Semi-Presidentialism and Democracy

Semi-Presidentialism and Democracy

9780230_242920_01_prexii.indd i

11/11/2010 5:05:07 PM

9780230_242920_01_prexii.indd ii

11/11/2010 5:05:07 PM

Semi-Presidentialism and Democracy Edited by

Robert Elgie Professor of Government and International Studies, Dublin City University

Sophia Moestrup Senior Program Manager, National Democratic Institute, USA

and

Yu-Shan Wu Distinguished Research Fellow, Institute of Political Science, Academic Sinica, Taiwan

9780230_242920_01_prexii.indd iii

11/11/2010 5:05:07 PM

Editorial matter, selection and conclusion © Robert Elgie, Sophia Moestrup and Yu-Shan Wu 2011 All remaining chapters © respective authors 2011 All rights reserved. No reproduction, copy or transmission of this publication may be made without written permission. No portion of this publication may be reproduced, copied or transmitted save with written permission or in accordance with the provisions of the Copyright, Designs and Patents Act 1988, or under the terms of any licence permitting limited copying issued by the Copyright Licensing Agency, Saffron House, 6-10 Kirby Street, London EC1N 8TS. Any person who does any unauthorized act in relation to this publication may be liable to criminal prosecution and civil claims for damages. The authors have asserted their rights to be identified as the authors of this work in accordance with the Copyright, Designs and Patents Act 1988. First published 2011 by PALGRAVE MACMILLAN Palgrave Macmillan in the UK is an imprint of Macmillan Publishers Limited, registered in England, company number 785998, of Houndmills, Basingstoke, Hampshire RG21 6XS. Palgrave Macmillan in the US is a division of St Martin’s Press LLC, 175 Fifth Avenue, New York, NY 10010. Palgrave Macmillan is the global academic imprint of the above companies and has companies and representatives throughout the world. Palgrave® and Macmillan® are registered trademarks in the United States, the United Kingdom, Europe and other countries. ISBN 978–0–230–24292–0 hardback This book is printed on paper suitable for recycling and made from fully managed and sustained forest sources. Logging, pulping and manufacturing processes are expected to conform to the environmental regulations of the country of origin. A catalogue record for this book is available from the British Library. A catalog record for this book is available from the Library of Congress. 10 9 8 7 6 5 4 3 2 1 20 19 18 17 16 15 14 13 12 11 Printed and bound in Great Britain by CPI Antony Rowe, Chippenham and Eastbourne

9780230_242920_01_prexii.indd iv

11/11/2010 5:05:08 PM

Contents List of Tables

vii

List of Figures

ix

Acknowledgments

x

Notes on Contributors 1

xi

Semi-Presidentialism: An Increasingly Common Constitutional Choice Robert Elgie

1

2

Clustering of Semi-Presidentialism: A First Cut Yu-Shan Wu

3

Variation in the Durability of Semi-Presidential Democracies Robert Elgie and Petra Schleiter

42

The Rules of Electoral Competition and the Accountability of Semi-Presidential Governments Jih-Wen Lin

61

4

21

5

Semi-Presidentialism in Western Europe Robert Elgie

81

6

Semi-Presidentialism under Post-Communism Oleh Protsyk

98

7

Semi-Presidentialism and Democratic Development in East Asia Benjamin Reilly

117

8

Semi-Presidentialism in Africa: Patterns and Trends Sophia Moestrup

9

The President is Not a Passenger: Portugal’s Evolving Semi-Presidentialism Carlos Jalali

156

Taiwan: Democratic Consolidation under President-Parliamentarism Yu-Shan Wu and Jung-Hsiang Tsai

174

10

134

v

9780230_242920_01_prexii.indd v

11/11/2010 5:05:08 PM

vi

11

12

13

Contents

Disintegrated Semi-Presidentialism and Parliamentary Oligarchy in Post-Orange Ukraine Kimitaka Matsuzato

192

Semi-Presidentialism and Moldova’s Flawed Transition to Democracy William Crowther

210

Semi-Presidentialism in the Weimar Republic: A Failed Attempt at Democracy Yu-Chung Shen

229

14

Haiti: Semi-Presidentialism in a Fractured Society Sophia Moestrup

246

15

Semi-Presidentialism: What Have We Learned? Robert Elgie, Sophia Moestrup, and Yu-Shan Wu

264

Bibliography

275

Index

293

9780230_242920_01_prexii.indd vi

11/11/2010 5:05:08 PM

Tables 1.1 The spread of countries with semi-presidential constitutions 1.2 Constitutional systems in Francophone Africa (as of March 31, 2010) 1.3 Constitutional systems in countries of the former Soviet Union and the former Yugoslavia (as of March 31, 2010) 1.4 Variation within semi-presidentialism 2.1 Clustering of semi-presidential countries 2.2 Parliamentarism and West European semi-presidential regimes 2.3 Operational subtypes under semi-presidentialism 2.4 Clustering and subtypes of semi-presidentialism 3.1 Duration and outcome of democratic episodes in semi-presidential regimes 3.2 The survival of semi-presidential democracies (Cox Proportional Hazards Model) 4.1 Free semi-presidential countries (2009) 4.2 Government types in free semi-presidential countries 4.3 Government formation in semi-presidential democracies: Results from the binary logistic regression 4.4 Regression of government durability of semi-presidential governments 5.1 The origins of West European semi-presidentialism 5.2 Presidential powers in West European semi-presidential countries 5.3 Party support in the executive in West European semi-presidential countries 6.1 Regime type and freedom house PR scores by geographic region 6.2 Regime type and electoral systems 6.3 Constitutional Regime Change (Original Post- Communist Regimes Only) 6.4 Countries with unstable constitutional regimes 7.1 Basic indicators for East Timor, Mongolia, and Taiwan 8.1 Semi-presidential countries in Africa, January 2010 8.2 Current and former semi-presidential electoral democracies in Africa, January 2010 8.3 Democratic breakdown and survival of president-parliamentary and premier-presidential systems in Africa, January 2010

8 10 11 18 23 26 31 33 50 54 68 70 71 74 84 87 88 101 108 111 114 124 135 137 143

vii

9780230_242920_01_prexii.indd vii

11/11/2010 5:05:08 PM

viii

8.4

8.5 10.1 10.2 10.3 10.4 11.1 13.1 13.2 14.1

Tables

Democratic breakdown and survival of different types of electorally generated semi-presidential governments in Africa, January 2010 Cross-tabulating authority patterns and government types The ROC’s political development and constitutional regimes Major parties’ seats in the legislative Yuan Semi-presidentalism in the ROC Measuring Taiwan’s presidential powers Cabinets of ministers in Ukraine, 2004–2009 The result of the 1919 parliamentary election in Weimar Germany The number of sitting days and the laws legislated by the Reichstag or by the Reich President Presidential powers in Haiti’s 1987 constitution

9780230_242920_01_prexii.indd viii

144 147 176 179 181 185 202 232 233 253

11/11/2010 5:05:08 PM

Figures 2.1 7.1 11.1 12.1 12.2 13.1

The impact of third-wave democratization on semi-presidentialism Political stability in East Timor, Mongolia and Taiwan, 1996–2006 President and legislatures under post-communism Support for leaders 2000–2008 For which party would you vote 2000–2008 Structure of the Weimar Constitution between the executive and the legislative

25 131 193 225 226 235

ix

9780230_242920_01_prexii.indd ix

11/11/2010 5:05:09 PM

Acknowledgments This book is based on a collection of papers presented at the conference on ‘Semi-Presidentialism and Democracy: Institutional Choice, Performance and Evolution’ held at Academia Sinica, Taipei, in October 2008. The conference aimed to take stock and expand our knowledge of semi-presidentialism, a constitutional form that has remained comparatively understudied despite its rapid spread in the late twentieth century. The conference covered all geographical areas where semi-presidential systems are found and explored the whole life cycle of the system, including adoption, operation, impact, and evolution in comparative and single-country context. Both the global coverage and the complete institutional approach were unprecedented. We would like to thank the Institute of Political Science at Academia Sinica (IPSAS) for kindly hosting the conference that brought together country experts and comparativists, political scientists, and constitutional lawyers, to share their research results. The conference thereby set the stage for significantly advancing our collective understanding of semi-presidentialism and democracy. We are also grateful to President Ma Ying-jeou of the Republic of China (Taiwan) for receiving the conference participants, and sharing his perspective on how semi-presidentialism should be practiced in Taiwan, thereby providing us with firsthand insights into the operation of semipresidentialism in a nascent democracy. During the Taipei conference, the able assistants at the IPSAS offered invaluable service. We would like to particularly thank Miss Chia-jung Lee, Miss Chun-Wei Su, and Mr. Chien-Wen Su for their warm help. Robert Elgie would like to thank the Irish Research Council for the Humanities and Social Sciences for the award of a Senior Research Fellowship in 2009–2010 that was a great help in the preparation of this book. The editors would like to thank Palgrave for commissioning the volume and particularly Alison Howson, Amber Stone-Galilee, and Liz Blackmore for their help in the preparation of the book. The Taipei conference and this resulting volume are the products of an interdisciplinary and international collaboration spanning several continents. To study a phenomenon that has spread to a quarter of the world’s countries, such teamwork is hugely helpful. We believe the present book lays the groundwork for further expanding this collaboration.

x

9780230_242920_01_prexii.indd x

11/11/2010 5:05:09 PM

Contributors William Crowther is Chair of the Department of Political Science at the University of North Carolina at Greensboro, and Director of the UNCG Center for Legislative Studies. He received his Ph.D. from The University of California at Los Angeles in 1986. He is the author of numerous publications on post-communist and communist politics, and on interethnic relations. Robert Elgie is Paddy Moriarty Professor of Government and International Studies at Dublin City University. He is the coeditor of the journal French Politics published by Palgrave and the coeditor of the Routledge/ECPR Studies in European Political Science series. He has published extensively on comparative politics, semi-presidentialism, and French politics. Carlos Jalali is Assistant Professor at the University of Aveiro, where he is also researcher at the Governance, Competitiveness and Public Policies Research Centre. He has published widely on Portuguese political institutions, parties, and electoral behavior in comparative perspective and is associate editor for political science of Análise Social, Portugal’s leading social science journal. Jih-Wen Lin is Research Fellow in the Institute of Political Science at Academia Sinica and Professor of political science at National Chengchi University. He received his Ph.D. in political science from the University of California at Los Angeles in 1996. His research interests cover formal analysis and comparative institutional design, and he has had articles published in Party Politics, Journal of Democracy, Electoral Studies, China Quarterly, Journal of East Asian Studies, and Issues & Studies, among others. Kimitaka Matsuzato is Professor at the Slavic Research Center, Hokkaido University, specializing in the politics of Ukraine, Moldova, the Baltic countries, South Caucasus, and unrecognized states. He published an article comparing semi-presidentialism in these countries in Demokratizatsiya (2006) and recently an article focused on Orthodoxy diplomacy around the unrecognized states of Abkhazia and Transnistria in Religion, State & Society (2009). Sophia Moestrup is Senior Program Manager at the National Democratic Institute (NDI) in Washington, DC. She has served as country director in Niger for the Danish Ministry of Foreign Affairs. She is coeditor with Robert Elgie of Semi-Presidentialism Outside Europe (Routledge, 2007) and Semi-Presidentialism in Central and Eastern Europe (Manchester University, Press 2008). xi

9780230_242920_01_prexii.indd xi

11/11/2010 5:05:09 PM

xii

Contributors

Oleh Protsyk is Senior Research Associate at European Centre for Minority Issues, Flensburg, Germany. His research interests include executive-legislative relations, political representation, and ethnic conflict regulation. His work has appeared in academic journals such as Comparative Politics, European Journal of Political Research, Europe-Asia Studies, Political Studies, and Post-Soviet Affairs. Benjamin Reilly is Professor of Political Science in the Crawford School of Economics and Government at the Australian National University. He is the author of six books and over 60 journal articles and chapters on issues of democratization, constitutional reform, party politics, electoral system design, and conflict management, and has advised many governments and international organizations on these subjects. His latest books are a study of democratization and political reform in Asia and the Pacific, Democracy and Diversity: Political Engineering in the Asia-Pacific (Oxford University Press, 2006), and an edited volume, Political Parties in Conflict-Prone Societies (United Nations University Press, 2008). He is currently a Visiting Professor at Johns Hopkins University’s School of Advanced International Studies (SAIS), in Washington DC. Petra Schleiter is Tutorial Fellow in Politics (St. Hilda’s College), and Lecturer in Politics (Department of Politics and International Relations), University of Oxford. Her research on comparative and Russian politics has been published in a range of journals including the American Political Science Review, Comparative Political Studies, British Journal of Political Science, European Journal of Political Research, and Post-Soviet Affairs. Yu-Chung Shen is Assistant Professor at Tunghai University, Taiwan. He has published articles on the subject of semi-presidentialism and the Weimar Republic in journals such as the Journal of Politics and Law. Jung-Hsiang Tsai is Assistant Professor at National Chung Cheng University, Political Science Department, Taiwan. He earned his Ph.D. in Political Science from Boston University. His research interests include comparative semipresidential studies, Sino-US relationships, and qualitative political methods. His works have been published in Crime, Law, and Social Change, and French Politics. Yu-Shan Wu is distinguished research fellow and director of the Institute of Political Science at Academia Sinica, and professor of political science at National Taiwan University. He has authored six books and edited six others, and published more than 100 journal articles and book chapters. His research interests are political and economic transitions in former socialist countries, democratic consolidation and constitutional engineering in nascent democracies, and theories of cross-Taiwan Strait relations. His area of focus includes Taiwan, mainland China, Eastern Europe, and Russia.

9780230_242920_01_prexii.indd xii

11/11/2010 5:05:09 PM

1 Semi-Presidentialism: An Increasingly Common Constitutional Choice Robert Elgie

There are now more than 50 countries in the world with semi-presidential constitutions. The first semi-presidential regimes were created in 1919 in Finland and Weimar Germany. The spread of semi-presidentialism was slow for a long time after this period, but during the third wave of democratization in the early 1990s semi-presidentialism became the constitutional form of choice for most new democracies. Today, countries continue to choose semi-presidentialism and for a variety of reasons. In 2007 Turkey passed a constitutional amendment that introduced a semi-presidential form of government as a way of strengthening the authority of the president. In contrast, in Kenya the conflict between the president and the leader of the opposition ended in March 2008 when they agreed to pass reforms that limited the president’s power and as a result created a semi-presidential system. Despite the increasingly common occurrence of this form of government, semi-presidentialism remains understudied, especially in comparison with its better-known presidential and parliamentary counterparts. In addition to Duverger’s (1980) initial work, a small number of scholars have addressed semi-presidentialism, most notably Sartori (1997), Shugart and Carey (1992), and Samuels and Shugart (2010), and there are three case study volumes specifically on the topic (Elgie ed. 1999; Elgie and Moestrup eds. 2007; Elgie and Moestrup eds. 2008), plus a number of related monographs and articles (e.g., Frison-Roche 2005; Skach 2005). More than that, there is still considerable confusion as to what constitutes semi-presidentialism, which countries can be considered semi-presidential, and whether semi-presidentialism can be studied in the same way as presidentialism and parliamentarism. The spread of semi-presidentialism makes it even more necessary to provide the answers to these very basic questions. This chapter begins by identifying the concept of semi-presidentialism and contrasting it with the more familiar concepts of presidentialism and parliamentarism. The chapter then maps the spread of semi-presidentialism since the appearance of the first semi-presidential regimes immediately 1

9780230_242920_02_cha01.indd 1

11/11/2010 5:05:49 PM

2 Robert Elgie

after the First World War. Finally, given the heterogeneous nature of semipresidential countries, the chapter proposes a way of incorporating semipresidentialism into the mainstream comparative politics research agenda.

The concept of semi-presidentialism The standard features of both presidentialism and parliamentarism are constitutional. Under presidentialism, the president is directly elected, or at least popularly elected by way of a US-style electoral college; the president is both the head of state and head of government (i.e., there is no separate office of prime minister); and both the president and the legislature serve for a fixed term. In contrast, under parliamentarism the head of state is not directly elected, being either a monarch or an indirectly elected president; there is a separate office of prime minister; and the prime minister and cabinet are collectively responsible to the legislature (i.e., the legislature has the power to approve the appointment of the prime minister and/or to remove the cabinet from office). The standard elements of semi-presidentialism are equally and no less constitutional. Under semi-presidentialism there is a directly elected, or popularly elected, president who serves for a fixed term; there is a separate position of prime minister; and the prime minister and cabinet are collectively responsible to the legislature. A country can be identified as presidential, parliamentary, or semi-presidential simply by referring to the most basic features of its constitution. Unlike presidentialism and parliamentarism, the concept of semipresidentialism has been subject to a considerable degree of confusion and controversy because, to date, most people have failed to define semipresidentialism in terms of its basic constitutional features. Instead, people have defined semi-presidentialism in terms of the actual powers of the president and/or prime minister in practice. For example, a key part of Duverger’s definition of semi-presidentialism states that for a country to be semi-presidential, the president must have ‘quite considerable powers’ (Duverger 1980: 166). Equally, Sartori stipulates that one condition for semipresidentialism is that the president “shares power with a prime minister, thus entering a dual authority structure” (Sartori 1997: 131). In this context, it is usually assumed that semi-presidentialism is where there are both a fairly powerful directly elected president and a fairly powerful prime minister. If the president is extremely powerful and the prime minister is relatively weak, then it is assumed that the country must be presidential, like Russia. In contrast, if there is a figurehead president and the prime minister is the key policy-making actor, then it is assumed that the country must be parliamentary, as in Austria. In this way, the decision to base the definition of semi-presidentialism on the actual powers of the president runs contrary to the standard way in which presidentialism and parliamentarism are defined and, thus, is

9780230_242920_02_cha01.indd 2

11/11/2010 5:05:49 PM

An Increasingly Common Constitutional Choice

3

puzzlingly inconsistent. More important, the decision to define semipresidentialism on the basis of the powers of the president and/or prime minister is methodologically problematic for two reasons. First, a definition based on the actual powers of the president and prime minister is not reliable. While it is possible to read a country’s constitution and make a reliable judgment about whether a country is presidential, parliamentary, or semi-presidential, it is not possible to examine the role of political actors and make a reliable judgment about whether a country has a fairly powerful president. Any such judgment must be subjective, not least because the power of political actors varies over time. If the concept of semipresidentialism is based on a subjective judgment about the power of the president and/or prime minister, then each scholar has free rein to decide what constitutes a fairly powerful president. As a result, the list of semipresidential countries varies from one person to another and the study of semi-presidentialism suffers because no two people will compare the same set of countries. The absence of a reliable definition of semi-presidentialism means that like is not be compared with like. This compromises the general debate about the effect of semi-presidentialism on political outcomes. Second, defining semi-presidentialism on the basis of presidential and prime ministerial powers creates a potential problem of case-selection bias. For example, Sartori requires a semi-presidential system to exhibit shared presidential/prime ministerial power (ibid.). However, if this requirement is adopted and a study of semi-presidential countries concludes that they have the potential to suffer from the problem of a presidential/prime ministerial rivalry or conflict, then it is largely because the only countries that have been chosen for study are those where the executive is divided in the first place. Thus, any conclusions about the effects of semi-presidentialism are unlikely to be robust. Therefore, to ensure a reliable definition and to avoid the selectionbias problem, the concept of semi-presidentialism needs to be understood in the same way as the concepts of presidentialism and parliamentarism. The concept should be defined on the basis of core constitutional features. Consistent with this approach, in this chapter, and indeed in this book as a whole, semi-presidentialism is defined as the situation where there is both a directly elected fixed-term president and a prime minister and cabinet who are collectively responsible to the legislature. The reliability of this definition means that in most cases the identification of countries with semi-presidential constitutions is very clear. However, constitutions can be idiosyncratic documents. In some cases, these idiosyncrasies raise classification issues, even when countries meet the basic constitutional requirements for semi-presidentialism. For example, in a small number of countries the president is formally classed as head of state and head of government, even though the constitution makes provision for a prime minister. These cases are potentially problematic because if

9780230_242920_02_cha01.indd 3

11/11/2010 5:05:49 PM

4 Robert Elgie

the president is head of government and the government is constitutionally responsible to the legislature, then, all else being equal, the president would appear to be responsible to the legislature too. This would violate the fixed-term president requirement and would mean that the country should not be classed as semi-presidential. A related example concerns the issue of individual prime ministerial versus collective cabinet responsibility. In many presidential countries it is possible for the legislature to dismiss ministers individually, but there is no prime minister and the government is not collectively responsible to the legislature. In other countries, the constitution refers to a prime minister who is individually responsible to the legislature, but there is no provision for collective cabinet responsibility. Should countries with this sort of constitutional stipulation be classed as semi-presidential? A further issue concerns the cases where the constitution states that the legislature must approve the appointment of the prime minister, but there is no provision for ongoing responsibility. In these cases, the prime minister’s responsibility to the legislature is a one-shot game. Once approved, the prime minister is protected from parliamentary censure. This situation raises the issue of whether the prime minister and cabinet are truly responsible to the legislature. If parliament has no opportunity to initiate the responsibility of the prime minister and cabinet, should a country be classed as semi-presidential? This chapter aims to identify the spread of semi-presidentialism over time. Therefore, answers need to be provided to these questions. In the rest of this chapter, a country is classed as semi-presidential only if it meets the basic constitutional requirements identified above. In addition, a number of subsidiary classification rules are also adopted. First, a country is classed as semi-presidential even if the president is formally designated as the head of the government or as the chief executive so long as there is a specific provision stating that the president serves for a fixed term and/or a clause indicating that the survival in office of the government is unrelated to that of the president. Thus, Mozambique is classed as semi-presidential because Article 147–3 of the 2004 revised constitution states that the president serves for five years, even though Article 146–3 states that the president is head of government and Article 207 states that the government is responsible to the National Assembly. Equally, Sri Lanka is classed as semi-presidential even though the president is a member of the cabinet because Article 43–2 explicitly states that the president remains in office even if the cabinet is dissolved. Second, a case is excluded from the list of semi-presidential countries if there is only a specific provision for individual prime ministerial and/or ministerial responsibility to the legislature rather than collective responsibility. Argentina is not classed as semi-presidential on this basis even though the constitution identifies the position of Chief of the Ministerial Cabinet, or prime minister, who is individually responsible to the Congress. By the

9780230_242920_02_cha01.indd 4

11/11/2010 5:05:49 PM

An Increasingly Common Constitutional Choice

5

same token, the Philippines from 1981 to 1986, where a similar position existed, is not classed as semi-presidential. Third, a country is classed as semi-presidential only if there is a general statement that the government is responsible to the legislature and/or if there is specific provision for parliament to pass a vote of no confidence. In contrast, a case is excluded from the list of semi-presidential countries if parliament is allowed only to consent to the appointment of the prime minister and/or to ratify the president’s desire to dismiss the prime minister. Uganda is excluded because there is only a requirement that parliament consent to the appointment of the prime minister. South Korea is excluded on the same basis and because Article 63 of the 1987 constitution states that parliament may only pass a recommendation for the dismissal of the prime minister and other ministers. In other words, the president may refuse to accept parliament’s recommendation and so there is no ongoing responsibility. Finally, countries such as Belarus from 1994–1996 are excluded because the constitution only provides the parliament with the power to approve decisions made by the president about the appointment and dismissal of the prime minister. Again, the legislature has no initiative in terms of invoking collective responsibility. In determining whether or not a country should be classed as semipresidential, these rules are very different in nature from decisions that are based on whether or not a president has “quite considerable powers” or whether a president and prime minister “share powers” in a dual authority situation. For example, to classify France or any other country as semi-presidential on the basis of the president’s and/or the prime minister’s powers is to make an interpretation of the French political process as a whole. Inevitably, this interpretation is contestable. Moreover, whatever interpretation is made, the grounds for making it are necessarily rooted in the French experience. As a result, these grounds are unlikely to be transferable to all the other countries where equivalent interpretations have to be made. This is why a definition of semi-presidentialism based on the powers of political actors is essentially unreliable and will inevitably produce a different list of semi-presidential countries from one scholar to the next. In contrast, to decide whether or not include a country as semi-presidential only if there is some provision for ongoing responsibility to the legislature is to invoke a reliable classification rule. The rule may be contested. However, the grounds on which particular cases are included and excluded from the list of semi-presidential countries are transferable across the whole set of countries under consideration. If the rule is changed, then the set of countries that are affected can be determined and the set adjusted accordingly. On the basis of the core constitutional features of semi-presidentialism identified above plus a small number of subsidiary classification rules, the next section charts the spread of semi-presidentialism over time.

9780230_242920_02_cha01.indd 5

11/11/2010 5:05:50 PM

6 Robert Elgie

The spread of semi-presidentialism To determine the spread of semi-presidentialism, it is necessary to decide when a country becomes semi-presidential and when it ceases to be semipresidential. Given the definition of semi-presidentialism used here is based on constitutional provisions, a country is considered to have become semi-presidential when it adopts a constitution or passes a constitutional amendment that establishes a semi-presidential constitutional framework as outlined in the previous section. By the same token, a country is considered to have ceased to be semi-presidential when the constitutional framework is changed such that the conditions for semi-presidentialism no longer apply. In addition, a country is deemed to have ceased to be semipresidential if the constitution is formally suspended, perhaps as the result of a military coup. The effect of these rules means that the starting date of semi-presidentialism can predate the first direct presidential election and/or the first parliamentary election under the new constitutional framework. So, in 2007 the Turkish constitution was amended to create a directly elected president immediately after the president was elected indirectly under the previous system. Thus, Turkey is counted as having a semi-presidential constitution from 2007, even though the incumbent president was elected indirectly and even though there may not be a direct presidential election for another five years. The same situation occurred in France in the 1960s. Here, the constitutional amendment that instigated the direct election of the president was passed in 1962, but the first direct election of the president did not occur until 1965. So, France is counted as semi-presidential from 1962, despite the absence of a directly elected president until some years later. In addition, it should be noted that even the most autocratic regimes usually have constitutions. These constitutions may be merely “token” documents. They may not, in practice, be strictly enforceable. Nonetheless, on paper, these countries will have decided to adopt either a presidential, parliamentary, semipresidential, or some other constitutional organization of power. In that sense, their systems are constitutionally classifiable. In this chapter, therefore, no assumption is made about whether or not a constitution is founded on the rule of law and/or whether the constitutional situation is justiciable. Instead, the criteria for a country to be counted as semi-presidential are based simply on whether or not a country has adopted a constitution with a semi-presidential framework. How many countries have adopted semi-presidentialism? The first semi-presidential constitution was adopted by Finland on July 17, 1919. This was closely followed by the constitution of Weimar Germany that was adopted on August 11, 1919. For many decades thereafter, the spread of semi-presidentialism was extremely slow. By the beginning of

9780230_242920_02_cha01.indd 6

11/11/2010 5:05:50 PM

An Increasingly Common Constitutional Choice

7

the 1960s there were only five countries with semi-presidential constitutions. The wave of independence among former British colonies in the 1950s and 1960s failed to result in any semi-presidential constitutions, with these countries naturally adopting UK-style parliamentary regimes for the most part. Perhaps more surprisingly, the equivalent wave of independence among former French colonies in the period 1959–1961 also failed, initially, to result in any semi-presidential constitutions, but rather a mix of parliamentary and presidential regimes. By the end of the 1960s there were only eight countries with semi-presidential constitutions, and by the mid-1980s there were still only eight such countries. The spread of semi-presidentialism took hold in the early 1990s during the wave of democratization at that time. In the period 1990–1992 inclusive 29 countries adopted semi-presidential constitutions, bringing the total to 39 overall. Indeed, this figure is somewhat conservative. For the purposes of this chapter, the start date for semi-presidentialism in countries of the former Soviet Union is deemed to be the point when they adopted their first post-independence constitution. Thus, semi-presidentialism in Armenia and Azerbaijan began in 1995, while semi-presidentialism in Ukraine began in 1996. Yet, many of these countries operated with de facto semi-presidential constitutions immediately following independence, when directly elected presidents were grafted onto systems that were governed by Soviet- era parliamentary-style constitutions. However, given the seeming absence of consolidated constitutional documents for these countries in the immediate post-independence period, the later dates are recorded here, not least because the key point about the sudden burst of semi-presidential constitution-making in the period 1990–1992 remains valid. The period since the early 1990s has seen the adoption of about one new semi-presidential constitution each year. This is a greater rate of adoption than at any time prior to 1990. This does not mean that the spread of semi-presidentialism is now inevitable and sooner or later all countries will become semi-presidential. Indeed, in 2009 Madagascar and Niger were removed from the list of semi-presidential countries as result of a coup in the former and the adoption, more than controversially, of a new constitution in the latter. Similarly, in January 2010 Angola adopted a presidential constitution, thus abandoning semi-presidentialism. It simply means that the spread of semi-presidentialism was not a once- off occurrence that was confined to the early 1990s. As of March 31, 2010 there were 53 countries with semi-presidential constitutions on the basis of the classification rules outlined above. (See Table 1.1). Which countries have adopted semi-presidentialism? Prior to the 1970s there were very few semi-presidential countries they were dispersed and seemingly unrelated. The adoption of semi-presidentialism in Austria in 1929 may have been associated with its operation in neighboring

9780230_242920_02_cha01.indd 7

11/11/2010 5:05:50 PM

8 Robert Elgie Table 1.1 The spread of countries with semi-presidential constitutions Year

No.

Gained

Lost

1919

2

Finland, Germany

–

1929

3

Austria

–

1934

1

–

Austria, Germany

1937

2

Ireland

–

1940

3

Cuba

–

1944

4

Iceland

–

1945

5

Austria

–

1952

5

South Korea

Cuba

1954

4

–

South Korea

1955

5

Cuba

–

1959

4

–

Cuba

1962

6

France, South Korea

–

1967

7

South Vietnam

–

1970

9

Burkina Faso (Upper Volta), Senegal

–

1972

8

–

South Korea

1974

7

–

Burkina Faso (Upper Volta)

1975

6

–

South Vietnam

1976

7

Portugal

–

1978

9

Burkina Faso (Upper Volta), Sri Lanka

–

1979

10

Comoros

–

1980

11

Peru

–

1981

11

Central African Republic

Central African Republic

1983

9

–

Burkina Faso (Upper Volta), Senegal

1985

8

–

Comoros

1987

9

Haiti

–

1988

9

Tunisia

Haiti

1989

10

Algeria

–

1990

17

Cape Verde, Congo-Kinshasa, Mozambique, Namibia, Poland, Romania, São Tomé e Príncipe

–

1991

27

Burkina Faso, Cameroon, Croatia, Gabon, Haiti, Macedonia, Mauritania, Niger, Rwanda, Senegal, Singapore

Niger

Continued

9780230_242920_02_cha01.indd 8

11/11/2010 5:05:50 PM

An Increasingly Common Constitutional Choice

9

Table 1.1 Continued Year

No.

Gained

Lost

1992

39

Angola, Bulgaria, Burundi, Central African Republic, Comoros, Congo-Brazzaville, Lithuania, Madagascar, Mali, Mongolia, Slovenia, Togo

–

1993

40

Niger, Russia

Rwanda

1994

43

Guinea-Bissau, Moldova, Yemen

–

1995

46

Armenia, Azerbaijan, Kazakhstan, Tanzania

Burundi

1996

49

Belarus, Chad, Taiwan, Ukraine

Niger

1997

47

–

Congo-Brazzaville, Congo-Kinshasa

1998

48

Kyrgyzstan

–

1999

50

Niger, Slovakia

–

2000

49

–

Comoros

2001

48

–

Moldova

2002

49

Timor-Leste

–

2003

48

Rwanda

Central African Republic, Kyrgyzstan

2004

49

Georgia

–

2005

51

Egypt, Central African Republic, Kyrgyzstan

Mauritania

2006

55

Congo-Kinshasa, Mauritania, Montenegro, Serbia

–

2007

56

Turkey

–

2008

56

Kenya

Mauritania

2009

55

Mauritania

Madagascar, Niger

Weimar Germany at that time. For the most part, though, there was little semi-presidential mimetism. Indeed, if anything, secondary sources suggest that, ironically, the US, with its long-standing directly elected president rather than its other constitutional features, was the main role model for constitution-makers in some of the earliest semi-presidential countries. For example, events in South Korea were strongly influenced by the US, even though the original version of the 1948 constitution did not include a directly elected president (Ahn 1997: 97–101). In the early 1970s the spread of the French constitutional model first became apparent. It should be remembered that France only adopted semipresidentialism in late 1962 when a constitutional amendment to directly elect the president was approved in a referendum. Therefore, when former

9780230_242920_02_cha01.indd 9

11/11/2010 5:05:51 PM

10 Robert Elgie

French colonies won independence in the period from 1959–1961 France was still parliamentary. This helps to account for the absence of semipresidentialism in Francophone Africa at that time. By the early 1970s, though, a number of countries in this region began to adopt new constitutions. With links to France still extremely strong, some of these countries adopted post-1962 French-style semi-presidential constitutions. Indeed, in the period from 1970–1989 nine different countries adopted semipresidential constitutions, six were former French colonies. In addition, the French influence in Haiti was at least one of the reasons why a semipresidential constitution was adopted there in 1987. Moreover, secondary sources indicate that the French model was the direct inspiration behind the adoption of semi-presidentialism in Sri Lanka in 1978 (Wilson 1980), even though there are considerable constitutional differences between the two systems. In the early 1990s the influence of the French model on Francophone Africa became particularly noticeable. In his speech to the Franco-African summit at La Baule in 1990, President Mitterrand explicitly linked the continuance of French aid to democratization (Mitterrand 1995: 328–341). In the end, the link was observed as much in the breach as in the observance, but in the early 1990s many former French colonies adopted new constitutions, or constitutional amendments, as a way of signaling, at least nominally, their willingness to accept this change in French policy. As a result, 13 of 23 Francophone African countries, including those in the Maghreb and three former Belgian Francophone colonies, now have semi-presidential constitutions. (See Table 1.2.) Indeed, of the remaining seven Francophone African countries, the Comoros, the Republic of Congo (Congo-Brazzaville), and Burundi adopted semi-presidentialism at some point after 1990 only Table 1.2 Constitutional systems in Francophone Africa (as of March 31, 2010) Semipresidentialism Algeria Burkina Faso Cameroon Central African Rep. Chad Dem. Rep. of Congo Gabon Mali Mauritania Rwanda Senegal Togo Tunisia

9780230_242920_02_cha01.indd 10

Presidentialism Benin Burundi Comoros Republic of Congo Côte d’Ivoire Djibouti Guinea Niger (since 2009)

Parliamentary monarchy

Transitional regime

Morocco

Madagascar

11/11/2010 5:05:51 PM

An Increasingly Common Constitutional Choice

11

to change their system once again at a later date. In total, more than onequarter of the total number of semi-presidential countries in the world are Francophone countries in Africa. In the early 1990s, two further politico- cultural influences also helped the spread of semi-presidentialism. The first was the diffusion of the Portuguese model of semi-presidentialism. In 1976 Portugal adopted a semi-presidential constitution. Around that time, most of Portugal’s overseas colonies also gained independence. However, most adopted single-party Marxist systems. In the early 1990s, these countries embarked on a process of democratization and adopted the Portuguese model of semi-presidentialism as their template. With the creation of the state of Timor-Leste in 2002, all six former Portuguese colonies had adopted this form of constitution, though, as noted above, Angola has since recently abandoned semi-presidentialism in favor of presidentialism. The second influence was the adoption of semi-presidentialism in the former Soviet Union and the countries of the former Yugoslavia. (See Table 1.3.) In the former Soviet Union, most countries immediately adopted directly elected presidents after the collapse of the USSR. In most of these cases, though, the essential parliamentary features of the Soviet era remained in place. These countries emerged, more by chance than design it seems, with constitutions that exhibited the basic features of semi-presidentialism. Indeed, even those countries that adopted presidentialism usually included some semi-presidential features in their constitutions. For example, Tajikistan and Uzbekistan all have prime ministers. Here, the legislature has to confirm the presidential appointment and dismissal of the prime Table 1.3 Constitutional systems in countries of the former Soviet Union and the former Yugoslavia (as of March 31, 2010) Semipresidentialism Armenia Azerbaijan Belarus Croatia Georgia Kazakhstan Kyrgyzstan Lithuania Macedonia Montenegro Russia Serbia Slovenia Ukraine

9780230_242920_02_cha01.indd 11

Presidentialism

Parliamentary Hybrid

Tajikistan Turkmenistan Uzbekistan

Estonia Kosovo Latvia Moldova

Bosnia

11/11/2010 5:05:51 PM

12 Robert Elgie

minister. Thus, these countries have some features of executive responsibility to the legislature, even if these features are insufficient to constitute semi-presidentialism as defined in this chapter. In the former Yugoslavia, five of the seven currently constituted countries have semi-presidential constitutions and there are even semi-presidential elements to Bosnia’s system. In total, over half of all the countries that currently have semi-presidential constitutions can be found in Francophone countries, Lusophone countries, and countries of the former Soviet Union and the former Yugoslavia, while nearly two-thirds of all countries in these regions/contexts have adopted semi-presidentialism as their basic constitutional form. That said, the spread of semi-presidentialism is not confined to a restricted set of countries in certain geographical areas and/or cultural contexts. Indeed, only in Central and South America and the Caribbean is semi-presidentialism absent from a large bloc of countries. Peru has had a prime minister since 1861 and a tradition of at least some form of collective cabinet responsibility to the legislature dating back to the 1933 constitution, Article 172. In addition, Haiti adopted a semi-presidential constitution in 1987. However, even though the adoption of semi-presidentialism has been debated in countries in this region, notably Brazil and, more recently, Mexico, semi-presidentialism has not been chosen. Elsewhere, semi-presidentialism has taken hold. There are long-standing semi-presidential countries in Western Europe, as well as more recent examples in Central and Eastern Europe. There are examples of semi-presidentialism in Asia (Mongolia, Taiwan, and previously South Korea and South Vietnam) and in former British colonies in East Africa (Kenya and Tanzania). In the Middle East, Egypt and Yemen have semi-presidential constitutions, and, although not an independent state, the constitution of the Palestinian Authorities is also semi-presidential. Thus, while concentrated in certain regions/cultural contexts, semi-presidentialism is now a worldwide phenomenon. Why have countries adopted semi-presidentialism? The presence of semi-presidentialism in so many Francophone and Lusophone countries, and countries of the former Soviet Union and the former Yugoslavia suggests that the spread of semi-presidentialism is primarily associated with particular regional and/or cultural contexts. However, the fact that semi-presidentialism is not confined to these regions/contexts is a sign that there are also independent reasons for the spread of this form of government. Moreover, even in the regions/contexts where semipresidentialism is particularly prevalent, there is no evidence to suggest that it was ever imposed on an unwilling country. In the case of Francophone Africa in the early 1990s, there was certainly an external incentive to adopt a democratic political system and to do so quickly, but there was no obligation to adopt semi-presidentialism. Instead, the presence of semi-presidentialism in many Francophone and Lusophone countries, and in many countries of

9780230_242920_02_cha01.indd 12

11/11/2010 5:05:52 PM

An Increasingly Common Constitutional Choice

13

the former Soviet Union and the former Yugoslavia is more associated with an extremely convenient process of institutional mimetism at a time when countries needed to democratize quickly and/or when there was constitutional chaos as a result of unexpected independence/statehood. In other words, the adoption of semi-presidentialism was the result of a particular constitutional contagion effect in the early 1990s. However, this effect was at best a necessary and not a sufficient condition for semi-presidentialism to emerge. More generally, there are other political reasons why semipresidentialism has been such an attractive constitutional option for so many countries, including those in Francophone and Lusophone areas, and in the former Soviet Union and the former Yugoslavia. First, semi-presidentialism represents a convenient compromise option when there are internal political divisions between those who support the adoption of presidentialism, on the one hand, and those who support the adoption of parliamentarism on the other. Assuming neither side can dominate the constitution-making process, then semi-presidentialism can represent an honorable outcome for both sides as each can claim to have won something for itself and also to have prevented the other side from achieving everything that it wanted. This motivation was particularly important during the process of transition in Central and Eastern Europe in the early 1990s, particularly in countries such as Lithuania and Slovenia. In addition, this motivation was also present in Francophone Africa in countries that held so- called “national conferences” in the early 1990s. These conferences brought together political parties and civil society decision makers. They were held in seven countries – Benin, Chad, Congo-Brazzaville, Gabon, Mali, Niger, and Togo – six of which adopted semi-presidentialism. In Congo-Brazzaville, Mali, and Niger, in particular, semi-presidentialism was not merely the result of external French incentives, but was also the result of internal political dynamics between supporters of competing constitutional options. Second, semi-presidentialism is also attractive to countries that wish to demonstrate their democratic credentials to the outside world. This motivation has manifested itself in two very different ways. In Iceland and Ireland, a directly elected president was deliberately chosen to mark a symbolic break with a colonial power that operated under a parliamentary monarchy. In these cases, figurehead presidents, but nonetheless popularly elected presidents, were grafted onto existing parliamentary systems as a sign that the newly sovereign country was “more democratic” than the former colonial monarchy. In other countries, semi-presidentialism has been attractive to powerful leaders, or autocrats, who wish to remain in office but who also wish to establish a veneer of democracy. Under semi-presidentialism, the fact that the president is elected by the people and that the government is responsible to an assembly that is elected by the people can be used as a sign that the regime is committed, ostensibly at least, to democratic principles. For example in Burkina

9780230_242920_02_cha01.indd 13

11/11/2010 5:05:52 PM

14 Robert Elgie

Faso the adoption of semi-presidentialism suited the interests of the powerful incumbent president, Blaise Compaoré. In the context of a dominant party regime with little expectation of an opposition electoral victory, it allowed the president to retain power, while at the same time claiming that the democratic credentials of the new constitution were strong. Again, the adoption of semi-presidentialism was more than just the result of pressure from an external political force. It was also the result of domestic preferences. Third, the adoption of semi-presidentialism is consistent with an unstable country wanting to reinforce executive authority and create conditions for strong leadership. This can be seen most clearly when countries adopt semi-presidentialism by passing a constitutional amendment to introduce the direct election of the president. This has happened in Egypt, France, Poland, Singapore, Slovakia, Sri Lanka, Taiwan, Turkey, and Yemen – in six of which – France, Poland, Sri Lanka, Taiwan, Turkey, and Yemen – there was a specific desire to increase the authority of the incumbent president specifically or the presidency in general by so doing. The French case is the most well-known in this regard. The Fourth Republic from 1946–1958 exhibited a high degree of prime ministerial turnover. In 1958, the Republic collapsed and Charles de Gaulle was asked to return to power. He agreed, subject to the passage of a new constitution with increased executive and particularly presidential powers. The 1962 constitutional amendment that created the direct election of the president was a way of ensuring that the presidentialisation of the political system would outlast the term of the first incumbent of the office. In Yemen, the direct election of the president replaced a more collegial presidential council and was a way of reinforcing the position of President Ali Abdullah Saleh in the context of ongoing conflict. Finally, in different circumstances, the adoption of semi-presidentialism is consistent with a country wanting to reduce, or be seen to reduce, the power of the president by increasing the role of the legislature. This can be seen most clearly when countries adopt semi-presidentialism by passing an amendment that creates the post of prime minister and/or that makes the cabinet collectively responsible to the legislature. This has happened in Belarus, Cameroon, Gabon, Georgia, Kenya, Senegal, Tanzania, and Tunisia. In all cases except Belarus, this change was a signal, disingenuous or otherwise, that the incumbent president would no longer play such an active role in the political system, or at least that there would be an institution to balance the power of the presidency. For example, in Cameroon and Gabon there were severe antigovernment protests in the early 1990s and the creation of a semi-presidential system was a way in which the incumbent president tried to placate political opposition. The adoption of semi-presidentialism in Georgia was a direct result of the so-called “rose revolution” that was a mass opposition movement to the highly presidentialized political regime in the country previously. When the old regime was overthrown, the new president created the post of prime minister as a way of signaling that there had been a fundamental change in

9780230_242920_02_cha01.indd 14

11/11/2010 5:05:52 PM

An Increasingly Common Constitutional Choice

15

the political system. In Kenya, the adoption of semi-presidentialism followed the disputed results of the December 2007 parliamentary election, the result of which the opposition claimed was rigged by President Kibaki and his supporters. The violence that followed the election was ended by a pact between President Kibaki and the opposition leader, Raila Odinga, which was based on a constitutional amendment creating a so-called “executive” prime minister and a government that was responsible to parliament. These scenarios suggest that the spread of semi-presidentialism is unlikely to be confined to restricted geographical areas or cultural contexts, such as Francophone and Lusophone Africa and countries of the former Soviet Union and the former Yugoslavia. Instead, semi-presidentialism has emerged in a variety of political situations. As a result, there are grounds to believe that the incremental spread of semi-presidentialism may continue in the future.

