
Bank.

Jones, Daniel, Stuart Bremer and David Singer 1996. “Militarized Interstate Disputes,
1816-1992: Rationale, Coding Rules, and Empirical Patterns.” Conflict Management and
Peace Science 15(2); 163:213.

Kemp, MC. and H. Wan. 1976, “An Elementary Proposition Concerning the Formation of
Customs Unions”, Journal of lntemational Economics, 6, 95-98.

Keohane, Robert. 1984. After Hegemony: Cooperation and Discord in the World Political
Economy. Princeton: Princeton University Press.

Kerremans Bart and Bob Switky, 2000. The Political Importance of Regional Trading
Blocs, Ashgate.

King Gary and Zeng L, 2001. “Explaining Rare Events in lntemational Relations,
lntemational Organization 55, 693-715.

Kono, Daniel et al. 1999. “Regional Regimes”, Unpublished Manuscript.

Krueger, Anne. 1995. “Free Trade Agreements versus Customs Unions”, NBER Working
Paper No. 5084, Cambridge, Mass.

Krueger, Anne. 1999. “Are Preferential Trading Arrangements Trade-Liberalizing or
Protectionist?”. Journal of Economic Perspectives 13 (4): 105-24.

Lawrence, Robert 1996. Regionalism, Multilateralism, and Deeper Integration. Washington
DC. Brookings Institution.

Lektzian David and Mark Souva. 2003. “Economic Sanctions and Democracy”, Journal of
Peace Research, 40 (6).

Leeds, Brett Ashley. 1999. “Domestic Political Institutions, Credible Commitments and
International Cooperation.” American Journal of Political Science 43 (4): 979-1002

124

Leeds, Brett Ashley and David Davis. 1999. “Beneath The Surface: Regime Type and
lntemational Interaction, 1953-78.” Journal of Peace Research 36:5-21

Lohmann Susanne and O’Halloran Sharyn, 1994. “Divided government and US. trade
policy: theory and evidence”, lntemational Organization 48,4, Autumn. 595—632.

Mansfield, Edward, Milner Helen, Peter Rosendorff. 1997. “Free to trade: democracies
and international trade negotiations”. Presented at Annual Meeting American Political
Science Association, Washington, DC, September.

Mansfield, Edward. 1998. “The proliferation of Preferential Trading Arrangements.”
Journal of Conflict Resolution 42 (5): 523-43.

Mansfield, Edward, Milner Helen, Peter Rosendorff, 2002. “Why Democracies Cooperate
More: Electoral Control and lntemational Trade Agreement’, International Organization,
56 (3), Summer, pp.477-513

Mansfield, Edward and Marc Busch, 1995. “The Political Economy of non-tariff barriers: a
cross-national analysis”, lntemational Organization 49,4. Autumn. 723-49.

Mansfield Edward and Helen Milner, 1997. The Political Economy of Regionalism. New
York: Columbia University Press.

Mansfield, Edward, and Rachel Bronson. 1997. “The Political Economy of Major-Power
Trade Flows,” in Edward Mansfield and Helen Milner (eds), The Political Economy of
Regionalism. New York: Columbia University Press.

Mansfield Edward and Helen Milner. 1999. “The New Wave of Regionalism”.
lntemational Organization 53, 3, Summer, pp.589-627.

Mansfield, Edward and Eric Reinhardt, 2002. “Multilateral Determinants of Regionalism:
The Effects of GATT/WTO on the Formation of Preferential Trading Arrangements,

lntemational Organization.

Maoz Zeev and Bruce Russett. 1993. “Normative and Structural Causes of the Democratic
Peace, 1946-1986.” American Political Science Review 87(3): 624-638.

Martin, Lisa. 2000. Democratic Commitments: Legislatures and lntemational Cooperation.
Princeton, N.J.: Princeton University Press.

Mattli, Walter. 1999. The Logic of Regional Integration: Eur0pe and Beyond. New York:
Cambridge University Press.

McGillivray Fiona. 1997 . “Party Discipline as a Determinant of the Endogenous Formation
of Tariffs.” American Journal of Political Science. Vol. 41, No. 2, April.

125

McGillivray Fiona and Alastair Smith, 1998. “Cooperating Democrats, Defecting
Autocrats”, Manuscript, Yale University.

Mendoza, Ronald and Chandrika Bahadur, 2002. “Toward Free and Fair Trade: A Global
Public Perspective”, Challenge, Sept-Oct.

Milner Helen and Rosendorff Peter. 1996. “Trade negotiations, information and domestic
politics.” Economic Politics. 8: 145-89

Milner Helen and Robert Keohane, 1996. “Internationalization and Domestic Politics:
Introduction,” in Robert Keohane and Helen Milner, Internationalization and Domestic
Politics, Cambridge University Press.

Milner Helen. 1997. Interests, Institutions, and Information: Domestic Politics and
lntemational Relations Princeton, N.J.: Princeton University Press.