Operationalising the concept of semi-presidentialism Maurice Duverger (1978, 1980, 1986) was the first to identify the concept of semi-presidentialism. In his work on this topic, Duverger’s sole aim was to explain why presidential power varies both across time within individual countries and across countries generally. At the time, the only countries with semi-presidential constitutions were in Western Europe. Talking about these countries, he stated, “the purpose of the concept of semi-presidential government is to explain why relatively homogeneous constitutions are applied in radically different ways” (ibid.: 177). To this end, he proposed four factors, “the actual content of the constitution, the combination of tradition and circumstances, the composition of the parliamentary majority, and the position of the president in relation to this majority” (ibid.). In other words, the definition of semi-presidentialism acted as Duverger’s case-selection mechanism; actual presidential power was his dependent variable; and he interacted four independent variables to explain the variations in presidential power within and across the six West European semipresidential cases in his set. Since Duverger’s work on the topic, the focus of inquiry has completely shifted. From the early 1990s onwards, semi-presidentialism has been explicitly compared with presidentialism and parliamentarism with the aim of determining which form of government is best for countries in the process of democratization. The standard academic consensus on semi-presidentialism was forged in this period and the consensus was hostile to this form of government. The advantages of semi-presidentialism were deemed to be threefold: a fixed-term president could provide stability in the context of a political crisis; a prime minister responsible to parliament could provide political flexibility; the presence of a dual executive with both a president and a prime minister could allow some degree of power-sharing between competing parties and avoid a potentially dangerous winner-takes-all scenario. The disadvantages of

9780230_242920_02_cha01.indd 15

11/11/2010 5:05:52 PM

16 Robert Elgie

semi-presidentialism were also threefold: the direct election of the president could encourage the rise of populist and autocratic leaders who felt that they had the legitimacy to act above the rule of law; the responsibility of the prime minister to the legislature could lead to governmental instability in the face of a fragmented legislature; the presence of a dual executive could institutionalize conflict within government, especially during periods of so-called cohabitation, leading either the president to assume power in order to break the deadlock, or to a stalemate in which the military decides to intervene to ensure effective government. In the context of these arguments, the main supporter of semipresidentialism as a model for new democracies was Giovanni Sartori (1997). However, even Sartori’s support was somewhat less than wholehearted: . . . my argument is not – in the context of parliamentarism – that semipresidentialism is “best” but, rather, that it is “more applicable” . . . And I further wish to underscore that this recommendation is not a strong one. Semi-presidentialism does leave us with unsettled problems. Nor do I deny that semi-presidentialism is a somewhat fragile system. (Sartori 1997: 137) Instead, Sartori recommends what he calls “alternating presidentialism” (ibid.: chap. 9). This system may be compatible with semi-presidentialism, but it is far from a ringing endorsement of this regime type. For the most part, academic opinion has been opposed to semipresidentialism. For example, Linz was worried about the presence of multiple actors within a semi-presidential executive, fearing that the complexity of the decision-making process would encourage the military to assume power to restore clarity (Linz 1994: 59). There was also a general worry about the effects of cohabitation, where the president and prime minister are from opposing parties and where the president’s party is not represented in the government. Linz and Stepan (1996: 286) feared that cohabitation would be destabilizing in the context of a country that was already facing the general problems of democratization. Given the potential for cohabitation is inherent under semi-presidentialism, this led most observers to recommend against this form of government. Recently, Samuels and Shugart (2010) have given qualified support to a particular form of semi-presidentialism. In general, though, the consensus against semi-presidentialism is still current and can be seen, for example, in the expert briefings relating to the formulation of the constitution of Afghanistan in 2003. (See, for example, Rubin 2003). While there is now a body of work on semi-presidentialism, the student of semi-presidentialism faces a dilemma. Much of the existing work is based on observations about countries where a Duverger-style or Sartori-style behavioral definition of semi-presidentialism is either explicitly used or is at least implicitly assumed. However, we have shown that there is a problem

9780230_242920_02_cha01.indd 16

11/11/2010 5:05:52 PM

An Increasingly Common Constitutional Choice

17

with operationalizing semi-presidentialism in this way because it leads to a case-selection bias. The solution, we argued, was to adopt a definition of semi-presidentialism on the basis of purely constitutional criteria. However, if we do so, then we face a different problem. In some countries with semipresidential constitutions the president is the main actor and the prime minister is required simply to carry out the president’s policy. Mozambique is an example. In other semi-presidential countries, the president is a figurehead and the prime minister enjoys great constitutional powers. Examples include Iceland and Ireland. In yet other countries, there is indeed a constitutional balance of power between the president and prime minister. France is the standard case. In other words, there is extreme variation within the set of countries with semi-presidential constitutions. Given this situation, we should not expect semi-presidentialism to be associated with particular outcomes. To put it another way, we should not operationalize semipresidentialism as a discrete explanatory variable. We should not compare semi-presidentialism with presidentialism or parliamentarism. The solution to the dilemma of studying semi-presidentialism is to avoid the selection-bias problem by adopting a definition that is based on the purely constitutional criteria and that reliably identifies a set of countries with semi-presidential constitutions, but to identify systematic institutional variation within the set of semi-presidential countries and to study the effects of this variation both within the universe of semi-presidentialism alone and between this set of countries and countries with presidential and parliamentary countries. There are various ways of capturing institutional variation within the set of countries with semi-presidential constitutions. The most common way is to use Shugart and Carey’s (1992) distinction between countries with president-parliamentary constitutions and those with premier-presidential constitutions. In president-parliamentary systems the prime minister is responsible to both the legislature and the president. In contrast, in premier-presidential systems the prime minister is responsible solely to the legislature. This strategy can be used to study the effects of variation solely within the set of semi-presidential countries. For example, both Protsyk (2005a, b) and Roper (2002) have also focused on the impact of institutional variation solely within the universe of semi-presidentialism. In addition, though, this strategy can be used to compare the effects of institutional variation more generally. Samuels and Shugart (2010) have identified the consequences of the variation between the two types of semipresidentialism and presidentialism and parliamentarism in terms of relations between presidents and parties. (Table 1.4 identifies the sub-types of semi-presidentialism on the basis of this distinction and using the constitutions currently in operation in semi-presidential countries). A further way of capturing the institutional variation within semipresidentialism is to measure the constitutional powers of presidents. There are various attempts at such measures and the inclusion of particular

9780230_242920_02_cha01.indd 17

11/11/2010 5:05:53 PM

18 Robert Elgie Table 1.4 Variation within semi-presidentialism

Country Algeria Armenia Austria Azerbaijan Belarus Bulgaria Burkina Faso Cameroon Cape Verde Central African Republic Chad Dem. Rep. of Congo Croatia East Timor Egypt Finland France Gabon Georgia Guinea-Bissau Haiti Iceland Ireland Kazakhstan Lithuania Macedonia Mali Mauritania Mongolia Montenegro Mozambique Namibia Peru Poland Portugal Romania Russia Rwanda São Tomé e Príncipe Senegal Serbia Slovakia Slovenia

Shugart and Carey 1= premier-presidential 0 = presidentparliamentary 1 1a 0 0 0 1 0 1 1 0 1 1 1b 1 1 1 1 0 0 0 1 0 1 0 1 1 1 0 1 1 0 0 0 1 1c 1 0 0 1d 0e 1 1 1

Siaroff (min. 1/max. 9) 7 6 1 7 7 3 6 n/d 6 6 n/d 6 4 4 n/d 2 7 7 6 6 6 1 3 5 4 4 7 7 4 2 8 7 7 3 3 5 7 n/d 8 7 2 2 1 Continued

9780230_242920_02_cha01.indd 18

11/11/2010 5:05:53 PM

An Increasingly Common Constitutional Choice

19

Table 1.4 Continued

Country Sri Lanka Taiwan Tanzania Togo Tunisia Turkey Ukraine Yemen

Shugart and Carey 1= premier-presidential 0 = presidentparliamentary 0 0 0 1 0 1 1f 0

Siaroff (min. 1/max. 9) 7 5 7 7 7 3 5 4

Notes: a Armenia was president-parliamentary until a constitutional amendment in 2005. b Croatia was president-parliamentary until a constitutional amendment in 2000. c Portugal was president-parliamentary until a constitutional amendment in 1982. d São Tomé e Príncipe was president-parliamentary until a constitutional amendment in 2002. e Senegal was premier-presidential until a constitutional amendment in 2001. f Ukraine was president-parliamentary until a constitutional amendment in 2006. Source: Siaroff (2003), plus author’s calculations for other countries, except for those marked n/d, that is, those that have never been democratic.

indicators rather than others is contestable. However, whatever indicators are used, the criteria are usually reliable because they are based on constitutional powers and the calculations are replicable. A recent example is Siaroff’s (2003) calculations of presidential powers. He identifies nine constitutional indicators of presidential power. He gives a value of 1 if the constitution includes the indicator and 0 otherwise. He then measures the powers of presidents cross-nationally within a range of 0 to 9. Given one of Siaroff’s indicators is whether or not the president is directly elected, the minimum score for a country with a semi-presidential constitution is 1. (Table 1.4 provides the Siaroff scores for countries with semi-presidential constitutions and illustrates the institutional variation within semi-presidentialism.) Again, this approach allows the study of variation solely within the set of semi-presidential countries as well as the comparison of countries with presidential, parliamentary and semi-presidential constitutions. For example, it might be hypothesized that countries with more powerful presidents are less likely to democratize successfully. This hypothesis could be tested either solely on countries with semi-presidential constitutions or more generally, and Siaroff provides scores for countries with presidential and parliamentary constitutions to facilitate the more general comparison. To sum up, the concept of semi-presidentialism should not be used to imply that semi-presidential countries operate in a particular way. To do so is merely to focus on a subset of countries with semi-presidential

9780230_242920_02_cha01.indd 19

11/11/2010 5:05:53 PM

20

Robert Elgie

constitutions. At the same time, the variety within the full set of semipresidential countries is so great that semi-presidentialism should not be expected to be associated with a particular set of outcomes. To study the full set of semi-presidential countries, the variation within semi-presidentialism needs to be captured systematically. This variation can be achieved by way of a dichotomous categorical variable like president-parliamentarism versus premier-presidentialism. A distinction along these lines facilitates the comparison of the effects of institutional variation between these sub-types of semi-presidentialism as well as comparisons between these sub-types and both presidential and parliamentary countries. The variation can also be achieved by way of a continuous ordinal variable based on Siaroff scores. Again, this sort of distinction allows comparisons to be made solely within the universe of semi-presidentialism as well as between countries with presidential, parliamentary, and semi-presidential constitutions.

Conclusion Semi-presidentialism is the situation where there is a directly elected fixedterm president and a prime minister and cabinet responsible to the legislature. This definition of semi-presidentialism is based upon a small set of constitutional features and allows the reliable identification of countries with semi-presidential constitutions and, for that matter, the reliable identification of countries with presidential and parliamentary constitutions as well. There was a dramatic spread of semi-presidentialism in the early 1990s. However, countries continue to adopt semi-presidentialism for a variety of political reasons, suggesting that the current number of countries with semi-presidential constitutions may increase further over time. A key feature of semi-presidentialism is the extreme variety of political practice within the universe of semi-presidential countries. However, to say that Ireland is semi-presidential or that Russia is semi-presidential is not to imply that Ireland and Russia operate in the same way or that they both operate like France. To operationalize a concept that incorporates countries where the constitutional situation is as varied as in France, Ireland, and Russia requires the identification of sub-types of semi-presidentialism. There are various ways in which such sub-types can be identified. By identifying the institutional variation within semi-presidentialism, comparisons can be made solely within the universe of semi-presidential countries or between the different types of semi-presidentialism and countries with presidential and parliamentary constitutions. This strategy provides a way of systematically introducing the concept of semi-presidentialism into the research agenda and advancing our knowledge of the effects of the institutional variety both within this increasingly popular form of government and between countries with this form of government and those with other forms.

9780230_242920_02_cha01.indd 20

11/11/2010 5:05:54 PM

2 Clustering of Semi-Presidentialism: A First Cut Yu-Shan Wu

Semi-presidentialism is a constitutional system that combines the core features of presidentialism and parliamentarism. It has a popularly elected president and a government headed by a prime minister who is responsible to the parliament. It was a phenomenon originally confined to Western Europe. It then spread to third-wave democracies in the 1990s. There were two major directions of its expansion: into the post-Leninist countries in Eurasia, and into the postcolonial world, particularly the Francophone and Lusophone countries.1 These three clusters of semi-presidential countries (West European, post-Leninist, postcolonial) comprise the majority of contemporary semi-presidential countries in the world. The way the system operates and performs is directly related to its expansion into different countries, with their varying institutional legacies and political traditions. As a result of its extensive spread, semi-presidentialism demonstrates a wide range of operational forms, with some bordering on parliamentarism, some on presidentialism, and yet others in between, or in constant fluctuation. Without a grasp of the clusters of countries that adopted a semi-presidential system and their main features, it is difficult to appreciate the extensively variable political phenomena associated with semi-presidentialism, and to identify the factors that cause such great variation. This chapter starts with the identification and differentiation of semipresidential countries in the world. We differentiate between democratic and nondemocratic regimes among the constitutionally semi-presidential countries and concentrate on the former to have a focused discussion.2 Next we proceed to cluster semi-presidential regimes, identify the entry mode of different clusters, spell out their operational subtypes, and establish the relationship between clusters and subtypes. By cluster we mean a group of countries that share an important institutional legacy, political tradition, or developmental history. By entry mode we mean the pattern of adoption of semi-presidentialism (upstream). By subtype of operation we mean the way in which the prime minister and the cabinet are held accountable, based on how they are appointed and dismissed (midstream). Subtype defines where 21

9780230_242920_03_cha02.indd 21

11/11/2010 5:09:10 PM

22 Yu-Shan Wu

the ultimate power of the state is located: in the hands of the president, the parliament, or both. Subtype is important by itself, for it identifies who has the ultimate power, and which political party is the ruling party. It is also important in affecting how the system performs, particularly under a divided government. In sum, we find semi-presidential countries in clusters with different entry modes and operational subtypes. How a specific semipresidential system operates has a lot to do with the cluster in which the system finds itself. The institutional legacy and political tradition prevalent in the cluster are important factors that we will look into to understand the divergent patterns of semi-presidentialism around the world.3

Global distribution of semi-presidentialism In 2009, there were 41 semi-presidential democracies in the contemporary world.4 They are readily clustered into three groups: established democracies, post-Leninist countries, and postcolonial countries (see Table 2.1). The last two groups are mostly nascent democracies. All the established semipresidential democracies are found in Western Europe and Scandinavia. They include Austria, Finland, France, Iceland, Ireland, and Portugal. PostLeninist semi-presidential countries are primarily in Eastern Europe and the former Soviet Union, but also sparsely in Asia. They can be subdivided into East European post-satellites, post-Soviet successor states, and Asian postLeninists. The former satellites are Bulgaria, Croatia, Macedonia, Poland, Romania, Slovakia, and Slovenia. The post-Soviets include Armenia, Georgia, Kyrgyzstan, Lithuania, Russia, and Ukraine. The Asian post-Leninists are Mongolia and Taiwan. Taiwan is unique in being the only post-Leninist, but not post- Communist country.5 The third cluster of semi-presidential countries is postcolonial. Most of them are either Francophone or Lusophone. The Francophone include Burkina Faso, Central African Republic, Gabon, Haiti, Madagascar, Mali, Niger, Senegal, and Togo; the Lusophone are Cape Verde, East Timor, Guinea-Bissau, Mozambique, Sao Tome, and Principe. Other postcolonial countries include Namibia, Singapore, Sri Lanka, Tanzania, and Yemen. The only semi-presidential country that does not fit into the three main clusters is Peru. As can be clearly seen in Table 2.1, the established democracies in Europe are the precursors of semi-presidentialism. They adopted the system between 1919 (Finland) and 1976 (Portugal). The post-Leninists and postcolonials are mostly third-wave nascent democracies that adopted the semi-presidential constitutional system in the 1990s. Among the 35 non-precursor countries, only two entered the semi-presidential club prior to the decade (Sri Lanka in 1978 and Peru in 1980), and three after the decade (East Timor in 2002, Georgia in 2004, and Kyrgyzstan in 2005). We can thus clearly identify two waves of expansion of semi-presidentialism: a gradual trend in Western Europe and Scandinavia from the 1910s to the 1970s, and a massive trend in

9780230_242920_03_cha02.indd 22

11/11/2010 5:09:10 PM

Table 2.1 Clustering of semi-presidential countries Cluster and subcluster

Country

Entry year Note

Western Europe

Austria

1929–1934 1945 1919 1962 1944 1937 1976

Finland France Iceland Ireland Portugal PostLeninist

East European Post- Satellites

Post- Soviet Successor States

Asian PostColonial

Bulgaria Croatia Macedonia Montenegro* Poland Romania Serbia* Slovakia Slovenia Armenia Azerbaijan* Belarus* Georgia Kazakhstan* Kyrgyzstan Lithuania Russia Ukraine Mongolia Taiwan

Francophone Africa Algeria* Burkina Faso

1992 1991 1991 2006 1990 1990 2006 1999 1992 1995 1995 1996 2004 1995 1998–2003 2005 1992 1993 1996 1992 1997

1989 1970–1974 1978–1983 1991 Cameroon* 1991 Central African 1981–1981 Republic 1991–2003 2005 Chad* 1996 Dem. Rep. of 1990–1997 Congo* 2006 Gabon 1991 Madagascar 1992–2009 Mali 1992 Niger 1991–1991 1993–1996 1999–2009 Senegal 1970–1983 1991

New nation

New nation

Nondemocratic Nondemocratic Nondemocratic

Nondemocratic

Nondemocratic

Nondemocratic Nondemocratic

Continued

9780230_242920_03_cha02.indd 23

11/11/2010 5:09:10 PM

24 Yu-Shan Wu Table 2.1 Continued Cluster and subcluster

Lusophone

Others

Country Togo Tunisia* Other Haiti areas Africa Angola* Cape Verde Guinea-Bissau Mozambique Sao Tome and Principe Other East Timor areas Egypt* Kenya* Namibia Rwanda* Singapore Sri Lanka Tanzania Yemen

Others

Peru Turkey*

Entry year Note 1992 1988 Nondemocratic 1987–1988 1991 1992 Nondemocratic 1990 1994 1990 1990 2002 2005 2008

Nondemocratic Entry recent and precarious

1990 1991–1993 Nondemocratic 2003 1991 1978 1995 1994 1980 2007

Not practiced

Note: * Non-democratic and dubious cases. See note 4.

Eurasia (primarily in Eastern Europe and former Soviet Union), as well as in the postcolonial world (particularly in Africa) in the 1990s (for a graphic view of the expansion of semi-presidentialism, see Figure 2.1). Even if we include all the nondemocratic and dubious cases, that is, the 15 asterisked countries in Table 2.1, the picture remains the same. In short, the semi-presidential countries in the world demonstrate heavy temporal and geographic concentration. This concentration will have a profound impact on the system.

Entry mode Among the three main clusters of semi-presidential countries, the established democracies in Western Europe adopted their constitutional order primarily to deal with political crisis (Finland, Weimar Germany, Austria, France, and Portugal)6 or to affirm national sovereignty (Ireland and Iceland).7 Those are situational requirements not rooted in any structural background. The entry mode is ad hoc and contingent.8 Post-Leninist semi-presidential regimes tell a different story. The dual power structure under the old system (general secretary cum president vs. premier

9780230_242920_03_cha02.indd 24

11/11/2010 5:09:10 PM

Clustering of Semi-Presidentialism: A First Cut

25

45 1

40 1

Number of SP Countries

35 19

30

18

25 20 15 14

15

6

6

10 1

1 1

6

6

5 0

3

2

4

5

4

2

1910s 1920s 1930s 1940s 1950s 1960s 1970s 1980s 1990s 2000s Year Established democracies

Post-Leninists

Post-colonials

Others

Figure 2.1 The impact of third-wave democratization on semi-presidentialism

who is theoretically responsible to the parliament) led easily to the semipresidential regime after the democratic transition (Blondel 1992). The postcolonial semi-presidential regimes have yet another entry mode. There the demonstration effect of the colonial powers is of great importance, particularly among the Francophone and Lusophone countries. Combined with those macro mechanisms (political exigencies, dual power structure, and demonstration effect) are political calculations by actors to maximize their interests. When both macro and micro factors are present, a semipresidential system is formed. The way in which a country enters semi-presidentialism has to do with its subsequent political development. For established democracies in Europe that adopted semi-presidentialism under ad hoc circumstances, cabinet answerability to the parliament is a cardinal principle of their constitutional order, a necessary expression of democracy. Prior to their entry into semi-presidentialism, those countries had accumulated a rich parliamentary experience.9 Looking at the figures presented in Table 2.2, we find the existence of a robust parliamentary tradition at the time of entry into semi-presidentialism for all of the Western European semi-presidential regimes. In these countries, semipresidentialism came much later than the establishment of parliamentary

9780230_242920_03_cha02.indd 25

11/11/2010 5:09:10 PM

26 Yu-Shan Wu Table 2.2 Parliamentarism and West European semi-presidential regimes

Country Austria Finland France Iceland Ireland Portugal Weimar Germany Average

Founding time of functional parliament 1867 (Reichsrat)a 1906 (Eduskunta)b 1870 (Chamber of Deputies)c 1918 (Althingi)d 1919 (Dáil Éireann)e 1911 (Chamber of Deputies)f 1871 (Reichstag)g

Entry into SP

Years of parliamentarism prior to entry

1929 1919 1962

62 13 92

1944 1937

26 18

1976

65

1919

48 46.3

Notes: a The parliament set up in Austria- controlled Cisleithania. b Installed by Russia in 1906 after the Russo-Japanese War. c Created after the Franco-Prussian War and the onset of the Third Republic. d Established by the 1918 constitution as the national parliament. e The revolutionary parliament of the unilaterally declared Irish Republic. f Set up by the First Republic that overthrew the monarchy. g Established after German unification.

tradition; the average lapse is 46.3 years. This shows that before the popular election of president, parliaments had represented people and fought for goals such as national independence or democratization for many decades. When a directly elected president came into existence, parliamentarism had been quite firmly established. The institutional move was never meant to shake the fundamental principle of parliamentarism, but to cope with contingencies. The semi-presidential order that emerged is by and large a parliamentary system, in line with the institutional legacy and political tradition of those countries, with the elected president expected to exercise power only under extraordinary circumstances, that is, when there is “failure of parliament.” The president is typically a titular head of state. When he or she wields real power, it is usually because his party holds a majority in the parliament, and the president exercises power as leader of the majority party (coalition), not as president per se. Post-Leninist semi-presidential regimes are mostly in Eastern Europe, except Mongolia and Taiwan. Prior to democratic transition, those countries were all Leninist party-states. The paramount leader in this type of system is the general secretary of the party or first secretary, or party chairman. He also usually assumed the position of state chairman or president of the country. The government, on the other hand, is headed by the prime minister, who is usually the number two power holder in the country. Theoretically

9780230_242920_03_cha02.indd 26

11/11/2010 5:09:11 PM

Clustering of Semi-Presidentialism: A First Cut

27

speaking, the government as headed by the premier is responsible to the parliament.10 However, as the parliament is not elected through competitive, multiparty elections and the party is in full control of the country, formal cabinet responsibility to the parliament does not translate into real parliamentarism. Nevertheless, one does find the existence of a dual power structure (Blondel 1992). As the country went through democratic transition, multiparty competition enlivened the parliament and turned it from a rubber stamp into a real power centre. The role of the president, on the other hand, also experienced some change. It is only natural that both the elite and the general public in many of those countries would assume that nothing short of the direct election of the president can be considered a real democracy, hence the provision of direct presidential elections and the emergence of a semi-presidential constitutional structure.11 As the presidents in those countries are the successors to the all-powerful general secretaries cum state chairmen, and the political culture heavily favors strongman politics, the post-Leninist semi-presidential regimes are inherently different from the established semi-presidential democracies in that the presidential powers are much stronger, and the parliamentary tradition much weaker. This is more the case in post-Soviet countries than in Eastern Europe where the parliamentary wing is a bit stronger. But even in East European semipresidential countries, the parliament cannot be compared with their counterparts in established democracies. Unlike the established democracies, but quite like the post-Leninist regimes, the postcolonial semi-presidential countries owed their constitutional format more to structure than to exigencies. Among the 19 postcolonial semi-presidential regimes, 14 are former colonies of either France of Portugal, both semi-presidential countries in Western Europe. For the postcolonials, the dominant entry mode is demonstration by former colonial powers.12 Close connections between the legal and political communities between the former colonies and colonial powers, including language, training of lawyers and politicians, legal literature, constitutional advisors, provided a convenient mechanism for semi-presidentialism à la France or Portugal to be introduced to the Francophone or Lusophone world when the tide of democratization arrived (Conac 2007: 79; Elgie and Moestrup 2007b: 242). Among the Francophone countries in Africa, their independence in 1960 originally did not bring about any genuine democracies. Instead one finds the emergence of a host of autocratic regimes. Toward the end of the 1980s, the third wave of democratization reached those Francophone countries, and their constitutional choice was naturally some form of French-style semi-presidentialism. French constitutional lawyers even drafted many of the constitutions in those countries that set up institutions of the Fifth Republic.13 As for Lusophone countries in Africa, their independence came about 15 years later than their Francophone counterparts. When they were fighting for independence under Portuguese colonial

9780230_242920_03_cha02.indd 27

11/11/2010 5:09:11 PM

28 Yu-Shan Wu

rule, most of the independence movements were inspired by Marxism, and the major force for independence was leftist guerrillas. With the democratic transition in Portugal in the mid-1970s (the Carnation Revolution), the African Lusophone countries secured independence but turned themselves into leftist dictatorships. When the tide of democratization reached those countries in the 1990s, it was natural that they would adopt some form of semi-presidentialism, following the Portuguese precedent.14 This mode of institutional choice by way of demonstration does not necessarily hold true for Asian postcolonial regimes such as Sri Lanka (although it did hold true for East Timor). However, the bulk of the postcolonial semipresidential regimes (63 percent) are in Africa where demonstration effect was the dominant factor in bringing about semi-presidentialism. As such, it is different from the entry modes of established democracies in Europe, and post-Leninist countries in Eurasia. In terms of the balance of power between president and parliament, the postcolonials are characterized by stronger presidents and weaker parliaments than the post-Leninists. The principle of government responsibility to parliament had not even been stipulated in their authoritarian constitutions, let alone put into practice. In most of those countries, one did not find the position of prime minister under the authoritarian rule.15 In sum, when the three semi-presidential clusters are compared, the established democracies cluster demonstrates the strongest parliamentary tendency, while the postcolonials have the strongest presidents. The post-Leninists are in between, with the East European post-satellites tilting toward a more balanced president-parliamentary relation, and the postSoviets toward a much stronger president. By way of political contingencies, dual power structures, demonstration effects, and political calculations on the ground, many countries in the three clusters chose semi-presidentialism. When the president and the parliamentary majority are from the same political party, things go smoothly. However, when the presidency and the parliament are controlled by different political parties, the Achilles heel of semi-presidentialism is exposed, and the system undergoes great stress. One then finds the different inherent tendencies in the three semi-presidential clusters and different operational subtypes. In those countries where parliamentarism is deep-rooted and the entry mode reinforced the institutional legacy, the solution is respect for parliamentary authority, and restriction of presidential powers. This is what we typically see in the established semipresidential democracies in Europe. The situation in post-Leninist and postcolonial countries is quite different, where the respect for one leader of the country is deep-rooted. The cabinet is held responsible to the president, at the expense of parliamentary authority. The idea that a government that does not enjoy support from the parliament should be removed from power is alien for the population in many of those nascent democracies. A particularly destabilizing situation may appear when the natural tendency in the

9780230_242920_03_cha02.indd 28

11/11/2010 5:09:11 PM

Clustering of Semi-Presidentialism: A First Cut

29

country is toward presidential control over the cabinet, while the constitution specifically requires cohabitation under divided government. Under such circumstances the system may find itself under great stress that may lead to disruption or even democratic breakdown.

Divergent subtypes The most fascinating aspect of semi-presidentialism is its wide range of operational patterns. Because semi-presidentialism contains elements of parliamentarism and presidentialism, it is only natural that these two tendencies come into conflict. How such conflicts are resolved or not resolved are key questions. Depending on how the system copes with such inherent conflict, one can find different modes of operation, or subtypes of semipresidentialism. More specifically, if the system in question tilts toward presidentialism, then the president commands the premier and the cabinet. If the system follows the principles of parliamentarism, then the ultimate power resides in the leader of the parliamentary majority party (or coalition), who is typically the prime minister. Where the real power lies in a semi-presidential system is the first and most important question. The answer is not self- evident as in a parliamentary or in a presidential system. One way to approach the issue is to determine if there is dual accountability. Since, by definition, under semi-presidentialism the prime minister and cabinet are responsible to the parliament, the main variable is whether the premier is also responsible to the president, that is, whether there is dual accountability (the Shugart- Carey (1992) dichotomy). This is how one differentiates between premier-presidentialism where there is only single accountability, and president-parliamentarism, where dual accountability exists. When operationalized, the Shugart- Carey dichotomy hinges on whether there is a constitutional provision of the presidential power to dismiss the premier. In this way, a particular presidential power (dismissing premier) is singled out to act as the defining feature of the dichotomous typology. This Shugart- Carey dichotomy is clear and highly related to many performance variables of semi-presidentialism, hence its prevalence in the literature. However, it does not directly point out the location of real power in a semi-presidential system. One can imagine a semi-presidential system that is legally president-parliamentary where the president restricts his role as if he were in a parliamentary system, that is, reigning but not ruling. It is also possible that in a semi-presidential system that is legally premierpresidential the president asserts himself in the formation and dismissal of the prime minister and the cabinet. The gap between institutions and behavior in semi-presidential systems was pointed out in Duverger’s seminal article of 1980 that introduced the concept to the English literature. There are also problems in the particular way in which the Shugart- Carey dichotomy

9780230_242920_03_cha02.indd 29

11/11/2010 5:09:11 PM

30 Yu-Shan Wu

is operationalized. Even though it is important for the president to be able to remove the prime minister if he is to hold the premier accountable, one can argue that the power to appoint the premier is of equal, or even greater, importance. Furthermore, it is not always clear whether the president has the power to dismiss the premier simply by reading the constitution. One may have to consult actual practice in order to determine if such power does exist. However, the clarity of the original dichotomy would be lost if we take behavior into account. It is clear that the Shugart- Carey dichotomy may be an important clue to where the real power of the system lies, but it is not the answer itself. The relation between the president and the parliamentary majority, the party system in the parliament, and the traditions and circumstances all come into play (Duverger 1980). Whatever the factors behind the locus of power in a semi-presidential system, it is important to recognize the different operational modes, or subtypes, of semi-presidentialism. The subtypes must be able to point directly to the locus of ultimate power, that is, who determines the prime minister and cabinet. They also need to be sensitive to the different relations between the president and the parliamentary majority, that is, united vs. divided government. Four subtypes can be identified. They are quasi-parliamentarism (QP), presidential supremacy (PS), alternation (ALT), and compromise (COM). Quasi-parliamentarism (QP) is the subtype in which the president yields to parliament under both united and divided government. In this subtype even when the president’s own party holds majority in the parliament, the president will not exercise any control of the government. It is up to the prime minister, a co-partisan of the president, to wield ultimate power by commanding the majority support in the parliament and acting as head of government. Except for QP, all the other subtypes are characterized by presidential control of the premier and government under unified government. The differences lie in the president-parliamentary relations under divided government. Alternation (ALT) is the subtype in which the president is an all-powerful leader when his party controls the parliament. However, when that dominant position is lost, as after an electoral defeat, then the president would yield to whatever majority formed in the parliament, and duly appoint its leader the prime minister. This is cohabitation. A president in the ALT sub-type can be either all-powerful or weak, depending on whether the country finds itself under unified or divided government, hence the designation “alternation.” Compromise (COM) is close to ALT except the president will keep control of specific cabinet posts and/or maintain certain executive powers (viz. power-sharing) even under divided government. This subtype specifies a particular distribution of power between the president and the prime minister when the president’s party is not the parliamentary majority. Presidential supremacy (PS) is a subtype in which the president appoints his favorite choice as prime minister whether the presidential party controls the parliament or not. The president insists on his appointing

9780230_242920_03_cha02.indd 30

11/11/2010 5:09:12 PM

Clustering of Semi-Presidentialism: A First Cut

31

authority even when he is opposed by a parliamentary majority. Table 2.3 delineates the four subtypes and offers empirical examples. All the above modes can be in equilibrium, hence stable, or disequilibrium, hence unstable. Typically a semi-presidential country would experience great stress when it moves from unified to divided government, especially for the first time. As the process of government formation is not always clearly stipulated in the constitution, and even if it is politicians strive to find leeway to their or their parties’ advantage, how the government is formed and sustained always entails conflict and stress. Different political actors may envision different subtypes as the model to follow, say the president aspires to PS while the parliamentary majority desires ALT. In such cases, a clash is almost inevitable. If the actors can settle on equilibrium that reflects their respective capabilities and aspirations, then the system can remain stable, although not without frictions that are contained within the overall system. However, if equilibrium cannot be found, and actors openly clash for control over government, then the system is in disequilibrium. Political order may break down as extra- constitutional measures or military actions may be taken to overpower political rivals. The country’s democratic survival may be at stake. The subtype is closely linked to the country’s cluster background. The underlying variable is the balance of power between the parliament and the

Table 2.3 Operational subtypes under semi-presidentialism Subtype

Under united government

Under divided government

Quasiparliamentarism (QP)

President yields to copartisan in parliament to form government

President yields to opposition majority leader in parliament to form government

Iceland

Alternation (ALT)

President appoints his favorite prime minister and form government

President yields to opposition majority leader in parliament to form government

France

Compromise (COM)

President appoints his favorite prime minister and form government

President and opposition majority leader distribute cabinet positions and/or divide executive power

Poland (under Lech Wałęsa)

Presidential supremacy (PS)

President appoints his favorite prime minister and form government

President appoints his favorite prime minister and form government

Russia

9780230_242920_03_cha02.indd 31

Case

11/11/2010 5:09:12 PM

32 Yu-Shan Wu

president. When that balance tilts toward the parliament, as is most obvious among the West European semi-presidential regimes, because of the entrenched parliamentary tradition, then quasi-parliamentarism is a likely outcome. Alternation and compromise are also possible. Besides institutional legacy and political tradition, a pro-parliamentary constitutional structure and consolidated party system may also come into play. If the balance of power tilts toward the president, as in postcolonial semi-presidential countries, because of strongman political culture and institutional legacy, then presidential supremacy is mostly likely to occur. ALT and COM are also possible. A pro-presidential constitutional structure and fragmented party system may also come into play. If the balance of power is in between the above two cases, with an institutional legacy, political tradition, constitutional stipulation, and/or party system not decisively favoring either the parliament or the president, then the middle options (alternation and compromise) are more likely. One can also find presidential supremacy. This is the case with the post-Leninists. Among them, the post-Soviet republics tilt toward the presidential end much more than the East European post-satellites. Table 2.4 shows the subtypes of all the semi-presidential countries in the world. In order to demonstrate the distribution of subtypes across clusters, we assign scores to subtypes. QP is assigned 1 point, ALT 2, COM 3, and PS 4. The higher the score, the greater is the president’s power in controlling the premier and the cabinet. We find that among the three main clusters, West European countries have an average of 1.5, the lowest among the three clusters. This is because the dominant subtype there is QP (quasiparliamentarism), with the only exceptions of France (ALT) and Finland (COM). The post-Leninists average 2.47, higher than the West Europeans, but lower than the postcolonials. Among the post-Leninists, the East European post-satellites have an average of 1.71, reflecting the fact that QP and ALT are the dominant subtypes there. The post-Soviet successor states have a higher average (3.17) than their East European counterparts. This is because PS is the prevalent mode among the post-Soviets, with the exceptions of Lithuania (QP) and Ukraine (PS turned ALT). For the Asian postLeninists, Mongolia is an outlier by adopting ALT as the operational mode, for its socioeconomic background and political tradition resembles more the post-Soviets than the post-satellites in Europe. Taiwan, on the other hand, chose PS and conforms to the pattern. The postcolonials are the cluster with the strongest presidential control over government. Its average score is 3.37. This is caused by the prevalence of PS as the operational mode. In sum, one finds the West Europeans score lower than the East Europeans, the postSoviets, and the postcolonials in that order. If we relax our definition, and include all the constitutionally semi-presidential regimes, then the contrast among the clusters becomes even sharper. The numbers in parentheses are the average scores of the countries in the clusters that include nondemocracies. We find the average for post- Soviets stands at 3.44, compared with 3.17

9780230_242920_03_cha02.indd 32

11/11/2010 5:09:12 PM

Clustering of Semi-Presidentialism: A First Cut Table 2.4

Clustering and subtypes of semi-presidentialism

Cluster and subcluster

Country

Subtype

Western Europe

Austria Finland France Iceland Ireland Portugal WE Average Bulgaria Croatia Macedonia Montenegro Poland Romania Serbia Slovakia Slovenia EE Average Armenia Azerbaijan* Belarus* Georgia Kazakhstan* Kyrgyzstan Lithuania Russia Ukraine PS Average Mongolia Taiwan

QP COM ALT QP QP QP

PostLeninist

East European Post- Satellites

Post- Soviet Successor States

Asian

QP ALT ALT n/a COM turned ALT COM n/a QP QP PS PS PS PS PS PS QP PS PS turned ALT ALT PS

PL Average PostColonial

33

Francophone

Africa

Algeria* Burkina Faso Cameroon* Central African Republic Chad* Dem. Rep. of Congo* Gabon Madagascar Mali Niger Senegal

Score 1 3 2 1 1 1 1.5 1 2 2 2 3 1 1 1.71 4 4 4 4 4 4 1 4 2 3.17 (3.44) 2 4 2.47 (2.72)

PS PS PS PS

4 4 4 4

PS PS

4 4

PS PS PS ALT PS

4 4 4 2 4 Continued

9780230_242920_03_cha02.indd 33

11/11/2010 5:09:12 PM

34 Yu-Shan Wu Table 2.4

Continued

Cluster and subcluster Francophone

Lusophone

Country

Subtype

Africa

Togo PS Tunisia* PS Other areas Haiti PS Angola* PS Africa Cape Verde QP GuineaPS Bissau Mozambique PS Sao Tome and ALT Principe Other areas East Timor ALT

Others

Overall

4 4 4 4 1 4 4 2 2

Egypt* Kenya Namibia Rwanda* Singapore Sri Lanka Tanzania Yemen

PS n/a PS PS QP PS PS PS

4 3 4 4 1 4 4 4 3.37 (3.56)

Peru Turkey

PS n/a

4

PC Average Others

Score

2.78 (3.04)

Note: * Non-democratic countries.

if we take in only semi-presidential democracies. The overall post-Leninist average is 2.72, compared with 2.47 for democratic regimes. For postcolonials, legally semi-presidential countries average 3.56, compared with 3.37 for democracies in the cluster. It is clear that by bringing in nondemocratic, legally semi-presidential regimes, the averages of post-Leninists and postcolonials rise to reflect the strong presidentialism in those countries. The difference between the post-Leninists/postcolonials, and the West Europeans, or between nascent democracies and established democracies also increases. The impact of cluster on subtype is made even clearer this way. It is interesting to note that when the dominant subtype is at variance with specific country’s actual mode of operation, then serious friction may arise, and the system becomes unstable. We can take an example of Ukraine, a country in the post- Soviet category where the dominant subtype is PS. The adoption of ALT as the operational mode after the Orange Revolution obviously did not bring stability to Ukraine. President Viktor Yushchenko and Premier Viktor Yanukovych clashed over a series of issues that almost led to armed conflict between the Ministry of Internal Affairs

9780230_242920_03_cha02.indd 34

11/11/2010 5:09:12 PM

Clustering of Semi-Presidentialism: A First Cut

35

troops and the presidential guards in 2007.16 This shows that in an area where the president is typically the dominant political figure it is difficult to practice a system that demands the president to yield to the parliamentary majority held by the opposition (viz. ALT). When cohabitation is forced on a resenting president, he is likely to fight by resorting to extraconstitutional measures. Similar cases can be found in Africa, where the dominant mode is also PS. When the constitution and the situation force the president to accept cohabitation, he rarely fails to put up a fight, either during the process of government formation, or after a majority divided government has been formed. In 1996 democracy broke down in Niger after one year of cohabitation. Sao Tome and Principe fared better for it weathered seven years of cohabitation, which is unique among African semi-presidential regimes. However, even there one finds a military coup in 1995. Mongolia is yet another outlier among predominantly PS systems. Its ALT regime was strained during cohabitation between 1996 and 2000, as President Natsagiin Bagabandi repeatedly struck down Great Hural’s candidates for premiership amidst a poisoned political atmosphere. That was also the time when a leading Mongolian politician, Sanjaasürengiin Zorig, was murdered. In East Timor, it was under cohabitation that an open rebellion occurred in 2006, amidst ferocious infighting between President Xanana Gusmão and the Fretilin Prime Minister Mari Alkatiri (Shoesmith 2007: 232–234). Ukraine, Niger, Mongolia, and East Timor are all cases in which their respective semi-presidential systems have adopted an operation mode that is at variance with the prevalent subtype of the cluster. As a result, instability or even democratic breakdown ensued.

Clustering of semi-presidentialism: A first cut Semi-presidential countries are not evenly distributed around the world. Nor did they emerge randomly in history. This chapter identifies semi-presidential countries (democratic and nondemocratic), and finds them in three clusters: West European, post-Leninist, and postcolonial. Besides geographical concentrations, there is also a temporal coincidence in their emergence. Originally entrenched in Western Europe, semi-presidentialism rapidly expanded into the post-Leninist countries in Eurasia, and into the postcolonial world, particularly the Francophone and Lusophone countries, in the 1990s, riding the third wave of democratization. The three clusters have their respective mode of entry and prevalent operational subtype. How a particular semi-presidential system operates and performs has a lot to do with its cluster background. West European countries adopted semi-presidentialism to deal with national exigencies, or to affirm national sovereignty, but they did so without undermining the parliamentary tradition deeply entrenched in the country. Post-Leninist countries adopted the system in an entirely different environment, as an adaptation of their dual power structure to democratic transition. Postcolonial countries have yet another mode of entry. The majority of

9780230_242920_03_cha02.indd 35

11/11/2010 5:09:13 PM

36

Yu-Shan Wu

them adopted the system by copying the constitutional order of France and Portugal, the erstwhile colonial powers that had adopted semi-presidentialism themselves. Besides macro factors, the political calculations of actors on the ground to maximize their interests also came into play. Semi-presidentialism appears in widely different forms. Based on the locus of ultimate power in the state structure and whether that locus shifts under different presidential-parliamentary relations, we differentiate among four subtypes: quasi-parliamentarism (QP), alternation (ALT), compromise (COM), and presidential supremacy (PS). The distribution of these four subtypes across semi-presidential countries is highly related to their clustering. Given that West European countries have the strongest parliamentary tradition, it is expected that the subtype emerging there would heavily tilt toward QP, and with some ALT and COM. Postcolonials are the other extreme. The weakest parliamentary tradition there suggests a heavy tilt toward PS, with minor cases of ALT or COM. Finally post-Leninists have a medium position, and find their East European wing tilting toward QP, ALT, and COM while the post-Soviet wing tilt toward PS. A 1- 4 scoring scheme proves our observation. We also find that if a particular country adopts an operational mode that is at variance with the dominant subtype of the cluster where the country finds itself, there is a high possibility that the system would become unstable, as testified by the cases of Ukraine, Niger, Mongolia and East Timor. This chapter is primarily interested in finding how semi-presidential countries cluster and how cluster features impact them. Although we know that semi-presidentialism is bound to expand beyond the three main clusters of countries, as it has already done on a small scale, we also know that the nature of the clusters would continue exerting influence on how semipresidentialism operates and performs in the countries that are found in the three main clusters.17 A starting point in the study of semi-presidential countries is to recognize how vastly different they are across clusters, and how similar they could be within clusters. Clustering provides a first cut in appreciating both the differences and similarities in the semi-presidential world. This chapter is a first cut.

9780230_242920_03_cha02.indd 36

11/11/2010 5:09:13 PM

9780230_242920_03_cha02.indd 37

11/11/2010 5:09:13 PM

5 5 4 1 5 6 2 4 6 1

5

5 2 6 3 5 1 1 4 4 4 6 1 1 5 5 2 3 2 2

6

6 2 6 3 6 1 1 5 3 4 7 1 1 6 6 2 3 2 2

NF F NF PF NF F F PF PF PF NF F F NF NF F PF F F

NF

NF NF PF F NF NF F PF NF F

–2 8 5 6 –6 10 9 –4 7 –1 –88* N/A 10 –6 –3 10 9 7 10

–1

2 –2 5 10 –7 –7 9 0 –4 N/A

6 2 6 3 6 1 1 6 3 3 7 1 1 6 5 1 3 2 2

5

6 6 5 1 6 7 1 5 6 1

5 2 6 3 5 1 1 4 3 4 6 1 1 5 4 1 3 2 2

4

5 5 4 1 5 6 2 3 6 1

NF F NF PF NF F F PF PF PF NF F F NF PF F PF F F

PF

NF NF PF F NF NF F PF NF F

–2 9 5 7 –3 10 9 –4 7 6 5 N/A 10 –6 3 10 9 7 10

–1

2 –2 5 10 –7 –7 9 0 –4 N/A

6 2 5 3 6 1 1 6 3 4 4 1 1 6 5 1 3 2 2

5

6 6 5 1 6 7 1 5 6 1

6 2 6 4 5 1 1 4 3 4 5 1 1 5 4 1 3 2 2

4

5 5 4 1 5 6 2 3 6 1

NF F NF PF NF F F PF PF PF PF F F NF PF F PF F F

PF

NF NF PF F NF NF F PF NF F

PR CL Status

6 6 5 1 6 7 1 5 6 1

PR

2007 Polity IV

2006

2005

Polity IV CL Status PR CL Status

Democracy test

Algeria* Angola* Armenia Austria Azerbaijan* Belarus* Bulgaria Burkina Faso Cameroon* Cape Verde Central African Republic Chad* Croatia Dem. Rep. of Congo* East Timor Egypt* Finland France Gabon Georgia Guinea-Bissau Haiti Iceland Ireland Kazakhstan* Kyrgyzstan Lithuania Macedonia Mali Mongolia

Appendix 2.1

–2 9 5 7 –3 10 9 –4 6 6 5 N/A 10 –6 4 10 9 7 10

–1

2 –2 5 10 –7 –7 9 0 –4 N/A

7 2 5 3 6 1 1 6 4 4 4 1 1 6 5 1 3 2 2

5

6 6 5 1 6 7 1 5 6 1

6 2 6 4 5 1 1 4 4 4 5 1 1 5 4 1 3 3 2

5

5 5 4 1 5 6 2 3 6 1

Polity IV PR CL

NF F NF PF NF F F PF PF PF PF F F NF PF F PF F F

PF

NF NF PF F NF NF F PF NF F

Status

2008

–2 9 5 7 –3 10 9 –4 6 6 5 N/A 10 –6 3 10 9 7 10

–1

2 –2 5 10 –7 –7 9 0 –4 N/A

Polity IV

7 2 6 3 6 1 1 6 4 4 4 1 1 6 5 1 3 2 2

5

6 6 6 1 6 7 2 5 6 1

6 2 6 4 5 1 1 4 4 4 5 1 1 5 4 1 3 3 2

5

5 5 4 1 5 6 2 3 6 1

–2 9 5 7 –3 10 9 –4 6 6 5 N/A 10 –6 3 10 9 7 10

–1

2 –2 5 10 –7 –7 9 0 –4 N/A

Polity IV

Continued

NF F NF PF NF F F PF PF PF PF F F NF PF F PF F F

PF

NF NF PF F NF NF F PF NF F

PR CL Status

2009

9780230_242920_03_cha02.indd 38

11/11/2010 5:09:13 PM

2005

2

3 2 4 1 1 3 1 3 5 5 3 3

2

2 3 5 1 1 3 2 4 6 6 3 4

5

Yemen

PF

F F PF F F PF F PF NF NF PF PF

F

F PF F F F F F NF NF

–2

8 ** –2 9 10 5 10 1 –4 –4 7 6

N/A

NA 6 6 9 10 10 9 6 –3

5

2 3 5 1 1 3 1 4 6 6 3 3

2

3 3 2 2 1 1 2 6 6

5

3 2 4 1 1 3 1 3 5 5 3 2

2

2 4 2 3 1 1 2 5 5

PF

F F* PF F F PF F PF NF NF PF F

F

F* PF F F F F F NF NF

–2

8 8 –2 10 10 6 10 1 –4 –4 7 7

N/A

8 6 6 9 10 10 9 6 –3

5

2 3 5 1 1 4 2 4 6 6 3 3

2

3 3 2 2 1 1 2 6 6

5

3 2 4 1 1 4 1 3 5 5 3 2

2

3 4 2 3 1 1 2 5 5

PF

F F PF F F PF F PF NF NF PF F

F

PF PF F F F F F NF NF

PR CL Status

Polity IV PR CL Status

Polity IV

2007

2006

–2

8 8 –2 10 10 6 10 1 –4 –4 7 7

N/A

8 6 6 9 10 10 9 5 –3

5

2 3 5 1 1 4 2 4 5 7 3 3

2

3 3 2 2 1 1 2 6 6

5

3 2 4 1 1 4 1 3 5 5 3 2

2

3 3 2 3 1 1 2 5 5

Polity IV PR CL

PF

F F PF F F PF F PF PF NF PF F

F

PF PF F F F F F NF NF

Status

2008

–2

8 8 –2 10 10 6 10 1 –4 –4 7 7

N/A

8 6 6 9 10 10 9 4 –3

Polity IV

5

3 3 5 1 1 4 2 4 5 7 3 3

2

3 3 2 2 1 1 2 6 6

5

3 2 4 1 1 4 1 3 5 5 3 2

2

3 3 2 3 1 1 2 5 5

PF

PF F PF F F PF F PF PF NF PF F

F

PF PF F F F F F NF NF

PR CL Status

2009

–2

8 8 –2 10 10 6 10 1 –4 –4 7 7

N/A

8 6 6 9 10 10 9 4 –3

Polity IV

Notes: * Non-democratic countries. Special Polity IV regime conditions are denoted by the following numbers: –66 denotes an “Interruption,” –77 denotes an “Interregnum,” –88 denotes a “Transition.” Resource: Polity IV Project: http://www.systemicpeace.org/polity/polity4.htm Freedom House: http://www.freedomhouse.org/template.cfm?page=1

5

2 4 3 3 1 1 2 5 5

CL Status

3 3 2 2 1 1 3 6 6

PR

Continued

Montenegro Mozambique Namibia Peru Poland Portugal Romania Russia Rwanda* Sao Tome and Principe Senegal Serbia Singapore Slovakia Slovenia Sri Lanka Taiwan Tanzania Togo Tunisia* Turkey Ukraine

Appendix 2.1

Clustering of Semi-Presidentialism: A First Cut

39

Notes 1. For the political transition in Eastern Europe and former Soviet Union, see Howard (1993), Dawisha and Parrott (1997a, 1997b, 1997c). For political transition in African countries, see Bratton and Van de Walle (1997), Carlson (1999). For a comparison between these two regions, see Beissinger and Young (2002). 2. Discussion of these two groups of countries has different meanings. For democratic semi-presidential regimes, we can explore how they choose this particular constitutional system (the upstream), how they operate it (the midstream), what the impact of the system is, whether those countries are likely to opt out of semi-presidentialism, and how the system is evolving (the downstream). For nondemocratic semi-presidential regimes, we can explore their democratic potential, and how their constitutional structure may facilitate or hinder possible democratization. In this chapter, we concentrate on those semi-presidential countries that satisfy minimum democratic requirements for a more focused investigation. 3. This emphasis on cluster is by no means any form of cultural determinism. It simply broadens the scope of our investigation to include institutional legacy and political tradition when we analyze the upstream and midstream variables in semi-presidential systems. Existing literature typically pays scant attention to the institutional legacy and political tradition of the counties that adopted semipresidentialism in understanding how the system works. 4. In Chapter 1, 53 semi-presidential countries were identified. Given our cutoff point for this study is 2009, we include Madagascar and Niger for the purposes of this paper even though they have since ceased to be semi-presidential. In addition, we exclude disputed semi-presidential states, such as the Palestinian Authority (Brown et al. 2004). On this basis, we differentiate between semi-presidential democracies and non- democratic nations that have prima facie semi-presidential constitutional structures. This differentiation is necessary because constitutional stipulations would have limited, if any, effect in the latter. We designate “nondemocratic” those countries that do not pass the minimal test of democracy, that is, “not free” on Freedom Rating (compiled by Freedom House) AND “below 6” under Polity IV, consistently from 2005 to 2009. There are 11 of them (see Appendix 2.1, with asterisks). They include Angola, which, as noted in Chapter 1, divested itself of semi-presidentialism in January 2010. In addition to these “nondemocratic” countries, we also exclude from our list Serbia and Montenegro, because they have been in existence for less than four years at the time of this writing, and Turkey, which changed the constitution to provide for direct presidential election and has not yet practiced the amendment. Finally, we exclude Kenya from our list for its adoption of a semi-presidential power-sharing scheme after the brutal ethnic conflict that followed the 2008 elections was precarious at best. The number of semi-presidential countries on our list is thus reduced to 41. In this chapter we will concentrate on those countries. 5. The main difference between Leninist states and Communist states is the former merely requires the party-state’s full political control of the society, while the latter adhere to the Communist ideology and insist on state-socialist economy besides political control. In Taiwan, the Kuomintang was heavily influenced by the Soviet Union. Its structure was Leninist ever since 1924. The KMT’s restructuring of the 1950s (gaizao) further strengthened its internal coherence and control over the Taiwanese society. However, the KMT and the Chinese Communist Party (CCP) have been at each other’s throat in their competition for political supremacy in

9780230_242920_03_cha02.indd 39

11/11/2010 5:09:14 PM

40 Yu-Shan Wu

6.

7.

8. 9.

10.

11.

12.

China. Hence the KMT developed an anti- Communist, but nevertheless Leninist structure. As for all the other post-Leninist countries, as they all had Communist regimes prior to democratization, post-Leninism was also post- Communism. The Right in those countries preferred even stronger single leader. However, the Left (not including the communists) usually preferred a purely parliamentary system, fearing the reemergence of autocratic rule. Semi-presidentialism was typically a political compromise between the Left and the Right at a time of national crises (Arter 1999: 51; Shen 2006). In Ireland and Iceland, semi-presidentialism was chosen primarily to assert national independence vis-à-vis former colonial masters (Gallagher 1999; Kristinsson 1999). Direct presidential election mainly served a symbolic purpose, not to fight national crises. For detailed account of the development of semi-presidentialism in Western Europe, see Chapter 6 of this volume. This parliamentary tradition does not mean that prior to their entry into semipresidentialism, those countries all had sovereign parliaments. Actually, except for France in 1962 and Austria in 1929, all the other SP regimes in Western Europe were established either at the time of independence (Finland, Ireland, and Iceland) or democratization (Weimar Germany and Portugal). However, this does not negate the fact that a parliament had been instituted to represent the people before democracy was ushered in, and that it had exercised great influence on the government For example, as stipulated in the 1977 Constitution of the Soviet Union, the Council of Ministers is appointed by the Supreme Soviet and shall tender its resignation when a new Supreme Soviet is elected (Article 129). The Council of Ministers is responsible to the Supreme Soviet, and to its Presidium when the Supreme Soviet is not in session (Article 130). As it turned out, most of the post- Soviet countries adopted semi-presidential constitution, except Estonia and Latvia, while in Eastern Europe semi-presidentialism had to compete with parliamentarism for institutional dominance (FrisonRoche, 2007: 57). One critical micro variable is whether there exists a dominant political force with its leader a highly probable winner of the direct presidential race. When such condition was met, whether it was post- communist party with strong presidential candidate (such as Boris Yeltsin in Russia and Ion Iliescu in Romania) or democratic movement with charismatic presidential hopeful (Lech Wałęsa in Poland and Vytautas Landsbergis in Lithuania), direct presidential election was likely to be instituted and the country moved into semi-presidentialism. However, whenever such condition was not present, such as in Hungary when the Hungarian Socialist Workers’ Party failed to play a dominant role in the transition process and its highly popular leader Imre Pozsgay could not be elected president, and in Czechoslovakia when the much adored Václav Havel failed to convince his Slovak allies to keep national unity, there was no direct presidential election and the post-Leninist country became a parliamentary democracy. Of course micro political factors need to accompany the demonstration effect to bring about a semi-presidential system. When the tide of democratization came, authoritarian rulers found it unavoidable to have dialogue with the democratic opposition. Starting with Benin, national conferences were held and plans of democratic transition were drawn. The authoritarian rulers insisted on popularly elected president with real power as the appropriate role for them to play in the democratized system. The opposition, however, was more interested in a powerful

9780230_242920_03_cha02.indd 40

11/11/2010 5:09:14 PM

Clustering of Semi-Presidentialism: A First Cut

13.

14. 15.

16. 17.

41

parliament that can hold the government responsible. Semi-presidentialism with direct presidential election and cabinet responsible to parliament was a perfect compromise that both the authoritarian elite and the democratic opposition could accept (Conac 2007: 85–86). Institutional demonstration plus micro political calculation made semi-presidentialism highly popular among the Francophone and Lusophone countries. Among all the 15 former French colonies in Africa, 11 adopted semi-presidentialism (see Table 2.4). One should also add the Republic of Congo (Congo Brazzaville) that was semi-presidential but later turned presidential. If we simply calculate democracies, then among the nine Francophone democracies in Africa, six are semi-presidential, one is former semi-presidential, and only one (Benin) is unquestionably presidential. Among the five Lusophone countries in Africa, four are now semi-presidential following the adoption of Angola’s presidential constitution in 2010. Even in Senegal, the harbinger of democracy in Africa, the position of prime minister was abolished in 1962 after the serious conflict between President Leopold Sedar Senghor and Premier Mamadou Dia. It was only when Senghor was preparing to restore democracy in the 1970s that the position of prime minister was reinstituted. See Matsuzato’s account in Chapter 12. This chapter does not intend to explore into the question of if new clusters of semi-presidential countries appear how would the legacies and traditions of the new clusters impact on the countries within. It can be hypothesized that such impact would indeed exist, but then it needs to be proven through empirical investigation.

9780230_242920_03_cha02.indd 41

11/11/2010 5:09:14 PM

3 Variation in the Durability of Semi-Presidential Democracies Robert Elgie and Petra Schleiter

This chapter examines why some semi-presidential democracies survive while others fail. The scale of suffering and loss of opportunity that followed the failure of some of these democracies remains staggering even at a historical distance: the death and destruction caused by the Nazi dictatorship in the aftermath of the Weimar Republic’s collapse, the squandering of resources that followed the collapse of democracy in Armenia in 1996, which allowed a small elite to benefit from economic growth while leaving an estimated 54 percent of Armenians struggling below the poverty level, and the havoc of civil war in Guinea-Bissau after the overthrow of democracy in 1998. While the consequences of the failure of semi-presidential democracies have often been devastating, the factors that enhance or reduce the durability of these democracies remain, as yet, poorly understood. The variation in the survival of semi-presidential democracies is striking. While some of these regimes, such as Finland, count among the longest-lived democracies in the world, others such as the Central African Republic and Kyrgyzstan, slide back to dictatorship after a few brief years of democracy. Why? To date there are no answers to this question. Although scholars have suggested a range of political and institutional factors that may enhance or detract from the survivability of semi-presidential democracies (Linz 1994; Linz and Stepan 1996; Duverger 1997; Pasquino 1997; Skach 2005), none of these potential explanations have yet been analyzed in a controlled comparative setting. This chapter addresses that gap. While there is a large and sophisticated amount of comparative literature on the survival of democracies, this literature has overwhelmingly focussed on the comparison between different types of democratic regimes – presidentialism, parliamentarism. and semi-presidentialism – and their impact on the durability of democracy (Gasiorowski 1995; Przeworski et al. 2000; Bernhard and Nordstrom et al. 2001; Bernhard and Reenock et al. 2004). Where they distinguish semi-presidential from parliamentary and presidential regimes, these studies find that there is no difference in the probability with which semi-presidential and parliamentary democracies consolidate or 42

9780230_242920_04_cha03.indd 42

11/11/2010 5:11:00 PM

Durability of Semi-Presidential Democracies

43

revert to authoritarianism (Svolik 2008: 162–163). As Cheibub and Chernykh put it, “[p]lausible as it may sound, the notion that democracies with a mixed constitution are more likely to become dictatorships [than parliamentary democracies] finds no support in the data” (Cheibub and Chernykh 2008). These are important findings, but they leave the question of what accounts for the substantial, and in human terms deeply consequential, variation in survival among semi-presidential democracies unaddressed. Semi-presidential democracies are characterized by the combination of a fixed-term popularly elected president with a prime minister and government who are responsible to the assembly (Elgie ed. 1999: 13), and work on this class of democracies has engaged extensively with the question of which factors enhance or reduce their survival prospects. Scholars have identified political factors that shape the relationship of the president to the assembly and government, and constitutional design as the critical influences on democratic survival, and much valuable work has been done in this field through case studies (Roper 2002; Schleiter 2003; Morgan-Jones and Schleiter 2004; Elgie and Moestrup eds. 2007), regional studies (Protsyk 2005a, b), and comparative work on the quality of democracy (Moestrup 2007; Elgie and McMenamin 2008a). But to date there is no controlled comparative study that examines the factors that condition the probability of survival among this class of democracies. This is no minor lacuna. The democratic transitions of the late 1980s and early 1990s turned semi-presidentialism into the constitutional format of choice for many new democracies (Shugart 2005a: 344). By 2002, 22 percent of the world’s democracies were semi-presidential (Cheibub and Chernykh 2008). The third wave of democratization brought semi-presidential regimes to regions around the world including Central and Eastern Europe (Bulgaria, Lithuania, Poland, and Romania), the Caucasus, Central Asia, and Asia (Armenia, Mongolia, Taiwan), sub-Saharan Africa (Mali, Niger, and Senegal), and Latin America (Peru). Given the growing popularity of semipresidentialism among democratizing countries, there is a pressing need for a better understanding of the conditions under which this type of democracy is more (or less) likely to survive. Our aim in this chapter is to address this gap. We do so by synthesizing the diverse elements of existing theory about the institutional and political conditions that affect the survival of semipresidential regimes. The resulting expectations are tested using survival analysis and an original dataset covering all of the 611 country years and 45 episodes experienced by semi-presidential democracies from 1919 to 2004. In so doing we link two literatures that have rarely intersected: the literature on semi-presidentialism, and the quantitatively oriented comparative literature on democratic survival. Our results demonstrate that political conditions pale in significance as determinants of the survival of these democracies when compared to constitutional design. Constitutional design

9780230_242920_04_cha03.indd 43

11/11/2010 5:11:00 PM

44

Robert Elgie and Petra Schleiter

has a major impact, and semi-presidential regimes that feature powerful presidents significantly lower the probability that democracy will survive.

Semi-presidentialism: Definition, literature, and hypotheses We adopt Elgie’s definition of semi-presidential regimes as characterized by the existence of “a popularly elected, fixed term president . . . alongside a prime minister and cabinet who are responsible to parliament” (Elgie 1999b: 13), which has considerable advantages over Duverger’s ambiguous classical definition. Duverger’s classical definition additionally requires that a president possess “quite considerable powers” (Duverger 1980: 161). This criterion has engendered extensive debates as to which powers count as “quite considerable,” and whether regimes should be classified according to the behavior or constitutional powers of presidents, resulting in the subjective and inconsistent classification of a range of countries. Elgie’s minimal institutional definition suffers from none of these problems and has for this reason become the preferred way to conceptualize semi-presidentialism in nearly all of the recent work in the field. In countries that adopt such constitutions, politics is predicated on a degree of power-sharing between the president and two other constitutional actors, the assembly and the government. This is because, given their direct election, presidents in semi-presidential democracies have a representative function, which they share with the separately elected assembly. Moreover, given the dual executive, the president and the government (which are responsible to the assembly) share executive functions to a greater or lesser degree. Analysts see the potential strength of these democracies in their flexibility to rely on the assembly majority or the president for support of the government and policy direction (Pasquino 1997; Sartori 1997), and their weakness in the potential for institutional conflict and gridlock between the president on the one hand, and the assembly and government on the other (Linz 1994; Linz and Stepan 1996; Skach 2005). The question that arises is under what conditions those weaknesses or strengths are accentuated so as to have effects on the survival of semi-presidential democracies? In the literature on semi-presidentialism there are two sets of approaches to this question, one focuses on constitutional design, the other on political circumstances. In this section we draw from these two perspectives a set of testable expectations. Clearly, there are other factors, too, that are likely to affect the survival of semi-presidential democracies, such as the economic situation, social structure, historical heritage, and region the country is located in. These will feature as controls in our analysis, but for now we focus on those variables that the literature on semi-presidentialism has identified as relevant in conditioning the success of power-sharing and survival of democracy under this type of democratic regime.

9780230_242920_04_cha03.indd 44

11/11/2010 5:11:00 PM

Durability of Semi-Presidential Democracies

45

Constitutional design Semi-presidential democracies feature much greater constitutional heterogeneity than parliamentary or presidential democracies (Siaroff 2003), and there is a broad consensus that variations in the constitutional design of these regimes have consequences for their survival prospects (Shugart and Carey 1992; Schleiter 2003; Protsyk 2005b; Elgie and Moestrup eds. 2007; Moestrup 2007a: 41). Broadly there is agreement that the absence of power sharing in semi-presidential constitutions, in a way which tilts the balance of power decidedly in favor of the president, is less likely to be successful – but the specific arguments take two distinct forms: Shugart and Carey (1992) first argued that the critical distinction among semi-presidential regimes lies in whether they make governments accountable to both president and assembly or to the assembly alone. Their argument is that semi-presidential regimes are likely to function well when control of the government is clearly assigned. This, they propose, is the case under premier-presidentialism, the constitutional subtype of semipresidentialism, in which the government is exclusively accountable to the legislature. By contrast, under president-parliamentary constitutions, which allow both the president and the legislature to dismiss the government, the dual control of the government is likely to generate institutional conflict between the president and the assembly, which may trigger potential crises of governability and eventually endanger democracy. Shugart and Carey argue that such “symmetry in [government] removal powers contributed to interbranch conflict in Portugal (1976–1982), and to chronic cabinet instability in Germany’s Weimar Republic as well. . . . [and encouraged] conflict and instability of the sort that led to the coup in Peru in 1968” (Shugart and Carey 1992: 118). There is significant preliminary support for Shugart and Carey’s analysis. Moestrup’s descriptive analysis of 83 young semipresidential democracies suggests that “on average premier-presidential regimes have better FH [Freedom House] scores (3.0 versus 3.9) and better chances of democratic survival” (Moestrup 2007a: 41). The first hypothesis that is derived from work on the constitutional variation among semipresidential regimes is therefore (all hypotheses are formulated subject to ceteris paribus): H1: The probability that semi-presidential democracies collapse is increased by president-parliamentary forms of semi-presidentialism. A second view is that a president’s legislative powers may be the decisive aspect of constitutional variation. Semi-presidential constitutions may grant presidents powers in a range of areas beyond government termination, most notably legislative powers such as the power to issue legislation by decree, veto powers, budgetary powers, powers to call referenda,

9780230_242920_04_cha03.indd 45

11/11/2010 5:11:01 PM

46

Robert Elgie and Petra Schleiter

and exclusive powers to introduce certain types of legislation (Shugart and Carey 1992: 150). The argument here is that presidents who enjoy extensive legislative powers may be less inclined to share their power. Rather, they may be able to rely on their constitutional prerogatives to marginalize the prime minister and the assembly (Colton and Skach 2005). In addition, extensive legislative (particularly decree) powers may enhance a president’s opportunities to subvert the rule of law, and should they do so, they may be more difficult to check (Schleiter and Morgan-Jones 2008). This potential is at the heart of Linz’s (1994: 48) fear that semi-presidentialism can come to resemble “a constitutional dictatorship.” As the constitutional powers of a president increase the risks that presidents pose to the survival of democracy can become more pronounced. The second hypothesis derived from work that focuses on constitutional variation within these regimes is: H2: The probability that semi-presidential democracies collapse increases with presidential legislative power. In practice, hypotheses 1 and 2 are likely to be more related than they appear to be at first glance because president-parliamentary constitutions, which give the president the power to dismiss the government, also tend to envisage a more dominant role for the president in governance, including legislation, than premier-presidential constitutions. The empirical evidence available to date suggests that semi-presidential systems that feature high aggregate levels of presidential power are associated with poorer democratic performance (Elgie and McMenamin 2008). Political circumstances A second set of scholars focuses on political circumstances rather than constitutional design as the decisive factor, which may render the power-sharing features of semi-presidential constitutions problematic. The arguments here concentrate on two types of politically complex situations, cohabitation and “divided minority government,” which are thought to make power sharing particularly difficult. The first argument is that semi-presidentialism can give rise to a uniquely difficult political situation where the president and prime minister have divergent policy preferences, so that the executive is “divided against itself” (Pierce 1991). This is most likely to be the case under cohabitation, the situation where the president and prime minister belong to opposing parties and the president’s party is not represented in government. Under cohabitation, the president is isolated within the executive. While presidents who find themselves in that situation may choose to accept the will of the legislature and coexist with a political opponent, they may seek to obstruct the government, creating a gridlock in which neither the president nor the prime minister are willing to compromise. Decision making may then be

9780230_242920_04_cha03.indd 46

11/11/2010 5:11:01 PM

Durability of Semi-Presidential Democracies

47

impeded by intra- executive conflict and legislative immobilism (Linz 1994: 55). Alternatively, if the constitution allows, the president may defy the legislature and dismiss the government in the knowledge that the legislature may oppose subsequent attempts to form a more pro-presidential government, which holds out the prospect of government instability until the next presidential or parliamentary elections. In both outcomes, cohabitation is thought to raise the risk of democratic collapse. Thus, the proposition that is derived from this work regarding cohabitation is that: H3: The probability that semi-presidential democracies collapse increases during periods of cohabitation. Presidents with extensive legislative or government related powers should be more able than constitutionally weak presidents to cause the deleterious outcomes, which writers who comment on cohabitation fear. Simply put, a president who lacks the veto power to obstruct a government’s legislative programme cannot cause legislative gridlock. Similarly, a president who lacks the power to dismiss the government cannot cause government instability. It is therefore not surprising that the empirical evidence with respect to cohabitation is contradictory. While Kirschke’s study of sub-Saharan Africa suggests that semi-presidential regimes working under divided executives have been more prone to collapse (Kirschke 2007: 1390), Elgie’s analysis of the link between cohabitation and democratic breakdowns around the world concludes that “there is insufficient evidence to make the claim that cohabitation is perilous under semi-presidentialism . . . Niger is the only country where cohabitation has ever been directly responsible for the collapse of democracy” (Elgie 2008: 63). Similarly Elgie and McMenamin, provide a controlled comparative study of the effect of cohabitation and other forms of divided government on the quality of semi-presidential democracies as measured by polity ratings, and conclude that “the theory that divided executives undermine democracy is unproven” (Elgie and McMenamin 2008). The second type of politically complex situation that can undermine the prospects for democratic survival is a “divided minority government,” which is defined as a situation where “neither the president nor the prime minister, nor any party or coalition, enjoy a substantive majority in the legislature” (Skach 2005: 15). The combination of a minority president with a minority government is thought to undermine effective governance by either part of the dual executive. As Skach puts it “divided minority government is, in effect, the combination of the potentially most problematic kind of presidentialism – divided government – with the potentially most problematic kind of parliamentarism – minority government” (Skach 2007: 104–105). Divided minority government, it is argued, “can predictably lead to an unstable scenario, characterized by shifting legislative coalitions and

9780230_242920_04_cha03.indd 47

11/11/2010 5:11:01 PM

48 Robert Elgie and Petra Schleiter

government reshuffles, on the one hand, and continuous presidential intervention and use of reserved powers, on the other” (Skach 2005: 17–18). It is thought to undermine effective power sharing, and to encourage the decline and subversion of the rule of law as presidents feel the need to take matters into their own hands. The fourth hypothesis is derived from Skach’s work: H4: The probability that semi-presidential democracies collapse increases during periods of divided minority government Again, it is presidents with extensive powers, not the constitutionally weaker presidents, who have the option of taking matters into their own hands by relying on presidential reserve powers, and it is not surprising, with respect to divided minority government alone, the evidence is contradictory. While Skach (2005) provides case study evidence to suggest that divided minority government was instrumental in the collapse of the Weimar Republic, comparative work by Cheibub and Chernykh (2008), suggests that this situation occurs too rarely to afford a basis for any certain conclusions, and Elgie (2008) finds no clear link between divided minority government and democratic collapse. In sum, the literature on semi-presidentialism identifies variation in the constitutional design of these regimes and politically complex situations as the factors that are likely to shape the probability of failures in powersharing under semi-presidential constitutions. There is a general consensus that failed power-sharing in semi-presidential regimes can engender a range of potential governance crises, including institutional conflict between the president and prime minister or the president and assembly, legislative immobilism, and government instability. Moreover, there is an underlying concern that when power-sharing fails, the checks and balances inherent in some or all of these constitutions may not be strong enough to restrain presidents who – by design, or as a result of frustrating crises of governance – decide to abuse their powers and to subvert the rule of law. In the sections that follow, we assess the hypotheses outlined above.

Data and variables In order to probe the propositions set out above, we examine the survival of all semi-presidential democracies from 1919, the year when the first constitution of this sort was adopted by both Finland and Weimar Germany, through the end of 2004. While some of semi-presidential democracies clearly are among the older, established democracies, semi-presidentialism as a constitutional choice remained relatively rare until the third wave of democratization, when it became the constitutional form of choice for many new democracies in particular during the 1990s. The vast majority

9780230_242920_04_cha03.indd 48

11/11/2010 5:11:01 PM

Durability of Semi-Presidential Democracies

49

of semi-presidential democracies around the world are therefore thirdwave democracies, and in choosing what data to work with, our decisions are driven by the nature of the regimes that third-wave transitions present us with. As Epstein et al. note, the third wave of democratization peopled the globe with partial democracies, and a good number of those are semipresidential (Epstein et al. 2006). Epstein et al. define partial democracies as democracies in which the chief executive is elected, but may face weak constrains, and his or her selection may not result from open and organized competition, but rather from lobbying by powerful political interests. Alternatively, the election itself could be manipulated or competition might lack effectiveness because political parties are highly factionalized. In studying democratic survival, partial democracies are arguably some of the most interesting democracies to understand: they are over four times less stable than full democracies, and movements in and out of the category of partial democracies account for the majority of transitions into and out of democracy globally (Epstein et al. 2006: 555). We therefore choose among the range of possible definitions and measures of the democracy the polity IV data, which scales political regimes from –10 (fully authoritarian) to +10 (fully democratic), and gives us the flexibility to include full and partial democracies. We apply the threshold of a polity score of +1 to identify our universe of cases. If a country’s score drops below this level we record the collapse of democracy. This yields a dataset that covers 611 country-years and 45 episodes of democracy that semi-presidential democracies around the world have experienced. Table 3.1 gives an overview of the data. As shown in Table 3.1, most of our cases have been continuously democratic, such as Finland since 1919, France since 1962, Croatia since 1999, and Senegal since 2000, but a number of the countries in our sample experienced multiple episodes of democracy; such as the Comoros Islands, Austria, Armenia, and Peru. Dependent and explanatory variables Our dependent variable is the duration and outcome (failure or survival) of each democratic episode measured in years. For the first Austrian episode of democracy from 1929 to 1933, for example, our dataset reveals a timeframe of four years followed by the collapse of democracy. For the second Austrian episode of democracy from 1945 to 2004 we show 59 years and the survival of democracy (see Table 3.1). Our discussion above identifies four variables that are thought to have potentially powerful effects on the survival prospects of semi-presidential democracies: president-parliamentarism, presidential legislative power, cohabitation, and divided minority government. All of these except presidential power are recorded as indicator variables (one if the characteristic is present, otherwise zero). We follow Shugart and Carey’s definition of the

9780230_242920_04_cha03.indd 49

11/11/2010 5:11:02 PM

Table 3.1 Duration and outcome of democratic episodes in semi-presidential regimes Country- episode Armenia 1 Armenia 2 Austria 1 Austria 2 Belarus Bulgaria Burkina Faso Central African Rep. Comoros 1 Comoros 2 Congo Brazzaville Croatia Cuba East Timor Finland France Germany Guinea-Bissau 1 Guinea-Bissau 2 Haiti Ireland Lithuania Macedonia Madagascar Mali Moldova Mongolia Mozambique Namibia Niger 1 Niger 2 Peru 1 Peru 2 Peru 3 Poland Portugal Romania Russia Senegal Slovakia Slovenia Sri Lanka Taiwan Tanzania Ukraine Totals

9780230_242920_04_cha03.indd 50

Period

Country years

Failure of democracy

1991–1996 1998–2004 1929–1933 1945–2004 1994–1995 1992–2004 1978–1980 1993–2003 1992–1995 1996–1999 1992–1997 1999–2004 1940–1952 2002–2004 1919–2004 1963–2004 1919–1933 1994–1998 1999–2003 1994–2000 1937–2004 1992–2004 1992–2004 1992–2004 1992–2004 1991–2000 1992–2004 1994–2004 1990–2004 1993–1996 1999–2004 1980–1992 1993–2000 2002–2004 1990–2004 1976–2004 1991–2004 1992–2004 2000–2004 1999–2004 1992–2004 1978–2004 1996–2004 2000–2004 1991–2004

5 6 4 59 1 12 2 10 3 3 5 5 12 2 85 41 14 4 4 6 67 12 12 12 12 9 12 10 14 3 5 12 7 2 14 28 13 12 4 5 12 26 8 4 13

Yes – Yes – Yes – Yes Yes Yes Yes Yes – Yes – – – Yes Yes Yes Yes – – – – – – – – – Yes – Yes Yes – – – – – – – – – – – –

611

16

11/11/2010 5:11:02 PM

Durability of Semi-Presidential Democracies

51

two regime subtypes and record president-parliamentarism if the government can be unilaterally dismissed by the president and is therefore dually responsible to both the president and the legislature. To measure the level of presidential legislative power we rely on those aspects of Siaroff’s nine-part index of presidential power, which record legislative powers (Siaroff 2003). Higher index scores indicate higher levels of presidential power. We record cohabitation when a president from one party holds power at the same time as a prime minister from a different party and when the president’s party is not represented in the cabinet. Therefore, cohabitation records situations where the president is isolated within the executive. If the president was classed as nonpartisan but the prime minister had a party affiliation, we did not count this situation as a case of cohabitation. To record divided minority government, we trace all periods of minority government and identify which of these were also periods of divided government, or cohabitation, so that the minority government was confronted with a president whose party was not included in the cabinet. Control variables Our choice of control variables is informed by the comparative literature on democratic survival and includes domestic political, economic, social, historical, and international political factors. The domestic political context in which constitutions operate (the nature of the party system in particular), has an impact on the governability and the survival prospects of democracies. In general, two party and moderate multiparty systems are perceived as less problematic than the combination of high party system fragmentation with popularly elected presidents (Mainwaring 1993). To capture the effects of different levels of party system fragmentation, we use the effective number of parliamentary parties (the Laakso-Taagepera index), divided into quartiles to generate three dummy variables. The dummies capture moderate multi-partyism (ENP 2.47–3.06), high levels of multi-partyism (ENP 3.07–4.37), and severe party system fragmentation (ENP 4.37–15.33). Because the least fragmented party systems (up to ENP 2.46) are thought to be least problematic, we choose to make this our omitted category. Higher levels of party system fragmentation should therefore raise the risk of democratic collapse relative to this category. Turning to the economy, there is extensive literature on the effect of economic development in promoting the survival of democracies (Przeworski and Limongi 1997; Przeworski et al. 2000; Boix and Stokes 2003; Robinson 2006). To control for the effect of wealth on democratic survival we include GDP per capita in 1990 International Geary-Khamis dollars and take the log of this variable because wealth is likely to have a diminishing marginal effect on the survival prospects of democracy. In addition economic growth has also been associated with the survival of democracy, while economic

9780230_242920_04_cha03.indd 51

11/11/2010 5:11:03 PM

52

Robert Elgie and Petra Schleiter

stagnation and contraction are thought to increase the probability of democratic breakdown (Gasiorowski 1995; Przeworski and Limongi 1997; Svolik 2008). We therefore control for growth, calculated on the basis of our annual GDP figures. Since the work of Dahl the nature of society has been thought to influence the chances that democracy will survive (Dahl 1971: 108). Notably, democracies with more homogenous populations are thought to have better chances of survival (Diamond et al. 1995: 42–43; Bernhard and Nordstrom et al. 2001). For this reason we control for ethnic fragmentation as measured by Alesina et al.’s index of ethnic fractionalization (Alesina et al. 2002). This index ranges from zero to one and higher values indicate greater levels of fragmentation. This takes us to a country’s historical legacies. A number of the countries in our sample are former colonies. The vast majority of studies on the subject argue that the effects of colonialism on the prospects of democratic stability are negative because it encouraged exporting dependent economic development, drawing political boundaries without regard for social structure and history, and the politicization of ethnicity (Bernhard and Reenock et al. 2004). To capture any effect of former colonial status on the survival of democracy after independence, we include a dummy variable that records former colonial status (one, otherwise zero). Several of our countries feature a second type of historical legacy that is thought to affect the chances of democratic survival: the experience of previous episodes of democracy. The effect of this legacy on the survival of democracy is difficult to gauge. On the one hand, elites and the populace in such countries have been introduced to the basic rules of democracy and this experience may aid the survival of subsequently established democratic systems, in particular if it was a significant duration. On the other hand, the previous experience was often one of democratic malfunction and failure, which, especially if it was prolonged, may serve to undermine support for democracy. To control for either type of effect we include the number of years spent under democracy during previous episodes. Finally, we consider international influences on the survival of democracy. The literature on democratic diffusion effects suggests that they operate most powerfully at the regional level (Starr and Lindborg 2003). If this literature were correct, then young democracies in regions that already feature a high proportion of democracies, such as Europe, would stand a better chance of survival than democracies in other parts of the world. To capture this effect we calculated the percentage of countries in the region that were democracies (e.g., had a polity IV score of ≥ +1). We based our calculation on the following regions: Asia, Central Asia and the Caucasus, Central and Eastern Europe (including Russia), North Africa and the Middle East (including Turkey), Sub-Saharan Africa, and Western Europe. Our data include no cases in Australasia.

9780230_242920_04_cha03.indd 52

11/11/2010 5:11:03 PM

Durability of Semi-Presidential Democracies

53

Empirical design Our model choice is guided by three considerations. First, because our aim is to examine the impact of a set of variables on the length of time for which a semi-presidential democracy survives before it breaks down, we employ survival analysis. Models within this statistical family make it possible to analyze the effect of covariates on the hazard rate, or the rate at which semi-presidential democracies fail at a particular time, given that they have survived up to that point. Second, given the long-running scholarly debate about democratic consolidation, a honeymoon effect for young democracies, and other time-dependent features of the hazard that democracies may face (Huntington 1991; Gasiorowski and Power 1998; Bernhard et al. 2003; Svolik 2008), we choose not to model the time dependency exhibited by our data directly, but instead opt for a Cox Proportional Hazards Model. The Cox Model is a semi-parametric model, which enables us to obtain estimates of the covariates of interest, while leaving the underlying distribution of failure times among semi-presidential democracies unspecified (Box-Steffensmeier and Jones 2004). Third, we take account of the possibility that democracy in some countries, say Peru (that has experienced three episodes of democracy to date), is inherently less stable than in others, such as Finland (that has been continuously democratic and semi-presidential since 1919), by estimating countryclustered standard errors. Of the 45 episodes of democracy that we observe, 16 ended in collapse, the remainder either ended for other reasons (for instance the transition to parliamentarism in Moldova), or were continuing to function as democracies when we ceased to observe them in 2004. Had it not been for the end of our study or the transition of these regimes to some other system of government, we would probably have observed these latter semi-presidential democracies to last longer. In the analysis that follows, the terminations of these democracies are therefore treated as unobserved, or censored in the language of survival analysis.

Results Table 3.2 presents the results of two models.1 Model 1 includes our full set of explanatory and control variables, while model 2 is a reduced model, which excludes those variables that do not reach statistical significance and fail to make a statistically significant contribution to the explanatory power of the model. The reduced model 2 therefore is our final model, which best fits the data. We report hazard ratios (which are the exponentiated coefficients) in Table 3.2 because of their intuitive interpretation. The hazard ratio gives the multiplicative effect that a one-unit change in a covariate has on the

9780230_242920_04_cha03.indd 53

11/11/2010 5:11:03 PM

54 Robert Elgie and Petra Schleiter Table 3.2 The survival of semi-presidential democracies (Cox Proportional Hazards Model) Model

1

2

Hazard ratio Hazard ratio (z-value) (z-value) Dependent variable: Survival/failure and duration of democracy 5.24*** H1 President5.24*** (2.67) parliamentarism (2.63) H2 Presidential power 0.89 (Leg.) (–0.40) H3 Cohabitation 1.00 (– 0.00) H4 Minority 2.07 government (0.86) 4.18** Effective number of parties 2.38 (2.20) (2.47–3.06) (1.31) 5.84*** Effective number of parties 4.79** (2.81) (3.07–4.37) (2.29) Effective number of parties 0.40 (4.38–15.33) (– 0.95) 0.38*** log GDP (real per capita) 0.23*** (– 3.64) (–2.46) 0.97** Growth 0.96*** (–1.96) (–2.59) Ethnic fragmentation 0.16 (–1.58) Former colony 0.85 (– 0.26) Number years spent under 1.02 Previous democratic episodes Proportion of democracies in region Log (partial) likelihood Number of regimes Number of years at risk Number of failures Linktest hat(2) Test of proportional hazards assumption Global test chi(2)

(1.16) 0.99 (–.36) –39.17 610 – 0.04 p=(0.760) 14.85 (13df) p=(0.317)

– 41.34 45 611 16 – 0.09 p=(0.589) 2.58 (5df) p=(0.765)

Note: *** Statistically significant at the .01 level, ** statistically significant at the .05 level, * statistically significant at the .10 level, based on countryclustered standard errors.

9780230_242920_04_cha03.indd 54

11/11/2010 5:11:03 PM

Durability of Semi-Presidential Democracies

55

baseline hazard. If a covariate produces a hazard ratio smaller than one, it lowers the failure rate of semi-presidential democracies, a hazard ratio larger than one indicates that a covariate raises the failure rate, and a hazard ratio of one indicates a coefficient equal to zero. The z-values report the statistical significance of the effects and are included in brackets below each hazard ratio. Statistically significant results are starred. Preliminary analysis reveals that there is no variation in outcome with respect to one variable: divided minority government, or the situation in which there is a minority government that does not include the president’s party. Not a single semi-presidential regime ever collapsed during a period of divided minority government. This is not to say that countries that suffered democratic collapse never experienced divided minority government. Germany’s Weimar Republic did, though not during the year of its collapse. It is therefore clear that H4 receives no support in the data although the effect of divided minority government on democratic survival cannot technically be estimated. However, to give the reasoning that underlies the argument the fullest possible assessment, we explore whether the lack of a government majority in the assembly is perhaps the real cause of the detrimental effect that Skach has attributed to divided minority government. Instead of divided minority government, the analyses that follow therefore include a variable that simply records minority government, or the situation in which the government has the support of less than a majority (50 percent +1) of the legislators. The full and reduced models show that constitutional variation among semi-presidential regimes has, as expected, powerful effects on democratic survival. Specifically, the scope of a president’s power, as shaped by whether or not the president has the power to dismiss the cabinet (H1), strongly and significantly influences democratic survival. Those constitutions that are president-parliamentary and make the government dually accountable to the assembly and the president powerfully reduce the durability of democracy, raising the risk of democratic breakdown by a factor of 5.24 compared to semi-presidential constitutions that make the governments accountable only to the assembly. Models 1 and 2 also demonstrate that controlling for the accountability of the cabinet, the level of presidential legislative power (H2) has no further effect on the durability of semi-presidential democracies, which is not surprising given that presidents who have the power to dismiss governments tend to be invested with significant legislative powers too. Neither cohabitation (H3), nor minority government (H4), have an effect on democratic survival. That is controlling for the manner in which the constitutional framework shapes and enforces power-sharing, political complexity has no further significant effect on the survival of semipresidential democracies, and both variables drop out of our final model 2. These are our central results, and we discuss them in more detail below. Turning to the control variables, moderate and fairly high levels of party system fragmentation have the expected effect of increasing the hazard of

9780230_242920_04_cha03.indd 55

11/11/2010 5:11:03 PM

56

Robert Elgie and Petra Schleiter

democratic collapse powerfully, by factors of 4.18 and 5.84 (model 2) respectively, when compared to situations of low party system fragmentation. Note that when semi-presidentialism is thought to be most vulnerable, extreme party system fragmentation, does not appear to be problematic. As is consistent with the results of other controlled comparative work wealth and economic growth have very significant effects in reducing the risk of collapse for semi-presidential democracies. Each additional percentage point of growth, for instance, reduces the risk of authoritarian reversal by 3 percent (model 2). A series of further control variables have no effect on the survival of democracy in semi-presidential regimes. These include historical status as a former colony, previous experience of democracy, ethnic fragmentation, and international context in terms of the percentage of democracies in the region. Turning first to the impact of former status as a colony, it is possible that this variable has no overall effect in our analysis because different colonial powers left varying legacies in their former dependencies (Bernhard and Reenock et al. 2004). Notably, Britain is often found to have had a less detrimental effect on its colonies than some of the other colonial powers. Former British dependencies, which feature no democratic breakdowns in our data, account for roughly one-third of the country-years that we observe in former colonies, and balance out the destabilizing effects of colonialism in some of the former French and Portuguese colonies, which feature eight collapses in our data. The coefficient for ethnic fragmentation too, fails to reach statistical significance. Again, this is not implausible because we would expect politically relevant features of social structure to be reflected in the effective number of political parties. Controlling for parliamentary fragmentation, ethnic fragmentation has no further effect. Nor has previous experience of democracy. As noted above this may be the case because the effects of valuable experience with democracy are counterbalanced by the detrimental effects of having experienced previous democratic failure. Finally, the proportion of democracies in a region, perhaps unsurprisingly, has no effect on survival of democracy once we control for wealth. We examine the specification of the models and the robustness of our central results in a number of different ways. To probe model specification, we test for nonlinearity in the regressors by including the square of the risk score as a regressor in models 1 and 2. We also test for violations of the proportional hazards assumption, on which the Cox model is based, using Schoenfeld residuals. The results of both tests are reported at the end of Table 3.2. None of the tests give reason to reject either model, nor is collinearity (tests not reported) a problem. To scrutinize the robustness of our central results from model 2 further we adopt a dual strategy. First, we examine to what extent our results are robust to the listwise exclusion of each country individually because the regime classification and constitutional scope of presidential power are

9780230_242920_04_cha03.indd 56

11/11/2010 5:11:03 PM

Durability of Semi-Presidential Democracies

57

disputed for some semi-presidential regimes such as Austria, Iceland, and Ireland (Duverger 1980; Shugart and Carey 1992). Second, we reestimate our model, using a Weibull instead of a Cox proportional hazards specification, to examine to what extent our estimates are affected by model choice. Our central result from model 2, that premier-presidential democracies face a very significantly reduced risk of collapse when compared to presidentparliamentary regimes, is completely robust under all of these specifications and never drops below the 90 percent confidence level.

Discussion Having examined the mechanisms that are believed to play a role in destabilizing semi-presidential democracies, we find support for the view that constitutional variation within semi-presidential regimes has a powerful effect on democratic survival. The critical decision in the design of semipresidential constitutions appears to be how far to tilt the balance of power in favor of the president. Our results suggest that giving a president extensive influence over the government through the power to dismiss it unilaterally significantly increases the risk of democratic collapse. This is not surprising given that the vast majority of semi-presidential democracies are young, third-wave democracies, and that our data includes some partial democracies: appropriate constraints on the executive in these settings appear to be of overriding importance. The scope of presidential legislative powers appears to have no additional impact on the survival of semipresidential regimes. We suspect that this is because the regime subtype distinction already absorbs some of the variation in legislative powers: president-parliamentarism correlates with higher levels of legislative power than premier-presidentialism. Finally, politically complex situations in and of themselves do not appear to raise the risk of democratic failure. Once we account for the president’s constitutional position, the complexities that cohabitation or divided minority government entail do not appear to have any further influence on the risk of democratic collapse. In this section we interpret our central finding in more detail. President-parliamentary constitutions, whose distinguishing feature is that they grant the power to dismiss governments to both the assembly and the president, very significantly and powerfully reduce the durability of democracy. As Shugart and Carey (1992) point out, these constitutions undermine power sharing because they enable the president and the assembly to replace the government at will. Because neither side is forced to compromise, the successful resolution of conflicts between legislators and the president over policy and control of the government is likely to be harder to achieve. In addition, these regimes typically give presidents generally greater powers (over legislation, the government, the army, the wider executive) than premierpresidential constitutions. Consequently, presidents who develop a preference

9780230_242920_04_cha03.indd 57

11/11/2010 5:11:03 PM

58

Robert Elgie and Petra Schleiter

for methods of governing that undermine the rule of law, accountability, and democracy – perhaps through enhanced control of the media, interference with elections, and the use of repression – have on average a wider panoply of constitutional powers available to pursue their aims. Put simply, presidentparliamentary regimes are more likely than premier-presidential regimes to lack sufficiently powerful checks on the president. Premier-presidential constitutions, by contrast seem to induce more successful power sharing, and seem to be able to process political conflict more successfully. The reason for this as Shugart and Carey (1992) argue, that these constitutions typically require a president who disagrees with an assembly majority to compromise. Because governments depend on the confidence of the assembly alone presidents cannot effect government change without assembly support, nor can they hope to keep a pro-presidential government in power against the will of a governing majority. As a result, premier-presidential constitutions appear to be more likely to preempt the escalation of political conflicts to the point where they prompt democratic breakdown. Moreover, for presidents with a preference for authoritarian methods of governing, the path to control of the governmental centre of power – to the ministry of the interior, the police, control of the media and media policy, and to the ministry of defense – can only lead via building assembly support. This appears to be on average a potent and effective check on a president’s opportunities to use censorship, the manipulation of elections, violence or repression to subvert the rule of law.

Conclusions In this chapter, we have shown that constitutional variation within semipresidential regimes has powerful effects on their survival, as the literature on semi-presidentialism had long suspected, but never yet demonstrated in controlled comparative work. Our results suggest that the survival prospects of the two subtypes of semi-presidential democracies, presidentparliamentary and premier-presidential regimes differ significantly, which has a bearing on the understanding of semi-presidentialism as a regime type, the study of democratic survival, and the normative literature on constitutional engineering. This section outlines what we see as the main implications of our work for each of these areas. Turning first to the literature on semi-presidentialism we provide a large-n controlled analysis of the expectations that scholars have generated about the factors that condition the survival of these democracies. Our work suggests that political conditions such as cohabitation and divided minority government have been overrated as causes of democratic collapse in much of the literature. By contrast, how the constitution structures incentives to powershare, appears to have overriding importance. Once sufficiently powerful

9780230_242920_04_cha03.indd 58

11/11/2010 5:11:04 PM

Durability of Semi-Presidential Democracies

59

institutional incentives to power-share are in place, semi-presidential regimes appear to navigate politically difficult situations without endangering democracy. But where the institutional incentives to power-share are weak semipresidential regimes appear to fail under a variety of political conditions, which may or may not coincide with periods of cohabitation or minority government. With respect to the comparative literature on the survival of democracies our work makes progress with integrating semi-presidential regimes into the mainstream comparative literature on democratic survival. Our results suggest that in order to understand the effects of this most heterogeneous of the main democratic regime types, analysts should pay attention to the constitutional variations among semi-presidential regimes. Thus future work on democratic stability should not just take account of the proliferation of semi-presidential regimes during the third wave of democratization it should also take account of the constitutional distinction between premierpresidential and president-parliamentary regimes in order to adequately capture the influence that the constitutions of 22 percent of the world’s democracies have on the durability of democratic rule. Turning to the normative literature on constitutional design, there seem to be two reasons for constitution drafters to choose semi-presidentialism: a desire for constitutional rules that require power-sharing, and the desire for flexibility through presidential action. Our results suggest that constitution drafters ought to be wary of leaning too far in the direction of investing the presidency with extensive powers in order to achieve flexibility, because powerful presidents appear to have a price in terms of democratic survival. Presidents who, by virtue of their authority over the government, have opportunities to assert control over the governmental centre of power while flouting the checks and balances and the type of power-sharing that semi-presidential constitutions envisage raise the risk of democratic collapse. The critical question in designing semi-presidential constitutions is what price to pay in the risk of democratic breakdown for the promise of more decisive presidential governance that enhancing a president’s authority over the cabinet appears to hold out. Our results suggest that the price of giving presidents the power to dismiss the cabinet is a 5.24-fold increase of the risk of democratic collapse. This should give pause for thought.

Appendix 3.1 Data sources Duration of Democratic Episodes, Number of Years Spent under Previous Democratic Episodes, and Proportion of Democracies in Region. These variables are based on Polity IV data (Marshall et al. accessed April 2008). President-Parliamentarism. The subtype of semi-presidentialism is recorded on the basis of the constitution of each country.