9

Milner Helen. 1997. “Industries, Governments, and the Creation of Regional Trade Blocs,’
in Edward D. Mansfield and Helen V. Milner (eds), The Political Economy of regionalism.
New York: Columbia University Press.

Milner Helen. 1999. “The political economy of international trade.” Annual Review of
Political Science. 2:91—114.

Milner Helen and Keiko Kubota. 2001. “Why The Rush to Free Trade: Democracy and
Trade Policy in the Developing Countries”, Annual Meeting of the American Political
Science Association, San Francisco, Aug 30-Sept 2.

Mohan Penubarti and Michael Ward. 2000. “Commerce and Democracy”, Working paper,
Center for Statistics and the Social Sciences. University of Washington.

Moravcsik, Andrew. 1997. “Taking Preferences Seriously: A Liberal Theory of
lntemational Politics.” lntemational Organization. 51:513.

Morrow James, Randolph Siverson, Taressa Taberes. 1998. “The Political Determinants of
lntemational Trade. American Political Science Review 92(3)

Nelson Douglas, 1988. “Endogenous Tariff Theory: A Critical Survey”, American Journal
of Political Science. 32 (3): 796-837.

Nielson, Daniel. 2003. “Supplying Trade Reform: Political Institutions and Liberalization
in Middle-Income Presidential Democracies”, American Journal of Political Science, Vol.

47, No. 3, July: 470-491.

O’Halloran, Sharyn. 1994. Politics, Process, and American Trade Policy. Ann Arbor:
University of Michigan Press.

126

Oneal, John R., and Bruce Russett. 1997. “T he Classical Liberals Were Right:
Democracy, Interdependence, and Conflict, 1950-1985.” lntemational Studies Quarterly
41 (June): 267-94.

Oneal, John R., and Bruce Russett. 1999. “Assessing the Liberal Peace with Alternative
Specifications: Trade Still Reduces Conflict.” Journal of Peace Research 36 (July).

Oneal, John R., and James L. Ray. 1997. “New Tests of Democratic Peace: Controlling
For Economic Interdependence: 1950-1985.” Political Research Quarterly 50
(December): 751-75.

Oneal, John R., Francis H. Oneal, Zeev Maoz, and Bruce Russett. 1996. “The Liberal
Peace: Interdependence, Democracy, and lntemational Conflict.” Journal of Peace
Research 33 (February): 624-38.

Ortiz Antonio Mena, 2000. “The Political Economy of Regionalism: A current Theorizin g
and Preliminary Observations on the G-3 FTA, CIDE Working Paper.

Oye, Kenneth. 1992. Economic discrimination and political exchange: World political
economy in the 1930s and 1980s. Princeton, NJ: Princeton University Press.

Pentland, Charles. 1973. lntemational Theory and European Integration. London: Faber.

Panagariya, Arvind. 2000. Preferential Trade Liberalization: The Traditional Theory and
New Developments. Journal of Economic Literature 38 (2): 287-331.

Pollins, Brian. 1989. “Does Trade Still Follow the Flag?”, American Political Science
Review 83.

Pomfret, Richard. 1988. Unequal trade: The econorrrics of discriminatory international
trade policies. Oxford: Basil Blackwell.

Putnam, Robert. 1988. “Diplomacy and Domestic Politics: The Logic of Two-Level
Games.” lntemational Organization. 42.

Remmer, Karen. 1998. “Does Democracy Promote Interstate C00peration?” lntemational
Studies Quarterly 42 (1): 25-52.

Riezman Raymond, 1999, “Can Bilateral Trade Agreements Help Induce Free Trade?”
CSGR Working Paper. No. 44. October. University of Warwick.

Reinhardt, Eric. 2000. “Aggressive Multilateralism: The Determinants of GATT/WTO
Dispute Initiation, 1948-1998”, Unpublished Manuscript.

Reinhardt, Eric. 1996. “Posturing Parliaments: Ratification, Uncertainty and lntemational
Bargaining”, PhD Dissertation (New York: Columbia University, December 11).

127

Rodrik Dani. 1994. The rush to free trade in the developing world. Why so late? Why now?
Will it last? In Voting for Reform: Democracy, Political Liberalization, and Economic
Adjustment, ed. Stephen Haggard, S Webb, pp. 61-88. New York: Oxford Univeristy Press.

Rogowski Ronald. 1987. “Trade and the variety of democratic institutions.” lntemational
Organization. 41: 203—24.

Rogowski Ronald. 1989. Commerce and Coalitions. Princeton, NJ: Princeton University
Press.

Rosendorff Peter and Helen Milner, 2001. “The Optimal design of lntemational Trade
Institutions: Uncertainty and Escape”, lntemational Organization 55 (4), Autumn, 829-857.