9780230_242920_04_cha03.indd 59

11/11/2010 5:11:04 PM

60

Robert Elgie and Petra Schleiter

Presidential Legislative Power. Measures the extent of a president’s legislative power based on Siaroff’s (2003) index. Cohabitation. Party affiliation of the president and government participation of the president’s party: www.worldstatesmen.org, and secondary sources: Africa Research Bulletin, Data Issue of the European Journal of Political Research and (Müller and Strøm 2000). Minority Government. World Bank Database of Political Institutions (“majority” variable) for the period 1975–2004 (Beck et al. 2001). For data prior to that time, we use secondary sources (Nousiainen 1971; Kolb 1990; Müller and Strøm 2000). GDP and Growth. GDP per capita in constant 1990 International GearyKhamis dollars and growth are based on Angus Maddison’s data http:// www.ggdc.net/Maddison/. For Timor-Leste and Iceland (not covered by Maddison) we employ data from the United Nations Statistics Division and http://www2.stjr.is/frr/thst/rit/sogulegt/english.htm#10 respectively. Ethnic Fragmentation. This variable is based on Alesina et al.’s (2002) index. For Sao Tome (not covered by the ethnic fragmentation index) we substitute Alesina et al.’s linguistic fragmentation index. Former Colony. We record status as a former colony at any point after 1918. Effective Number of Parliamentary Parties. This variable is drawn from Gallagher (http://www.tcd.ie/Political_Science/staff/michael_gallagher/ ElSystems/) supplemented by figures for early Central and East European elections from (Birch 2003). Missing data were calculated using http:// africanelections.tripod.com/ and http://psephos.adam- carr.net/.

Note 1. All statistical analyses were performed using STATA (version 9.0).

9780230_242920_04_cha03.indd 60

11/11/2010 5:11:04 PM

4 The Rules of Electoral Competition and the Accountability of SemiPresidential Governments Jih-Wen Lin

For all democracies, government accountability can be measured by how a government is formed and how long it endures. The most accountable government is one that speaks for all people and can be dismissed as soon as it fails to carry out this mandate. By these criteria, there is a tradeoff among constitutional systems. Governments in a parliamentary system are responsible to the parliament, but the same mechanism gives the governing party a justification to discipline its legislators and control political power even if it performs poorly. In a presidential system, the people elect a president as the head of government. They elect legislators for fixed terms, but responsibility can be blurred if the president is not supported by a legislative majority. Semi-presidentialism is more complicated. The people elect a president who appoints a prime minister, relying on the confidence of the parliament to survive.1 Thus defined, a semi-presidential government has two principals: the parliament and the president. This situation not only makes the semi-presidential government qualitatively different from its counterpart in other constitutional systems, but also creates a serious problem of accountability: what if the president and the legislative majority are incongruent? Intuitively, one would think that governments in a presidentialparliamentary system, in which the president enjoys the freedom to dismiss the prime minister, carry a dual responsibility to the president and the parliament whereas those in the premier-presidential system, in which the president cannot freely unseat the prime minister, are mainly accountable to the parliamentary majority (Shugart and Carey 1992).2 This chapter shows that government accountability in semi-presidential regimes is more complicated than can be explained by these two subtypes of semi-presidentialism. The analysis focuses on how governments are formed and replaced in semi-presidential regimes, and a key explanatory variable is whether there is a stable and coherent legislative majority. For this reason, 61

9780230_242920_05_cha04.indd 61

11/11/2010 5:11:46 PM

62 Jih-Wen Lin

we cannot fully understand the accountability of semi-presidential governments without knowing how legislators are elected. The next section presents a theory to explain why the rules of electoral competition affect accountability. The third section verifies this theory by examining the formation and duration of governments in semi-presidential democracies. The fourth section uses three case studies to illustrate how the theory works in reality. The concluding section addresses the implications of the empirical findings on the accountability of semi-presidential governments.

Why do the rules of electoral competition affect accountability? Which factors affect the formation and survival of semi-presidential governments, hence their accountability? Is it the president’s power to appoint and dismiss prime ministers? Maurice Duverger provides a clue. He posited that in semi-presidential countries where coherent and stable majorities are found there is a disparity between the constitution and practice, and that a president becomes dominant when s/he controls this majority (Duverger 1980: 182–186). If the existence of a coherent and stable majority determines the president’s practical power, does it affect government accountability? If it does, what happens if no such majority exists? Which factors explain the existence of a stable and coherent majority? The first question concerns why legislative cohesion may affect government formation. To establish the relationship between these two variables, we consider the following scenarios. First, when the constitution restrains the president’s power in government formation, the more cohesive the legislative majority, the greater the likelihood that a majority government will be formed, and vice versa. Second, when the constitution allows the president to appoint the prime minister without legislative participation, the president has an incentive to form a minority government under his/her control if he/ she is confronted with a strong opposition majority. Whether or not the president is able to fulfill this goal depends on his/her capacity to overwhelm the opposition. A president colliding with a strong legislative opposition runs the risk of jeopardizing his/her policy goals if he or she insists on forming a minority government; he or she may also undermine the electoral fortunes of his/her party if he or she decides to dissolve the parliament and initiate a snap election. When the opposition parties are fragile, the president has a better chance of forming a minority government or even of splitting up the opposition alliance and assembling a pro-president majority. All scenarios suggest that legislative cohesion has a significant impact on government formation. The president’s power to appoint the prime minister affects government formation only if the parliament is weak. The second issue follows the first: what factors facilitate the formation of a coherent and stable legislative majority? Thanks again to Duverger, a

9780230_242920_05_cha04.indd 62

11/11/2010 5:11:46 PM

Rules of Electoral Competition 63

key determinant is the electoral system used for the legislative election, for four reasons. First, electoral systems shape the number of viable parties, which is the most significant factor that affects legislative cohesion. It is obvious that the more parties there are, the less likely they will form a stable coalition. A legislature can be fragmented in such a way that all parties in the majority coalition enjoy veto power. Conversely, a legislature dominated by a single party is most stable because no other party can upset the decisions made by this party. As for the exact relationship between legislative cohesion and the number of parties, Duverger proposed that “the simple-majority single-ballot system (henceforth called single-member simple plurality, SMSP) favors the two-party system” and that “the simple-majority system with second ballot (two-round systems, TRS) and proportional representation (PR) favor multipartism” (Duverger 1963: 217, 239). A further distinction should be made between TRS and PR: multipartism resulting from TRS encourages coalition-building, while the low winning threshold of PR gives parties little incentive to engage in electoral cooperation. We therefore expect legislative cohesion to be higher under single-member systems – including SMSP and TRS – than under proportional systems. Second, the capacity for coalition-building varies by electoral system. According to the median voter theorem (Downs 1957), two parties competing under a single-member system have the incentive to converge on the voters’ median policy position so that each can attempt to gain a majority of the votes. Even if more than two parties join a single-member race, supporters of small parties are encouraged to vote strategically for large parties, thereby creating a centripetal tendency of electoral competition. Such an incentive does not exist under proportional systems, in which the best campaign strategy is to garner support from a particular set of voters. Assuming that centrist legislators are more likely to build coalitions, we expect singlemember electoral systems to be more conducive to the building of coherent legislative majorities than proportional systems. Third, intraparty cohesion is also associated with electoral systems. When only one winner is to be elected, political parties have good reason to discipline their members because the threshold of winning is high. At the national level, delegates elected under the single-member rules have no reason to take different policy positions unless they represent constituencies with contradictory interests. Intraparty cohesion should also be high under PR because parties have the justification to penalize members who may damage the party’s image. The same logic suggests that intraparty cohesion is weakest if multiple seats are to be allocated by the candidates’ vote shares (henceforth called candidate-based multimember systems, including single nontransferable voting (SNTV), and single transferable voting (STV)). Under these systems, the threshold of winning decreases with district magnitude, giving some candidates an incentive to defect from the party; politicians are

9780230_242920_05_cha04.indd 63

11/11/2010 5:11:46 PM

64 Jih-Wen Lin

motivated to cultivate personal votes because voters vote by candidate, and seats are allocated by candidates’ vote shares. Lastly, the costs and risks of legislative elections differ by electoral system, which in turn affect the ability of legislators to pass a censure motion. We can reasonably assume that the smaller the role played by political parties in the election, the greater the burden on candidates; hence, individual legislators will be more reluctant to face an early election. Because parties play more important roles in proportional representation than in candidate-based multimember systems, the campaign costs of individual candidates should be higher in the latter systems than in the former one. The role of political parties in the single-member races depends on the institutionalization of the party system: when most voters vote by party, individual candidates bear limited campaign costs; when the party system is not institutionalized, voters tend to vote by candidate, making the competition personal. The same logic applies to mixed-member systems with a significant portion of single-member seats. As for the calculation of political parties, the most critical factor is the number of seats to be changed. It is well-known that seat distribution does not shift much between elections under PR or candidate-based multimember systems unless some parties expect a major change in their vote share. Singlemember systems, by contrast, are likely to produce a significant swing in seat shares even if vote shares do not change much. When elections are partycentered, therefore, political parties face a greater risk under single-member systems than under proportional systems. But the same logic implies that single-member systems give opposition parties a strong incentive to pass a vote of no-confidence if they are expecting an upward swing in their vote shares. If a censure motion requires the endorsement of both political parties and their legislators, only PR favors its passage in most circumstances. Summarizing the analysis above, we find the single-member electoral systems most conducive to the formation of a stable and coherent legislative majority. These systems increase the chances of the legislature being dominated by a disciplined party (or coalition) that has the capability to boycott government bills or impeach the prime minister. Proportional systems, in particular those allocating seats by the vote shares of candidates, are more likely to create a multiparty system where coalition building is costly and majorities are incoherent. Even if PR encourages parties to initiate a snap election by passing a censure motion, the new parliament may not be much different from the old one. Because the likelihood of a majority government decreases with the proportionality of the electoral system and is minimized when legislators are elected by candidate-based multimember systems, we can derive a hypothesis (the coding of which will be discussed later) stating the impact of electoral systems on government formation in semi-presidential regimes: H1: In semi-presidential regimes, the likelihood of majority government decreases with the proportionality of the electoral system used for parliamentary elections.

9780230_242920_05_cha04.indd 64

11/11/2010 5:11:46 PM

Rules of Electoral Competition 65

The next two hypotheses test whether the likelihood of a minority government increases with the powers of the president when he or she faces an antagonistic legislative majority. These hypotheses are based on the belief that the greater the president’s power, the more likely for him/her to be the de facto political leader in charge of government formation and the more difficult for the opposition majority to counter the president’s move. Here we test two hypotheses. H2 examines the consequence of a particular institutional design by using the president’s power to appoint the prime minister as the independent variable. H3, based on the belief that the more constitutional powers a president enjoys the greater his/her political influence, examines how the president’s general powers explain government formation. H2: In semi-presidential regimes, the probability of majority government is lower in cases in which the president can unilaterally appoint the prime minister. H3: In semi-presidential regimes, the probability of majority government decreases with the number of the president’s constitutional powers. The second dimension of accountability is the survivability of government, and is subjected to the influence of many factors, such as the legislative support the government receives, the government’s governing capability, and the stability of the political system. A less noticeable but important variable is the likelihood of the current government being replaced by a new legislative majority. Clearly, the higher this probability, the shorter the life of a government is likely to be. Conversely, a government – even if not representing the majority of seats – may survive a long time if no legislative majority can be assembled to form a new government. Due to the fact that a new majority usually results from a new election or the rebuilding of legislative coalitions, two institutional variables may influence a government’s duration: the electoral system and the electoral cycle. As explained above, single-member systems are more likely than other systems to produce a clear and coherent majority. An opposition party winning the majority of seats under a single-member electoral system is given a strong mandate to form a new government and terminate the life of the incumbent government. The situation is quite different if the legislative election uses proportional systems – these systems rarely allow a single party to capture the majority of seats, and a government is usually a coalition of several parties. Because proportional systems rarely produce a significant change in seats, legislative elections have limited impact on government formation. A prime minister heading a coalition government can mitigate popular discontent by adjusting the partisan composition of his/her cabinet without stepping down, a move that is difficult to take if a

9780230_242920_05_cha04.indd 65

11/11/2010 5:11:46 PM

66 Jih-Wen Lin

single party controls the government. We therefore expect governments to last longer in proportional systems. H4: In semi-presidential regimes, government duration increases with the proportionality of the electoral system. This hypothesis applies to parliamentary systems as well, but semipresidentialism is peculiar in that voters can also demonstrate their will in presidential elections. As pointed out by Samuels and Shugart (2010), in presidential and semi-presidential democracies, direct presidential elections change the way political parties are held accountable to the people. But semi-presidentialism is different from presidentialism in that the legislature is allowed to sack the head of government, creating a complicated structure of accountability. The electoral cycle, the timing of the legislative and presidential elections, therefore becomes another variable that affects a government’s duration. Given the observation that concurrent elections, where presidential and legislative elections take place at the same time, are more likely to induce a pro-president legislative majority and that the nonconcurrent electoral cycle decreases this probability (Shugart and Carey 1992; Jones 1995; Shugart 1995), we hypothesize that nonconcurrent cycles shorten government duration in which the president can freely dismiss the prime minister. The reason is that presidents supported by a legislative majority have no need to dismiss the prime minister, while cohabitation gives presidents an incentive to displace the prime minister by initiating a snap election. H5: In semi-presidential regimes in which the president can dismiss the prime minister, government duration is longer in concurrent electoral cycles than in nonconcurrent electoral cycles. So far we have made theoretical arguments by analyzing the dynamics of government accountability in semi-presidential regimes. The following sections will test these hypotheses by examining what has happened in semi-presidential regimes around the world.

Data and analysis This section will test the above hypotheses by examining the formation and duration of semi-presidential governments. The cases were selected using the following procedure. First, to be as comprehensive as possible, all countries with a semi-presidential system, as defined by this book, are considered. Second, to discern the institutional effects, the opposition legislators should be free from political threats, and elections should be open and fair. We therefore screen out countries that are not considered free. By

9780230_242920_05_cha04.indd 66

11/11/2010 5:11:47 PM

Rules of Electoral Competition 67

this criterion, the data to be analyzed only include countries marked by Freedom House as “Free” (in 2009).3 Third, countries in which the president is defined by the constitution as the head of government are also excluded because they function like presidential countries in which the parliament’s censure motion is not a very effective means to affect government accountability. Free semi-presidential countries instituting their president as the head of government include Namibia and Peru. This selection process leaves 19 countries. Table 4.1 lists these countries and how they appoint the prime minister and elect legislators. We begin by analyzing government formation. In this chapter, a “government” is defined as the appointment of a new prime minister.4 To be consistent, governments formed when a country is not considered free are excluded from the dataset. In this way, the number of governments is 252, which covers almost all governments formed in free semi-presidential countries. Following this definition, a majority government is defined as one where the total number of seats controlled by the government parties is at least half of the total legislative seats (of the Lower House for bicameral countries); a minority government holds less than half of the seats. Majority governments can be further distinguished into a consolidated majority (if it includes the president’s party) and a divided majority (if it does not). In the same way, whether a minority government is consolidated is determined by whether it includes the president’s party. Table 4.2 shows the distribution of government types in our dataset. At first glance, the data meets the theoretical expectations. On average, the percentage of majority government in cases using single-member systems is 77.3 percent (or 90 percent if a short-lived Ukraine minority government is dropped), while the average percentage of majority government in cases adopting candidate-based multimember systems (SNTV or STV) is 37.6 percent. The average percentage of majority government in cases using proportional systems is 80.4 percent with a high variance. Single-member electoral systems do seem to be conducive to the formation of majority governments, in sharp contrast with the high frequency of minority governments in countries using candidate-based multimember systems. To avoid the ecological fallacy, we investigate how the formation of each government is affected by the variables discussed above. In addition to institutional factors such as the rules of electoral competition and presidential powers, the following variables may also influence government formation. First, learning may enhance coalition-building by reducing incomplete information and the transaction costs of bargaining. This variable is approximated by the sequential order of the government formed under semi-presidentialism. Second, given the belief that countries with a parliamentary tradition are more likely to delegate government formation to parliamentary parties, we examine whether the odds of majority government are higher in Western Europe, the birthplace of parliamentarism. Third, the

9780230_242920_05_cha04.indd 67

11/11/2010 5:11:47 PM

68 Jih-Wen Lin Table 4.1 Free semi-presidential countries (2009) Country

The constitutional procedure of prime minister designation

Austria Bulgaria

The Chancellor is appointed by the president. The prime minister is nominated by the president and elected by the National Assembly. Cape Verde The Chairman of the Council of Ministers (prime minister) is nominated by the president and elected by the National Assembly. Croatia The prime minister is appointed by the president and approved by the Assembly. Finland The prime minister is appointed by the president (until 1999). Since 2000, the prime minister is elected by the Eduskunta (parliament). France The prime minister is appointed by the president. Iceland The prime minister is appointed by the president. Ireland The prime minister is nominated by the House of Representatives and appointed by the president. Lithuania The prime minister is appointed by the president and approved by the Seimas (parliament). Mali The prime minister is appointed by the president. Mongolia The prime minister is nominated by the president and appointed by the State Great Hural (parliament). Poland The prime minister is nominated by the president and approved by the Sejm (parliament). Portugal The president appoints the prime minister after consultation with the political parties of the Assembly. Romania The president appoints the prime minister on the basis of a vote of confidence by the parliament. Sao Tome e The prime minister is selected by the National Principe Assembly and approved by the president. Slovakia The prime minister is appointed and recalled by the president of the Slovak Republic. Slovenia The prime minister is nominated by the president and elected by the National Assembly.

Electoral system PR PR

PR

PR PR

TRS, PR (1986) PR STV

MMM

TRS Block vote (1992– 1995; 2008–), TRS (1996–2007) PR

PR

PR, MMP (2008–) PR PR PR

Continued

9780230_242920_05_cha04.indd 68

11/11/2010 5:11:47 PM

Rules of Electoral Competition 69 Table 4.1 Continued Country

The constitutional procedure of prime minister designation Electoral system

Taiwan

The prime minister is appointed by the president.

Ukraine

SNTV. MMM (2008–) The prime minister is nominated by the president MMM, and approved by the parliament (before 2006). PR (2006–) The parliament nominates and approves the prime minister (after 2006).

Source: The constitution and election laws of each country.

Polity IV score of democratic institutionalization is also considered to measure the impact of democratization on government formation. These variables are abbreviated and coded as follows: 1. MAG: Whether the government is a majority government (1) or not (0). 2. PPM: Whether the president can unilaterally appoint the prime minister without legislative participation (1) or not (0). 3. PCP: The Siaroff score of a president’s constitutional power. The higher the score, the greater the number of constitutional powers a president has. 4. PRO: The proportionality of the electoral system, 1 = single-member systems; 2 = mixed-member majoritarian systems; 3 = mixed-member proportional systems; 4 = proportional systems (PR, SNTV, STV). 5. SEQ: The sequential order of governments formed in a free country since semi-presidentialism was established. 6. DEM: The Polity IV score of democratic institutionalization. 7. WEU: 1 = Western European country; 0 = non-Western European country. Some variables need further explanation. For PPM, the following situations are coded 0: the legislature elects the prime minister, the appointment of the prime minister requires legislative approval, or the prime minister is nominated by the legislature and appointed by the president. For countries in which the president is constitutionally bound to consult the parliament when appointing the prime minister, PPM is coded 1 because the legislature is not given any formal power to reject the president’s decision. PCP, measured by the number of constitutional powers reserved for the president, has several analytical advantages: it can be treated as a control variable for PPM or an independent variable indicating the president’s general powers; it helps us compare semi-presidential countries seen by some as parliamentary (such as Austria and Iceland, which have very low PCP) with other cases. PRO is a continuous variable coded by the weight of proportional representation in a country’s electoral system. The two extremes, coded 1 and 4, respectively, are

9780230_242920_05_cha04.indd 69

11/11/2010 5:11:47 PM

70 Jih-Wen Lin Table 4.2 Government types in free semi-presidential countries Country

Time Con. Con. Div. Div. Majority Minority Electoral period majority minority majority minority Gov. (%) Gov. (%) system

Austria Bulgaria Cape Verde Croatia Finland France 1

1945– 1991– 1992– 2000– 1945– 1959– 1985; 1988– France 2 1986– 1987 Iceland 1947– Ireland 1944– Lithuania 1992– Mali 1992– Mongolia 1 1992– 1995 2008– Mongolia 2 1996– 2007 Poland 1997– Portugal 1976– Romania 1 1991– 2007 Romania 2 2008– Sao Tome e 1991– Principe Slovakia 1998– Slovenia 1991– Taiwan 1 1997– 2007 Taiwan 2 2008– Ukraine 1 2005 Ukraine 2 2006–

25 2 4 0 18 20

1 1 0 1 4 3

1 2 1 3 8 2

0 1 0 1 4 0

96.30 57.14 100 60 76.47 88

3.70 28.57 0 40 23.53 12

PR PR PR PR PR TRS

0

0

1

0

100

0

4 10 3 5 1

1 4 0 0 0

20 4 6 0 1

2 5 2 0 0

88.89 60.87 81.82 71.43 100

11.11 39.13 18.18 0 0

5

0

3

0

100

0

2 2 2

2 1 2

2 7 0

2 6 2

50 56.25 37.5

50 43.75 62.5

0 4

0 1

1 7

0 0

100 92.86

0 7.14

MMP PR

0 0 1

0 0 6

3 9 0

0 1 0

100 90 14.29

0 10 85.71

PR PR SNTV

1 0 2

0 2 0

0 0 0

0 0 0

100 0 100

0 100 0

MMM MMM PR

PR PR STV MMM TRS Block vote TRS PR PR PR

Note: (1) Cons. Gov. = consolidated government; Div. Gov. = divided government. (2) Governments formed when a country is “not free” or “partially free” are excluded. (3) Cases with incomplete or inconsistent information are excluded.

single-member systems and pure PR systems. The mixed-member majoritarian systems are coded 2 because the distribution of the single-member seats and the PR seats are counted separately. The mixed-member proportional systems, allocating all seats in proportion to the parties’ PR vote share, are coded 3 rather than 4 because small parties are disadvantaged in the singlemember races. DEM is an additive eleven-point scale (0–10) indicating the competitiveness and openness of executive recruitment, the constraints on the chief executive, and the competitiveness of political participation.

9780230_242920_05_cha04.indd 70

11/11/2010 5:11:48 PM

Rules of Electoral Competition 71 Table 4.3 Government formation in semi-presidential democracies: Results from the binary logistic regression Model 1 Variables

Model 2

Coefficient

S.E.

Sig.

Coefficient S.E.

Sig.

PCP PPM PRO MAS CBM SEQ DEM WEU CONSTANT

–0.262 0.260 – 0.455 – – 0.073 – 0.021 –0.062 3.413

0.093 0.421 0.179 – – 0.029 0.057 0.476 0.924

0.005 0.537 0.011 – – 0.012 0.715 0.896 0.000

– 0.238 – 0.068 – 0.865 –0.981 0.067 – 0.031 0.247 1.878

-2 Log Likelihood Nagelkerke R-squared

197.760

–

–

195.444

–

–

0.175

–

–

0.190

–

–

0.095 0.012 0.467 0.884 – – 0.486 0.075 0.494 0.047 0.030 0.025 0.093 0.737 0.507 0.626 1.027 0.068

Note: Dependent variable = MAG.

Among the variables discussed above, the proportionality of the electoral systems can be redefined to characterize their qualitative features. Using PR as the baseline category, the following dummy variables are defined for the second model: 1. MAS: majoritarian-leaning systems, including SMSP, TRS, mixed-member majoritarian systems, and block vote.5 2. CBM: candidate-based multimember systems, including SNTV and STV. Whether the government is a majority government or not is a categorical dependent variable. The logistic regression is thus an appropriate statistical method to estimate how much the explanatory variables are associated with the odds of majority government. The binomial logistic regression (Table 4.3) shows the following results. First, electoral systems exert a significant impact on government formation. As shown in Model 1, the log odds of majority government being formed decreases by 0.455 as the proportionality of electoral systems increase by 1. Using PR as the baseline category, Model 2 shows that the log odds of majority government increases by 0.865 if the electoral system becomes majoritarian, and decreases by 0.981 if candidate-based multimember systems are used. These results confirm H1, which says that the likelihood of majority government decreases with the proportionality of the electoral system. The second finding is that government formation is more sensitive to the president’s general powers than to his/her power in the appointment of prime ministers. The sign of PPM is positive in Model 1 but negative in

9780230_242920_05_cha04.indd 71

11/11/2010 5:11:49 PM

72 Jih-Wen Lin

Model 2, suggesting this variable has a dubious effect on government formation.6 By contrast, the impact of PCP is consistently negative and statistically significant. These results imply that a president supported by a pro-president minority government may still find it hard to carry out his/her mandate if he/she lacks veto power, cannot issue decrees or dissolve parliament at any time, enjoys little power in reserve domain, or is not allowed to chair cabinet meetings. By the same logic, the greater a president’s constitutional powers, the smaller the capacity of the opposition majority using its legislative powers to deter the president from forming a minority government. Third, SEQ is statistically significant in both models but DEM and WEU are not. Together, these results indicate that majority government is facilitated by a country’s experience with semi-presidentialism rather than its location or the degree of democratic institutionalization. A plausible explanation is that SEQ captures the likely impact of democratic institutionalization and parliamentary tradition for two reasons: first, the West European cases have a longer experience with semi-presidentialism in which the frequency of majority government increases over time, making SEQ a good proxy of their history; second, our cases are “free” semi-presidential countries, which reduces the variance of DEM and its explanatory power. The second part of the empirical analysis seeks to explain the durability of semi-presidential governments. As theorized above, the life of a semipresidential government is sensitive to the birth of a new legislative majority, which makes the electoral system a reasonable explanatory variable. Because semi-presidentialism gives some presidents the freedom to dismiss the prime minister and nonconcurrent electoral cycles are likely to deprive the president of a supportive majority, electoral cycle and semi-presidential regime type are plausible explanations. The following variables are added to test H4 and H5. 1. PRE: regime type, including president-parliamentary system (1) or premier-presidentialism (0). 2. DUR: the durability (days divided by maximum days) of the current government. 3. CYC: electoral cycle; the parliamentary election is held beyond six months before or after the presidential election (1) or not (0). 4. PRE×CYC: the product of PRE and CYC. The definition of PRE follows Shugart and Carey’s (1992: 23–24), which states that the president appoints and dismisses cabinet ministers in a president-parliamentary regime (see Chapter 1 for the list of presidentialparliamentary countries).7 Thus defined, PRE helps measure how much government formation is affected by the president’s power of dismissal (H2). DUR counts the durability of a government by using the maximum days it

9780230_242920_05_cha04.indd 72

11/11/2010 5:11:49 PM

Rules of Electoral Competition 73

can last as the denominator. CYC considers an electoral cycle concurrent if a parliamentary election is held within six months before or after the presidential election under the assumption that the longer the gap between the two elections, the greater the likelihood voters will vote for different parties.8 Since presidents in the president-parliamentary regimes have the power to dismiss the prime minister, PRE×CYC represents the interaction between regime type and electoral cycle. In line with the theory depicted above, we expect DUR to decrease with PRE, CYC, and the product of the two. Additionally, DUR should be longer for majority governments (MAG) because they are supported by a legislative majority. Moreover, DUR should increase with the proportionality of electoral systems (PRO) because proportional systems are less likely to create a new and cohesive majority to replace the government. Since governments in semi-presidential regimes bear a dual responsibility to the parliament and the president when the latter is powerful, we expect DUR to decrease with PCP. Namely, the more powerful a president is, the higher his/her capacity to dismiss a prime minister. Table 4.4, which reports the results of the two models treating electoral cycle and regime type differently, largely confirms the hypothesis. Note that when explaining government duration, electoral systems are coded by their proportionality rather than qualitative differences because we are interested in the likelihood of the current government being replaced by a new majority, and the more proportional an electoral system is, the more difficult for this majority to emerge. In both models, the coefficient of PRO is statistically significant. In fact, we can safely tell from the coefficients (0.046 and 0.050) that government durability increases by at least 4.6 percent as the proportionality increases by 1. Concerning electoral cycle and regime type, neither model reveals significant results. For other variables, PCP has a negative and significant effect on DUR, a result implying that a powerful president not only increases the likelihood of minority government but also reduces the durability of a government. SEQ is again a significant and positive factor for government duration, suggesting that learning may reduce the bargaining cost in the formation and maintenance of governments. Interestingly, government duration is shorter for the Western European (WEU) regimes and decreases as DEM goes up. This is not a surprising result because the parliamentary tradition of Western Europe facilitates the building of legislative majorities, and democratic institutionalization is partly characterized by a country’s constraints on the chief executive. Overall, the empirical analyses demonstrate that electoral systems exhibit a strong impact on the beginning and end of semi-presidential governments, hence their accountability. In particular, the majoritarian systems are conducive to the formation of majority governments and the emergence of new majorities that can replace the old ones. Namely, semi-presidential

9780230_242920_05_cha04.indd 73

11/11/2010 5:11:49 PM

74 Jih-Wen Lin Table 4.4 Regression of government durability of semi-presidential governments Model 1 Variables

Coefficient

PCP PPM PRO MAG CYC PRE PRE×CYC SEQ DEM WEU CONSTANT F

– 0.022 – 0.022 0.046 0.113 0.054 0.014 – 0.008 0.003 – 0.043 0.274 2.983

Adjusted R-squared

0.089

Model 2

S.E.

Sig.

0.010 0.054 0.018 0.052 0.053 0.046 – 0.003 0.003 0.068 0.107 –

0.030 0.693 0.013 0.030 0.308 0.754 – 0.013 0.281 0.528 0.012 –

–

–

Coefficient – 0.025 0.008 0.050 0.124 – – 0.061 0.008 0.003 - 0.065 0.273 3.849 0.109

S.E.

Sig.

0.010 0.052 0.018 0.050 – – 0.059 0.003 0.003 0.062 0.100 –

0.014 0.871 0.005 0.015 – – 0.300 0.014 0.241 0.297 0.007 –

–

–

Note: Dependent variable = DUR.

countries adopting majoritarian systems are more likely to have governments representing the majority of legislators and responsive to the challenge of new majorities. Additionally, how a semi-presidential constitution positions the president also affects government accountability: the more constitutional powers a president enjoys, the more likely for him/her to form a minority government and shorten the life of a government. This is an understandable result because a powerful president is usually treated as the supreme political leader who decides how the government is formed and dismissed. For how the contextual factors affect government accountability, experience with semi-presidentialism has greater explanatory power than democratic institutionalization or the location of a regime do. With this big picture in mind, the next section presents three case studies that illustrate how the theoretical and statistical models work in reality.

Three case studies For a closer look at how the aforementioned factors affect the accountability of semi-presidential governments, France, Finland (before 2000), and Taiwan are worth comparing. These countries have been using different electoral formulas over a long period of time – TRS in France, PR in Finland, and SNTV in Taiwan – and their constitutions do not require that parliament participate in the appointment of the prime minister. We can thus control the rules of government formation and isolate the effect of electoral systems. Although

9780230_242920_05_cha04.indd 74

11/11/2010 5:11:50 PM

Rules of Electoral Competition 75

Taiwan is a new democracy and does not have a parliamentary tradition, France and Finland experienced periods in which their parliament was fragmented and social expectations for a strong presidency were high (Elgie 1999a: 73). At the aggregate level, these three cases display variant probabilities of majority government – 88.46 percent for France, 76.47 percent for Finland (1945–2000), and 14.29 percent for Taiwan (under SNTV). Interestingly, the average government duration of each of these three countries is in the reverse order: 0.39 for France, 0.45 for Finland, and 0.51 for Taiwan. These figures are consistent with the statistical findings shown above. In the following, each case study addresses three issues: the president’s constitutional powers, the relationship between the electoral system and legislative cohesion, and the actual process of government formation and government duration. France The presidents of the French Fifth Republic (1959 to present) appoint a prime minister without parliamentary approval and are active in cabinet reshuffling (Thiébault 2000: 512–515). The French president, as the chair of the Council of Ministers, is also an important policy-maker. However, rarely were minority governments formed in the Fifth Republic. To be more specific, a minority government was never formed when the majority of seats were controlled by the opposition alliance; instead, the presidents always appointed an opposition prime minister in these circumstances. To explain the pattern of France’s government formation, we begin by examining the role played by its electoral system. TRS, also called the two-ballot single-member system, is the electoral system of the National Assembly (Lower House) of the French Fifth Republic. Except in 1986, this system has been used in all National Assembly elections as well as the presidential elections.9 Under this electoral system, any parliamentary election candidate receiving more than 12.5 percent of the votes of the registered electorate in the first round can stand in the second round, and whoever wins the most votes in the second round is elected. In two ways, TRS explains why France has a cohesive parliamentary majority. First, despite their competition in the first round, parties with similar ideologies usually cooperate in the second round so as to minimize the chances of an opposition coalition winning the seat; they can also establish pre-electoral alliances so their supporters are given a clear signal of how to make their choice (Thiébault 2000: 498–502). Either case createa two bloc-politics, with one of them controlling the majority of seats. Second, the single seat design encourages parties in the same coalition to strengthen their interparty cohesion in order to promote their favorite policies. Due to these factors, the National Assembly always sees the left or the right alliance holding the majority of seats. The predominance of majority governments in the Fifth Republic can thus be attributed to the cohesive majority alliances of the National Assembly.

9780230_242920_05_cha04.indd 75

11/11/2010 5:11:50 PM

76

Jih-Wen Lin

The best evidence is the three cohabitations: whenever a National Assembly election gives the opposition alliance a new majority, the president names an opposition leader prime minister. Jacques Chirac of the Rassemblement pour la République (RPR) was appointed prime minister after right-wing parties won the majority of seats in 1986, five years after the socialist Mitterrand began his first term; Mitterrand appointed the RPR’s Edouard Balladur prime minister in 1993 when the right-wing coalition won a landslide victory; Lionel Jospin, a socialist, was selected prime minster by President Chirac in 1997, when the left-wing alliance took over the majority.10 The enormous seat swings resulting from TRS and the nonconcurrent electoral cycle (before 2002) make it difficult for prime ministers to govern for long: a new election can easily deprive their coalition of a majority, and a new president will have a mandate to establish a supportive majority. These institutional factors explain why the average duration of French Fifth Republic governments is only 0.39, even though most of them represented the majority of legislators. Taiwan Taiwan’s constitutional reform of 1997 gave the Legislative Yuan (parliament) the power to cast a vote of no- confidence in the prime minister but revoked legislative confirmation of the appointment of the prime minister, making the country’s constitutional system semi-presidential. The constitutional powers of Taiwan’s president are more limited than his French counterpart: the president does not participate in cabinet meetings and dissolves the Legislative Yuan only if a vote of no- confidence is passed; national security affairs, the reserved domain of the president, are administered by the Executive Yuan (government) rather than the office of the president. Despite the president’s limited power, the pattern of Taiwan’s government formation is the opposite of France’s: when SNTV was used (before 2008), six out of the seven governments were minority despite the existence of an uncompromising opposition majority in the Legislative Yuan. As explained, the prevalence of Taiwan’s minority governments is strongly correlated with SNTV, a candidate-based multimember system. Since the Legislative Yuan was opened for total reelection in the early 1990s, SNTV was used every three years to elect the members of the legislature. SNTV is a plurality system under which voters cast only one ballot for a particular candidate, votes received by one candidate cannot be transferred to other candidates, and a district usually elects multiple winners. Such a system encourages candidates to cultivate personal votes, overloads individual candidates with a heavy campaign burden, reduces the reelection rate, weakens party discipline, and produces limited seat swings because of its proportional nature.11 The actual process of government formation shows how SNTV constrained opposition legislators’ actions. When the pro-Taiwan independence

9780230_242920_05_cha04.indd 76

11/11/2010 5:11:50 PM

Rules of Electoral Competition 77

Democratic Progressive Party (DPP) candidate Chen Shui-bian defeated the Kuomintang (KMT) candidate in the presidential election of 2000, the KMT still held an absolute majority of seats in the Legislative Yuan. To counter the KMT’s boycott, President Chen appointed a former KMT general prime minister. This was not a majority government, according to the KMT, because the DPP turned down the opposition’s request to hold a cross-party negotiation of government formation. Nevertheless, the opposition legislators did their best to discredit the government but never attempted a vote of no- confidence. Some KMT members asserted that the government should be voted down if the president refused to negotiate, but this proposal was not even put on the legislative agenda. This pattern persisted during Chen’s presidency, and each of the prime ministers he appointed formed minority governments. Ironically, the same electoral system extends the life of Taiwan’s governments. The high campaign cost and low reelection rate under SNTV makes the president unafraid of a censure motion against the minority government he assembled; the opposition parties have little to expect from an early election because seat distribution does not change much under this electoral system. As a result, the government’s life is terminated only when the president finds it politically necessary to appoint a new prime minister. Another outcome predicted by our theory is that Taiwan’s governments survive longer than its French counterparts even though Taiwan’s president enjoys far less constitutional powers. A plausible explanation is that the fewer the president’s constitutional powers, the greater the importance of the government, hence the durability of the latter. Finland Finland has a long semi-presidential history. The constitution of Finland (the 1919 Constitution Act) gave its president considerable powers to counter revolutionary pressures when the nation declared independence (Paloheimo 2001: 271). Under the 1919 Constitution, the president had the initiative of government formation and the constitution did not require the parliament to confirm the president’s appointment of the prime minister (Nousiainen 2000: 268). Until 1991, Finnish presidents could actively dissolve the parliament and call for an early election. The president enjoyed reserved domain in foreign affairs, presented bills to parliament, had veto power, and issued decrees. These stipulations remained largely intact until a new Constitution Act came into effect in 2000. The pattern of government formation, however, had undergone a significant change. Under the old Finnish constitution, the percentage of minority governments is 23.53, which is higher than France but lower than Taiwan’s, but no minority government was formed after 1977.12 This pattern can be explained by the dual effects of PR, the electoral system used in Finland’s parliamentary elections. Legislative cohesion under proportional systems is balanced by two consequences: first, multipartism

9780230_242920_05_cha04.indd 77

11/11/2010 5:11:51 PM

78 Jih-Wen Lin

can weaken interparty cohesion; and second, the risk and cost of elections give legislators an incentive to pass a censure motion. It follows that legislative cohesion will increase when political parties improve their coalitionbuilding capabilities. As illustrated below, this is what happened in Finland, where open list PR has been used to elect its legislators since 1906. Some historical nuances aside, Finland’s parliamentary parties can be divided into three major groups – left, center, and right, and two of them usually constitute a majority. Whether these parties are able to dominate government formation depends on their coalition-building capabilities. According to the veto-player theory (Tsebelis 2002), the shortening of ideological distance and the restriction on minority vetoes give political parties greater incentives to form coalitions. Regarding ideology, the stances of the Finnish political parties witnessed a significant change in the past century. Built on the ruins of civil war, Finland unfolded its history with antagonistic partisan rivalries, and fragile, ephemeral governments followed. The inability of parliamentary parties to build effective coalitions made the Finnish president a pivotal player in government formation. During the Cold War, the need to maintain a peaceful relationship with the Soviet Union enhanced the political influence of the Finnish president, the decision maker in foreign affairs. The impact of the external environment was so strong that its transformation inevitably catalyzed a change in Finland’s domestic politics. Since the late 1970s, Europe gradually replaced the Soviet Union as Finland’s major external concern, a trend that undermined the president’s authority and increased the importance of ministerial governance. This transition also made the political parties more practical and conciliatory, as the left-right division became less relevant for many issues. Finland’s interparty cohesion was further strengthened by an institutional reform. According to Finland’s 1919 Constitution Act (Section 66), one-third of all deputies could vote to postpone a law until it was supported by a majority following a subsequent election. The abolition of this rule in 1992 greatly increased the voting power of the large parties (Mattila 1997) and reinforced the convention that, following each parliamentary election, the leader of the largest party would be appointed prime minister and mandated to take charge of government formation. Under PR, a parliamentary election is unlikely to create a seat swing large enough to signal victory for a particular coalition. Instead, government formation is to be negotiated among the parties. As a result, government duration is largely determined by intra- coalition cohesion. In Finland, the enhanced coalition-building capability thus prolonged government duration: between 1945 and 1977 the average duration is 0.43; since 1977 the average increases to 0.69. In sum, although PR has always been Finland’s electoral system, the consequences it produces have changed over time and can be used to explain the life cycles of Finnish governments.

9780230_242920_05_cha04.indd 78

11/11/2010 5:11:51 PM

Rules of Electoral Competition 79

Conclusion For democratic countries, the accountability of a government is gauged by how it begins and how long it lasts. The greater the number of people a government represents and the quicker it responds to changes in public opinion, the more accountable it is. By this standard, factors affecting the accountability of governments formed in a semi-presidential regime deserve special attention because these governments are not only accountable to the lawmakers but also to the president, who appoints the prime minister. Accordingly, government accountability in semi-presidential regimes is associated with the political importance of the president. The president’s powers, although important, are not sufficient to explain the formation and duration of semi-presidential governments. In this chapter, a model is built to show that, even if the president enjoys the power to single-handedly appoint the prime minister, a cohesive parliament can still prevent a minority government from being formed by credibly threatening the president with a censure motion or legislative boycotts. The model also explains why the electoral system is the major source of a country’s legislative cohesion. More specifically, the majoritarian-leaning systems are more likely than other electoral systems to create a stable and cohesive majority, hence a majority government. The effect of the proportional systems is the opposite. The rules of electoral competition also have a strong impact on government duration: insofar as the life of a government decreases with the likelihood of it being replaced by a new and strong majority, government duration should be shortest under majoritarian systems because of the massive seat swings they create. That the formation and duration of semi-presidential governments are sensitive to the cohesion of the parliament is consistent with a recent study by Tavits (2009), who claims that having a president directly or indirectly elected makes little difference if the government rests on the confidence of the parliament to survive. To conclude, semi-presidential regimes using majoritarian systems to elect their legislators have a higher probability of being governed by a majority government, but the same systems also make governments susceptible to the challenge of a new majority. A majoritarian electoral system certainly has the disadvantage of under-representing minority interests, making the “majority” government accountable only to those who support the large parties. By this standard, the proportional systems are more responsive because they favor the representation of small parties. However, proportional systems have their drawbacks when used in semi-presidential regimes, such as allowing the president to override a fragmented legislature and making governing parties insensitive to the outcome of parliamentary election. It is indeed difficult for semi-presidential democracies to accommodate responsiveness and accountability.

9780230_242920_05_cha04.indd 79

11/11/2010 5:11:51 PM

80 Jih-Wen Lin

Notes 1. There are so many sub-types of semi-presidentialism that one scholar proposed rejecting this concept (Siaroff 2003). To be inclusive, this chapter follows Elgie’s (2007) definition. 2. For this reason, the presidential-parliamentary system is criticized for the blurred responsibility of the government. In practice, constitutional arrangements concerning governmental power and responsibility are sometimes vaguely stipulated (Baylis 1996) and depend on convention to work. 3. For the non-free semi-presidential countries, the book edited by Elgie and Moestrup (2007) provides useful discussions. 4. This definition allows us to focus on the president’s appointment of the prime minister. Consecutive cabinets headed by the same prime minister will be counted as one government unless they are separated by a parliamentary election or a reshuffling that changes the majority/minority status of the government. 5. A multi-member district system, block voting (sometimes called plurality-atlarge voting), is highly disproportionate. Using plurality rule to rank the vote share of the groups, the largest group can easily have the candidates on its slate win every seat. 6. The result is similar if we recode “prime minister’s appointment requires consultation with parliament” as 0. 7. The list of presidential-parliamentary regimes can be found in Shugart and Carey (1992, 118; 160), Moestrup (2007: 42), Elgie and Moestrup (2007: 243), and this book. 8. This is a loose definition making no distinction among concurrent, honeymoon, counter-honeymoon, and midterm elections. Distinguishing electoral cycles by these categories is analytically inefficient because we need to define three dummy variables. 9. Since 1962, the French president has also been elected by TRS. If no candidate receives a majority of votes in the first round, the top two candidates will compete in the second round. This system is thus slightly different from the current electoral system of the National Assembly. 10. The National Assembly election leading to the first cohabitation (1986–1988) actually used PR rather than TRS. It is unlikely for a single PR race to alter the two-bloc competition that TRS created, and it happens that the right won the majority of seats in this election. But PR does reduce the gap between the two alliances: in 1986, the seat share of the right surpassed the left by 13 percent; in 1981, the left outnumbered the right by 37.08 percent. 11. For an overview of the features and consequences of SNTV, see Grofman et al. eds. 1999. 12. The caretaker governments are excluded. For details, see the website of the Finnish government: http://www.government.fi/tietoa-valtioneuvostosta/ hallitukset/en.jsp.

9780230_242920_05_cha04.indd 80

11/11/2010 5:11:52 PM

5 Semi-Presidentialism in Western Europe Robert Elgie

Western Europe is the birthplace of semi-presidentialism and the study of semi-presidentialism. The first two countries to adopt semi-presidential constitutions were Finland and Weimar Germany in 1919 and for decades, semi-presidentialism was confined almost exclusively to this region. From an academic perspective, Western Europe was the first region in which semi-presidentialism was studied (Duverger 1978; Nogueira Alcalá 1986; Ceccanti et al. 1996; Duverger 1996; Elgie ed. 1999; Pasquino 2007) and there has been considerable work on individual West European semi-presidential countries. In this chapter, we do not provide another description of the origins and basic functioning of West European semipresidential systems. Instead, we explore the extent to which we can draw general lessons about semi-presidentialism from a study of semipresidentialism in this region. We show that the experience of West European semi-presidentialism, and particularly semi-presidentialism in France, has shaped the study of semi-presidentialism generally. However, we argue that we need to be very careful about drawing any conclusions concerning the general effects of semi-presidentialism from the study of the West European experience alone. In the first part of the chapter, we identify the semi-presidential countries of Western Europe. In the second part, we show that there is considerable variation within West European semi-presidentialism. In the third part, we show that the study of semi-presidentialism was shaped by the study of West European and, in particular, French semi-presidentialism. We conclude that the study of Western Europe illustrates that semipresidentialism should not be used as a single explanatory variable, but that we must focus on the effects of variation within semi-presidentialism. We also conclude that the experience of individual West European cases can be used to illustrate particular issues that are potentially relevant to other cases, but that we have to be careful about generalizing from such a small sample.

81

9780230_242920_06_cha05.indd 81

11/11/2010 5:13:23 PM

82 Robert Elgie

Semi-presidential countries in Western Europe The definition of semi-presidentialism and the identification of semipresidential countries have been contested ever since Duverger’s founding work (Elgie 2007a). The study of West European semi-presidentialism is no exception in this regard. There has always been a certain amount of disagreement as to which West European countries should be classified as semipresidential. For example, Duverger (1980) identified six semi-presidential countries in this region (Austria, Finland, France, Iceland, Ireland, and Portugal), plus Weimar Germany as an historical example. However, in a much- cited article, Stepan and Skach (1993) identified only two such countries (France and Portugal). More recently, Matthew Shugart (2006) identified three West European semi-presidential countries (Austria, France, and Portugal). In this book, semi-presidentialism is defined as the situation where a country’s constitution makes provision for both a directly elected president and a prime minister and cabinet that are collectively responsible to the legislature. This definition makes no reference to the powers of the president as a defining feature of the concept. Therefore, even countries with figurehead presidencies have the potential to be classed as semi-presidential if the other elements of the definition are met. This makes the identification of semi-presidential countries relatively straightforward because no judgment call needs to be made as to whether a president is sufficiently powerful for a country to be classed as semi-presidential. All the same, in order to identify both the set of West European semi-presidential countries and the period when semi-presidentialism began in these countries, a few clarifications need to be made. In Finland the 1919 constitution established the popular election of the president rather than direct election. The president was elected by a 300person US-style electoral college who had been directly elected by the people. In 1987 the constitution was amended. As a result, people voted both for the president directly and for electoral college representatives. In the event that no candidate won more than 50 percent of the popular ballot, the electoral college would choose the president. In the 1988 presidential election, the electoral college selected the president. In 1991 the constitution was amended again and the electoral college was abolished. Since the 1994 election, the president has been elected in a two-ballot French-style system. In this chapter, we assume that the electoral college procedure was the equivalent of direct election and classify Finland as semi-presidential from 1919. This interpretation is consistent with the scholarship on Finland (e.g., Arter 1985; Paloheimo 2003). In Iceland and Ireland the president is directly elected, but parties have regularly agreed to stand only one candidate at the election. In this event, the sole candidate, usually an incumbent president, is elected unopposed

9780230_242920_06_cha05.indd 82

11/11/2010 5:13:23 PM

Semi-Presidentialism in Western Europe

83

and there is no poll. In Iceland the president serves for a four-year term and there have been five contested elections and ten uncontested elections since the constitution was adopted in 1944. For example, in 2008 the incumbent president was unopposed and the scheduled election did not take place. In Ireland the president serves a seven-year term and there have been six contested elections and six uncontested elections since the constitution was adopted in 1937. Despite the regular absence of presidential elections, we classify both countries as semi-presidential since the adoption of each country’s constitution. This is because the absence of a direct election has been caused by contingent political factors rather than by any constitutional amendment. In addition, we identify the start of semi-presidentialism as the year when a country’s constitution was adopted or when a semi-presidential amendment was made to the constitution, even if there were transitional arrangements for the election of the first president. For example, in Austria in 1945, Finland in 1919, Iceland in 1944, and Weimar Germany in 1919, the first president was elected indirectly by parliament. However, because these elections were constitutionally sanctioned transitional arrangements, we classify all of these countries as semi-presidential from the time when their constitutions were adopted rather than from the time when the first direct election of the president was held. In addition, in Austria in 1929 a constitutional amendment was passed that made provision for the direct election of the president. In fact, though, no such election took place prior to the suspension of the constitution in 1933. Even so, because the constitution was formally semi-presidential for this period, we classify Austria as semipresidential from 1929 to 1933. Finally, in France the direct election of the president was brought about by a constitutional amendment in 1962. This provision was not included in the original version of the Fifth Republic’s constitution in 1958. In fact, it was deliberately excluded (Elgie 1999a). So, the start of semi-presidentialism in France dates from 1962. On the basis of the definition adopted in this book and given these clarifications, there have been eight periods of semi-presidentialism in Western Europe, in seven different countries, six of which currently have a semipresidential constitution: Austria 1 (1929–1933), Austria 2 (1945–), Finland (1919–), France (1962–), Germany (1919–33), Iceland (1944–), Ireland (1937–), and Portugal (1976–).

Variation within West European semi-presidentialism The study of semi-presidentialism was initially motivated by Duverger’s (1978; 1980) desire to understand why countries that shared a basic constitutional structure operated in such different ways. While Duverger identified certain constitutional features that were common to a small set of West European countries, his interest in these countries was based on the

9780230_242920_06_cha05.indd 83

11/11/2010 5:13:23 PM

84

Robert Elgie

observation that in practice these countries operated in very different ways. In other words, there was no West European model of semi-presidentialism. Following Duverger and generalizing, we argue that the study of semipresidentialism needs to start from the assumption that we should not study semi-presidentialism as a discrete explanatory variable. Instead, we need to identify the variation within semi-presidential countries and study the effects of such variation. In a West European context, the variation within semi-presidentialism can be identified in a number of ways. Here, we focus first on the various origins of semi-presidentialism, second on the variation in presidential powers, third on the variation in party composition of the executive, and fourth on the government’s support in the legislature. The origins of West European semi-presidentialism are diverse (Table 5.1). We could chart this diversity in a variety of ways – whether semipresidentialism came about as a function of a constitutional amendment or a new constitution entirely; whether it was introduced while a country was already democratic or during a period of democratization, and so on. While these questions and others are legitimate lines of investigation, we focus on the presidential context of the founding circumstances. We do not need a strong theory of path dependence to suggest that the origins of institutions have the potential to leave a lasting effect. Indeed, Duverger (1980: 180–182) argued that the short-term circumstances concerning the adoption of semi-presidentialism had an effect on the longer-term exercise of presidential power. In a West European context, we find that there were three occasions when a directly elected president was expected to bring about or reinforce a trend Table 5.1 The origins of West European semi-presidentialism Country

Founding context

Austria 1

Ideologically motivated desire to weaken parliament where social democrats had strength. Also, reform passed after two failed attempts to elect a government. So, desire for executive leadership. A strong presidency was associated with prior collapse of democracy. First president elected by parliament and was a figurehead. Direct election of the president was a compromise. President adopted many of the Czar’s powers from the previous regime. The reform was designed to reinforce an already presidentialized system. The president was expected to be a figurehead like the monarch under former Danish rule. First president was elected by parliament. Historic and strong leader remained as prime minister. First president was nonpolitical and was the sole nominee, so no expectation that the president would be powerful. Strong president inherited from autocratic regime. Strong president designed as a counterweight to parliament.

Austria 2 Finland France Iceland Ireland

Portugal Weimar

9780230_242920_06_cha05.indd 84

11/11/2010 5:13:24 PM

Semi-Presidentialism in Western Europe

85

toward executive leadership. This was most notably the case in France in 1962, when the direct election of the president was a very deliberate attempt to institutionalize a strong presidency in a divided country. In addition, we might also see the two historic examples – Austria 1929 and Weimar 1919 – as cases where the direct election of the president was introduced as a counterweight to what was expected to be, or already perceived to be, a strong parliament. While there is little evidence that the Austrian president was particularly influential from 1929–1933, in France the president remains the fulcrum of the political process and in Weimar the president’s powers were exercised controversially particularly from 1925–1933. Arguably, the context in which semi-presidentialism was introduced helped to legitimize and assure the exercise of presidential power in these cases. By contrast, in three cases – Austria (1945–), Iceland and Ireland – a directly elected president was introduced in the context where the incumbent was expected to be a figurehead. In Ireland, the opposition criticized the idea of a directly elected president on the grounds that they feared the incumbent prime minister, Éamon de Valera, would seek election as president and proceed to introduce an authoritarian regime. In fact, there is no evidence that he ever had any presidential ambitions (Keogh and McCarthy 2007). Instead, he remained as prime minister and the presidency emerged as a purely symbolic institution. In this context, similar to the situation in Austria and Iceland, the origins of semi-presidentialism were not associated with any desire for presidential influence and this helps to explain why the presidency has remained a secondary institution. The cases of Finland and Portugal also resemble each other. As in Iceland, directly elected presidents assumed leadership roles that were broadly consistent with those under the previous regime. However, unlike Iceland, these regimes were formerly headed by strong figures. Thus, relatively powerful presidents emerged. In contrast to all of the other cases, though, the role of the president in Finland and Portugal has changed considerably over time. Thus, the founding context helps to explain the initial distribution of presidential power in these two countries, but not its subsequent development. Overall, while in most instances there is evidence that the origins of semi-presidentialism help to explain the subsequent exercise of presidential power, we also find considerable variation in the origins of West European semi-presidentialism. So, in this regard it is difficult to identify a standard West European model of semi-presidentialism. Another element of variation within West European semi-presidentialism concerns the constitutional powers of presidents. There are various ways of measuring presidential power. An early attempt was made by Duverger himself. In Échec au Roi (1978), Duverger selected 14 areas of presidential power and identified whether the constitution provided a country’s president with the power to decide matters in each area, whether there was merely a power to block the decision of another actor, or whether the president had no

9780230_242920_06_cha05.indd 85

11/11/2010 5:13:24 PM

86 Robert Elgie

power at all. Duverger did not provide a sum of the different scores, but he did rank countries on the basis of the president’s constitutional powers. He presented this ranking in a table that was reproduced in his European Journal of Political Research article (1980: 179). It is impossible to replicate Duverger’s method as he seems to weight the different powers unequally and he provided no information about the weightings. All the same, in descending order of power, he identified Finland and Iceland as the countries where the president had the strongest constitutional powers; there was then a block comprising Weimar, Portugal, and Austria where the president was quite powerful; next came France where the president was more of a regulator than a decider (ibid.: 23); finally, he identified Ireland as the country with the weakest president. (It should be remembered that Duverger was considering both the Finnish and Portuguese cases before their constitutions were changed to reduce the power of the president.) While Duverger’s method is impossible to replicate, he clearly showed that there was great variation in the constitutional powers of West European semi-presidential presidents. Since Duverger, there have been various attempts to measure presidential power more systematically. The first was provided by Shugart and Carey (1992: 155). Their scores are set out in Table 5.2. Here, the average score for the six West European countries is 6.1, but there is a wide range with Ireland scoring 0 and Iceland scoring 11. If Weimar is added, then the average score increases to 7.5 and Weimar becomes the country with the strongest president scoring 16. Given that Finland’s constitution has been amended considerably since Shugart and Carey’s initial calculations, we replicated the exercise on the basis of the six current examples of West European semipresidentialism (Table 5.2). Now, the average score has decreased to 4.8, but there is still the same range of scores. In both exercises the French score is around the average for the set of countries as a whole. Another attempt to measure presidential powers has been conducted by Siaroff. (Table 5.2). He identifies nine such powers. For West European semi-presidential countries, the average is 2.7, but this time France no longer has an average score. Instead, France emerges as the country with strongest president scoring 7 and both Austria and Iceland have the weakest presidents scoring just 1. The average score is 2.7 and 3.1 when Weimar and Austria 1 are included.1 In addition to calculating the powers of presidents, Shugart and Carey (1992) also distinguished between president-parliamentary and premierpresidential forms of semi-presidentialism. President-parliamentarism is where the prime minister is responsible to both the president and the legislature. Premier-presidentialism is where the prime minister is responsible solely to the legislature. All else being equal, we would expect presidents in president-parliamentary regimes to be more willing to act independently of the legislature because they have the opportunity to dismiss the prime minister. By contrast, in premier-presidential regimes presidents must work with (or through) the prime minister and cabinet as the government

9780230_242920_06_cha05.indd 86

11/11/2010 5:13:24 PM

Semi-Presidentialism in Western Europe

87

Table 5.2 Presidential powers in West European semi-presidential countries pres-parl = 0 Shugart and prem-pres = 1 Carey (1992) Austria 2 Finland France Iceland Ireland Portugal

Shugart and Carey (2008)a

Siaroff b

0 1 1 0 1 1

5 8 4 8 0 6

4 1 5 11 0 7.5

1 2 7 1 2 3

Average (6)

–

5.2

4.8

2.7

Austria 1 Weimar

0 0

– 12

– –

4 5

Average (7/8)

–

6.1

–

3.1

a

Notes: Author’s calculations. b Siaroff’s original scores (except Ireland where Siaroff’s mistake has been corrected) plus author’s calculations where missing.

is accountable only to the legislature. Using this distinction, there have been four president-parliamentary examples in Western Europe, including the two prewar cases, and four premier-presidential examples (Table 5.2). However, the distinction between these subtypes of semi-presidentialism is perhaps less helpful in the West European case than in other regions because it includes the Austrian and Icelandic, which are the two most notorious cases in the world in terms of the gap between the wording of the constitution – highly presidentialized – and the actual operation of political power – unequivocally parliamentary. Nonetheless, we might still postulate that this distinction explains part of the difference between politics in Weimar and the early years of the Portuguese system and politics in Fifth Republic France or Finland. Thus, even when we reduce the classification criteria to just one variable – whether or not the prime minister is also responsible to the president – we still find considerable variation within West European semi-presidentialism. In short, in terms of presidential power there is no model of West European semi-presidentialism. The final elements of variation that we identify concern the president’s relationship with the rest of the executive and the executive’s relationship with the legislature. Like Duverger, we would expect both relationships to effect the operation of semi-presidentialism. The president’s power is likely to be a function of whether he or she has support within the executive and the government’s authority will be at least partly determined by whether it can rely on the support of the legislature. Thus, if there really is a West European model of semi-presidentialism, we would expect a similar relationship to occur across the set of countries under consideration. To assess the president’s relationship with the rest of the executive and the executive’s

9780230_242920_06_cha05.indd 87

11/11/2010 5:13:25 PM

88

Robert Elgie

relationship with the legislature we focus on cohabitation, divided executive, minority government, and divided minority government. To calculate the percentage of time that each semi-presidential country has operated under each situation, we adopt a yearly unit of observation from 1919 to 2006. If a country experiences any of these situations for more than six months of a calendar year, then we count the whole year as an example of this situation. We take information concerning basic party affiliations from www.worldstatesmen.org, and we use a wide range of secondary sources to acquire other information. In terms of the party relationship within the executive, we focus on cohabitation and divided government. We define cohabitation as the situation where the president and prime minister are from opposing parties and where the president’s party is not represented in the cabinet. So, cohabitation corresponds to the situation where, like Mitterrand from 1986–1988, the president is isolated from any party support within the executive. Semipresidentialism is commonly associated with cohabitation. On the basis of the definition adopted here, what is the incidence of cohabitation in West Europe semi-presidentialism?2 The figures show that West European semipresidential countries tend to experience cohabitation around 22 percent of the time (Table 5.3). In this region, five of eight examples have experienced cohabitation very close to the average figure. So, there is some commonality across the set of semi-presidential countries in this regard. However, there is still considerable variation. The outliers are the two Austrian cases, where it has been rare, and Portugal, where it has occurred more than 50 percent of the time. Table 5.3 Party support in the executive in West European semi-presidential countries Divided executive – president and PM are Minority governCohabitation – from different parties ment – government president’s party is but the president’s does not enjoy not represented in party is represented in majority support in the government the government the legislature (% of period) (% of period) (% of period) Austria 2 Finland France Iceland Ireland Portugal

11.3 19.5 20.5 22.6 24.3 54.8

56.5 43.7 4.5 19.4 7.1 9.7

6.5 26.4 11.6 6.5 35.7 41.9

Average (6)

22.8

26.7

20.8

Austria 1 Weimar Average (8)

0.0 13.3 22.1

20.0 13.3 26.1

0.0 46.7 21.5

9780230_242920_06_cha05.indd 88

11/11/2010 5:13:25 PM

Semi-Presidentialism in Western Europe

89

We define a divided executive as the situation where the president is from one party and the prime minister is from another party, but where the president’s party is represented in the cabinet. In this case, there is divided executive in the sense that the president and prime minister are from different parties, but the president’s party is part of the governing coalition and so cohabitation is avoided. In France, this situation occurred from 1974–1976 when President Giscard d’Estaing from the Independent Republicans (and then the UDF) was faced by Prime Minister Jacques Chirac from the Gaullist party. The government included representatives of the president’s party as well as the Gaullists. A divided executive is also commonly associated with semi-presidentialism. What is the overall incidence of divided executives in West European semi-presidentialism?3 The figures show that there is tremendous variation (Table 5.3). On average, semi-presidential countries in this region have experienced divided government just over 25 percent of the time. However, France has experienced this situation only 4.5 percent of the time, whereas it has occurred in Austria 56.5 percent of the time. The final measures of variation concern the executive’s relationship with the legislature. Here, we identify periods of minority government and divided minority government. Minority government is where the government fails to enjoy majority support in the legislature. To identify periods of minority government, we use the World Bank’s Database of Political Institutions (DPI). This data set records a variable called “Majority,” which indicates whether the government enjoys majority support in the lower house of the legislature. The DPI data set goes back to 1975. We have used secondary sources to complete the data for all years. On this basis, what is the incidence of minority government in West European semi-presidentialism? (Table 5.3). We find that there has been minority government over 21 percent of the time. Again, though, there is huge variation, with Austria, France, and Iceland scarcely ever experiencing this situation, but it’s much more common in Ireland, Portugal, and Weimar. In addition, we also identify periods of “divided minority government.” This concept was identified recently by Cindy Skach (2005). She defines this situation as the case where “neither the president nor the prime minister, nor any party or coalition, enjoys a substantive majority in the legislature” (ibid.: 15). She identifies this situation as problematic and says that it “can predictably lead to an unstable scenario, characterized by shifting legislative coalitions and government reshuffles, on the one hand, and continuous presidential intervention and use of reserved powers, on the other” (ibid.: 17–18). We operationalize divided minority government as the combination of cohabitation and minority government. We find that there have been very few years of divided minority government. It occurred in Finland in 1927, 1929, 1948–1950, and 1972; in Ireland from 1948–1950; in Portugal in 1980, 1986–1987 and 2002; and in Weimar in 1920.4 Thus, while there is some variation across the eight countries under consideration, we can

9780230_242920_06_cha05.indd 89

11/11/2010 5:13:25 PM

90 Robert Elgie

conclude that this is the only variable for which there is very little variation across West European semi-presidential countries. To put it simply, divided minority government is a very rare phenomenon everywhere. All told, we find that on the basis of virtually all indicators, there is considerable variation within West European semi-presidentialism. This finding tells us two things. First, there is no West European model of semi-presidentialism, and this is not surprising. It is the assumption from which Duverger began the study of semi-presidentialism in the 1970s. The absence of a West European model means that we should not compare West European semi-presidentialism in toto, with semi-presidentialism in any other region(s). More than that, unless the variation within West European semi-presidentialism is anomalous, it suggests that if we wish to study semi-presidentialism generally, then we need to operationalize the variation within semi-presidential countries from the start. Rather than trying to explore whether semi-presidentialism leads to discrete outcomes, we need to explore whether particular types of semi-presidentialism lead to certain outcomes while other types of semi-presidentialism lead to different outcomes. Second, we have also found that France rarely emerges as a typical case of West European semi-presidentialism. Indeed, there is no typical West European semi-presidential country. If France is the archetypal semipresidential country; it is not so on the basis of the general West European experience. Instead, if France is archetypically semi-presidential, then it must be because it shares characteristics with countries outside this region. It is beyond the scope of this chapter to explore this issue. However, we might reasonably be skeptical as to whether this is really the case.

The study of West European semi-presidentialism In this section, we focus on the study of West European semi-presidentialism and outline the ways in which scholars have compared West European semipresidential countries. We emphasize the limits of existing studies. We then assess the study of individual semi-presidential countries in Western Europe, with a particular focus on the study of semi-presidentialism in France. We show that the French case has generated very different conclusions about the strengths and weaknesses of semi-presidentialism. The comparative study of semi-presidentialism in Western Europe Until the early 1990s, the vast majority of semi-presidential countries were found in Western Europe. There had been brief experiences with semi-presidentialism in Cuba in the 1940s and in South Korea and South Vietnam in the 1960s and early 1970s. Also, in the 1970s a small number of African countries adopted semi-presidential constitutions, though usually only for a short period of time. However, by 1987 there were still only eight countries in the world with semi-presidential constitutions, six of

9780230_242920_06_cha05.indd 90

11/11/2010 5:13:26 PM

Semi-Presidentialism in Western Europe

91

which were in Western Europe (Peru and Sri Lanka being the other two). In this context, it is not surprising that Duverger focused solely on the West European experience when writing about semi-presidentialism in the 1970s and 1980s. (Duverger 1978, 1980, 1986). Moreover, given Duverger’s intellectual influence, it is also not surprising that the limited literature on semipresidentialism during this period also focused solely on Western Europe (Bartolini 1984; Nogueira Alcalá 1986). This literature usually focused on all six cases of West European semi-presidentialism, often plus the Weimar Republic, although Duverger clearly privileged the study of France even within the context of his comparative work. In the early 1990s the third wave of democratization generated a lively debate about the advantages and disadvantages of different regime types. This debate centered mainly on presidentialism versus parliamentarism, but there was some discussion of semi-presidentialism. Inevitably, the earliest contributions to this literature also focused almost exclusively on the West European experience when drawing more general conclusions about the benefits or otherwise, usually the latter, of semi-presidentialism. However, what is common to this literature is that it focused only on a subset of West European semi-presidential countries. For example, Linz’s (1994) discussion of semi-presidentialism confined itself exclusively to Western Europe and, particularly, to semi-presidentialism in France and Weimar. Similarly, Sartori’s (1997) conclusions were based almost entirely on West European semi-presidentialism, particularly Finland, France, and Portugal. In their work, Shugart and Carey (1992) did discuss the Sri Lankan case alongside the same three West European examples, but their main focus was on pure presidential and parliamentary countries. Overall, the conclusions of the work on semi-presidentialism in this period were derived almost entirely from observations of the West European experience, and then often from only a limited number of West European semi-presidential countries. Since this time, the focus on West European semi-presidentialism has declined somewhat. In the late 1990s there was a vibrant debate in Italy that included a number of comparative studies of all West European countries or at least most of them (Sartori 1995; Ceccanti et al. 1996; Pegoraro and Rinella 1997; Rinella 1997; Giovannelli 1998; Canepa 2000). More recently, Cindy Skatch (2005) has compared the experience of Weimar and France. This literature apart, there has been much less focus on the comparative study of West European semi-presidentialism. Indeed, in the last few years Pasquino’s (2007) study of semi-presidentialism in France, Portugal, and Poland remains one of the few studies that compares more than one West European semi-presidential country and its remit extends to Eastern European as well. In general, the trend in recent times has been either to include West European countries in larger-n studies (Elgie 2008), or to focus on individual countries (Elgie ed. 1999). The study of West European semipresidentialism as a discrete unit of analysis has all but disappeared.

9780230_242920_06_cha05.indd 91

11/11/2010 5:13:26 PM

92 Robert Elgie

The emergence of the opportunity to study plenty of new semi-presidential countries in Eastern Europe, the former Soviet Union, and elsewhere is one reason for the decline in the specific focus on Western Europe. However, the development of political science as a much more rigorous discipline has also had a significant effect on the decline in the study of West European semi-presidentialism. The basic problem in this regard is that we have insufficient numbers to undertake a cross-sectional statistical analysis of West European semi-presidentialism alone. This means that if we wish to study all examples of semi-presidentialism in this area, we have to compare them with semi-presidential countries in other regions and/or with examples of other regime types in Western Europe and elsewhere. We cannot compare variation within West European semi-presidentialism on the basis of crosssectional data alone. A further problem concerns a statistical analysis based on panel data. The eight cases of West European semi-presidentialism generate a maximum of 376 observations in terms of semi-presidential years from 1919 to 2006. For some research questions there is little variation in terms of the standard variables. For example, if we hypothesize that presidential power has an effect on democracy, which we assess in terms of Polity IV measures, then we have 312 observations, because Polity does not provide a score for Iceland. If we dichotomize the Polity scores into the standard democracy (+1 to +10 inclusive) versus autocracy (–10 to 0 inclusive) categories, then we have only two years of autocracy in semi-presidential Western Europe. Even if we dichotomize the 310 democracy observations into years of full democracy (+8 to +10 inclusive) vs. partial democracy (+1 to +7 inclusive), then there are still only 36 years of partial democracy. In other words, any statistical analysis based on panel data would soon run out of degrees of freedom. While there may be some studies of West European semipresidentialism based on panel data that generate statistically meaningful results, many of the usual issues with which semi-presidential scholars are concerned – the success or failure of democracy, the quality of democracy and so on – cannot be studied in the West European context alone. In addition, as the discipline has advanced, the deficiencies of some of the existing work on Western Europe have become apparent. This is particularly true for Duverger’s own work. He aims to explain the variation in the practice of presidential power in all six contemporary West European cases. He identifies three factors that may explain this variation: the founding context of semi-presidentialism, the constitutional powers of the president, and the president’s relationship with the majority (Duverger 1978, 1996). However, the problem with this work is that none of the variables are identified rigorously. For his dependent variable, Duverger identifies a number of “modalities,” including a hegemonic presidency and a limited presidency (Duverger 1996: 518–587). Unfortunately, these modalities are poorly specified and a contemporary researcher would be expected to identify them much more rigorously. For his explanatory variables, we have already seen

9780230_242920_06_cha05.indd 92

11/11/2010 5:13:26 PM

Semi-Presidentialism in Western Europe

93

that Duverger’s measure of presidential powers is not replicable. In addition, his conceptualization of the founding context lacks any theoretical foundation and while his analysis of the president’s relationship with the parliamentary majority is well specified as it relates to France, the same is not true when it comes to the other five cases. In short, and with the benefit of hindsight, Duverger’s main contribution to the study of semi-presidentialism was the identification of the concept itself. Moreover, his identification of diversity within the set of West European countries remains the foundation on which rigorous studies of semi-presidential countries are based. However, his inductive approach to the study of West European semi-presidentialism has been much less successful at explaining cross-national variation than more recent deductive models. The comparative study semi-presidentialism in Western Europe In addition to comparative work, there are also many studies of individual West European countries. There are individual chapters on each West European country in a number of volumes (Duverger 1986; Elgie ed. 1999). Also, there are discrete articles on individual countries (e.g., Arter 1985; Martins 2006; Magalhães 2007; Nousiainen 2001). In this context, France has been studied the most. Indeed, France has emerged as the archetypical semi-presidential country. The concept of semi-presidentialism is usually synonymous with the French system. For example, Lijphart (2004: 102) calls France “the best-known example of semi-presidentialism” and criticizes this form of government on this basis. In this subsection, we focus on the impact of the study of France on the study of semi-presidentialism generally. The French experience has affected the study of semi-presidentialism in two very different ways. First, it encouraged some scholars to support the adoption of semi-presidentialism. This work is usually founded on an examination of the situation in France prior to the 1962 constitutional reform in comparison with the situation until the early 1980s. In essence, the story is very simple. Under the Fourth Republic (1946–1958), there was extreme governmental instability with the 25 government in 12 years. This was caused by the fragmentation of the party system and resulted in a leadership vacuum. With the creation of the Fifth Republic there was much more governmental stability and a bipolarization of the party system. Obviously, the change from a proportional to a majoritarian electoral system was partly responsible for the change. However, another factor was the adoption in 1962 of a directly elected president, who was able to provide an executive focus for the system. For some observers of the French case, the resulting system was more stable, democracy was more consolidated, and the quality of governance improved. The experience of the French case had a particular impact in Italy in the late 1990s. The similarity between the Italian situation at this time and the situation in France in the early 1960s was a key factor in this regard.

9780230_242920_06_cha05.indd 93

11/11/2010 5:13:26 PM

94 Robert Elgie

Reflecting on the debate at this time, one of the proponents of the semipresidential model, Gianfranco Pasquino, writes: for reasons having to do with the political culture of the two countries, the electoral proportional system, and the type of party systems, in Sartori’s classification, both belonging to the category of polarized pluralism, the French and the Italian political system could definitely be considered the two most similar systems. Therefore, my conclusion was that, semi-presidentialism having succeeded in modernizing and improving the working and the performance of the institutions of the French Republic, it was reasonable to expect that a similar improvement would follow its implementation in the Italian political context. (Pasquino 2005: ?)5 Given the near collapse of the Italian political system in the 1990s, the issue of constitutional reform was firmly on the political agenda and semipresidentialism emerged as one of the most commonly proposed solutions. In turn, this situation generated a considerable academic interest in semipresidentialism. Indeed, there is now probably more written in Italian specifically on West European semi-presidentialism than in either English or French. With the constitutional debate in Italy now largely at an end, the interest in semi-presidentialism seems to have declined somewhat. However, it is clear that the French case had a major impact in generating support for the introduction of semi-presidentialism. By contrast, the French experience has also affected the study of semipresidentialism by encouraging some scholars to oppose the adoption of semi-presidentialism. There are two ways in which this has occurred. The first focuses on the initial decades of the Fifth Republic. This period saw a presidentialization of the political system first under de Gaulle and then his successors. While many in Italy saw the return of executive authority as a positive development, other observers were concerned that a semipresidential constitution might generate a highly presidentialized form of government. For example, Lijphart’s continuing objection to semipresidentialism is associated with the problem of hyper-presidentialization. He states that semi-presidential systems “actually make it possible for the president to be even more powerful than in most pure presidential systems” (Lijphart 2004: 102). Also, Linz worries that semi-presidentialism can come to resemble “a constitutional dictatorship” (Linz 1994: 48). Given this potential for personalized leadership, many comparativists were concerned about recommending semi-presidentialism for new democracies. The second way in which the French experience has affected the study of semi-presidentialism is by focusing on events in France from the mid1980s onward. In 1986 France experienced its first period of cohabitation. There was political opposition to this situation – former prime minister and

9780230_242920_06_cha05.indd 94

11/11/2010 5:13:26 PM

Semi-Presidentialism in Western Europe

95

future presidential candidate, Raymond Barre, argued that it was damaging to presidential authority and that the president should resign. In fact, President Mitterrand had little incentive to do so and by making prime minister Chirac’s life as difficult as possible for the next two years he regained popularity and was reelected in 1988. In this context, the French experience has loomed particularly large. For obvious reasons, even when he was writing comparatively, Duverger focused overwhelmingly on the French case (Duverger 1978). Moreover, Sartori’s (1994) largely positive attitude toward semi-presidentialism was also heavily influenced by the French experience. By contrast, Lijphart’s (1992) influential critique of semi-presidentialism, particularly his notion of semi-presidentialism alternating between presidential and parliamentary phases (ibid.: 8), was entirely generated from observations of the French case. Equally, Stepan’s negative attitude toward semi-presidentialism was also strongly influenced by the French experience (Stepan and Suleiman 1995). In fact, we suspect that the overwhelming academic consensus against semi-presidentialism that formed by the mid-1990s was shaped by two very influential pieces of work on France that combined both of these criticisms. The first was Suleiman’s (1994) chapter in the highly influential volume edited by Linz and Valenzuela. Suleiman emphasized the strength of the French president, but he also underlined the rather shallow political foundations on which this strength was based. The second was Roy Pierce’s (1991) largely negative description of the first French experience of cohabitation from 1986–1988. For writers such as Linz and Stepan, a strong president was an inherently dangerous prospect and the notion that a country might also have to deal with the instability caused by shifting parliamentary and presidential majorities and the conflict that this could generate within the executive was sufficient to demonstrate to the satisfaction of most people that semi-presidentialism was a problematic constitutional choice and one that newly democratized countries should avoid. Overall, we argue that the general consensus against semi-presidentialism that formed in the early 1990s was heavily influenced by the study of semipresidentialism in Western Europe. More than that, we suggest that the consensus against semi-presidentialism was heavily influenced by the study of a small number of West European countries and, indeed, mainly by the study of the French experience in this region. This is a very small “n” on which to base any general conclusions. Nonetheless, the arguments identified in the early 1990s remain the bedrock of the academic wisdom concerning semipresidentialism.

Conclusion Western Europe is the cradle of semi-presidentialism. Duverger first identified the concept with reference to this region. The key intuition that

9780230_242920_06_cha05.indd 95

11/11/2010 5:13:26 PM

96 Robert Elgie

semi-presidentialism should not be operationalized as a discrete variable but that we need to examine the effects of different types of semipresidentialism was inherent in Duverger’s early work on countries in this area. More generally, most of the basic arguments for and against semipresidentialism were defined solely on the basis of reference to experiences in this area. In particular, the French experience shaped the early thinking about semi-presidentialism and much of that wisdom remains intact today. Obviously, though, there are problems associated with generalizations from such a small number of cases. In particular, we have to be careful as to how well the idiosyncrasies of the French case can be applied to countries more generally. Moreover, even though there is great variation within West European semi-presidentialism, we do not find examples of the sorts of semi-presidentialism that now occur in places like Central Asia and other parts of the former Soviet Union as well as certain parts of Africa, particularly Southern and Eastern Africa. To say the least, Western Europe does not provide a full picture of the semi-presidential world. At a certain point, there was little option but to examine the West European experience inductively and make generalizations about semi-presidentialism elsewhere. Now, though, we would do better to look beyond Western Europe for our insights into semi-presidentialism. Indeed, we would do better to adopt a deductive approach to the study of semi-presidentialism. Having done so, we should test our hypotheses upon as wide a set of countries as possible, including West European semi-presidential countries. Thus, the West European experience should remain central to our understanding of semi-presidentialism, but only in the context of broader studies of this form of government.

Notes 1. Siaroff’s score for Ireland is incorrect. Here, the president does not have the right to veto legislation, yet Siaroff’s measures indicate that this is the case. In Ireland, the president has the right to send a bill to the Supreme Court for its constitutionality to be judged. This is the equivalent of Article 61 of the 1958 French Constitution rather than Article 10, which outlines the French president’s limited veto power. So, Siaroff’s score for Ireland should be 2, not 3. This is the score recorded in this paper. 2. If the president does not have or claim a party affiliation, then we exclude the possibility of cohabitation. 3. Again, if the president does not have or claim a party affiliation, then we exclude the possibility of divided government. 4. It is possible that we have still not captured the incidence of divided minority government as Skach defines it. If the DPI records a minority government, it may be the case that there is cohabitation (i.e., the president’s party is not represented in the government) and that the president’s party enjoys majority support. However, given semi-presidentialism is defined on the basis that the government is responsible to the legislature, it is highly unlikely that this scenario would

9780230_242920_06_cha05.indd 96

11/11/2010 5:13:26 PM

Semi-Presidentialism in Western Europe

97

occur. Therefore, we are confident that we can use the DPI indicator as the basis of an attempt to calculate the incidence of divided minority government. 5. Pasquino’s recent comments on a semi-presidential blog confirm his interpretation of the situation: http://www.semipresidentialism.com/The_Semi-presidential_ O ne/ Blog / E nt r ie s/2 0 0 8/6/19_ C ou nt r ie s _ t hat _ have _ ac t ive ly_ debate d _ introducing_SP_(1)_-_Italy.html

9780230_242920_06_cha05.indd 97

11/11/2010 5:13:27 PM

6 Semi-Presidentialism under Post-Communism Oleh Protsyk

Constitutional engineering has been an important issue in the post-communist world over the last two decades. The process of post-communist transition encouraged the proliferation of different institutional frameworks for exercising political power. Due to this continuing process of experimentation with institutional templates, the region became an important ground for studying the causes and consequences of institutional innovations in the design of political systems. An institutional configuration providing for a popularly elected president and a prime minister and cabinet responsible to the legislature, which is referred to in this volume as a semi-presidential constitutional design, became a common choice among the countries of the region. The questions that scholars ask about semi-presidentialism in the postcommunist world are similar to those motivated by the long tradition of research on the role of formal institutions in politics. A key set of questions deals with the effects of institutions on the prospects for democratic consolidation. In this sense, debates about the virtues and drawbacks of semipresidentialism in the post-communist world build on older debates, most famously popularized by Linz (1990, 1994), about the merits of presidential and parliamentary systems. There has been a considerable divergence of views in these debates about the effects of constitutional choices (Mainwaring and Shugart 1997; Reynolds 2002; Cheibub 2007). Similar differences characterize the discussions about how consequential the variations in post-communist institutional designs have been for explaining the diverging trajectories of transition that former communist bloc countries have experienced. This chapter is organized as follows. First, I provide some basic information on constitutional regime types in the post- communist region, focusing on the nature and evolution of semi-presidential regimes. Second, I provide a short overview of the literature on post- communist semi-presidential regimes in relation to democratic consolidation. Third, I put forward some ideas about causal mechanisms that help explain the patterns of correlation between the strength of presidential powers and democratic underperformance/breakdown that this literature identifies. Finally, the chapter briefly 98

9780230_242920_07_cha06.indd 98

11/11/2010 5:14:05 PM

Semi-Presidentialism under Post- Communism

99

addresses the question of whether semi-presidentialism provides an equilibrium institutional solution by examining patterns of constitutional stability in semi-presidential regimes.

Constitutional choices in post-communist transition A semi-presidential institutional framework, in its different configurations, has been the preferred constitutional choice among the countries of the former communist bloc. If one classifies the first post-communist constitutions adopted by these countries into one of the three basic democratic regime types – keeping in mind that the analytical utility of having a semipresidential type remains much more disputed in the literature than the utility of differentiating between presidential and parliamentary types – the semi-presidential category turns out to be by far the most populated. Of 29 former communist bloc countries, only eight adopted parliamentary constitutions at the start of transition: Albania, Bosnia-Herzegovina, Federal Republic of Yugoslavia, Latvia, Czech Republic, Estonia, Hungary, and Slovakia. Another six adopted a presidential framework as their first post-communist constitutions: Azerbaijan, Belarus, Georgia, Tajikistan, Turkmenistan, and Uzbekistan. The first constitutions of the remaining fifteen countries shared the attributes of semi-presidentialism. This set of countries includes Armenia, Bulgaria, Croatia, Kazakhstan, Kyrgyzstan, Lithuania, Macedonia, Moldova, Mongolia, Poland, Romania, Russia, Slovenia, and Ukraine. Over the course of the almost twenty year period since the initial constitutional choices were made, constitutional reforms altering the basic institutional setup took place in a number of countries. These changes are addressed in some detail later in this chapter. It’s important to note that the overall distribution of regime types has not changed dramatically: semipresidential regimes remain the most frequent category among the former communist bloc countries. This endurance of semi-presidential regimes refutes the original expectations that semi-presidential constitutional choices would be short-lived, “transitional” arrangements. It also explains the sustained scholarly interest in the topic, which has significant practical relevance for constitution- and policy-makers across the region. It is important to note that the term “region” is used here with many reservations. It has become increasingly problematic to speak about postcommunist countries as a region. These countries are moving in radically different directions in terms of their political, economic, and social systems. This variation reflects some underlying structural differences that preceded communist rule. The rule itself, however, could be conceptualized as a similar experimental treatment applied to a set of countries with very different types of societal legacies. Constitutional rules can also be thought of as a sort of treatment applied to societies that faced the need to modify their political systems after the

9780230_242920_07_cha06.indd 99

11/11/2010 5:14:05 PM

100 Oleh Protsyk

collapse of communist rule. This treatment was far from identical, especially in cases where semi-presidential constitutional arrangements were chosen. While scholars routinely point out the existence of a substantial variation inside presidential and parliamentary regime types, especially large differences in institutional configuration of semi-presidential regimes make many authors question the analytical utility of the concept of semipresidentialism. For the concept to have any analytical purchase as a scope condition or independent variable, recognizing and summarizing this variation is essential. There have been attempts to summarize this variation using constitutional powers of presidents as a basis for classification (Shugart and Carey 1992; Frye 1997; Metcalf 2000; Beliaev 2006). Shugart and Carey’s (1992) distinction between president-parliamentary and premier-presidential systems can also be considered as one of the enduring typologies of semi-presidential regimes. The degree of presidential control over the cabinet is central in their definitions of these two systems. Under the president-parliamentary system, both president and legislature have a right to dismiss the cabinet unilaterally. Under the premier-presidential system, this power belongs exclusively to the legislature. The question about where the authority over the cabinet survival rests is critical for the functioning of the political system. In terms of patterns of post- communist constitution making, presidents rewarded with powers of cabinet dismissal were also granted strong legislative and other nonlegislative powers. While not addressing the need to have a multidimensional concept of presidential powers, Shugart and Carey’s typology nevertheless provides one useful proxy for distinguishing between semipresidential regimes with constitutionally strong and weak presidents. The initial distribution of premier-presidential and presidentparliamentary regimes in the post- communist region and subsequent changes to this distribution are presented in Table 6.1. This table indicates that the premier-presidential constitutional framework was chosen more frequently than the president-parliamentary one during the first two decades of transition. Another important feature of this distribution is the absence of president-parliamentary regimes in Central Europe, a region that faced most favorable conditions for democratization. This feature of distribution highlights some of endogeneity problems in studying the effects of institutional choices in post- communist transition. Discerning the effects of these choices on countries’ ability to democratize has been an important focus of scholarly writing on post- communist semi-presidentialism.

Democratic performance issues in the literature on postcommunist semi-presidentialism The central finding from the literature on post- communist semipresidentialism is the negative effect on democratic performance when

9780230_242920_07_cha06.indd 100

11/11/2010 5:14:05 PM

Semi-Presidentialism under Post- Communism

101

Table 6.1 Regime type and freedom house PR scores by geographic region Central Europe

Balkans

President-parliamentary President– parliamentary Croatia 1990–2000 ↑ Premier-presidential 3.89 Bulgaria 1991–2008 ↓ 1.56 Lithuania 1992–2008 • 1.07 Poland 1992–2008 ↓ 1.13 Romania 1991–2008 ↓ 2.56 Slovakia 1999–2008 • 1.00 1.59 Presidential – Parliamentarian Latvia 1991–2008 • 1.59 Slovakia 1992–1999 ↓ 2.17 Hungary 1990–2008 ↓ 1.12 Czech Republic 1992– 2008 • 1.00 Estonia 1992–2008 ↓ 1.33 1.55

Premier-presidential Croatia 2000–2008 • 2.00 Macedonia 1991–2008 • 3.38 Serbia 2007–2008 • 3.00 Yugoslavia 2000–2003 • 3.00 Slovenia 1990–2008 ↓ 1.12 2.50 Presidential –

Western CIS

Caucasus and Central Asia

Presidentparliamentary Belarus 1996–2008 ↑ 6.36 Russia 1993–2008 ↑ 4.57 Ukraine 1996–2004 ↑ 3.57 4.83

Presidentparliamentary Armenia 1995–2005 ↓ 4.75 Georgia 2004–2008 • 3.25 Kazakhstan 1993– 2008 • 6.00 Kyrgyzstan 1993– Premier-presidential 2008 ↑ 5.07 Moldova 1994–2000 4.77 ↓ 2.80 Premier-presidential Ukraine 2004–2008 Armenia 2005–2008 • 3.00 • 4.67 2.90 Mongolia 1992–2008 ↓ 2.07 Presidential 3.37 Belarus 1994–1996 ↑ 5.00

Parliamentarian Parliamentarian Albania 1998–2008 ↓ Moldova 2000–2008 3.22 ↑ 2.86 Bosnia-Herzegovina 1995–2008 ↓ 4.42 Montenegro 2007–2008 • 3.00 Yugoslavia 1992–2000 ↓ 5.86 Serbia-Montenegro 2003–2007 • 3.00 3.84

Presidential Azerbaijan 1995– 2008 • 6.00 Georgia 1995–2004 • 3.63 Uzbekistan 1992– 2008 • 7.00 Turkmenistan 1992– 2008 • 7.00 Tajikistan 1994–2008 ↓ 6.15 5.91 Parliamentarian –

Legend: The arrows indicate whether or not the PR rating increased or decreased during the period, with a decrease being “more free” and vice versa. FH scores reported in the table are averages of annual ratings for the entire length of regime. To get a “pure” FH score for a given regime, the dates used are one year into the regime and one year before it ended to avoid overlaps with other regime types. The • indicates no change. If a regime lasted less than 3 years, all scores (not just one year into it and before) were used. Note: Regimes with PR Scores of 1–2.5 are considered Free. 2.5–5.5 – Partially Free. 5.5–7 – Not Free. FH scores only used until 2007, 2008 figures not yet released. Source: www.freedomhouse.org

choosing a semi-presidential system with a constitutionally powerful president. This finding is far from being fully articulated and unanimously supported. It is a common theme that appears in various writings on the topic; a proposition that is supported by evidence generated by qualitative and quantitative studies. These writings are reviewed only briefly here,

9780230_242920_07_cha06.indd 101

11/11/2010 5:14:05 PM

102 Oleh Protsyk

without giving justice to the variety of topics and issues discussed in this literature. Rich and detailed accounts of the democratic performance of postcommunist semi-presidential regimes are contained in many country case studies collected in a series of volumes edited by Elgie and Moestrup (Elgie ed. 1999; Elgie and Moestrup eds. 2007; Elgie and Moestrup eds. 2008a). While these studies do not easily lend themselves to simple and crisp generalizations – much of case-based findings about institutional effects on democratic performance are context- dependent and far from being causally straightforward – they suggest the existence of a relationship between having a regime with a constitutionally strong president and a country’s inability to consolidate democratic gains. Thus summarizing the case study evidence from the volume devoted exclusively to the post- communist semipresidentialism, Elgie and Moestrup talk about negative effects of political systems that are described as “highly presidentialised semi-presidentialism and the balanced presidential-prime ministerial semi-presidentialism (Elgie and Moestrup 2008b: 257).” These studies are much more indeterminate about the effects of regimes with constitutionally weak presidents, although the above-cited chapter posits that “the parliamentary-like form of semi-presidentialism” is associated with some positive consequences. Attempts to discern institutional effects of this type of semi-presidentialism as compared to parliamentarism also face some peculiar challenges. Some scholars consider the former type of regimes to be nothing more than parliamentary regimes with a popularly elected president. In this view, the fact of popular election, in the absence of strong constitutional prerogatives, does not make much difference in presidential behavior. Tavits (2009) assembles a large body of evidence suggesting that few differences in behavior of popularly elected and indirectly elected Central European presidents could be attributed to a mode of election. While one can argue about whether evidence collected by the author actually matches theoretical claims made, the study highlights a particularly problematic status of this type of semi-presidentialism as an analytical category. While not directly addressing the questions of democratic performance, a number of studies point to the problems with semi-presidential regimes that can have an indirect effect on democratic performance. Roper (2002) finds that those premier-presidential regimes that are considered to be “the most presidential” have the greatest level of cabinet instability. Protsyk (2005) finds little empirical support for the claim that a premier-presidential institutional setting can have a conflict-mitigating effect during the early stages of democratic consolidation. Besides the case study research there have been a number of quantitative studies attempting to discern the effects of semi-presidential constitutional choice on democratic performance. These studies contain evidence that attribute some blame for bad democratic performance to constitutionally

9780230_242920_07_cha06.indd 102

11/11/2010 5:14:06 PM

Semi-Presidentialism under Post- Communism

103

strong presidents. These findings are rather robust; such studies routinely introduce a variables to control for the effects of social, economic, and structural factors that are usually given a causal primacy in explanatory accounts of differences in post- communist democratic performance. Relying on a large cross-region dataset, Moestrup (2007a) finds that semipresidential regimes performed worse than parliamentary regimes in the former Soviet Union where presidents tend to be constitutionally strong but not in Eastern Europe where powers of presidents are much weaker. In a study that focuses only on post- communist countries, Beliaev (2006) reports that political regimes with stronger executive powers of the president exhibit worse democratic performance and are less able to consolidate as democracies. The latter study also notes the lack of attention to delineating the mechanisms through which the impact of institutional choices on democratic consolidation realizes. The remainder of this chapter discusses some of these mechanisms.