Sherry Bennett Quinones and Scott Gates. 1995. “Economic Risk and the Politics of
Protectionism, lntemational Interactions. 21 (1), 63-83.

Schiff Maurice and Alan Winters, 1998. “Dynamics and Politics in Regional Integration
Arrangements: An Introduction”, The World Bank Economic Review. Vol. 12. No. 2: pp.
177-95.

Schiff Maurice and Alan Winters, 2003. “Regional Integration and Development”, Oxford
University Press.

Shugart Soberg and John M. Carey. 1992, “Criticisms of Presidentialism and Responses,”
in Presidents and Assemblies: Constitutional Design and Electoral Dynamics, Cambridge.
Cambridge University Press, pp. 28- 54.

Seokwoo Kim, 1995. The Political Economy of Protection in Developing Countries: 1977-
1990, Dissertation (University of North Carolina at Chapel Hill)

Seokwoo Kim, 1998. The Political Economy of lntemational Trade, Han Wool Academy.

Schultz, Kenneth. 2001. “Looking for Audience Costs.” Journal of Conflict Resolution
45(1): 32-60.

Simmons Beth. 2000. “International law and state behavior: commitment and compliance
in international monetary affairs.” American Political Science Review 942819—35.

Tsebelis George. 1999. “Veto Players and Law Production in Parliamentary Democracies:
An Empirical Analysis”, American Political Science Review, September.

Tsebelis, George. 1995. “Decision Making in Political Systems: Veto Players in

Presidentialism, Parliamentarism, Multicameralism and Multipartyism”, British Journal of
Political Science 25:289-325.

128

Tomz Michael. 2002. “Democratic Default: Domestic Audiences and Compliance with
lntemational Agreements.” Paper prepared for delivery at the 2002 Annual Meeting of the
American Political Science Association, Boston, August 29-September 1.

Urata Shujiro, 2001. “Regionalism, the New WTO Round and Japan’s New Trade Policy”,
Journal of Japanese Trade and Industry. Japan Economic Foundation, October.

Urata Shujiro, 2002. “Globalization and the Growth in Free Trade Agreements.” Asia
Pacific Review, Vol. 9, No. 1, 2002.

Vamvakidis Athanasios, 1999. “Regional Trade Agreements or Broad Liberalization:
Which Path Leads to Faster Growth?” IMF Staff Papers Vol. 46, No. 1 (March)
lntemational Monetary Fund.

Vanhanen, Tatu. 2000. A New Dataset for Measuring Democracy, 1810-1998. Journal of
Peace Research, Vol.37, No. 2: 251-265.

Verdier Daniel. 1998. “Democratic convergence and free trade?”, lntemational Studies
Quarterly. 42: 1—24.

Verdier Daniel, 1994. Democracy and lntemational Trade: Britain, France, and the United
States, 1860-1990 Princeton: Princeton University Press.

Wei Shang-J in and Jeffrey Frankel, 1998. “Open Regionalism in a World of Continental
Trade Blocs”, IMF Staff papers, Vol. 45, No.3, lntemational Monetary Fund.

Weaver Kent and Bert Rockman, 1993. “Assessing the Effects of Institutions,” in Weaver
and Rockman, eds. Do Institutions Matter? Government Capabilities in the United States
and Abroad, Washington, DC: The Brookings Institution.

Whalley, John. 1998. Why do countries seek regional trade agreements? In The
regionalization of the world economy, edited by Jeffrey Frankel. Chicago: University of
Chicago Press.

William Dixon and Bruce Moon, 1993. “Political Similarity and American Foreign Trade
Patterns,” Political Research Quarterly 46, 10-1.

Winters, Alan. 1996, “Regionalism versus Multilateralism,” Policy Research Working
Paper No. 1687, The World Bank, Washington DC.

Winters, Alan. 1998, “Regionalism and the Next Round”, in J. Schott (ed.), Launching New

Global Trade Talks - An Action Agenda, Special Report 12, Institute for lntemational
Economics, Washington DC. (September).

129

Wonjae Hwang. 2001. “Political Institutions and Political Survival: Correcting
Measurement Problems and Expanding the Arguments”, Unpublished paper, Michigan
State University.

World Bank. 1987. World Development Report. New York: Oxford University Press.

World Trade Organization. 1995. Regionalism and the World Trading System. Geneva:
WTO.

Yeats, A. 1997. “Does MERCOSUR’s trade performance justify concerns about the effects
of regional trade agreements?” PPR Working Paper 1729, World Bank, Washington DC.

Zorn, Cri stopher. 2001. “Generalized Estimating Equation Models for Correlated Data: A

Review With Applications,” American Journal of Political Science. Vol 45. No.2, April,
470-490.

130

 

 

   

l

           
              

1 .

\

:ESRARIE
l

l H

02504 3955

  

will
tlll

IHlll
A Hf .
ill
>

ll

fillGAN

W
1293

 

l.