Presidential leadership and democratic consolidation: Linkage mechanisms This section examines how the variation in concentration of constitutional powers in the hands of post- communist presidents affected the dynamics of political contestation in post- communist countries. This inquiry is facilitated by providing a brief overview of the democratic record of post- communist countries. Such an overview illustrates some of the findings reported in the previous section and gives the reader a better sense of how the values of key variables used to construct the argument are distributed. Table 6.1 provides one possible summary of the democratic record. It lists cases by constitutional regime type and groups them by geographic region, which can be considered as one rough proxy measure of structural variables that affect countries’ ability to democratize. Table 6.1 reports Freedom House’s political rights scores, which are used as an indicator of competitiveness of political regimes. For each of the cases Table 6.1 provides an average score for the period of regime duration and also indicates the direction of change in the scores over time (by subtracting the value of a score at the start of the regime from the score for the last available year). Higher scores signal worse democratic performance. Two important key findings for our discussion of the effects of semipresidential subtypes on democratic performance emerge from this exercise. First, premier-presidential regimes did not perform worse than pure parliamentary regimes in either of the two geographic regions where both types of regimes are concentrated. In Central European cases, the summary scores for premier-parliamentary and parliamentary were nearly identical, while in the Balkan cases the group of premier-presidential regimes performed significantly better than the group of parliamentary regimes. It is

9780230_242920_07_cha06.indd 103

11/11/2010 5:14:06 PM

104 Oleh Protsyk

also highly significant that over time each of the regimes that belonged to one of these groups either improved their democratic record or maintained the same score as it had at the time of regime inauguration. The second finding points to a substantially worse performance of president-parliamentary regimes in comparison to both premier-presidential and parliamentary regimes. The record of presidential regimes, which are geographically concentrated in the region of Caucasus and Central Asia, is even worse than the record of president-parliamentary regimes. Many of the presidential regimes started as fully authoritarian regimes, which explains why scores for some of them did not worsen further over time. Bad performance in the case of president-parliamentary regimes cannot be primarily attributed to the effects of geographic region. Both president-parliamentary and premier-presidential regimes are found across three out of four geographic regions specified in Table 6.1. President-parliamentary regimes did worse than premier-presidential regimes in each of the respective regions. Moreover, the democratic performance of the majority of presidentparliamentary regimes became worse over time, which, as mentioned above, is not the case for either premier-presidential or parliamentary regimes. We now discuss how exactly president-parliamentary framework affects countries’ ability to democratize, which is narrowly understood here as the ability to ensure competitiveness of political process. The key argument here is that endowing presidents with strong legislative and nonlegislative powers under conditions of a weak system of checks and balances leads to the proliferation of authoritarian practices. A counterfactual, which is always implicit in this type of argument, is that having a constitutionally weak president or parliamentary-type executive would have resulted in lower levels of power concentration in the hands of the executive and over the long term would have been beneficial for strengthening democracy. Two key mechanisms of how power concentration in the hands of presidents leads to the consolidation of authoritarian practices are discussed in the following pages. The first is the presidential (ab)use of executive decrees to monopolize political power. The second is the president’s deliberate employment of strategies directed at preventing party system development. The resulting interaction of a constitutionally strong presidency with weak parties helped to sustain and reproduce a pattern of authoritarian rule across a large number of post- communist countries. The thesis presented here is a revision of the classical Linz argument about the perils of presidentialism (Linz 1994). Although accepting the general premise of the argument made by Linz’s critics, who state that neither of the constitutional systems is inherently less conducive to democratic consolidation (Mainwaring and Shugart 1997), this section argues that a strong presidency under the specific conditions of a post- communist transition had a systematic adverse effect on efforts to establish democratic rules. Linz’s argument, however, is substantially revised here in order to make it

9780230_242920_07_cha06.indd 104

11/11/2010 5:14:07 PM

Semi-Presidentialism under Post- Communism

105

applicable to the post- communist context, which differed from the context of democratization in other parts of the world in many important respects. While the discussion initiated by Linz focused on the presidential role in undermining the survival of already existing democratic regimes, the postcommunist transition, especially in regions outside Central Europe, was all about building democratic institutions from scratch. Presidents in the Latin American context, which is the primary empirical reference point in Linz and his critics’ discussion, faced opposition political parties, institutionalized legislatures, and established judiciaries. All of these institutions contributed to the emergence of varying degrees of (albeit imperfect) checks and balances on Latin American presidents. Due to a peculiar legacy of the previous regimes, the first post- communist presidents, especially in regions outside Central Europe, faced few of these constraints in their efforts to preside over the construction of new political regimes. Executive decree authority and usurpation of political power Constitutions in many presidential and president-parliamentary systems grant presidents the power to issue executive decrees with legal power. There is a variation in the scope and limits of these powers but usually they provide presidents with the unilateral power of lawmaking. Whether these powers help to resolve collective action problems within the legislature or lead to power usurpation is one of the central questions in the literature on decree authority. The most comprehensive comparative study of presidential decree powers to date, the 1998 edited volume Executive Decree Authority comes out strongly in favor of the first interpretation (Carey and Shugart 1998). This benign view of the role of decree powers is, however, based primarily on a review of Latin American and US experience. Examining the presidential usage of decree powers in the post- communist context reveals less benign purposes and consequences of the use of executive decree authority. Presidents in presidential and president-parliamentary post- communist regimes relied extensively on decree powers for a variety of purposes (Protsyk, 2004; Haspel et al. 2006). Addressing policy problems, which is the focus of the 1998 volume cited above, does not exhaust the range of purposes for which decree powers were used by post- communist presidents. When the term “policy” is understood in the narrow sense of setting up rules and regulations in various areas of society, public policies were not the only focus of presidential decree making. This was especially the case during the formative years of new political regimes. Presidential decrees were actively used for other purposes such as delineating the jurisdiction of various government institutions and agencies; establishing lines of subordination and routes of decision-making in public bureaucracies attached to various branches of government; and specifying appointment procedures to a large number of government positions. Statutory provisions dealing with appointment procedures, for example,

9780230_242920_07_cha06.indd 105

11/11/2010 5:14:07 PM

106 Oleh Protsyk

had the final say in appointment matters to the presidential office either by providing presidents with the power to appoint officials single-handedly or by entitling them to confirm bureaucratic appointments in cases where the nomination powers were legally reserved for other institutional actors. Subsequent decisions on individual appointments that required presidential intervention were formalized in the form of presidential decrees. Presidential efforts to accumulate a high level of legal control over appointments and the manner in which these appointment powers were subsequently used are especially interesting for understanding the dynamics of power usurpation in regimes with constitutionally strong presidents. This is because appointments play a critical role in linking and reinforcing formal rules of subordination with informal norms of personal loyalty. Control of appointments and, most crucially, dismissals, contributed substantially to presidents’ ability to use government for political purposes. The critical importance of appointment matters for the president’s efforts to consolidate political power is reflected in the high share of appointment decrees in the total presidential decree output. Following the victory in the 1994 elections, Belarus’ president Lukashenko, for example, focused his efforts on staffing various public and semipublic institutions with his loyalists. Out of a total of 287 decrees issued during the first six months of his presidency, 180 decrees dealt with appointment matters. A content analysis of these decrees reveals that they were used to secure the appointment of loyalists to key positions regardless of whether the president had a formal legal right to make an appointment to a given political or civil service office. Among Lukashenko’s first decrees, for example, were decrees appointing the heads of state television and main government newspapers, which were positions that were clearly outside the scope of legal presidential competencies in appointment matters at the time of his election (Protsyk 2008). A large number of appointments in total decree output and similar tensions over the legal control of appointment powers in Russia and Ukraine are documented in the literature as well (Protsyk 2004). Although not policy decisions in a direct and immediate sense, appointment decrees, as well as decrees establishing government agencies or assigning policy jurisdictions, were instrumental in presidential efforts to amass ever more power. This accumulation of formal and informal power allowed presidents to use the government to restrict the competitiveness of political systems. The various ways in which authority was abused and public bureaucracies compromised is well analyzed in the literature. Practices that social scientists believe are sanctioned by the presidential office include selective law enforcement, arbitrary application of administrative norms and regulations, use of state ownership as a means of exerting political influence, and the overall politicization of public bureaucracies and management of state-run enterprises (McFaul et al. 2004).

9780230_242920_07_cha06.indd 106

11/11/2010 5:14:07 PM

Semi-Presidentialism under Post- Communism

107

The less benign view of the effects of executive decree authority presented here rests on the realization of the ineffectiveness of the system of checks and balances developed in much of the post- communist region. Even when the institutional and political environment in which post- communist presidents found themselves is compared with what is generally considered as weak checks and balances systems in Latin America, the performance of the post- communist institutions provided inferior checks on the executive. One telling illustration of this situation is the post-1993 practice of the Russian Constitutional Court to refuse considering cases where the constitutionality of presidential decrees is challenged (Protsyk 2008). In all areas of decision-making by decree presidents generally faced little opposition from the legislature, courts, or media. The subservient position of public institutions was partly a product of the president’s ability to shape the design of these institutions and their rules of operation. The very framework of the president-parliamentary constitutional design, which often provides the president with powers to dissolve the legislature, undermined the ability of legislative bodies to provide effective checks on presidential behavior. The overall weakness of checks and balances is also a product of the weakness of societal actors, first of all, political parties. Institutions designed to be part of the checks and balances system cannot perform their functions if operation of these institutions is not backed by the existence and support of organized political actors. Understanding the relationship between the president-parliamentary constitutional design and party system development is briefly addressed below. Constitutionally strong presidents and party system underdevelopment At the time of the initial choice of post- communist institutions, preferences for a strong presidency were combined with preferences for candidate rather than party oriented electoral systems. Single-member district (SMD) systems or mixed systems with a strong SMD component were a favorite choice of politicians designing electoral systems in president-parliamentary regimes. The existence of clientelistic networks inherited from the past provided the majority of incumbent politicians with the necessary resources to be successful in the SMD competition. Fully proportional electoral design proposals, which threatened to undermine these networks and deprive local powerbrokers of their key advantage, were discarded by the incumbent majorities across president-parliamentary regimes. Table 6.2 lists details about the sequence of electoral systems adopted by post- communist countries, and reflects this prevailing set of preferences among politicians in presidentparliamentary regimes. As seen in Table 6.2, none of the president-parliamentary regimes started the transition under PR electoral rules for legislative elections. The same

9780230_242920_07_cha06.indd 107

11/11/2010 5:14:08 PM

Caucasus and Central Asia

Western CIS

Balkans

Central Europe

Table 6.2 Regime type and electoral systems Premier-presidential Bulgaria 1991–2008 Lithuania 1992–2008 Poland 1992–2008 Romania 1991–2008 Slovakia 1999–2008 Parliamentarian Latvia 1991–2008 Slovakia 1992–1999 Hungary 1990–2008 Czech Republic 1992–2008 Estonia 1992–2008

(1990) Mixed x 1 round • (1994) PR x 4 rounds (1992) Mixed (50%) x 4 rounds (1991) PR x 6 rounds (1990) PR x 5 rounds (2002) PR x 2 rounds (1993) PR x 5 Rounds (1990) PR x 4 rounds (1990) Mixed (54%) x 5 rounds (1990) PR x 6 Rounds (1992) PR x 5 Rounds

President-parliamentary (1992) Mixed (43%) x 1 round • (1995) Mixed (63%) x 1 round Croatia 1990–2000 (2000) PR x 3 Rounds Premier-presidential (1990) SMD x 2 rounds • (1998) Mixed (29%) x 1 round • Croatia 2000–2008 (2002) PR x 3 rounds Macedonia 1991–2008 (2007) PR x 1 round Serbia 2007–2008 (2000) PR x 1 round Yugoslavia 2000–2003 (1990) PR x 5 rounds Slovenia 1990–2008 Parliamentarian (1991) SMD x 1 • (1996) Mixed (29%) x • (1997) Mixed Albania 1998–2008 Bosnia-Herzegovina 1995–2008 (26%) x 1 • (2001) Mixed (29%) x 2 (1996) PR x 4 rounds Montenegro 2007–2008 (2006) PR x 1 round Yugoslavia 1992–2000 Serbia-Montenegro 2003–2007 (1992) PR x 2 rounds - No elections took place, Federal Parliament selected from republican parliaments. President-parliamentary Belarus 1996–2008 (2000) SMD x 3 Rounds Russia 1993–2008 (1993) Mixed (50%) x 4 rounds • (2007) PR x 1 round Ukraine 1996–2004 (1994) SMD x 1 round • (1998) Mixed (50%) x 2 rounds Premier-presidential Moldova 1994–2000 (1994) PR x 2 rounds Ukraine 2004–2008 (2006) PR x 2 rounds Presidential Belarus 1994–1996 (1995) SMD x 1 Round Parliamentarian Moldova 2000–2008 (2001) PR x 2 rounds President-parliamentary Armenia 1995–2005 Georgia 2004–2008 Kazakhstan 1993–2008 Kyrgyzstan 1993–2008 Premier-presidential Armenia 2005–2008 Mongolia 1992–2008

(1995) Mixed (21%) x 1 round • (1999) Mixed (43%) x 1 round (2004) Mixed (64%) x 1 round • (2008) Mixed (50%) x 1 round (1994) Mixed (13%) x 4 rounds • (2007) Mixed (90%) (1995) SMD x 3 rounds • (2007) PR x 1 round (2003) Mixed (57%) x 1 round • (2007) Mixed (69%) x 1 round (1992) SMD x 4 rounds • (2008) SMD *Note: 2008 Electoral Code is ambiguous

Continued

9780230_242920_07_cha06.indd 108

11/11/2010 5:14:08 PM

Semi-Presidentialism under Post- Communism

109

Table 6.2 Continued Presidential Azerbaijan 1995–2008 Georgia 1995–2004 Uzbekistan 1992–2008 Turkmenistan 1992–2008 Tajikistan 1994–2008

(1995) Mixed (20%) x 2 rounds • (2004) SMD x 1 round (1992) Mixed (64%) x 4 rounds (1995) SMD x 3 rounds (1994) SMD x 4 rounds (1995) SMD x 2 rounds • (2005) Mixed (35%

Sources: http://www.electionguide.org/; Sarah Birch. “Electoral Systems and Political Transformation in Post- Communist Europe,” Palgrave MacMillan 2003.

was true in the case of presidential regimes. SMD or mixed systems (with no more than 50 percent of PR seats) were the only types of electoral rules adopted by president-parliamentary and presidential regimes at the start of the transition. A PR electoral system, which provides a much stronger impetus for party development, was, on the other hand, a frequent choice among parliamentary and premier-presidential regimes from the very beginning of the post- communist period. Presidents in president-parliamentary and presidential regimes found that these initial decisions to adopt SMD or mixed systems were well suited to their goals of power accumulation. Regardless of whether they participated or not in decisions about the choice of initial electoral rules, presidents opposed proposals for changing these rules. Weak party systems posed little threat to presidential dominance, which explains the presidents’ preference for keeping these electoral systems in place. Although presidents tried to build their own party machines from the beginning of the 1990s, they consistently used veto threats and the legislative majorities they controlled to prevent the switch to PR systems. The temporal dimension is important in discussions of the evolution of presidential preferences regarding electoral systems. As Table 6.2 indicates, PR systems or mixed systems with a high share of PR seats were introduced in the late 2000s by president-parliamentary regimes in Russia and Kazakhstan. These changes were endorsed by presidents for a number of reasons. One critical factor was that the gradual strengthening of authoritarian practices reduced over time the regime opposition’s ability to compete in the political arena. By the late 2000s the Russian and Kazakh opposition was so weakened that it could not take advantage of the introduction of a PR system. Presidents’ support for PR was thus a result of presidents’ confidence in the ability of proregime political parties to fully dominate the electoral field. Under the conditions of such domination, a PR system was seen by the presidents and elites loyal to them as a way of reducing the transaction costs of doing politics. Throughout most of the post- communist transition, however, presidents relied on SMD and mixed systems to produce weak parties, large numbers of independent MPs, and undisciplined legislatures. Presidents used

9780230_242920_07_cha06.indd 109

11/11/2010 5:14:08 PM

110 Oleh Protsyk

selective incentives or sanctions to ensure the compliance of the majority of independents with presidents and pro-presidential parties’ legislative agendas. Legislative majorities constructed in such a way provided an endorsement for presidential choices of cabinet appointments and legislative policies. Precisely because these majorities were the endorsers rather than the authors of appointment and policy decisions, legislative activity under president-parliamentarism contributed little to the emancipation of political parties. Finally, presidential efforts to undermine opposition with all available means (including political repression and violence) hurt the ability of a majority of political parties to raise money and recruit cadres for political activity. The deliberate use of state administrative and law enforcement apparatus against the opposition sent the public a message that non-sanctioned political activities were associated with high personal risks. As a number of recent studies indicate, fears of state retribution deter political participation and undermine parties’ ability to self- organize (McMann 2006; Schedler 2006).

Constitutional stability under semi-presidentialism While the level of democracy is a critically important outcome, it is not the only criterion employed to judge the performance of different constitutional systems. Constitutional instability is also of great interest to social scientists. Constitutional reforms are highly important events in themselves because they provide valuable information about the ability of different constitutional frameworks to manage political conflicts and cope with various challenges that political systems face. This section compares the stability of semi-presidential and other types of regimes found in the post- communist region. The question of whether a semi-presidential constitutional framework proved to be as stable as other types of constitutional systems is explored; then the discussion turns to the causes of instability, and some of the effects that constitutional change in semi-presidential systems had on the functioning of political systems. Table 6.3 lists all political regimes that were established by the adoption of the first post- communist constitution in each of the countries in the region. Table 6.3 does not list regimes that were subsequently formed in some of the countries. This provides a clear comparison of how the initial political regimes fared in terms of constitutional stability. Table 6.3 splits these regimes into two categories: the upper row includes those regimes that were stable, which means they did not experience constitutional modifications altering constitutional system type. The bottom row lists “failed” regimes, which are those regimes that experienced a change of constitutional system type. Table 6.3 also provides information on the democratic status of

9780230_242920_07_cha06.indd 110

11/11/2010 5:14:09 PM

Semi-Presidentialism under Post- Communism

111

Unstable regimes

Stable regimes

Table 6.3 Constitutional Regime Change (Original Post- Communist Regimes Only) Central Europe

Balkans

Premier-presidential Bulgaria 1991–2008 (0) Lithuania 1992–2008 (0) Poland 1992–2008 (1) Romania 1991–2008 (1)

PremierPresidentpresidential parliamentary Macedonia 1991– Russia 1993– 2008 (0) 2008 (0) Slovenia 1990– 2008 (0)

Parliamentarian Latvia 1991–2008 (0) Hungary 1990–2008 (0) Czech Republic 1992– 2008 (1) Estonia 1992–2008 (0)

Parliamentarian Slovakia 1992–1999 (0)

Western CIS

Parliamentarian Albania 1998– 2008 (0) BosniaHerzegovina 1995–2008 (0)

Presidentparliamentary Croatia 1990– 2000 (0)

Presidentparliamentary Ukraine 1996– 2004 (0) PremierParliamentarian presidential Yugoslavia Moldova 1994– 1992–2000 (0) 2000 (0) Presidential Belarus 1994– 1996 (0)

Caucasus and Central Asia Presidentparliamentary Kazakhstan 1993–2008 (2) Kyrgyzstan 1993–2008 (4) Premierpresidential Mongolia 1992– 2008 (1) Presidential Azerbaijan 1995–2008 (0) Uzbekistan 1992–2008 (2) Turkmenistan 1992–2008 (2) Tajikistan 1994– 2008 (2) Presidentparliamentary Armenia 1995– 2005(0) Presidential Georgia 1995– 2004 (0)

Legend: Free is in plain text, based on Fredomhouse.org scores of 1–2.5 /// Partially Free is in italics, based on FH PR score of 2.5–5.5 /// Not Free is in bold text, based on PR 5.5–7. (x) – Where X is the number of constitutional changes that have some effect on presidential powers (but do not lead to a change of constitutional system type).

each of the cases by using different fonts for names of democratic, partially democratic, and authoritarian regimes. Table 6.3 reveals significant cross-regional variations in the failure rate of constitutional regimes. Higher shares of political regimes were stable in Central Europe and the Caucasus/Central Asia regions than in the Balkan and Western Commonwealth of Independent States (CIS) regions. The group of Central European countries performed best in terms of constitutional

9780230_242920_07_cha06.indd 111

11/11/2010 5:14:09 PM

112 Oleh Protsyk

stability. Only one country out of nine in this group changed its constitutional system type. The Western CIS region has the worst performance record in this respect. Three out of four countries changed their initial choice of constitutional system type. A pattern of cross-regional variation identified in Table 6.3 provides some grounds for linking constitutional stability with levels of democracy. Two geographical regions with a better record of constitutional stability (Central Europe and Caucasus/Central Asia) occupy opposite ends on the democracy continuum. In other words, the most democratic and least democratic regimes in the post- communist world experienced the smallest number of constitutional system changes. Regimes with intermediate levels of democracy, which are over-represented in two other geographic regions (Balkans and Western CIS), had a higher level of constitutional instability. Partial democracies are most likely to experience constitutional change and this finding holds across different regions and constitutional system types. Among partial democracies that underwent a constitutional system change are presidential, president-parliamentary, and premier-presidential regimes. Only a group of partial democracies with a parliamentary form of government, including Albania and the highly idiosyncratic case of BosniaHerzegovina, avoided changes to the type of constitutional system. The roots of constitutional instability of regimes with intermediate levels of democracy can be traced to a number of factors. One underlying cause is linked to intra- elite bargaining that followed brief periods of high political mobilization. This mobilization was based on popular demands for further democratization and its most immediate trigger was electoral falsification. So- called “color” democratic revolutions led to constitutional regime changes in Georgia and Ukraine. Similar events led to constitutional revisions in the case of Kyrgyzstan. These revisions, however, were inside the boundaries of what is considered to constitute a case of presidentparliamentary regime. Thus Kyrgyzstan appears in the upper rather than the lower part of Table 6.3. In the absence of such mobilizational events as “color” revolutions, a slow process of democratization had a similar destabilizing effect on the setup of a constitutional system. A combination of domestic and international pressures for further democratization could be seen as a major cause of constitutional reforms in Armenia and Croatia. In both countries, presidential dominance in the political process, made possible by a presidentparliamentary constitutional design, was perceived as the main obstacle to achieving progress in democratic reforms. The impact that external pressures and diffusion of norms can have on countries’ choices of domestic institutions is well illustrated by these cases. Most accounts of the 2005 constitutional reform in Armenia, for example, agree on the critical role that pressure from the Council of Europe played in forcing the incumbent elites

9780230_242920_07_cha06.indd 112

11/11/2010 5:14:09 PM

Semi-Presidentialism under Post- Communism

113

to undertake such a reform. In the case of Croatia, the impact of European institutions was even more comprehensive. Another factor that explains constitutional change in partially democratic regimes is the failure of the semi-presidential framework to reconcile presidential and legislative claims of control over the executive. This was the case in Moldova, where legislators chose to abolish the institution of a popularly elected presidency following a series of clashes over distribution of power, policy, and appointment matters with the incumbent president (Roper 2002). At the time of this writing, a similar scenario may occur in Ukraine, where the recently established premier-presidential regime is in crisis. President Yushchenko’s determination to have a major say in executive matters faces increasingly radical opposition from a number of parties that control a constitutional majority of seats in the national parliament. Moldova at the time of the 2000 constitutional reform and current day Ukraine are both cases of premier-presidential regimes that grant the president only limited constitutional powers. In the absence of strong partisan support in the legislature, presidents in this type of constitutional regime face a losing battle if they decide to challenge the legislative control of the executive. As the Moldovan case demonstrates, a change of constitutional system type might be an outcome of conflict generated by such a challenge. Table 6.3 somewhat overestimates the degree of constitutional stability in the post- communist world. Many regimes in the upper row, which lists constitutionally stable regimes, also went through waves of constitutional revisions. Although the revisions did not amount to a constitutional system change, some of them significantly modified the presidential or legislature’s powers. The number in parenthesis next to each regime case is an attempt to capture a number of such revisions. In cases when legal transformations altering a constitutional system type did take place, what were the new constitutional systems put in place to replace the old ones? Table 6.4 provides details on these transformations. It lists all the cases reported in the bottom row of Table 6.3 and indicates how they evolved constitutionally. Several patterns are revealed in the data presented in Table 6.4. It highlights the finding that partial democracies were most likely to experience a constitutional system change. There was also one case of an authoritarian regime (Yugoslavia 1992–2000) and one case of a fully democratic regime (Slovakia 1992–1999) that went through institutional transformations that constitute, by this chapter’s criteria, a change of constitutional system. In the case of Yugoslavia, there were two changes of constitutional system type recorded. Both presidential and semi-presidential types of partial democracies went through constitutional changes. The number of constitutional changes was higher in the case of semi-presidential regimes but this is due to the

9780230_242920_07_cha06.indd 113

11/11/2010 5:14:10 PM

114 Oleh Protsyk Table 6.4 Countries with unstable constitutional regimes Presidential Georgia Presidential (1995–2004) 3.63 → President-Parliamentary (2004–2008) 3.25 Belarus Presidential (1994–1996) 5.00 → President-Parliamentary (1996–2008) 6.36 President-parliamentary Armenia President-Parliamentary (1995–2005) 4.75 → Premier-Presidential (2005–2008) 4.67 Croatia President-Parliamentary (1990–2000) 3.89 → Premier-Presidential (2000–2008) 2.00 Ukraine President-Parliamentary (1996–2004) 3.57 → Premier-Presidential (2004–2008) 3.00 Premier-presidential Moldova Premier-Presidential (1994–2000) 2.80 → Parliamentary (2000–2008) 2.86 Parliamentary Slovakia Parliamentary (1992–1999) 2.17 → Premier-Presidential (1999–2008) 1.00 Yugoslavia* Parliamentary (1992–2000) 5.86 → Premier-Presidential (2000–2003) 3.00 → Parliamentary (2003–2007) 3.00 Legend: Regimes in plain text are “free,” regimes in Italics “partially free,” and regimes in bold “not free.” Calculations based Freedomhouse.org PR scores. Using www.freedomhouse.org scores, countries with PR scores of 1–2.5 are considered free; 2.5–5.5 are partially free; 5.5–7 are considered not free. FH scores only used until 2007, 2008 figures not yet released. Note: * Yugoslavia was called Union of Serbia and Montenegro after 2003. Due to space limitations the annex with the list of constitutional revisions is not included here but it is available upon request from the author.

fact that a semi-presidential constitutional framework (either in presidentparliamentary or premier-presidential type) was the most frequent choice of the initial constitutional framework in the group of partial democracies. Trajectories of constitutional changes in partially democratic regimes differed across initial types of constitutional systems. All three presidentparliamentary systems listed in Table 6.4 went on to become premierpresidential. As discussed earlier, these are the cases where democratization pressures led to constitutional changes weakening formal presidential powers. Comparing the Freedom House scores for each of the regimes indicate that democratic performance in each of the three countries has proven to be better under the new constitutional system. In the cases of Armenia and Ukraine the improvements are minimal but this is partly due to the short time period the new regimes have been in place. Modest democratic gains indicate the beginning of a trend of consistently better democratic performance under the new institutional framework. Two regimes that started as presidential were transformed into presidentparliamentary regimes. The impetus for these transformations was very different in each of the countries. The increasingly undemocratic regime of president Lukashenko in Belarus pushed for constitutional changes in order

9780230_242920_07_cha06.indd 114

11/11/2010 5:14:10 PM

Semi-Presidentialism under Post- Communism

115

to consolidate presidential dominance by increasing the president’s formal powers, which were limited under the initial presidential constitutional framework adopted in 1994. A constitutional system change in Georgia, on the other hand, was a product of a so-called “color” democratic revolution. The presidential powers as a result of this change were also strengthened. Despite the different nature of events that led to the adoption of a presidentparliamentary framework in Belarus and Georgia, the underlying logic of constitutional changes was similar in both cases. Constitutional transformations reflected the preferences of politicians interested in strengthening the formal powers of the presidential office. The fact that power-seeking politicians in both cases considered president-parliamentary systems to be more appealing than the classical presidential system highlights the need to rethink our understanding of a status of presidential office in presidential and president-parliamentary systems. Table 6.4 also reports two cases of transformation of parliamentary systems into premier-presidential systems. Both the sequence of events and motivation of political actors were different in these two cases. In the Slovakian case, a key decision to introduce popular election of the president was a product of parliamentary parties’ inability to agree on an acceptable candidate. The legislature then decided to amend the constitution to transfer decision-making power on this issue back to the electorate. A similar constitutional change in Yugoslavia reflected very different political circumstances and calculations. At the time when constitutional changes were introduced, Yugoslavia was on the opposite end of the continuum of democratic performance from Slovakia. Introduction of a popularly elected presidency was a strategic move on the part of an authoritarian incumbent president to extend his term in office. A second change of the Yugoslav constitutional system, which followed fewer than three years later, took place in an already partially democratic environment and reflected party actors’ preferences for having a pure parliamentary system at the federal level of a reorganized union of Serbia and Montenegro (Todosijevic 2004). The dynamics of constitutional changes indicate the heterogeneity of causes and the variety of political circumstances in which constitutional transformations took place. One important conclusion is that partially democratic regimes are especially susceptible to the revision of constitutional norms. Changing the constitutional rules of the game appears to be partial democracies’ frequent response to political imperatives generated by the movement to or from genuine democracy. Another conclusion is that the deepening of the process of democratization is not compatible with the president-parliamentary constitutional choice. The distribution of formal constitutional powers under this constitutional system type strongly favors the president and generates discontent among other institutional actors such as legislatures and parties. This discontent is managed by the presidents when they face organizationally

9780230_242920_07_cha06.indd 115

11/11/2010 5:14:10 PM

116 Oleh Protsyk

weak political parties. A process of gradual strengthening of political parties increases the potential for conflict in these regimes.

Conclusion This chapter provides an overview of the evolution and performance of semi-presidential regimes in the post- communist world. Post- communist constitutional design experiments generated a rich and diverse experience of institutional interactions in a large number of transitional polities. This experience provides grounds for assessing how institutional frameworks affect polities’ ability to democratize, resolve political conflicts, and maintain regime stability. The chapter stresses the importance of conceptual differentiation among subtypes of semi-presidentialism. It argues that institutional variation leads to remarkable differences in the performance of semi-presidential regimes. Presidents’ ability to restrict the competitiveness of the political regimes over which they preside proved to be highly correlated with the strength of formal legislative and nonlegislative powers of presidents. A presidentparliamentary constitutional design is highly detrimental for the prospects of democratic consolidation and conflict resolution. Legal empowerment of presidents under this type of constitutional framework generate patterns of power relations that are not compatible with a genuine democratic process. President-parliamentary regimes are able to deepen democracy only by means of a transition to a premier-presidential constitutional framework. Both president-parliamentary and premier-presidential regimes in the post- communist world face a similar difficulty that many democratizing presidential regimes in other parts of the globe face. This difficulty is the challenge of reconciling the presidency with multipartism. The growing maturity of post- communist multiparty systems challenges popularly elected presidents in both types of semi-presidential regimes. While premierpresidential regimes have proven to be able to endure and maintain their institutional distinctiveness in the face of such a challenge, the constitutional framework of president-parliamentary regimes does not constitute an equilibrium institutional solution for democratizing multiparty polities.

9780230_242920_07_cha06.indd 116

11/11/2010 5:14:10 PM

7 Semi-Presidentialism and Democratic Development in East Asia Benjamin Reilly

East Asia contains three of the world’s youngest semi-presidential democracies: Mongolia, Taiwan, and East Timor. In addition to having a semipresidential constitutional structure, each of these countries also represents a relatively unusual case of democratization: Taiwan is one of East Asia’s famous “tiger” economies and the world’s only Sinitic democracy, but faces an ongoing crisis of nationhood; Mongolia is one of the few unambiguous examples of a successful transition to democracy and a market economy in the postcommunist world; while East Timor is both East Asia’s poorest nation and its newest democracy. As such, each represents an important test case for assessing the effect of semi-presidentialism upon democratic development. It is worth noting at the outset that two other East Asian states, South Korea and Singapore, are also sometimes identified as having semi-presidential systems. However, these are both questionable cases for students of semipresidentialism. South Korea features a strong executive president with the power to appoint the prime minister, who cannot then be removed by the legislature. Singapore is at the other extreme, with a Westminster-style parliamentary government and an elected but essentially figurehead president. Both countries thus represent challenges in terms of their correct classification, and in practice are most often categorized as cases of straight presidentialism and parliamentarism, respectively. Singapore is also far from being a genuine democracy. For these reasons, both of these cases are excluded from further discussion in this chapter. By contrast, Mongolia, Taiwan, and East Timor are clearly semi-presidential, and are frequently cited as standout examples of successful democratization in East Asia. However, each has also been troubled by extended periods of political instability, deadlock, and outright violence. None has suffered a complete democratic collapse (although East Timor has come close), but all remain unconsolidated democracies. My argument in this chapter is that the semi-presidential constitutional model adopted in each country 117

9780230_242920_08_cha07.indd 117

11/11/2010 5:15:44 PM

118 Benjamin Reilly

has contributed directly to this lack of consolidation by exacerbating the impacts of political crisis, particularly during periods of divided government. Accordingly, this chapter will assess the impact of semi-presidentialism upon democratic development in these three, relatively unlikely cases of East Asian democratization.

East Asian models of semi-presidentialism The decision to adopt semi-presidential systems of government in East Timor, Mongolia, and Taiwan was made in the context of their respective transitions to democracy. Semi-presidential constitutions were adopted by constitutional revisions in Mongolia in 1992 and Taiwan in 1997, and as part of East Timor’s foundation constitution in 2002. In each case, these new or revised constitutions explicitly provided for a directly elected president with more- or-less significant powers, and a prime minister chosen by the legislature. As such, each is clearly semi-presidential in the sense of having “a popularly elected, fixed-term president existing alongside a prime minister and cabinet who are responsible to parliament” (Elgie 1999b: 13). However, the powers granted to these directly elected presidents differ considerably in each case. For instance, Taiwan’s constitution grants extensive powers to the president, including the ability to appoint (and, by extension, dismiss) the prime minister and cabinet. By contrast, East Timor’s president is a largely symbolic figure with no role in cabinet formation, whose most important power is as supreme commander of the armed forces – a provision that had great importance during the internal conflict of 2006, discussed below. Mongolia sits somewhere in between, with the president and the parliament both having influence over prime ministerial appointments. Using Shugart and Carey’s (1992: 23–25) standard classification, Taiwan is an example of “president-parliamentarism,” while East Timor, with its relatively constrained presidential powers, is clearly “premier-presidential.” Mongolia’s presidential powers are more substantial than East Timor’s but also fall under the “premier-presidentialism” label (and indeed have become more so as a result of constitutional amendments introduced in 2001 that removed the president’s erstwhile powers to block prime ministerial appointments). In none of these countries do presidential powers approach those of the French Fifth Republic, the archetypal model for many discussions of semi-presidentialism. Each country’s form of semi-presidentialism also differs in other ways. In Taiwan, the president is head of state and commander of the armed forces; represents the nation in foreign relations; and appoints the heads of four of the five yuan (branches of government), including the prime minister and the cabinet. This makes the president potentially very powerful, especially during periods of unified single-party rule. Presidential powers in East Timor and Mongolia are much less extensive, with the prime minister prevailing

9780230_242920_08_cha07.indd 118

11/11/2010 5:15:45 PM

Semi-Presidentialism and Democratic Development

119

in most affairs of state. East Timor’s president is symbolically extremely important but has relatively few constitutional powers and (unlike Taiwan) cannot appoint or dismiss the prime minister. Nonetheless, as analysts predicted, the president’s symbolic status as the “father of the nation” has been extremely powerful at times of political crisis (Shoesmith 2003). Mongolia’s president also has considerable powers, although less than previously. The 1992 constitution allowed the president to veto government bills and appointments (including that of the prime minister). An extended period of cohabitation in the late 1990s led to a drawn-out conflict between the two main parties over the balance of presidential and prime ministerial powers, as successive parliamentary nominees for prime minister were rejected by the president. It was not until 2001, when both the presidency and the parliament were again controlled by the same party, that Mongolia finally passed constitutional reforms that reduced the president’s ability to reject the parliament’s wishes. This compromise also allowed members of parliament to serve in the nation’s cabinet (which is possible in both Taiwan and East Timor).

Background While it would be hard to find three more different Asian states than East Timor, Mongolia, and Taiwan, from the point of view of democratic development they share a number of characteristics. First, while not without problems, all three are genuine democracies, characterized by free elections and an open media and judiciary. As one indication of this, each have been ranked among the most democratic countries in Asia on the annual Freedom House scores of political rights and civil liberties since the turn of the century. Each have also met Huntington’s (1991: 266–267) “two-turnover” test of democratic achievement, although again (as discussed below) in rather different ways. Second, all three of these states have been consistently seen as relatively unlikely cases of democratization – cases in which countries essentially beat the odds, whether we are talking about a history of foreign occupation and conflict (East Timor), a Sinitic culture and ever-present external threat (Taiwan), or the dismal legacy of Soviet-style communism (Mongolia). In addition, democracy in each country faces major ongoing challenges: Taiwan’s international status remains unresolved; Mongolia is still struggling to establish a viable market economy; while East Timor has suffered violent internal conflicts and remains one of the world’s poorest countries. The comparative politics literature suggests that such circumstances represent major stumbling blocks to democracy. As such, these three cases are all, in their way, unusual positive cases for students of democratization. In addition to their democratic endurance, the practice of democratic governance in East Timor, Mongolia, and Taiwan also has other common

9780230_242920_08_cha07.indd 119

11/11/2010 5:15:45 PM

120 Benjamin Reilly

features. All display unusually high levels of popular participation and relatively strong party systems, for instance. Indeed, on the basis of recent election results, East Timor, Mongolia, and Taiwan may be in the process of developing nascent two-party systems which, potentially, may come to resemble the alternating government-and- opposition party models of many consolidated democracies (Reilly 2007). This is relevant for discussions on the functioning of semi-presidentialism, as it means that party system fragmentation – a frequent cause of instability and gridlock in many new democracies – is not a significant factor affecting political stability or democratic consolidation. Despite this, democracy in all of these countries has often functioned poorly, with recurrent problems of gridlocked government, political instability, and politically motivated violence. My argument in this chapter is that these recurring pathologies are due, in part, to their semi-presidential constitutional structures, particularly the propensity of such systems to deliver splits within the executive where the president and the prime minister are drawn from different parties. Such periods of divided government have had a pronounced negative impact on political stability and effectiveness in each country, weakening the consolidation of democracy in all three cases. It is worth stating at the outset that this is not a new argument – indeed it reflects one of the most consistent critiques of semi-presidentialism in the academic literature. As a recent analysis of constitutional design in Weimar Germany and the French Fifth Republic concluded, “one of the main lessons from this experience, with respect to semi-presidentialism as a model for export to democratizing countries, was that everyday politics was affected by the power struggle between the president and prime minister” (Skach 2005: 112). This power struggle has clearly hampered democratic consolidation in both East Timor and Mongolia. The situation in Taiwan was somewhat different, being an example of minority government from 2000–2008 when the president’s party was in opposition in the legislature but maintained control of the premiership (in part due to a constitutional amendment in 1997 that enabled the president to appoint a premier without legislative approval). There too, however, political gridlock and instability have been prevailing problems. Despite this, the lessons of these and other Asian cases of semipresidential government have yet to be integrated into the comparative literature, much of which continues to rely predominantly on the experience of a few familiar Western cases such as France. As Elgie (1998: 298) notes, there is “a great need to examine the politics of semi-presidential regimes outside the established democracies.” Yet even some of the most systematic academic attempts to test such critiques, such as Elgie’s 2008 article cited above that looks at whether the drawbacks of semi-presidentialism have been exaggerated, fail to include cases such as East Timor or Taiwan

9780230_242920_08_cha07.indd 120

11/11/2010 5:15:45 PM

Semi-Presidentialism and Democratic Development

121

in their analysis, limiting the relevance of their conclusions for East Asia in particular. This chapter addresses this problem. As developing democracies, East Timor, Mongolia, and Taiwan are unquestionably important test cases for debates about the impact of semi-presidentialism on democracy. The most serious recent challenges to democratic stability in each of these countries has occurred in part due to conflicts between presidents and prime ministers during periods of cohabitation, or between the president and the legislature during periods of divided government. In this respect, the experience of these three countries is very much in line with the academic consensus against semi-presidentialism due to its propensity for intra-administration conflict, institutional uncertainty, and immobilized politics.

Democratic development A starting point for selecting the country cases examined in this chapter is the distinction between autocratic and democratic forms of government. My primary focus is on democratic and democratizing states, as it is only in democracies that constitutions and other institutional variables are really consequential for political outcomes (hence the exclusion of lessthan- democratic cases such as Singapore). As I have detailed in a recent book-length survey of political engineering in the Asia-Pacific region, constitutional reform in general and the choice of semi-presidentialism in particular go hand in hand with democratization in East Asia (Reilly 2006). The number of East Asian regimes that can be considered to meet the basic Schumpeterian definition of democracy – that is, governments that are chosen via open and competitive elections (Schumpeter 1947: 269) – has grown significantly over the past two decades. While at the end of the Cold War only Japan could lay claim to the title of an “established” East Asian democracy, the years since then have ushered in a new era of liberalization and democratization across the region. Major transitions from authoritarian rule toward democracy began with the popular uprising against the flagrantly corrupt Marcos regime in the Philippines in 1986, the negotiated transitions from single-party governments in South Korea and Taiwan in 1987, and the collapse of communism in Mongolia in 1990, before moving on to the resumption of civilian government in Thailand in 1992, the United Nations intervention in Cambodia in 1993, the fall of Indonesia’s Suharto regime in 1998, and the international rehabilitation of East Timor that culminated in 2002. Despite backsliding in some countries (e.g., Thailand), more East Asian governments are today chosen through competitive and freely contested elections than ever before. As well as being among Asia’s genuine democracies, East Timor, Mongolia, and Taiwan share another distinctive characteristic: none had any prior experience of democratic governance before their transitions from authoritarian

9780230_242920_08_cha07.indd 121

11/11/2010 5:15:45 PM

122 Benjamin Reilly

rule. Prior to their democratic transitions, each was under the influence of large foreign powers – be it Russia in relation to Mongolia, Indonesia in East Timor, or China’s claim to sovereignty over Taiwan. This is not a propitious starting point for a transition to democracy; indeed, in different ways, each country lacked some of the essential preconditions for successful democratization. Nonetheless, each has succeeded to the extent that successive free elections and peaceful changes of power have occurred. Indeed, as noted above, each country meets Huntington’s “two-turnover” test of democratic consolidation: that is, when the party or group that takes power in an initial election loses a subsequent election and turns over power, and those election winners then peacefully turn over power to the winners of a later election. Mongolia, which has had four turnovers of power since independence, clearly qualifies on this score, as does Taiwan, especially with the election of Ma Ying-Jeou to the presidency in 2008. East Timor is marginal, but could also be included, if we assume that the 2001 constituency assembly elections, which marked the end of UN administration there, is counted as a handover of power. Taiwan was the first of this trio to democratize, with the incumbent Kuomintang (KMT) party playing a crucial role in opening up the political system to competitors. The KMT – the Chinese nationalist government-inexile from the civil war – which had ruled Taiwan unchallenged since the end of the Second World War, itself initiated the decisive steps toward political liberalization in the late 1980s. Opposition parties such as the Democratic Progressive Party (DPP) were legalized, media restrictions lifted, and constitutional reforms enacted, resulting in the direct election of the Taiwanese parliament, the Legislative Yuan, for the first time in 1992. Voters rewarded the KMT’s skilful handling of this transition by reelecting them to office – an unusual outcome for a transition from authoritarian rule. Further constitutional reforms saw the first direct presidential elections under Taiwan’s revised semi-presidential constitution held in 1996, which was also won by the KMT. It was not until March 2000 that a clear transfer of power across party lines occurred, when DPP candidate Chen Shui-bian scored a narrow victory in presidential elections, ending more than five decades of unbroken KMT rule and ushering in a fractious period of minority government between the DPP-led executive and the KMT- dominated legislature. Chen’s narrow reelection in 2004 saw this period of minority government extend for a total of eight years. Mongolia – a poor, landlocked, post- communist state enjoying few of the facilitating conditions for democracy – has to date experienced six competitive elections and successive transfers of power since the fall of communism. In contrast to Taiwan’s sequenced transition from authoritarian rule, democracy in Mongolia arrived much more suddenly, with the collapse of the Soviet Union leading to an economic crisis and regime change in 1990. In a truncated and at times chaotic transition toward a liberalized polity and

9780230_242920_08_cha07.indd 122

11/11/2010 5:15:46 PM

Semi-Presidentialism and Democratic Development

123

market economy, Mongolia’s former ruling communist party organization, the Mongolian People’s Revolutionary Party (MPRP), took over 80 percent of the seats in the Great Hural (parliament) in the country’s first- ever multiparty elections in July 1990, and followed this with an even more resounding victory in subsequent elections held two years later. That same year, 1992, saw Mongolia adopt a new, semi-presidential constitution, and the following year the country’s president, Punsalmaagiyn Ochirbat (a former MPRP stalwart) was elected for the first time by popular vote (Fish 1998a). It was not until 1996 that this period of MPRP dominance ended. Mongolia’s 1996 parliamentary elections saw the surprise victory of the new Democratic Coalition, a union of liberal parties, which captured 50 of the parliament’s 76 seats, sweeping the MPRP out of power. But in the subsequent presidential elections of 1997, Natsagiyn Bagabandi showed that the MPRP was anything but finished as a political force, sailing to a resounding victory. This ushered in a three-year period of cohabitation in Mongolia, with the president and prime minister from opposing parties and the president’s party not represented in government. This unstable situation lasted until 2000, when the MPRP returned to government with an overwhelming parliamentary election victory, winning 72 seats in the 76-member parliament with around 52 percent of the popular vote and restoring unified single-party rule. The emergence of an independent and democratic East Timor was even more unlikely. A long-neglected Portuguese colony, it was invaded by Indonesia in November 1975 following a declaration of independence by the leftist Revolutionary Front for an Independent East Timor (Fretilin). For the next 23 years, a low-level guerrilla war between the Indonesian military and the disputed entity’s resistance forces ensued. But in 1998 the Asian economic crisis hit Indonesia, prompting the collapse of the longruling Suharto regime. In the chaotic months that followed Suharto’s successor, former Vice-President B. J. Habibie, made a series of snap decisions that opened the door for a referendum on East Timor’s future. Under pressure from the international community, Habibie surprisingly agreed to a UN-sponsored referendum to settle the question of East Timor’s status. A popular referendum on the question (the first ever held in the territory) was duly held in 1999, with almost 80 percent of the electorate rejecting Habibie’s offer of autonomy, in effect registering a decisive vote for independence. This unleashed a wave of violence, destruction, and looting by pro-Indonesian militias. Frantic international negotiations resulted in an Australian-led regional military intervention that restored basic security and placed East Timor under the control of the UN. Two years later, in its final act before handing back power to the East Timorese people, the UN organized East Timor’s first free elections on August 30, 2001 – to a constituent assembly, which then drafted a constitution for the new state. The constitution, based on Portugal’s post-1982 constitution,

9780230_242920_08_cha07.indd 123

11/11/2010 5:15:46 PM

124 Benjamin Reilly

installed a semi-presidential system of government, with a dominant prime minister and a popularly elected president possessing vague but potentially significant powers. With the drafting process complete, the constituent assembly was transformed into the nation’s first democratic parliament. To complete this transformation, presidential elections were held in April 2002, and were won with a massive 87 percent of the popular vote by the charismatic former resistance leader Xanana Gusmão. Parliamentary elections, however, had seen a landslide victory by Fretilin, led by Gusmão’s rival Mari Alkitiri – effectively setting up a period of divided executive government (Soares et al. 2003). How successful have these various transitions proved to be in consolidating democracy? One way to answer this question is to look at the comparative rankings given to each country by independent bodies such as Freedom House, which conducts an annual survey of political freedom and civil rights around the world. Table 7.1 presents some basic geographic, demographic, economic, and political data for these states, including the latest Freedom House scores. These are measured on a scale of 1 to 7, with 1 representing the highest degree of freedom and 7 the lowest. As Table 7.1 shows, both Mongolia and Taiwan are accorded the highest rankings on these scores, while East Timor is currently listed as “Partly Free,” predominantly due to its ongoing problems of political violence. In sum, our three key cases deal with semi-presidential systems where there have been extended periods of divided government between rival presidents and prime ministers. While the president’s powers vary, and are constrained to greater (e.g., East Timor) or lesser (e.g., Taiwan) degrees, the symbolic political role of the president remains paramount in all cases. Some see this as a virtue. Elgie and Moestrup (2007: 245), for instance, argue that “the Timor-Leste and Mongolian cases show that a president whose role is usually more symbolic than participatory can help to stabilize the political process in times of political crisis. These cases suggest that

Table 7.1 Basic indicators for East Timor, Mongolia, and Taiwan

Country

Population 2004

East 925,000 Timor Mongolia 2,590,000 Taiwan 22,700,000

Land area (sq km)

Per capita GNI ($US)

14,900

840

156,700 36,200

1,000 16,495

Freedom house Year of Year semiranking demopresidentialism 2008 cratization adopted 3.5 (Partly Free) 2.0 (Free) 1.5 (Free)

2001

2002

1990 1987

1992 1996

Source: Asian Development Bank, Key Indicators 2008 (www.adb.org/statistics); Freedom House, Freedom in the World 2008 (www.freedomhouse.org).

9780230_242920_08_cha07.indd 124

11/11/2010 5:15:46 PM

Semi-Presidentialism and Democratic Development

125

semi-presidentialism may not be as inherently problematic as its critics suggest and, indeed, that a particular form of semi-presidentialism may even have some conflict- dampening advantages.” Regardless of whether this is true – and I have doubts – even with relatively constrained formal powers, in practice presidents in East Timor, Mongolia, and Taiwan have exercised far more political power during periods of divided government, particularly in time of national crisis, than one would expect from a narrow reading of their country’s constitution. In each case, the travails of divided government put the president and the prime minister in direct competition not just for power but for legitimacy. The travails of divided government placed immense pressures on the young and embryonic democratic systems of Mongolia, Taiwan, and especially East Timor.

Problems of divided government: Instability, polarization, and violence In his survey of comparative constitutional design, Giovanni Sartori – one of semi-presidentialism’s best-known proponents – characterized it as a “fragile system,” noting “The split majority problem still haunts . . . the semipresidential experience” (1994: 137). The recent experience of divided government in Mongolia, Taiwan, and East Timor supports this conclusion. In each case, familiar problems of split executives, dual power-bases, and competition between rival officeholders acted as powerful negative factors for democratic consolidation, particularly during the early years of transition from authoritarian rule. Deep and sometimes crippling internal divisions have been the reality of semi-presidentialism in each country for significant periods of their recent democratic experience. Mongolia has had the shortest period of divided government, with the cohabitation from June 1996 to July 2000. East Timor’s period of divided government can be assessed as lasting from April 2002, when Xanana Gusmão won the country’s first presidential election, through to Mari Alkitiri’s resignation and replacement by Jose Ramos-Horta as prime minister in July 2006. Taiwan’s experience of minority government was a constant throughout Chen Shui-bian’s presidency, from 2000 to 2008, as the DPP controlled executive power but the opposition “pan-Blue” coalition, led by the formerly-ruling KMT, maintained a majority in the legislature. These periods of government division placed a great deal of stress on each country’s embryonic political institutions and appeared to undermine the consolidation of democracy in each country. One reason for this is that in each case, the initial period of division came early in the country’s democratic experience, and many of the political actors had no real sense of how to deal with the uncertainties of political incongruence. Moreover, division brought with it a series of political problems that undermined political institutionalization, turning national politics into a competition among

9780230_242920_08_cha07.indd 125

11/11/2010 5:15:46 PM

126 Benjamin Reilly

powerful individuals. This led to familiar patterns of political polarization, instability, and violence emerging in each country, although at markedly different levels. While the severity of these differed considerably – the crisis in East Timor, for instance, was clearly a different magnitude to that of Taiwan or Mongolia – the incidence of these problems of political polarization, instability, and violence can be compared in each country to those times when unified government was in place. Mongolia Mongolia was the first of our trio to experience political cohabitation between rival parties. In Mongolia, the surprise 1996 parliamentary victory of the Democracy Coalition – a broad collection of democracy activists, intellectuals, and reformers – set up a period of direct confrontation between the legislature and the incumbent MPRP government of President Natsagiyn Bagabandi. The extent of this confrontation and failure to make cohabitation function was illustrated by Bagabandi’s successive rejection of the legislature’s nominees for the post of prime minister (he vetoed one Democratic nominee for prime minister seven times in a row). Under Mongolia’s 1992 constitution, the parliament could override this presidential veto only by a two-thirds majority vote, a level of representation that the Democratic Coalition (unlike previous governments) did not possess. Cohabitation lasted until 2000, and most of this period was spent in severe political gridlockdeadlock, with caretaker governments marking time for most of this period. This ongoing political and constitutional battle also led, unsurprisingly, to a high degree of political instability, despite the government’s majority in parliament. 1998 alone saw four different executives formed, mostly due to the president’s willingness to obstruct the new government by refusing to ratify its prime ministerial nominee – a power that President Bagabandi used no less than 11 times during the 1996–2000 cohabitation period. The end of divided government in 2000 did not immediately end this period of gridlock: an ongoing constitutional standoff between the parliament and the constitutional court saw the period of systemic crisis continue for another year. The resolution of this protracted period of political instability and stalemate finally came in 2001, with the passage of constitutional amendments that reduced the president’s power to appoint the prime minister and cabinet. These require the president to appoint the parliament’s choice of prime minister within five days, and to then reach agreement with the prime minister on the cabinet within a week, or else risk the prime minister taking the issue directly to parliament. Together with concurrent reforms stipulating that parliament be dissolved if no decisions are made within 45 days (a response to the internal disagreements that plagued the Democratic Coalition government), these reforms have rebalanced the relationship between the president and parliament.

9780230_242920_08_cha07.indd 126

11/11/2010 5:15:47 PM

Semi-Presidentialism and Democratic Development

127

The reductions in presidential powers plus the extended period of united government since then led to more stable politics in Mongolia, at least until the aftermath of the 2008 parliamentary elections. These elections, won convincingly by the MPRP, led to accusations by the Democratic Party that the vote was rigged, and subsequent protests against the election results turned violent, with five protesters killed and a state of emergency declared. Following the protests, the Democratic Party was included in the government coalition as a junior partner. The most recent presidential elections, in 2009, were won by the Democratic Party, ushering in another period of divided executive government with the MPRP continuing to hold the premiership. Despite these recent upheavals, Mongolia’s democratic experience has for the most part been remarkably peaceful. The main exception to this, apart from the events of 2008, was during the 1996–2000 cohabitation period, which saw the country’s first and so far only political killing. In 1998, S. Zorig, one of the country’s most prominent democracy activists and parliamentarians, was found murdered in his apartment. Zorig, who was Minister for Infrastructure at the time of his death, had been one of the leaders of Mongolia’s democracy movement. His murder remains unsolved, although it was widely presumed to be a political assassination (Finch 2002). On balance, semi-presidentialism does not appear to have been a positive factor for the consolidation of democracy in Mongolia. Indeed, Mongolia is an example of how the inherent contestation within semi-presidential systems over the control of executive power, particularly during periods of cohabitation, can pose a threat to democratic development. As such, Mongolia’s experience undermines claims by authors such as Steven Fish who praise semi-presidentialism as “a boon to democratization in Mongolia” (2001: 331). Rather, as Moestrup and Ganzorig emphasise, the Mongolian experience reinforces the reality that semi-presidentialism presents “particular challenges for young democracies in terms of institutionally- driven conflict due to alternative presidential and parliamentary readings of the constitution” (2007: 182). Taiwan Taiwan’s semi-presidential constitution is the outcome of its phased transition to democracy in the late 1980s and subsequent constitutional amendments, which strengthened the president’s hand, particularly the adoption of direct presidential elections in 1997. The emergence of minority government between 2000–2008 under Chen Shui-bian’s presidency presaged a sharp increase in political polarization on several levels: between the two party blocks, the “pan-blue” KMT and the “pan-green” DPP; between the president and the parliament; and – perhaps most seriously – between the population at large, as campaigning politicians emphasized distinctions between islander and mainlander identities that could be mobilized as

9780230_242920_08_cha07.indd 127

11/11/2010 5:15:47 PM

128 Benjamin Reilly

electorally-potent social divisions. The regionalized nature of party support served to deepen these divisions, particularly between Fukien Taiwanese and the rest. Taiwan came, quite quickly, to exhibit many of the distinctive sociopolitical characteristics of a society polarized around ethnic issues. Chen’s presidency featured eight years of minority government dividing the executive and the legislature. This deepened Taiwan’s already significant political polarization, coloring nearly every aspect of politics. As one commentator put it, the division of executive and legislative powers between the DPP and the KMT “engendered political gridlock, which hurt the economy and caused deterioration in social relations. It permeated the decision-making processes, and affected domestic politics as well as military planning and formulating foreign policy” (Copper 2003: 326). Chen’s overt championing of the need for a more independent international identity was popular domestically, but caused concerns in both China and the US, the two countries that effectively control Taiwan’s international space. Taiwan’s deepening ethno-nationalist division was in part a reflection of this overt intervention of international factors into the domestic arena, as well as a predictable outcome of an increasing political emphasis on issues of identity, particularly at election time. Chen’s eight-year administration saw a succession of cabinets formed and then reshuffled. Early cabinets featured a majority of opposition members in the executive (including Chen’s first premier, Tang Fei, a former KMT general), while later cabinets were more conventional DPP- dominated coalitions, but still faced a hostile parliamentary majority. Although the DPP was the largest party in the legislative Yuan for most of this period, it never had a legislative majority, creating a situation of deadlock in which the KMT and its allies usually controlled enough seats to block DPP-supported legislation. With a government divided between a president and a parliamentary majority from opposing political parties, Taiwan entered a period of policy paralysis and confusion, departing from its “developmental state” model of earlier years (Wu 2007a). The 2000–2008 minority government also saw Taiwan’s most serious contemporary experience of political violence. In the last days of the 2004 election campaign, Chen and his vice presidential running mate Lu Hsiu-lien were both injured in an apparent assassination attempt. Their subsequent election victory, by the narrowest of margins, led some opponents to claim that the attacks were deliberately staged to help the government. While the truth remains unclear and may never be known, the 2004 shootings serve as another example of the problems of politically-motivated violence intruding into electoral politics during periods of divided government. Taiwan’s semi-presidential political system, it seems, had become prone to conflict. In addition, there is abundant evidence that the stability and effectiveness of Taiwan’s political system deteriorated during the period from 2000–2008. The duration of each cabinet – a key indicator of political stability – declined

9780230_242920_08_cha07.indd 128

11/11/2010 5:15:47 PM

Semi-Presidentialism and Democratic Development

129

by more than 50 percent, from an average of 36 months prior to 2000 to just 16 months since then. At the same time, governments also became less effective, passing fewer bills and losing a majority of votes in the legislature. These included crucial budget and defense bills, which were consistently blocked by the opposition. As Yu-shan Wu (2007b: 201–218) has argued, Taiwan’s experience highlights the fact that semi-presidentialism is a system that is easy to choose but difficult to operate. East Timor In East Timor, the assumption of a semi-presidential government in May 2002 ushered in an ongoing power struggle rooted in decades- old personal antagonisms between the country’s key independence leaders. The results of the first- ever presidential election that year pitched East Timor into an ongoing period of divided executive government that resembled cohabitation, with a president from outside the governing party confronting a unified government.1 The Fretilin- dominated parliament was led by a prime minister, Mari Alkatiri, who was both personally and politically opposed to President Xanana Gusmão. Gusmão, who was elected with no specific party endorsement, was widely revered for his role as a guerrilla leader during the decades-long quest for independence, and made no secret of his growing discontent with Alkatiri and Fretilin’s increasingly autocratic governance style. This personal antagonism exacerbated the country’s deep historical and regional divisions, fomenting ethnic clashes and tension within the bureaucracy and armed forces. These tensions mounted over the next three years and came to a head during 2006, when Dili and surrounding regions descended into mass civil violence. The unrest had begun in March, when roughly one-third of the East Timorese defense force were dismissed for striking in protest over poor working conditions, low pay, and claims of ethnic discrimination. As gangs took up arms on the streets, the disciplined forces split along ethnic lines. Ongoing clashes between the army and the police saw more than 30 people killed and 150,000 Dili residents displaced, many of whom remain in temporary camps to this day. An Australian-led contingent of 2,200 foreign troops was deployed to ensure security in the lead-up to the 2007 elections. The collapse in public order brought to a head the long-running tensions between Gusmão and Alkitiri, with the president supporting the fired soldiers and endorsing public calls for Alkatiri’s resignation, based in part on allegations that he had armed a hit squad to kill his political opponents. The unrest also led Gusmão to utilize his limited but symbolically important presidential powers and declare a state of emergency (although there were questions about the constitutional legality of this move). This further deepened the high-level standoff between the president and prime minister, ultimately forcing Alkatiri to step down in June 2006 – although not

9780230_242920_08_cha07.indd 129

11/11/2010 5:15:47 PM

130

Benjamin Reilly

before an extended period of brinkmanship in which Gusmão threatened to tender his own resignation if Alkitiri did not go first. In his place, Nobel Prize winner and former foreign affairs minister Jose Ramos-Horta (who had quit the cabinet in response to Fretilin’s continued support for Alkitiri) was appointed as the new prime minister, and a new UN mission in East Timor was established in August to help stability and national reconciliation, in preparation for the 2007 presidential and parliamentary elections. These elections were relatively peaceful and resulted in a major political realignment. In an unusual move, the election outcome saw Gusmão and Ramos-Horta essentially swapping jobs: Ramos-Horta was elected president in May while Gusmao stood at the head of the opposition CNRT party ticket for the June parliamentary elections. Fretilin, whose vote share declined by almost half, from 59 percent at the 2001 elections to just 29 percent, was the major loser. Despite this, Fretilin remained the largest party in parliament, a situation that led to another extended period of political uncertainty in the postelection period, in which Gusmão became prime minister at the head of a coalition government. The new government was based around Gusmão’s CNRT party and excluded Fretilin, which became the main opposition. Gusmão’s appointment as prime minister nominally brought the period of severely divided executive government in East Timor to a close, although the partisan identity of the president and the prime minister remains separate. Supporters of semi-presidentialism sometimes claim that it can provide for elements of power-sharing, thus making it a good option for transitional or conflict-prone polities (Sartori 1994; Fish 2001). The East Timorese experience suggests that while power sharing between rival political elites can indeed be facilitated by semi-presidential structures, it does not follow that this will necessarily play a role in reducing conflict. Indeed, the East Timor experience suggests that semi-presidential structures can serve to enhance social divisions in the event of institutional conflict. As Sophia Moestrup has observed, “rather than promoting cooperation in divided polities, semi-presidentialism with its dual executive can serve to institutionalize political divisions, thereby polarizing future political conflict . . . the political struggle between President Gusmão and Prime Minister Alkatiri, driven by ideological differences, made the difficult task of establishing an effective and democratic government for the young state even more daunting” (2007a: 32).

Conclusion The experiences of the three countries discussed in this chapter highlight one of the recurring problems of semi-presidential constitutions in new democracies. In each case, the advent of institutional division coming soon after the introduction of new constitutional arrangements and in the early

9780230_242920_08_cha07.indd 130

11/11/2010 5:15:47 PM

Semi-Presidentialism and Democratic Development

131

years of a transition to democracy deepened political tensions, instability, and confusion, and stymied the institutionalization of the established rules and procedures necessary for democracy to take root and function effectively. In each case, stability began to improve once the period of division ended and the office of the president and prime minister again came from congruent parties. One means of illustrating the problems of political stability in these three cases is the World Bank’s Governance Matters database, which includes an aggregate measure of “political stability” for all states, drawn from a combination of public and private sources. For our purposes, this “political stability” measure combines indices of politically-motivated internal and external violence in a given country with a separate measure of government durability such as the government’s ability to carry out its declared program, and to stay in office. Figure 7.1 compares these indicators for each of our three country cases over the past decade. Figure 7.1 provides a clear view of how periods of institutional division have affected political stability in each of our countries. In East Timor, for example, political stability declined sharply between 2000 and 2002 as the UN administration prepared the country for independence, and then declined further, with the lowest levels reached in 2006 when the standoff between Alkatiri and Gusmão came to a head. In Mongolia, the stability measures were at their lowest during the 1997–2000 period of cohabitation, but rose with the resumption of single party rule. In Taiwan, too, stability

Political stability 100 90 80 70 60 50 40 30 20

Mongolia Timor-Leste Taiwan

10 0 1996

1998

2000

2002

2003

2004

2005

2006

Country’s percentile rank (0–100) Figure 7.1 Political stability in East Timor, Mongolia and Taiwan, 1996–2006 Source: Kaufmann D., A. Kraay, and M. Mastruzzi 2007.

9780230_242920_08_cha07.indd 131

11/11/2010 5:15:47 PM

132

Benjamin Reilly

measures were at their highest in the late 1990s but declined with the election of President Chen Shui-bian in 2000, which ushered in Taiwan’s period of minority government, with the lowest levels reached in 2006. These indicators help illustrate my key point: while semi-presidentialism has its benefits, it places unusual strains on new democracies. In particular, periods of institutional division can put great stress on countries that have not yet developed established practices of political coexistence. In addition, the uncertainties of constitutional law in situations of divided power create their own problems: in East Timor, Mongolia, and Taiwan, disagreements over which particular office had which particular constitutional powers was a recurring source of conflict. No constitution can codify all situations that officeholders are likely to face, meaning that even the most thorough constitutional text will inevitably have some grey areas. This is a particular problem for semi-presidential constitutions, as it is precisely those grey areas of uncertainty that are likely to create conflict. Of course, the evidence from this study only relates to one region – East Asia – and it is possible that our cases are not representative of other regions. A large-n study by Moestrup (2007a), for instance, found that Asian semipresidentialism performed worse than either presidential or parliamentary systems, in contrast to other regions such as Eastern Europe, where its record was superior to alternative constitutional structures. I am not arguing that semi-presidentialism will inevitably be a poor constitutional choice. Semipresidential systems have a number of positive attributes, not least the limits they place on untrammelled executive power and their capacity to sustain a degree of sharing and balancing of responsibilities between rival elites. As such, they are clearly an attractive option for those seeking to construct power-sharing deals in times of political crisis, as the recent internationally backed peace settlement in Kenya instituting a semi-presidential model shows. Some authors go so far as to argue that the political incongruities created by dual executives can be a positive force for developing norms of democracy. For example, Elgie and Moestrup maintain that “In Timor-Leste, President Gusmão has provided an important check on the executive power of the one-party government, and played the role of arbiter on several occasions. President Ochirbat in Mongolia similarly voiced his criticism of the MPRP government while simultaneously facilitating negotiations with the opposition outside the framework of parliament. These were both presidents with quite limited powers who leveraged their positions to provide a constructive counterbalance to the government, thereby smoothing the functioning of the system” (2007a: 244–245). By contrast, I maintain that a more plausible assessment of these cases is that the benefits of semi-presidentialism, while real, were outweighed by its negative consequences. Indeed, the structural features of semipresidentialism amplify some of the most common and pressing problems

9780230_242920_08_cha07.indd 132

11/11/2010 5:15:48 PM

Semi-Presidentialism and Democratic Development

133

faced by all new democracies: the need to build trust between government and opposition members and their supporters, the need to institutionalize accepted political practices and routines of appropriate behavior, and the need to ensure loyalty among the disciplined forces such as the military and the police. Political standoffs, divided loyalties, and constitutional uncertainties can undermine all these objectives. As such, the message of this chapter – in keeping, it should be said, with much of the literature on this topic – is that semi-presidentialism presents a real risk to political stability in developing democracies, particularly during periods of divided government.

Notes My thanks to Ho-pong To for his helpful comments on an earlier version of this chapter; the usual disclaimer applies. 1. According to the standard definition of semi-presidentialism used in this book, East Timor was not strictly an example of cohabitation as the president was nonpartisan.

9780230_242920_08_cha07.indd 133

11/11/2010 5:15:48 PM

8 Semi-Presidentialism in Africa: Patterns and Trends Sophia Moestrup

Of the more than 50 countries in the world that currently have a semipresidential constitution (see Chapter 1 for a complete list), over one-third (23) are found on the African continent. Of the 53 member countries in the African Union, more than half (27) currently are or recently were semipresidential (Table 8.1). Although semi-presidentialism has been particularly prevalent in Francophone and Lusophone Africa, the list also includes Anglophone and Arabic speaking countries. A closer analysis of the performance of semi-presidentialism in Africa has broad relevance not only for Africanists. Of the 29 semi-presidential electoral democracies listed by Freedom House in 2010 (Freedom House 2010), six are in Africa. Nine more countries in Africa are former semi-presidential electoral democracies, including Burundi, Comoros, Congo-Brazzaville, and Niger, which have since become presidential. How semi-presidentialism performs in Africa must necessarily impact our broader assessment of the suitability of this regime type for nascent democracies. Theoretical debates about the impact of different regime types on nascent democracies have become much more nuanced over the past few years, particularly as semi-presidentialism has received greater scholarly attention. The rise of the study of semi-presidentialism has led to a closer analysis of the factors that influence the performance of semi-presidential democracies. Scholars have looked at variables such as authority patterns, partisan strengths in parliament, and combinations of these. Africa is also receiving greater attention as the study of semi-presidentialism has expanded from Europe to cover nascent democracies in the rest of the world.1 There is a lack of scholarly consensus on whether semi-presidentialism is likely to be beneficial for divided societies. Lijphart is not in favor of semi-presidentialism as a regime choice for societies with deep ethnic and other divisions. He considers semi-presidentialism “only a slight improvement over pure presidentialism” as “the zero-sum nature of presidential elections remains” (2004: 102). He states that parliamentary systems “offer the optimal setting for forming a broad power-sharing executive,” thereby 134

9780230_242920_09_cha08.indd 134

11/11/2010 5:16:29 PM

Semi-Presidentialism in Africa

135

Table 8.1 Semi-presidential countries in Africa, January 2010 Electoral democracies Burundi* Cape Verde Comoros* Guinea-Bissau Mali Namibia Sao Tome e Principe Senegal

Former electoral democracies

No transition to electoral democracy

Burkina Faso Central African Rep. Congo-Brazzaville* Madagascar* Mauritania Mozambique Niger*

Algeria Angola Cameroon Chad Dem. Rep. of Congo Egypt Gabon Kenya Rwanda Tanzania Togo Tunisia

Note: * No longer semi-presidential. Source: Freedom House (2010).

facilitating the establishment of democratic government in deeply divided societies (ibid.: 101). Contrary to this view, Blondel argues that by allowing for power sharing between two major societal groups, dual leadership systems can contribute “to ensuring the loyalty of politically relevant groups” (Blondel 1992: 167). Roeder also supports the view that semi-presidential regimes are well suited for societies that have suffered under ethnopolitical crises. Interestingly, he argues that it is because semi-presidentialism (like presidentialism) provides for a separation of powers, thereby creating “opportunities for alternative majorities to emerge in the legislative and executive branches” (Roeder 2005: 71). The argument is that this helps create a basis for the emergence of cross- cutting societal cleavages. This chapter contributes to this scholarly debate by unpacking the factors that have influenced democratic survival of nascent democratic semipresidential regimes in Africa. Following an overview of the spread of democratic semi-presidential regimes across Africa, the chapter lays out the framework for the subsequent analysis of patterns and trends in the performance of these nascent democracies. The chapter proceeds to a discussion of the paths these countries have followed since the first transition to democracy, emphasizing the influence of differences in presidential powers and in partisan strength in parliament. The analysis privileges the identification of trends over the details of individual country cases.

Semi-presidentialism in Africa Africa is notorious for the rule of “big men.” By some accounts, the formal constitutional framework matters little for the performance of African

9780230_242920_09_cha08.indd 135

11/11/2010 5:16:29 PM

136

Sophia Moestrup

political systems, where leaders rely on personal patronage and clientelistic networks to maintain themselves in power (Hyden 2006). However, as Posner and Young (2007) point out, this view overlooks the significant degree of institutionalization of political power that has occurred in Africa since the early 1990s. The contrast is striking between the percentage of African leaders that left power through institutionalized means or natural death between 2000 and 2005 (about 80 percent), compared to the 1960s (approximately 35 percent) (Posner and Young 2007: 128). Similarly, in his study of elections in Africa between 1990 and 2003, Lindberg finds that “[m]ore than half of all elections in Africa during the study period were free and fair” (Lindberg 2006: 69). This chapter focuses on those countries that at one point or another made the transition to become an electoral democracy.2 Where no such full transition to democracy ever took place, it is difficult to make an argument for the impact of semi-presidentialism, as the institutional rules have not been allowed to play themselves out fully. The following discussion is thus centered on the impact of semi-presidentialism on democratic survival, not on the effects of this regime type on the eventual transition to democracy of semi-authoritarian countries. Interestingly, as shown in Table 8.2, all 15 countries to be studied further are located in sub-Saharan Africa. To date, none of the North African semipresidential countries have completed a democratic transition. The countries listed below mostly crossed the threshold to become electoral democracies in the early 1990s, though there are a few outliers, notably Burkina Faso, Mauritania, and Senegal. Burkina experienced a short, aborted transition to democracy at the end of the 1970s, before the “third wave” (Huntington 1991) truly hit Africa following the fall of the Berlin Wall. Senegal and Mauritania, on the other hand, had a late transition to democracy, compared to the rest of the region. Data for GDP per capita show that democracy has so far survived among the richest (Cape Verde and Namibia) as well as the poorest (Mali) countries in the sample. The poorest country (Burundi) has reverted to democracy, contrary to Congo-Brazzaville with a GDP per capita almost ten times as high (though both countries are now presidential). Similarly, literacy rates appear to be a poor indicator of the chances of democratic survival: compare a 23 percent literacy rate in Mali with an 86 percent literacy rate in Congo-Brazzaville. Ten of the countries in Table 8.2 have experienced an authoritarian reversal, while the residual five have remained electoral democracies since their first transition. Of the countries where democracy broke down, three have since reverted to the status of electoral democracies (Burundi, Comoros, and Guinea-Bissau); two (Burundi and Comoros) have interestingly done so with the adoption of a new presidential constitution, thus leaving the experience with semi-presidentialism behind. Congo-Brazzaville and Niger also now have presidential constitutions.

9780230_242920_09_cha08.indd 136

11/11/2010 5:16:30 PM

Semi-Presidentialism in Africa

137

Table 8.2 Current and former semi-presidential electoral democracies in Africa, January 2010

Country

GDP per Adult Semicapita literacy presidential (PPP $) 2006 regime, 2006 % period

Burkina Faso 1,084

26

Burundi

333

59

Cape Verde CAR

2,833 679

83 49

Comoros

1,152

74

CongoBrazza. GuineaBissau

3,550

86

1990– 1981, 1991– 2003, 2005– 1979–1985, 1992–2000 1992–1997

467

63

1994–

Madagascar

1970–1974, 1978–1983, 1991– 1992–1995

878

71

1992–2009

Mali

1,058

23

1992–

Mauritania

1,890

55

739

44

1991–2005, 2006–2008 2009– 1990–

4,819

88

1990–

612

30

Sao Tome

1,534

88

1991, 1993–1996, 1999–2009 1990–

Senegal

1,592

42

Mozambique Namibia Niger

1970–1983, 1991–

Freedom Electoral House democracy, score period 20103

Current regime (P=pres., SP=semipres.)

1978–1980

SP

4 PF

1993, 2005– 4.5 PF*

P

1991– 1 F* 1993–2003 5 PF

SP SP

1996–1999, 3.5 PF* 2002– 1992–1997 5.5 NF

P

1994–1999, 4 PF* 2000–2003, 2005– 4 1992–2009 5 PF 1992–

2.5 F*

P SP

SP SP

2006–2008 5.5 NF

SP

1994–2009 3.5 PF

SP

1990–

2 F*

1993–1996, 4.5 PF 1999–2009

SP P

1991– 5

2 F*

SP

2000–

3 PF*

SP

Note: * Indicates current status as an electoral democracy. Source: Freedom House (2010), UNDP (2008), Elgie Chapter 1.

Analytical framework Studies of the performance of semi-presidential democracies in terms of democratic survival have generally focused on the impact of different distributions of presidential powers, on the one hand, and on the importance of different electoral outcomes, on the other. The first approach is characterized by an analysis and comparison of constitutionally mandated presidential powers, while the latter focuses on the presence or absence of a

9780230_242920_09_cha08.indd 137

11/11/2010 5:16:30 PM

138

Sophia Moestrup

parliamentary majority supporting the president. Some scholars have also looked at the combined effect of different authority patterns and parliamentary majorities, which is the approach followed in the present chapter. Presidential powers and authority patterns A focus on constitutionally mandated powers allows us to assess the impact of institutional design on behavioral outcomes, keeping in mind that such impact will vary depending on extra- constitutional factors such as partisan alignment or elite interaction (Shugart 2005: 327). Shugart and Carey (1992) have devised a scoring system to measure presidential power in terms of legislative and nonlegislative powers. They have argued for the need to subdivide semi-presidential regimes into two categories, premier-presidential and president-parliamentary regimes, given the distinct differences in authority patterns between the two. Shugart (2005) has further elaborated on this distinction, classifying a number of semipresidential countries as premier-presidential or president-parliamentary, depending on whether the prime minister and cabinet are accountable only to parliament or accountable to both parliament and president (boiling down to whether the president in a semi-presidential regime can unilaterally decide to dismiss the prime minister). Semi-presidential regimes with strong presidential powers are posited to suffer from some of the same weaknesses of presidential systems, notably winner-take-all elections and the risk of personalization of presidential power (Lijphart 2004). President-parliamentary systems in particular are considered more vulnerable to authoritarian reversal than premier-presidential systems, because of the opportunities for deadlock as the president seeks to impose his or her choice of prime minister (Shugart and Carey 1992: 120–121). The results of the multivariate analysis presented in Chapter 4 confirm the dangers of authority patterns that favor the president, indicating that president-parliamentary regimes are much worse for democracy than premier-presidential ones. Where powers are more evenly balanced between president and prime minister/parliament, the concern is that disagreements between the president and the prime minister and parliament can also easily degenerate into political gridlock, with protracted power struggles between the two executives, irrespective of whether they belong to the same party or coalition, or whether they find themselves in a situation of cohabitation (Elgie 2005: 107–108). According to Tsai (2008), this is particularly likely to happen where a strong president faces a strong parliament, which tends to breed intransigence on the part of the two institutions with separate electorally mandated legitimacy.6 In contrast, the expectation is that in countries where power is concentrated with the prime minister and where the president serves more of a representative role, there is a greater chance of democratic survival. This is

9780230_242920_09_cha08.indd 138

11/11/2010 5:16:31 PM

Semi-Presidentialism in Africa

139

most likely to happen in premier-presidential systems (Elgie 2005: 105–107; Tsai 2008). Authority patterns are here believed to be clearer, with little incentive for the personalization of presidential power. The president is expected to play more of a regulatory role. Below is an overview of presidential legislative and nonlegislative powers in African semi-presidential countries, based on the scoring system developed by Shugart and Carey (1992). The countries are similarly classified into premier-presidential and president-parliamentary regimes, following Shugart (2005). Partisan strength in parliament As Skach (2005: 3–4) rightly points out, a classification of semi-presidential regimes strictly based on presidential powers as provided for by the constitution overlooks the fact that presidential power is, in fact, dynamic and dependent on the president’s partisan strength in parliament. Whether the president is backed by a parliamentary majority will greatly influence his or her actual powers. In situations of cohabitation between a president and a prime minister from opposite sides of the political spectrum and where the president’s party is not represented in the government, for example, the French president cedes centre stage to the prime minister. Conversely, when the French president is backed by a parliamentary majority, he is able to exercise more power than a catalogue of constitutional powers would suggest possible (ibid.: 99). As outlined by Skach, the following three types of governments are possible: consolidated majority (where the president is backed by a majority in parliament); divided majority, or cohabitation (where the prime minister is backed by a majority in parliament in opposition to the president and where the president’s party is not represented in the government); and divided minority (where neither the president nor the prime minister controls a legislative majority and where they may be from different parties). The danger of cohabitation has been the classical argument against semipresidentialism. Linz (1994, 1997) argues that cohabitation is likely to be even more pernicious in a semi-presidential regime than divided government in a presidential one, leading to a competitive diarchy. Similarly, Stepan and Suleiman (1995: 399) see “the possibility of constitutional conflict between two electorally legitimated executive [as] the central problem” of semi-presidentialism. Despite the prevailing consensus as to the dangers of cohabitation, Elgie and McMenamin (2008) do not find evidence to support this conventional wisdom in their study of the impact of divided executives on democratization in mixed (semi-presidential) systems. Looking at data from 47 countries over a 30-year period, they found no statistically relevant impact of cohabitation. More recently, Skach (2005) and Elgie (2008) have focused instead on the perils of divided minority government. In this situation, without a stable

9780230_242920_09_cha08.indd 139

11/11/2010 5:16:31 PM

140

Sophia Moestrup

governing majority, legislative immobilism can easily result in presidential intervention beyond the limits prescribed by the constitution (Skach 2005: 18). This situation is particularly likely to occur in countries with noninstitutionalized party systems. Elgie (2008) notes that divided minority governments have empirically been much more frequent than situations of cohabitation in nascent semi-presidential democracies. In this view, majority government is likely to be the most propitious for democratic survival. Coherent majorities allow the executive to act with purpose in the pursuit of political and economic reform, which is particularly important for democratizing countries (Skach 2005: 116). In contrast to this view, Lijphart posits that consolidated majorities are likely to hinder the establishment of democratic government in divided societies. He argues that the dangers of semi-presidentialism are enhanced by the fact that it is “possible for the president to be even more powerful than in most pure presidential systems” when the president is backed by a parliamentary majority (Lijphart 2004: 102). In Lijphart’s view, this would serve to undermine rather than consolidate democracy in countries with deep societal divisions, where he argues some form of power-sharing is essential. Combining the analysis of institutional design and electoral outcomes It is important to separate institutional and extra- constitutional variables, when developing subtypes of semi-presidential regimes. Thus, contrary to Tsai (2008), I would argue that cohabitation should not be considered a subtype of semi-presidentialism, but a situation in which all semi-presidential systems can theoretically find themselves, depending on shifts in partisan alignment.7 The actual performance of a semi-presidential regime in a situation of cohabitation is likely to depend in part on the distribution of constitutional powers, and in part on patterns of elite interaction that have developed over time. In their thorough analysis of the democratic performance of different subtypes of semi-presidential systems, Elgie and McMenamin (2008) found all subtypes to be equally impervious to the negative effects of cohabitation. They distinguish between presidential-like, parliamentary-like and dual executives, depending on whether the president or the prime minister dominates the executive or whether both have significant independent powers. However, in their classification of these different subtypes, actual power relations prime over constitutional provisions. On the basis of this classification it is difficult to make accurate inferences as to the interaction effect of institutional design and divided executives. Combining two different patterns of presidential authority (premierpresidential versus president-parliamentary systems) and three electoral outcomes (majoritarian government, cohabitation and divided minority government, following Skach, 2005), results in six logically possible categories.

9780230_242920_09_cha08.indd 140

11/11/2010 5:16:31 PM

Semi-Presidentialism in Africa

141

By cross tabulating the effect of different combinations of authority patterns and electoral outcomes on the democratic survival of nascent African semi-presidential democracies, we can analyze the existence of emerging patterns and trends.

Patterns and trends in the performance of semi-presidentialism in Africa In this section, a presentation of differences in presidential powers and authority patterns in semi-presidential democratic constitutions across Africa is followed by an analysis of differences in electorally generated types of government. There is also a discussion of the correlation between authority patters on the one hand and partisan strength in parliament on the other; and democratic breakdown. The section focuses in particular on the possible interactive effect of different authority patterns and partisan alignments and the presence of general patterns and trends. Presidential power and authority patterns Appendix 8.1 shows the distribution of presidential powers across our 15 countries using Shugart and Carey’s (1992) scoring system to facilitate crossregional comparison. It is difficult to avoid some element of interpretation, as constitutional texts vary in their clarity. Measuring presidential powers is far from an exact science, as Shugart (2005) admits. To check the validity of the coding, the scores are compared to those of Shugart (2005) and van Cranenburgh (2007). The latter has developed her own scoring system to conduct an analysis of the institutional basis of presidential power in 29 Anglophone and Francophone presidential and semi-presidential countries in Africa. In general, there are only minor deviations, one being the classification of Senegal as a premier-presidential system by Shugart (2005), and another van Cranenburgh’s scoring of presidential powers to dismiss the cabinet in Mali and Niger. In the first case I agree with van Cranenburgh that in Senegal the president has full authority to dismiss the cabinet (making the regime president-parliamentary), while in the latter I agree with Shugart that Mali and Niger (prior to the 2009 constitutional change) should be classified as premier-presidential systems, with limited to no authority for the president to dismiss the prime minister on his or her own initiative.8 It is interesting to note that the president does not have reserved domains of legislation in any of the countries listed in Appendix 8.1. Also, decree powers are very limited across our country cases, when defining decree power as “the ability of the president to make law” (as opposed to regulatory decrees, Shugart and Carey 1992: 151). There are generally strong limits even on emergency powers, as the assembly in most cases must remain in session and prolong a state of emergency beyond a short time period (generally two weeks).

9780230_242920_09_cha08.indd 141

11/11/2010 5:16:31 PM

142

Sophia Moestrup

Overall, nonlegislative powers are stronger than legislative powers across our country cases. In fact, compared to semi-presidential systems in Central and Eastern Europe, nonlegislative powers of presidents in African semipresidential democracies are impressively high. For example, while Belarus and Russia – the highest scoring in the region and both classified as “not free” by Freedom House – score a 9 and an 8 respectively (Elgie and Moestrup 2008), countries like Sao Tome and Namibia (both classified as “free”) score 9 and 10, respectively. Particularly striking are the more limited powers of the president to appoint the cabinet and to dissolve parliament in Central and Eastern Europe, compared to Africa.9 We also note, however, that nonlegislative powers in premier-presidential systems, as should be expected, are generally lower than in president-parliamentary ones. This observed pattern of comparatively high presidential powers in African semi-presidential democracies overall, but with significant variations across countries, is confirmed by recent research by Kim and Bahry (2008). In a study modeling variables that impact the “interruption” of presidencies in third wave democracies (including both presidential and semi-presidential regimes), Kim and Bahry found that “the level of presidential constraint is lower on average in Africa, and more variable across countries, than in Latin America.” They go on to state that this “difference may help to explain why African presidents are more likely to be removed by the military than by pressure from other branches of government” (ibid.: 19). In her study of variations in presidential power across Anglophone and Francophone African countries, van Cranenburgh (2007: 11–12) interestingly found that there is no significant difference in average levels of presidential powers between presidential and semi-presidential regimes. Returning to our country cases, Table 8.3 shows that president-parliamentary systems have been more frequent than premier-presidential ones among African semi-presidential democracies (10 out of the 15 countries in our sample were or are today president-parliamentary). President-parliamentary regimes have also been more prone to democratic breakdown than premierpresidential countries. Out of the 10 semi-presidential democracies that experienced an authoritarian reversal, eight were president-parliamentary systems (Burkina Faso, Burundi, the Central African Republic, Comoros, Guinea-Bissau, Mauritania, Madagascar, and Mozambique) and only two were premier-presidential (Congo-Brazzaville and Niger). Of the five semipresidential countries that have thus far remained democratic, since their first transition to democracy, two have a president-parliamentary system (Namibia and Senegal) and three have a premier-presidential system (Cape Verde, Mali, and Sao Tome). The breakdown rate has been 80 percent for presidentparliamentary systems and 40 percent for premier-presidential ones. The countries where democracy broke down have had higher presidential powers (an average of about 15.2) compared to those where democracy has survived (an average score of 12). Given that the measurement of presidential

9780230_242920_09_cha08.indd 142

11/11/2010 5:16:32 PM

Semi-Presidentialism in Africa

143

Table 8.3 Democratic breakdown and survival of president-parliamentary and premier-presidential systems in Africa, January 2010 President-parliamentary system

Premier-presidential system

Breakdown of democracy

Burkina Faso (1980), Burundi (1993)*, CAR (2003), Comoros* (1999), Guinea-Bissau (1999, 2003), Madagascar (2009),10 Mauritania (2008), Mozambique (2009), → 8 countries, 9 cases of democratic breakdown

Congo Brazzaville* (1997), Niger (1996, 2009)* → 2 countries, 3 cases of democratic breakdown

Survival of democracy

Namibia, Senegal → 2 countries

Cape Verde, Mali, Sao Tome and Principe (2003 const. amendment) → 3 countries

Note: * No longer semi-presidential.

powers is far from an exact science and that the numbers of countries we are looking at is relatively small, these numbers are just an indication of general trends. When comparing the scores of presidential powers in president-parliamentary versus premier-presidential regimes we see that, as expected, the scores of the former tend to be higher than those of the latter (approximately 16 versus 9). There is little difference in the average score of president-parliamentary regimes that broke down compared to those that did not (16.4 versus 16). On the other hand, the average score of premierpresidential regimes that did break down was higher (12) than the average across premier-residential systems that have remained democratic (9.3). The higher score of presidential powers found in countries in our sample where democracy broke down is thus linked to two factors: 1) the higher proportion of president-parliamentary regimes among the countries that suffered an authoritarian reversal, and 2) the relatively higher scores of presidential powers in the premier-presidential regimes that broke down. These findings confirm the general expectations from the literature: that a power distribution skewed in favor of the president is likely to endanger a nascent semi-presidential democracy; and that premier-presidential regimes appear to stand a higher chance of consolidating their democratic gains than do president-parliamentary ones. From Appendix 8.1 and Table 8.3 we see four cases that appear particularly promising for further investigation, as they go counter to these general trends. These include two countries (Namibia and Senegal)11 that though president-parliamentary, have remained democratic during the period of study; one country (CongoBrazzaville) that, while having a premier-presidential regime with fairly limited presidential powers, did suffer an authoritarian reversal; and one

9780230_242920_09_cha08.indd 143

11/11/2010 5:16:32 PM

144

Sophia Moestrup

premier-presidential regime (Niger), where democracy broke down for a second time in 2009, despite amendments to the constitution to reduce presidential powers following the previous coup in 1996. Electorally generated types of semi-presidential governments in Africa Is there a correlation between type of government and democratic breakdown or survival? Our expectation from the discussion above was that cohabitation and minority governments are unlikely to favor a consolidation of democratic gains, but the jury is out regarding the impact of majority governments. As we can see from Table 8.4, there are 12 cases of breakdown, as democracy broke down twice in Niger and Guinea-Bissau. Similarly, there are seven cases of democratic survival, given that Mali and Sao Tome have gone through long periods of two different government types. Table 8.4 appears to validate the poor reputation of divided minority governments, with three out of four instances of such governments breaking down. Majority governments seem somewhat prone to democratic breakdown, with eight out of 13 cases of majority government breaking down. In contrast, there is no pattern of greater democratic breakdown or survival for cohabitation governments. Appendix 8.2 confirms that by far the most frequently found type of government in our country cases is majoritarian, whether in the form of one-party or coalition governments. Though there have been only about half as many instances of one-party majoritarian as coalition government, one-party majoritarian governments have outlasted coalition governments in by about 20 years (approximately 70 years versus approximately 50 years when we add up the duration of all majoritarian governments). Of the eight

Table 8.4 Democratic breakdown and survival of different types of electorally generated semi-presidential governments in Africa, January 2010 Majority government

Cohabitation

Divided minority government

Breakdown of democracy

Burundi (1993), Comoros Niger (1996) (1999), Congo Brazzaville → 1 case (1997), Guinea-Bissau (1998), Madagascar (2009), Mauritania (2008), Mozambique (2009), Niger (2009) → 8 cases

Burkina Faso (1980), Central African Rep. (2003), Guinea-Bissau (2003) → 3 cases

Survival of democracy

Cape Verde, Mali, Namibia, Sao Tome12 Sao Tome, Senegal → 1 case → 5 cases

Mali13 → 1 case

9780230_242920_09_cha08.indd 144

11/11/2010 5:16:32 PM

Semi-Presidentialism in Africa

145

cases of democratic breakdown of majoritarian governments, five happened during periods of one-party government (Burundi, Comoros, Guinea-Bissau 1998, Madagascar, and Mozambique) and three during periods of coalition governments (Congo-Brazzaville, Mauritania, and Niger 2009). The vulnerability to democratic breakdown of one-party majoritarian as compared to majority coalition governments, is in line with Lijphart’s concern about the need for some element of power-sharing in divided societies. Most minority governments happened during interim periods between presidential and legislative elections, where a president may chose to appoint a minority government until fresh legislative elections determine the new balance of power in parliament (e.g., Sao Tome and Principe, Senegal, Madagascar, and Guinea-Bissau). In Guinea-Bissau, this proved to be very problematic and contributed to the coup against Kumba Yala in 2003, as the president kept postponing parliamentary elections and government action ground to a halt (Azevedo and Nijzink 2007: 149–151). Other minority governments are more similar to the situation of the Weimar Republic described by Skach (2005), where, in the absence of a cohesive majority in parliament, the president opts for the appointment of a cabinet led by technocrats. Such governments have been rare, among our cases. The only two cases lasting more than a few months include the Central African Republic (1999–2003) and Mali (2002–2007).14 Of the two, the Central African Republic did experience an authoritarian reversal during President Patassé’s second term in office. Central African President Patassé chose to govern through a minority government15 rather than to share power through a coalition government. This approach definitively contributed to the zero-sum nature of Central African politics. However, it is difficult to ascribe the breakdown of democracy to the disfunctions of minority governments in the country. Patassé managed to alienate powerful losers who retained significant military power and were not afraid to use it. This process began during his first term in office, where his majority coalition government was rescued several times by external military intervention (Mehler 2005). In Mali, on the other hand, President Touré, who did not have any formal party affiliation, was able to govern through a technocratic cabinet during his first term in office, thanks to a prevailing national consensus backing him as the “hero” of the democratic transition in 1991 (Villalón and Idrissa 2005). His first term in office interestingly served to boost the consolidation of the Malian party system, with the rise of a formal opposition and clearer policy delineation among political parties since the beginning of his second term in office, in 2007. The least frequent type of government is cohabitation. Only Niger and Sao Tome and Principe have experienced this type of government for more than one month. Cape Verde had one month of overlap between newly elected President Pires of the PAICV and an MpD prime minister issued from a parliamentary majority from the 1995 legislative elections. Subsequent

9780230_242920_09_cha08.indd 145

11/11/2010 5:16:33 PM

146

Sophia Moestrup

legislative elections provided Pires with a majority in parliament, thereby ending the interim period of cohabitation. Niger and Sao Tome and Principe provide for an interesting contrast. Government deadlock during the one-year cohabitation in Niger was a significant contributing factor to the breakdown of democracy in early 1996 in that country; continued squabbling between president and prime minister provided a disloyal military a golden opportunity to return to power (Moestrup 2007b). In contrast, Sao Tome and Principe weathered seven years of cohabitation from 1994 to 2001 and still managed to maintain a Freedom House status of “free.” This is not to say that government has always run smoothly in Sao Tome and Principe, witness frequent tensions within the political class and two attempted coups or armed insurgencies, in 1995 and 2003 respectively. Government instability in Sao Tome and Principe has been much greater since de Menezes came to power in 2001 and ended the long period of cohabitation by the initiation of a string of coalition governments. Thus, during the seven years of cohabitation, Sao Tome and Principe had only four prime ministers, while the country has had eight prime ministers since 2001 – twice as many in roughly the same number of years.16 Repeated conflict between the president and the government and shifting party alliances have caused political instability (Seibert 2004: 16), without, however, leading to the breakdown of democracy – yet. As noted by the World Bank (2005: iii) “political and civil institutions remain fragile” in Sao Tome and Principe. Unfortunately, government instability has not abated with the passage of a constitutional amendment that effectively moved Sao Tome and Principe from the category of presidentparliamentary to that of premier-presidential countries in January 2003. When discussing cohabitation, one should also consider cases that did not happen, meaning situations where the president preempted cohabitation by dissolving parliament, as happened in Congo-Brazzaville in late 1992. Tensions over whether President Lissouba had the constitutional right to dissolve parliament were resolved by forming a government of national unity until parliamentary elections could take place in May 1993. Following the elections, the president again formed a coalition government to reduce tensions and end armed clashes between militia groups supporting the three main political leaders. By that time it was too late, however. The logic of confrontation and winner-take-all politics spun out of control as presidential elections in 1997 approached, ending with a coup d’état against Lissouba (Clark 2005). Discussion of patterns and trends Now we look at the interaction between authority patterns and different government types to see if any patterns emerge. Table 8.5 illustrates that the breakdown of democracy in a number of our country cases was in a sense “over- determined.” In Niger in 1996, though a premier-presidential system,

9780230_242920_09_cha08.indd 146

11/11/2010 5:16:33 PM

Semi-Presidentialism in Africa

147

Table 8.5 Cross-tabulating authority patterns and government types Majoritarian Presidentbroke down parliamentary

survived Premierpresidential

broke down

survived

Burundi, Comoros, GuineaBissau (1998), Madagascar, Mauritania, Mozambique Namibia, Senegal

Cohabitation

Divided min. Burkina, CAR, Guinea-Bissau (2003)

Sao Tome

Congo Niger (1996) Brazzaville, Niger (2009) Cape Verde, Mali, Sao Tome

Mali

cohabitation and a president with strong powers led to such severe gridlock as to delegitimize the fledgling democracy and provide an opportunity for a military comeback. Similarly, in Burkina Faso, the Central African Republic and Guinea-Bissau 2003, presidents with significant powers over cabinet formation and dismissal opted for governing through minority governments, thereby contributing to polarizing politics and setting the stage for an authoritarian reversal. On the other hand, majoritarian premier-presidential systems provided the opportunity for the nascent democratic institutions in Cape Verde and Mali to take root, while avoiding overly personalizing politics around the president. Sao Tome and Principe remains a puzzling case that merits further investigation. It is possibly a case where, given the small population of the country (approximately 150,000) and the strategic importance of its oil reserves, political developments in the country evolve within the constraints of geopolitical interests in the region, with Nigeria playing an important role in defusing the 2003 coup (Seibert 2004). Previously, in discussing the impact of different authority patterns and variances in presidential powers, we found four cases that merited further attention, as their democratic performance went counter to what would have been predicted, based on their authority structures: Congo-Brazzaville, Niger 2009, Namibia, and Senegal. Does a closer look at the type of government these regimes evolved under help explain their democratic outcomes? As indicated above, the threat of cohabitation and the attempts by Lissouba to avoid that situation help explain the fact that CongoBrazzaville, though a premier presidential system with limited presidential powers, still broke down. The case of Niger’s authoritarian reversal in 2009 is more complex. Following the restoration of democracy in 1999,

9780230_242920_09_cha08.indd 147

11/11/2010 5:16:33 PM

148

Sophia Moestrup

the president’s veto and nonlegislative powers were reduced, theoretically setting the stage for a clearer distribution of powers between president and prime minister. Moreover, the country avoided cohabitation and minority governments, with majority coalition governments having a clear majority partner, the Mouvement National de la Societé de Développement (MNSD), backing the president and prime minister, both from the MNSD. However, power struggles emerged within the party between incumbent President Mamadou Tandja and Prime Minister Hama Amadou. Tandja’s determination to cling to power beyond the end of his second term in December 2009, led to the overturn of democracy through an unconstitutional referendum. The August 4, 2009 referendum passed a new presidential constitution giving Tandja three more years in office during a “transition period” and the right to unlimited reelections. Niger is an example of the personalization of power negatively affecting not only president-parliamentary but also premier-presidential systems. What about the two cases of president-parliamentary systems that survived? Interestingly, they both fall into the category of majoritarian government. What, if anything, distinguishes the majoritarian president-parliamentary systems that broke down from those that did not? When taking a closer look at Senegal we see, unfortunately, that the country is moving away from the consolidation of democratic gains, as evidenced by its downgrading by Freedom House from “free” to “partly free” in 2009. Rivalry between the president and prime minister, though from the same party, has contributed to government instability and has resulted in the jailing of a former prime minister under allegations of corruption (IRIN 2006), like in Niger. President Wade has been highly intolerant of the rise of potential challengers, as he prepares the ground for his son, Karim Wade, to become his successor. What about Namibia? Not only have the governments in Namibia consistently been majoritarian since the transition to democracy, they have been single-party majoritarian governments with the same party remaining in power throughout the whole period. The country is characterized by the presence of a former liberation movement turned political party (SWAPO). This party, like the PAICV in Cape Verde (a premier-presidential regime), has by the nature of its origins greater internal cohesiveness than most political parties in our other country cases. This could render the country less vulnerable to the personalization of power stemming from the presence of a powerful president. However, the PAIGC in GuineaBissau, also a former liberation movement, has fractured internally, in part due to the divisive role that President Vieira played (Azevedo and Nijzink 2007). Mozambique, where FRELIMO has dominated the political scene since the transition to democracy in 1994, lost its status as electoral democracy in 2009 (Freedom House 2010). Similarly, the continued dominance of SWAPO in Namibia does raise concerns about “increasing power concentration and a concomitant lack of toleration, bargaining, and compromise” (van Cranenburgh 2006: 602). Overall, it would indeed seem

9780230_242920_09_cha08.indd 148

11/11/2010 5:16:34 PM

Semi-Presidentialism in Africa

149

that president-parliamentary systems are unlikely to facilitate a consolidation of democracy in Africa.

Conclusion Semi-presidentialism in Africa has generally been characterized by neither power-sharing nor separation of powers. The countries where the prospects for consolidating democratic gains look the most promising are the minority few where the president’s ability to dismiss the prime minister singlehandedly is limited. It is evident from the analysis of the countries in our dataset, that semi-presidentialism has not been a panacea for healing deep societal rifts in Africa. African semi-presidential regimes are generally characterized by considerable nonlegislative powers, while presidents’ decree powers are minimal overall. The temptation for presidents is thus not to rule by decree in a situation of divided government, but to avoid such situations by dismissing the government and/ or dissolving parliament (e.g., Niger, Guinea-Bissau, and Congo-Brazzaville), by coopting “independents” and individual opposition members (as in the Central African Republic), or by increasing the president’s powers over cabinet formation (Madagascar). Either course of action will serve to further polarize the political system, providing an opportunity for authoritarian reversal. To understand the performance of African semi-presidential democracies, it is helpful to consider the interactive effect of authority patterns (whether the regime is to be considered president-parliamentary or premier-presidential) and electorally mandated government outcomes (whether one finds majoritarian, cohabitation, or divided minority governments). Logically, minority governments are most prone to happen and to be detrimental to democratic survival in president-parliamentary systems where the president has considerable nonlegislative powers, namely powers over cabinet formation and the ability to dissolve parliament, while cohabitation governments in such systems are unlikely to happen at all. On the other hand, majoritarian presidentparliamentary governments, particularly when single-party ones, can be too “effective” for their own good, in a society driven by deep cleavages. Thus, for example, in Burundi in 1993, “the rapidity with which [policies] were carried out made the government appear dangerously unpredictable in the eyes of many Tutsi, particularly those in the military” (Sullivan 2005: 90). In premier-presidential regimes, minority governments, like in Mali, are likely to be indicative of the prevalence of a certain amount of consensus in support of the president, as the president would otherwise have a limited ability to impose his choice of prime minister. Cohabitation has been rare, but its occurrence or threat thereof did contribute to the breakdown of democracy in Niger and Congo-Brazzaville. Coalition governments, while majoritarian, have often proven to be fickle (as in Niger and Sao Tome), contributing to government infighting and instability. The most promising combination appears to be one of premier-presidentialism with one-party

9780230_242920_09_cha08.indd 149

11/11/2010 5:16:34 PM

150 Sophia Moestrup

majoritarian government at least in the initial few years following the transition to democracy. This provides a certain element of stability, while minimizing the dangers of a powerful president gradually overextending his powers. Unfortunately, the breakdown of democracy in Niger in 2009 illustrates that premier-presidential systems with majoritarian governments are not immune to the dangers of the personalization of presidential power. Based on the analysis presented here, there are a few country cases that merit further observation and study of the political dynamics at work. If developments in Senegal and Namibia confirm the negative effects of a combination of majoritarian governments and president-parliamentary authority patterns, they will serve to further validate Lijphart’s concern about the dangers of excessive power concentration in young semipresidential democracies. On the other hand, Sao Tome and Principe merits further study, as the country has moved from a president-parliamentary to a premier-presidential system, after weathering seven years of cohabitation. The country would seem to confirm the positive impact of power- dividing on the consolidation of young democracies. Finally, democratic developments in Comoros and particularly Burundi under their new presidential regimes will be interesting to compare and contrast to their failed democratic experiments under semi-presidentialism.

Constitutions Burkina Faso 1977, copy kindly provided by Robert Elgie Burundi 1992, copy kindly provided by Robert Elgie Cape Verde 1992, http://www.oceanalaw.com Cape Verde 1999, http://www.parlamento.cv/constituicao/const00.htm Central African Republic 1994, 2004, http://www.oceanalaw.com Comoros 1996, http://www.oceanalaw.com Congo Brazzaville 1992, http://www.oefre.unibe.ch/law/icl/home.html Guinea-Bissau, copy kindly provided by Robert Elgie Madagascar 1992, http://www.oefre.unibe.ch/law/icl/home.html Madagascar 1998 amendment, http://www.oceanalaw.com Mali 1992, du Bois de Gaudusson et al. (1998), Vol. 2 Mauritania 2006, http://www.oceanalaw.com Mozambique 1990, http://www.oceanalaw.com Mozambique 2004 amendment, http://confinder.richmond.edu/country.php Namibia 1990, http://www.oceanalaw.com Namibia 1998 amendment, http://www.oceanalaw.com Niger 1996, Raynal (1993) Niger 1999, http://www.assemblee.ne/texteslegaux/cons99.htm Sao Tome and Principe 1990 amendments, http://www.oceanalaw.com Sao Tome and Principe 2003 amendments, http://www.parlamento.st/ Senegal 1998 amendments, du Bois de Gaudusson et al. (1998), Vol. 2 Senegal 2001, http://www.oceanalaw.com

9780230_242920_09_cha08.indd 150

11/11/2010 5:16:34 PM

9780230_242920_09_cha08.indd 151

11/11/2010 5:16:34 PM

Pattern of authority

4 0 3

12

11

23

4

0 1

2

6

8

pres-parl

4

1 4

1 0

1

2 3 2 0

Burundi 1992

0 1 0 0

pres-parl

Nonlegislative powers Cabinet formation Cabinet dismissal Censure Dissolution of parliament Total legislative powers Total nonlegislative powers Total

Legislative powers Package veto Partial veto Decree Exclusive introduction of legislation Budgetary policy Referenda

Burkina Faso 1977

Appendix 8.1 Presidential powers

21

10

11

0 3

4

3

1 4

3 3 0 0

premier-pres pres-parl

12

6

6

0 3

0

3

1 2

1 2 0 0

6

5

1

0 4

0

1

1 0

0 0 0 0

pres-parl premierpres

14

11

3

0 3

4

4

1 2

0 0 0 0

presparl

13

11

2

0 3

4

4

0 0

2 0 0 0

premier-pres

2

1

1

0 1

0

0

1 0

0 0 0 0

pres-parl

17

12

5

0 4

4

4

1 4

0 0 0 0

premierpres

6

6

0

0 3

0

3

0 0

0 0 0 0

Mali 1992

Continued

Central GuineaAfr. Rep. Congo Bissau Madagascar Cape Verde 1994, Comoros Brazz. 1996 1992, 1995 Madagascar 1992, 1999 2004 1996 1992 amd. amd. 1998 amd.

9780230_242920_09_cha08.indd 152

11/11/2010 5:16:35 PM

4 4 2 1 4 11

4

4

2 3

7

13

15 pres-parl

0 2

1 4

20 pres-parl

2 0 0 0

0 0 2 0

12 pres-parl

10

2

0 2

4

4

0 0

2 0 0 0

Source: See individual constitutions listed at the end of this chapter.

Legislative powers Package veto Partial veto Decree Exclusive introduction of legislation Budgetary policy Referenda Nonlegislative powers Cabinet formation Cabinet dismissal Censure Dissolution of parliament Total legislative powers Total nonlegislative powers Total Pattern of authority

Mozambique Namibia Mauritania 1990, 2004 1990, 1998 2006 amd. amd.

Appendix 8.1 Continued

18 premierpres

7

11

0 3

0

4

1 4

2 3 1 0

Niger 1992

12

1

0 4

4

4

0 0

1 0 0 0

9

1

0 3

2

4

0 0

1 0 0 0

Sao Tome 2003 amd.

12 13 10 premier- pres-parl premierpres pres

3

9

0 3

0

0

1 4

1 2 1 0

Niger 1999

Sao Tome 1990 amd.

16 pres-parl

11

5

2 1

4

4

1 0

3 0 1 0

Senegal 1998 amd.

20 pres-parl

11

9

0 3

4

4

1 4

3 0 1 0

Senegal 2001

Semi-Presidentialism in Africa

153

Appendix 8.2 Majority, cohabitation and minority semi-presidential governments in African electoral democracies, January 2010 Country Burkina 1978– 1980 Burundi 1993 Cape Verde 1991– Central African Rep. 1993–2003 Comoros 1996– 1999 Congo Brazza. 1992–1997 Guinea-Bissau 1994–1998 Guinea-Bissau 1999–2003 Guinea-Bissau 2005– Mali 1992–

Madagascar 1992–2009 Mauritania 2007–2008 Mozambique 1994–2009 Niger 1993– 1996 Niger 1999– 2009 Namibia 1990–

Sao Tome e Principe 1991–

Senegal 2000–

Min. Govt.

Democratic breakdown Source

1979

Yes

Elgie (2008)

Yes

Yes

Reyntjens (1996) Kirschke (2007) Mehler (2005)

1996–1999

Yes

Elgie (2008)

1992–1997*

Yes

Clark (2005)

1994–1998

Yes

Azevedo and Nijzink (2007) Azevedo and Nijzink (2007) Azevedo and Nijzink (2007)

Maj. Govt.

Cohabit.

1993 1991–2001, 2001– 1993–1998*

2001 February 1999–2003

1999–2001*, 2001–2002* 2005–2008 2009 Jan– July*, 2009 March March1992–1994*, 1994–2002, 2007*– 1992–1997*, 2002–2009 May–August 2008* 1994– 1993–1995*

No

2001, 2002– Yes 2003 2008 No August–Dec 2002–2007

No

Villalon and Idrissa (2005)

1997–1998, 2002 2007–2008 May

Yes

Marcus (2005), Elgie (2008) AFP, May 6, 2008 Manning (2007) Moestrup (2007b) Moestrup (2007b) van Cranenburgh (2006) Elgie (2009a), Elgie (2010)

1995–1996

Yes No Yes

1999–2009*

Yes

1990–

No

1991–1994*, 1994– 2001–2002 2002–2004*, 2001, 2005–2009* 2004– 2005, Jan. 2010– 2000–2001*, 2001 April 2001*–

No

No

Galvan (2001)

Note: * indicates coalition government. Source for all countries: African Elections Database (http://africanelections.tripod.com).

9780230_242920_09_cha08.indd 153

11/11/2010 5:16:35 PM

154

Sophia Moestrup

Notes 1. See for example the four chapters on Africa (Guinea-Bissau, Mozambique, Madagascar and Niger) in Elgie and Moestrup (2007); Kirschke (2007); and van Cranenburgh (2007). 2. An “electoral democracy” as defined by Freedom House (2010) must have satisfied the following criteria: “1) A competitive, multiparty political system; 2) Universal adult suffrage for all citizens (with exceptions for restrictions that states may legitimately place on citizens as sanctions for criminal offenses); 3) Regularly contested elections conducted in conditions of ballot secrecy, reasonable ballot security, and in the absence of massive voter fraud, and that yield results that are representative of the public will; 4) Significant public access of major political parties to the electorate through the media and through generally open political campaigning. . . . Freedom House’s term ‘electoral democracy’ differs from ‘liberal democracy’ in that the latter also implies the presence of a substantial array of civil liberties.” 3. Average of political and civic liberties scores. Freedom House ranks countries scoring 1–2.5 as “free” (F), 3–5 as “partly free” (PF), and 5.5–7 as “not free” (NF). A country can still be classified as an “electoral democracy,” though it is ranked as “partly free” (e.g., Guinea-Bissau). 4. In March 2009, President Vieira was murdered by military elements intent on avenging the killing of General Tagme Na Waie. This was not an overturn of the democratic regime, however. Constitutional order was maintained, and the provisions for the replacement of the dead president were followed, without interference from the military. Like the killing of President Kennedy in 1963, the murder of the head of state does not per se constitute a regime breakdown. 5. This is another borderline case. Sao Tome has suffered two coups/military insurgencies, in 1995 and 2003, that each lasted about a week. Both were resolved with the restoration of the constitutional order, and Sao Tome maintained its Freedom House score of “free” for both years. 6. Tsai develops a categorization of five different subtypes of semi-presidential systems based on a combination of constitutional powers, custom and partisan alignment in parliament. By using behavioral outcomes as the underlying principle for his classification, Tsai’s design makes it difficult to separate out the independent impact of institutional and extra- constitutional variables, a point I shall revert to below. 7. Tsai (2008) provides an interesting suggestion for categorizing semi-presidential systems into five subtypes: president-dominant; premier-dominant; parliamentdominant; cohabitation; and balance-of-powers. His criteria of classification include the constitutional powers of the president and parliament, and the party situation in parliament. In reality, however, Tsai focuses on behavioral outcomes in his classification. Thus, by definition, balance-of-powers systems is the category where deadlock can lead to democratic breakdown. This leads to circular reasoning. 8. In both constitutions, it is stated that the president dismisses the prime minister upon the presentation by the prime minister of the resignation of the government. While the president may be able to pressure a prime minister from his own party to resign, this is unlikely to happen in a situation of divided majorities, where the opposition controls a majority in parliament. The authority to dismiss the cabinet is therefore not an independent presidential power.

9780230_242920_09_cha08.indd 154

11/11/2010 5:16:36 PM

Semi-Presidentialism in Africa

155

9. This is, of course, under the assumption that the reading of constitutional texts has not been biased across the two regions. I have sought to be conservative in my coding of presidential powers. For example, where the president can appoint a prime minister without the requirement for formal investiture, but the government must submit its program to parliament for approval within a given time frame, at which time it can be toppled by a vote of no- confidence, I have discounted the president’s power of cabinet formation somewhat (coding it as a 3 instead of a 4). Nijzink et al. (2006) do not reduce presidential powers to reflect this limitation, in their coding of African presidential powers. Their coding of presidential powers to censure parliament contributes to further increasing the scores for nonlegislative presidential powers compared to those in Appendix 8.1. This is apparently due to a “double counting” of presidential powers when considering the parliament’s ability to censure the government and the president’s ability to dissolve parliament. More discussion on codings of presidential powers would be welcome, to further advance the study of semi-presidentialism. Suffice to say, I have attempted to be conservative in my scoring of presidential nonlegislative powers, and still, they appear much more significant than in other regions, including countries in the former USSR. 10. Madagascar changed from a premier-presidential to a president-parliamentary regime form with the constitutional amendment of 1998. 11. Sao Tome and Principe, though earlier a president-parliamentary system, has become premier-presidential since the 2003 constitutional amendment. 12. Sao Tome is listed under “majority government” which has been the modal form of government on the islands since the transition to democracy in 1991, but also under “cohabitation” as the country did survive seven years under this form of government, contrary to what we would expect based on the predominant views in the literature. 13. Mali is similarly listed in two places. While “majority government” has been the most frequent government type in Mali since the transition to democracy in 1992, the country did spend five years in a situation of “divided minority government,” without this being detrimental to the consolidation of democratic gains in the country. 14. Further analysis of historical data could lead to include Burkina Faso in this group. 15. Patassé managed to secure the backing of a fluid parliamentary majority for his government thanks to the cooptation of independents and individual opposition members (Mehler 2005, 141–142). 16. See the list of heads of government of Sao Tome and Principe at http:// en.wikipedia.org.

9780230_242920_09_cha08.indd 155

11/11/2010 5:16:36 PM

9 The President is Not a Passenger: Portugal’s Evolving SemiPresidentialism Carlos Jalali

A cursory glance at the semi-presidential dimension of Portuguese politics suggests stability. Since democratization, in the aftermath of the April 1974 “carnation revolution,” Portugal has had four elected presidents. The first three served two full five-year terms, with high levels of public approval. The fourth incumbent was elected in 2006, and his presidency (in mid-2009) resembled that of his predecessors, with approval ratings that maintained the president as the most popular and best-regarded figure in Portuguese politics. This stability is also evident in the functioning of Portugal’s semipresidentialism, which is the focus here, with the balance of power within the executive favoring the prime minister for the past 30 years. However, this apparent calm on the surface masks a considerably less constant undercurrent. Portugal’s semi-presidentialism has seen relevant changes over the past 30 years, both in formal terms and in terms of its practice. Moreover, as will be shown, the president is by no means a toothless figure within the executive, even if political leadership rests with the prime minister. This chapter assesses Portugal’s evolving semi-presidentialism over six key dimensions. First, the role that semi-presidentialism played in democratic consolidation and how, on balance, it appears to have helped democratization. The second focuses on the interaction between the party system and the presidential dimension, which helps explain the continued predominance of the prime minister within the executive. The third pertains to the formal changes to the constitution, which, while impacting on presidential prerogatives, did not eliminate presidents’ considerable constitutional powers. The subsequent three dimensions concern the practice of semipresidentialism over time. First, how (and why) have presidents used their non-legislative (constitutional) powers – most notably, the powers related to appointing and dismissing the government and/or dissolving parliament? 156

9780230_242920_10_cha09.indd 156

11/11/2010 5:18:10 PM

Portugal: The President is Not a Passenger 157

Second, how have presidents used their legislative (constitutional) powers – an often underexplored dimension of semi-presidentialism that sheds important light in the Portuguese context? The third concerns the media and public agenda-setting power that presidents have, and how this interplays with presidents’ constitutional prerogatives. As will be shown, this constitutes a relevant weapon in the president’s armoury. While the practice of semi-presidentialism falls short of a presidentialized executive, Portuguese presidents are not merely ceremonial, even after the 1982 constitutional revision, which removed many of the president’s more substantial “non-legislative” powers. As will be demonstrated, presidents are able to influence policy-making, and their role is considerably greater than that of indirectly elected presidents, even when facing single-party majority governments. Presidents may not be the “drivers” in the executive, but their systematic use of “legislative” and “agenda-setting” powers (and occasionally of “non-legislative” ones) means that they are far from powerless “passengers.”

The definition and context of semi-presidentialism in Portugal Portugal’s transition to democracy not only heralded the third wave of democracy, but also foreshadowed what was to become very much an institutional feature of the third wave: semi-presidentialism. Yet the growing ubiquity of semi-presidentialism has not removed misgivings toward the concept and its application, with considerable divergence as to what is semipresidentialism and which countries fit this category (Elgie 2004). Portugal aptly illustrates this definitional problem, with some authors suggesting that only from 1976 to 1982 could it be construed as semipresidential, the constitutional revision of 1982 having reduced presidential powers to “little more than the normal powers of a normal parliamentary presidents” (Pereira 1984; Moreira 1989; Sartori 1997: 129). This position has been also strong in Portugal (see Araújo 2003: 87–88), and Araújo goes so far as concluding that semi-presidentialism is a “weak and poorly rigorous” concept (ibid.: 106). While this position has not gone uncontested – for a thorough refutation, see Neto and Lobo (2009) – it does reflect the definitional uncertainties of semi-presidentialism. Here we define semi-presidentialism as “the situation where a popularly elected, fixed term president exists alongside a prime minister and cabinet who are responsible to parliament” (Elgie 1999b: 13). This dispositional definition avoids the ambiguity and subjectivity that the relational dimension, based on the “actual powers of political actors” introduces; and permits distinguishing between the institutional framework of a regime and the power relations that can occur within it (Elgie 1998, 2004). By this definition, Portugal has been a semi-presidential regime throughout the post-1974 democratic period.

9780230_242920_10_cha09.indd 157

11/11/2010 5:18:10 PM

158 Carlos Jalali

What explains Portugal’s adoption of semi-presidentialism? One popular explanation suggests that the choice of semi-presidentialism aimed at providing an institutional framework capable of withstanding potential crises, the directly elected president serving as a “safety-valve” for situations where government and legislature failed (for an overview, see Cruz 1995: 224). However, such considerations played at best a secondary role. Semipresidentialism was more a child of (and response to) the immediate concerns of the complex revolutionary dynamic of Portugal’s transition to democracy than a choice borne out of institutional foresight. The common denominator between the various factions behind the military movement (MFA, Armed Forces Movement) that ended the Estado Novo was the opposition to the regime’s intractability over the decade-long independence wars in its then African colonies. Unable to alter policy from within, the army was finally to dismantle the Estado Novo. Yet there was little forethought or agreement on what should replace it (Carrilho 1985: 467). The coup thus gave way to a convoluted revolutionary context, in which regime-choice – how Portugal should be organized and governed – remained undecided. As such, regime- choice became the central conflict in Portugal’s revolutionary period, until the successful counter- coup of November 25, 1975 settled on a liberal democratic regime. The adoption of semi-presidentialism must be placed within this context. The radicalization of the Portuguese revolution after March 1975 led to a first pact between the Armed Forces Movement and the emerging political parties (known as the platform of constitutional agreement or MFA-parties pact) of April 1975. This envisaged the indirect election of the president by an electoral college composed of members of an MFA assembly (which included the military Council of the Revolution) and the deputies of the constituent assembly.1 The direct election of the president was formalized in the second MFAparties pact of February 1976, which reflected the weakening of the radical military wing after November 25, 1975. Although military tutelage was still enshrined in the pact, with an “implicit clause” that the first president would come from the military (Pereira 1984; Pires 1989: 291), its extent was substantially reduced. This helps to explain why the indirect presidential election was abandoned: the direct election of the president became necessary to legitimize military tutelage after its loss of revolutionary legitimacy with the November 25 counter- coup (Cruz 1995: 228–229). Concurrently, the parties had an interest in ensuring the inclusion of the army in the emerging institutional framework, providing a measure of stability to a highly fraught democratization. Semi-presidentialism was thus a direct result of Portugal’s transition process. This begs the question of its impact on the consolidation of democracy. On balance, the presidency appears to have helped more than it hindered the process of democratic consolidation. Portugal’s democratic consolidation

9780230_242920_10_cha09.indd 158

11/11/2010 5:18:10 PM

Portugal: The President is Not a Passenger 159

was a gradual, evolving, and contested process: the regime-choice conflicts of the revolutionary transition did not suddenly disappear with the November 1975 or the subsequent 1976 elections. In this context, the presidency was an important (albeit not sufficient) element in consolidation, in terms of its attitudinal, behavioral, and structural (Schedler 2001) dimensions. In terms of the first, the presidency quickly established itself as the most popular and trusted institution in Portuguese politics, not least because of the supra-partisan note set early on by its first incumbent, Ramalho Eanes. This provided a consistent anchoring of trust in the newly nascent democracy’s institutions that contrasted with fractious attitudes toward other political institutions, notably the government and the legislature, and helped generate support for the new regime at the mass level. In terms of behavior, the presidency acted as a (mostly) efficient safety valve in the context of the instability of the initial years of Portuguese democracy. As detailed below, Eanes helped to broker a new government after the first minority constitutional government lost parliamentary confidence in 1978. Equally, dissolution powers provided a solution to parliamentary majority breakdown – notably in 1979, when dissolution generated a considerably more stable government after the five governments of the first constitutional legislature. As outlined below, this role of the presidency was not universal – Eanes did not always use his powers in a way that contributed to stability. However, on one key dimension, the presidency contributed toward democratic consolidation: maintaining military consensus with the democratic regime despite the latter’s evolution meaning an increasingly weaker military role. This aspect cannot be underplayed: Portugal’s democratic consolidation ultimately ensured civilian control of the armed forces, with consensus on this from both military and civilian elites. That such outcomes are neither automatic nor easy is evidenced by the experience of other third-wave democracies, such as Spain (namely, the 1981 military coup). Eanes was a military figure – a general who had risen in the ranks in the revolutionary period – and he retained active military status as president, accumulating the roles of president and chief of staff of the armed forces. His support for the regime and his acceptance of key moments in the civilianization of Portuguese politics were decisive elements in democratic consolidation. The most salient case of this was the 1982 constitutional revision, even if it was not the only one.2 This aimed at removing the military enclaves in the new democracy’s institutional structure, and is widely seen as signalling the consolidation of Portuguese democracy.3 Despite his misgivings toward this revision – which removed military oversight from the political system and weakened the presidency – Eanes did not use his formal (or informal) powers to impede it, and thus helped pave the way to the civilianization of Portugal’s democratic institutions. Moreover, the provisions of this revision (further analyzed below) also contributed to the structural dimension

9780230_242920_10_cha09.indd 159

11/11/2010 5:18:11 PM

160 Carlos Jalali

of democratic consolidation, notably in terms of entrenching the “institutional foundations” of Portuguese democracy. Indeed, the 1982 revision was the only major constitutional redistribution of power in the political system after democratization, with the subsequent stability (and general acceptance) of the new institutional structure a good indicator of consolidation.

Presidents yield: Semi-presidentialism and the party system The government has been the central locus of political leadership throughout the democratic period, and remains the key political prize for parties. This perception appears to be shared by voters (Magalhães 2008), as reflected by higher turnouts for legislative elections than for presidential ones (particularly where an incumbent president is up for reelection). Two related factors help account for this secondary position. First, the genetic imprint in Portugal’s adoption of semi-presidentialism, which removed the presidency from party system competition until 1986.4 Second, the inability to translate presidential majorities into the parliamentary arena. The presidency of Eanes is central to a comprehension of the role of semipresidentialism in the Portuguese party system, and helped to establish a pattern of governmental dominance within the executive. Rising from relative obscurity within the Armed Forces Movement with the events of November 25, 1975, where he was the operational military commander in the counter-coup that essentially secured West European-style liberal democracy for Portugal, his choice as presidential candidate stems largely from this episode.5 Eanes was backed to the presidency in 1976 by (inter alia) the two dominant parties of the party system, the centre-left Socialists (PS) and the centre-right Social Democrats (PSD). This support reflected the conditions of the “implicit military clause” behind the adoption of semi-presidentialism. Eanes was backed as a “non-political” supra-partisan figure, and his presidency was aimed at including the military in the new regime. Although relatively popular while in office – particularly against a backdrop of a divided assembly and thus of generalized governmental instability – the “military clause” that gave him the presidency also largely precluded Eanes from presidentializing the executive. With little voice within the parliamentary parties and lacking a partisan vehicle of his own, his attempts at presidentialization (notably through the appointment of “presidential governments” in 1978–1979) were blocked by his erstwhile partisan supporters. Only at the end of his second term, when he was already a lame duck, did he organize the Party of Democratic Renewal (PRD), and even then only indirectly. But by then the die was cast, and the PRD was a political vehicle for Eanes after the presidency, “the civilianisation of his political designs” (Morlino 1998: 72). The Eanes presidency – and the implicit military clause behind it – meant that the presidential level was largely offside in terms of party system

9780230_242920_10_cha09.indd 160

11/11/2010 5:18:11 PM

Portugal: The President is Not a Passenger 161

competition for the first decade of Portugal’s democracy, with legislative elections the central political contest. It also helps explain why cohabitation did not produce the instability and deadlock predicted by Stepan and Suleiman (1995: 413) in new democracies. First, their prediction that such deadlocks might invite military intervention was limited by the military’s stake in the political system, be it through the presidential “military clause,” or (until 1982) through the Council of the Revolution. Second, the incongruent majorities in the presidency and parliament were assuaged by the asymmetry in their respective party system wherewithal: Eanes, as a nonpartisan president, could not count on party support in case of a conflict with parliament, as the 1978–1979 presidential governments demonstrated. Without party support – and thus without the possibility of controlling parliament – the payoffs to presidential conflictualism were limited. In a sense, then, cohabitation – defined by Elgie (2004: 328) as “the situation where a president from one party or coalition shares power with a prime minister from an opposing party or coalition” – became, during Eanes’ tenure, as the situation where a partisan government shared power with president without a party or coalition. The “military clause” lapsed in 19866, and the ensuing civilianization of the presidency was accompanied by the full integration of the presidential level in the party system. However, this civilianization has not altered the pattern of predominance of the government in executive terms. While the three presidents that succeeded Eanes all had historical ties with either of the two dominant parties – indeed, all three formerly led their parties – none of them was able to translate his presidential majority into a parliamentary one. The presidency of Mário Soares (1986–1996) marked the definitive civilianization of the political system. Soares’ first term was characterized by an apparently cooperative attitude toward the government (Frain 1995: 656–665), with a corresponding “above politics” and “above party” stance designed to secure reelection that proved very popular with the electorate and became part of the modus operandi of his successors. This moderate tone and institutional balance may be considered somewhat surprising, given that Soares had been the undisputed leader of the PS, faced by a minority Social Democrat government, in an assembly where the PRD held the balance. It is less so if we take into account the nature of Soares’ election in 1986 – a narrow victory, with less than 200,000 votes separating him from his right-wing opponent in the decisive run- off, and based on extremely grudging and contingent Communist second ballot support, unlikely to be repeated if the Communist Party could avoid it.7 Soares’ second term in office was bolstered by having won reelection with over 70 percent of the vote in the first ballot. However, the more interventionist tone of Soares’ second term (developed below) fell well short of contesting the executive dominance of government (still led by

9780230_242920_10_cha09.indd 161

11/11/2010 5:18:11 PM

162 Carlos Jalali

the PSD, but now with a single party majority), and the president was always careful to cultivate popular support and the perception of being “above party.” The Jorge Sampaio (1996–2006) presidency neatly demonstrates the differences between the Portuguese and French experiences of semipresidentialism. Although the presidential and governmental majorities coincided for much of Sampaio’s time in office (1996 to 2002), the dominant element within the executive remained the prime minister, as Sampaio did not control the Socialist near-majorities in parliament. The internal party balance is relevant here. Sampaio was a defeated leader of his party – he was unable to prevent the PSD’s second absolute majority of 1991, and was defeated in the subsequent 1992 Socialist party congress by António Guterres, who became prime minister in 1995. Thus, Sampaio did not dominate his party, even if an important sampaísta faction remained within it. A second factor reinforced this effect. Sampaio’s election had been preceded by the large PS victory in the 1995 legislative elections, which reinforced Guterres’ internal hand. He could hardly flourish against his own party and against a leader who had succeeded precisely where Sampaio had failed. Party system dynamics have largely precluded presidential leadership within the executive. Congruent presidential and parliamentary majorities are rare; and when they have occurred, the crucial parliamentary majority “belonged” to the prime minister. In periods of incongruent majorities – the norm over the past three decades of democracy – the ability of a president to gain control within the executive is all the more limited. At the same time, the Portuguese experience does not confirm Bahro et al. (1998) prediction that semi-presidential presidents are “void of influence” and “reduced” to the position of their indirectly- elected counterparts when facing an opposing parliamentary majority. As will be shown, presidents wield considerable influence within the political system, even when facing a government backed by a solid (and rival) parliamentary majority. Presidential vs. prime ministerial influence within an executive thus appears to be better captured as a continuum rather than a dichotomy (as suggested by Sartori 1997).

The evolving constitutional framework Constitutional rules provide the de jure underpinning for de facto balances of power within semi-presidential executives (Elgie 1999c: 16–17). This section evaluates the constitutional resources available to the president, how these have changed over time – notably, with the amendments of 1982 – and how they affect the nature of Portuguese semi-presidentialism. The original constitutional framework of the presidency was – like much of the 1976 constitution – the product of compromises between the parties and

9780230_242920_10_cha09.indd 162

11/11/2010 5:18:12 PM

Portugal: The President is Not a Passenger 163

the MFA as well as among the parties themselves. The system that emerged was a hybrid, a cross between the Weimar and French Fifth Republic constitutions. Its key elements (in 1976) were: (a) President had the power to dissolve parliament, subject to approval of the Council of the Revolution (Article 136[e]). (b) President had extensive veto power, including pocket veto (Articles 137 and 139), and could preventively refer decrees and bills to the Council of the Revolution (Articles 277 and 278). (c) Government was politically responsible to both the assembly and the president (Article 193). (d) President nominated and dismissed the prime minister (Articles 136[f] and 190[1]) and ministers on prime ministerial proposal (Article 136[g] and 190[2]). (e) Council of Ministers meetings were chaired by the prime minister, and could only be chaired by the president at prime ministerial request (Article 136[h]). (f) Government was to conduct the general policy of the country and was the supreme body of public administration (Article 185[1]). (g) Government had legislative competence, via decree-laws [Article 201]. (h) President chaired the Council of the Revolution (Article 136[a]) and was supreme commander of the armed forces (Article 137[1a]). (i) Given his role in the Council of the Revolution, the president was given considerable powers for guaranteeing the regular functioning of democratic institutions and “of the fulfilment of the constitution and fidelity to the spirit of the Portuguese Revolution” (Article 142). Comparing this with the constitutional framework of Fifth Republic France, it is clear that the constitutional texts are not a sufficient explanatory variable in accounting for differing practices (Duverger 1986: 14–15). In neither case is the head of state purely a symbolic figurehead, nor is he the chief executive. Indeed, certain elements of the 1976 constitution suggest a relatively stronger presidency. Whereas in France the government is nominally responsible to the national assembly only, in Portugal government was politically responsible to both the assembly and the president. This responsibility was somewhat asymmetric – positively affirmed confidence in the government on the part of the assembly was not necessary for the maintenance of the government, whereas (by implication) that of the president was (Miranda 1986: 136). The powers of the Portuguese president were modified by the 1982 constitutional amendment. This was aimed at removing military tutelage, and consequently touched upon presidential prerogatives. With few exceptions (notably Matos 1983, 1986; Morais et al. 1984), most studies agree that this revision resulted in an overall weakening of the president (Neto and Lobo

9780230_242920_10_cha09.indd 163

11/11/2010 5:18:12 PM

164 Carlos Jalali

2009), at the extreme suggesting that only the period from 1976 to 1982 was semi-presidential (Sartori 1997). This revision did increase presidential power in some areas, largely due to the abolition of the Council of the Revolution and consequent redistribution of its powers. The most notable element here is the power to dissolve parliament, no longer constrained by the Council of the Revolution. The revised constitution also obliged the prime minister to keep the president informed on matters pertaining to the country’s internal and external policy; and it substantially widened the areas where a presidential veto could only be overridden by a two-thirds majority in parliament. However, temporal constraints on the president’s power to dissolve were introduced; and the pocket veto was removed, as was Article 142. Although government remained responsible to both the president and the assembly (Article 193), the government’s political responsibility was limited to the assembly (Article 194[1]). The president’s power to dismiss government was also constrained – he could now only do so “when necessary to ensure the regular functioning of democratic institutions, after consultation with the Council of the State” (Article 198[2]). The 1982 constitutional amendment also saw a strengthening of the assembly in its political oversight of government (Morais et al. 1984: 92–98). Thus, the potential presidentialization of government was further constrained by lowering the requirement for overthrowing a government from two votes of no- confidence in thirty days to only one; and by the requirement that all governments, including presidential-appointed ones, present their programme for assembly approval within 10 days of appointment, prior to this their actions being “limited to the functions strictly necessary for the management of public affairs” (Article 189), with rejection of a government’s programme entailing its dismissal. Subsequent amendments were primarily concerned with other aspects of the constitution, but one important addition to presidential power must be noted. A 1989 amendment added the mechanism of referendums to the Portuguese political system. As in France, referendums are proposed by the assembly or the government, but the decision as to whether to hold them belongs to the president. Thus, there is potential for conflict between president and government, as a president may refuse to call a referendum proposed by the government if he feels that the issue does not warrant it or lacks “relevant national interest.” However, the area of jurisdiction of referendums is constrained and thus likely to prevent alterations in the political balance of power. The constitution explicitly excludes from referendum matters pertaining to a) alterations to the constitution; b) budgetary, financial, or tax matters; and c) inter alia, legislation affecting the election of holders of organs of sovereignty or their statutes. Referendums thus cannot be used to provoke changes in the political system comparable to France in 1962 or Italy in the early 1990s.

9780230_242920_10_cha09.indd 164

11/11/2010 5:18:12 PM

Portugal: The President is Not a Passenger 165

Using non-legislative powers: From failed presidential governments to the “atomic bomb” Relational analyses of semi-presidentialism frequently base the existence of “considerable presidential powers” on a president’s powers over government formation and survival. It is thus relevant to examine how Portuguese presidents have used their “non-legislative” armoury over the past three decades. As will be seen, presidents have used the full range of their nonlegislative powers, albeit with quite distinct outcomes. Interestingly, it is the more recent usage of these powers that potentially most widen a president’s de facto powers, reflecting how presidential patterns remain more susceptible to change than might be expected given the duration of Portugal’s semi-presidentialism. The Eanes presidency marked the most distinct examples of presidential involvement in government formation – and helped define its limits also. Eanes’ interpretation of his role in government formation came relatively early on, when he helped broker the new regime’s second constitutional government, a coalition between the Socialists and the right-wing CDS8 installed in January 1978, after the first constitutional minority PS government collapsed. His response to the collapse of the PS- CDS government (after little more than six months) was to set the executive balance of power decisively in favor of the government. Eanes sought greater presidential control by appointing extra-party “governments of presidential inspiration.” Eanes’ gamble was relatively straightforward: to exploit the divisions within and between the parties, using his own support in the electorate and among political élites. Given the highly fractious nature of the democratic parties, particularly the PS and PSD (largely a consequence of the progressive clarification of their real positions as democracy became consolidated), presidential administrations might divide and undermine the PS and PSD, or at least their leaders (Lobo 2000: 158–159). Thus, Eanes would either be able to force party support for the government, and also for himself, or, if the parties preferred to confront him, he might hope to draw on popular support against them, as did General de Gaulle in 1962. Instead, the parties united against Eanes, and decisively blocked this strategy, aware of the dangers (to them) of giving his government breathing space.9 The first “presidential government” lasted a mere 85 days, its programme rejected by the Socialists and CDS. Its successor, led by Mota Pinto, was to last longer, essentially due to the tactical support of the Socialists (who hoped to weaken the PSD prior to the next legislative elections, as Mota Pinto was a PSD dissident). However, the partisan rug was gradually pulled from under its feet from March 1979 onward, with the government’s proposals rejected in parliament (not least, its proposed budget) and it fell three months later. The third government represented an admission of defeat by

9780230_242920_10_cha09.indd 165

11/11/2010 5:18:12 PM

166 Carlos Jalali

Eanes, with the president first dissolving parliament (as the parties wanted) before appointing Maria de Lurdes Pintasilgo as interim prime minister. The failure of the “presidential governments” of 1978–197910 set a pattern of low presidential interventionism over government formation. It is fair to say that opportunities for such interventionism have also diminished, as governments became more stable and frequently sustained by parliamentary majorities. Looking at the period from August 1980 to 2009, only 8.2 years featured a minority government; the rest split between single party majorities (12.7 years) and two-party majority coalitions (8.8 years). Presidents had the opportunity to play a more interventionist role in the four occasions when governments ended prematurely, and invariably opted to dissolve parliament rather than back (and be tied to) alternative governments. The most salient example of this occurred in the aftermath of the 1987 motion of censure, which led to the collapse of the Cavaco Silva-led PSD minority government. Soares rejected a proposal from his Socialist party to form a coalition government with the PRD (with Communist parliamentary support), instead opting to dissolve parliament (when the legislature still had two more years to run), even if that was almost certain to hand victory to the PSD, which had a strong lead in the polls at that point. Instances of presidential influence on government formation are thus largely limited to “safe,” popular situations, and contingent on the political balance of power. This is exemplified by both the Soares and Sampaio presidencies. The first prevented the PSD from appointing its new leader Fernando Nogueira as prime minister immediately prior to the 1995 legislative election, a decision facilitated by the PSD’s declining popularity (later confirmed by a drop of over 15 percentage points in those elections visà-vis the 1991 legislatives). Sampaio imposed on prime minister Guterres the resignation from government of two close allies of the latter, Armando Vara (assistant minister for youth and sport) and Luís Patrão (former chief of staff of Guterres and secretary of state for home affairs) after their alleged involvement in a financing scandal in 2000.11 Their resignation was a popular one, and Sampaio was able to push for it given Guterres’ falling popularity in a context of deteriorating economic conditions and the proximity of the presidential election. In a different context, Guterres might well have resisted such presidential pressures. If the use of the power of appointment in government has grown increasingly constrained, the opposite trend can be seen in terms of the power to dissolve parliament. Portuguese “political wisdom” has dubbed this the “atomic bomb” in the presidential armoury – a metaphor that emerged to capture not only the “impact” of a dissolution, but also the unlikelihood of its use. The comparative experience lent credence to this latter dimension. The practice of semi-presidentialism in Europe has generally delimited parliamentary dissolutions to situations where there has been a breakdown of agreement within the assembly. Finland’s post-war

9780230_242920_10_cha09.indd 166

11/11/2010 5:18:12 PM

Portugal: The President is Not a Passenger 167

experience exemplifies this clearly: all four dissolutions of parliament – one in 1953 by Paasikivi, three under Kekkonen in 1962, 1971, and 1975 – took place either when a majority broke down, or when the government did not have a parliamentary majority (as in 1962) (Paloheimo 2001: 101; Raunio and Wiberg 2003: 308–309). This was also the norm behind Portugal’s parliamentary dissolutions of 1979, 1982, 1985, 1987, and 2001. Only in France has dissolution occurred against a functioning government with a stable parliamentary majority, and such dissolutions have consistently occurred only when the president’s popular mandate is more recent than the assembly’s. As Schleiter and Morgan-Jones (2005: 23) put it, “semi-presidential institutions tend to encourage deference to the more recent electoral mandate.” However, the “nuclear option” was used under unexpected circumstances in December 2004, when president Sampaio dissolved parliament even though there was a coherent and apparently stable parliamentary majority (a coalition on the right between the PSD and the CDS), and the assembly had a more recent popular mandate than the president. The dissolution power was used to remove the government of Pedro Santana Lopes, and was essentially due to the government’s unpopularity. The resulting February 2005 election reaffirmed support for Sampaio’s decision. As expected, the election produced an alternation in government, with the return of the Socialists to power after three years of a centre-right coalition administration. The Socialists’ 45 percent share of the vote gave them a comfortable seven- deputy parliamentary majority – their first since democratization – and the elections resulted in an historic rout for the Portuguese right. The PSD obtained 28.7 percent of the vote, its lowest share since 1983; the CDS also declined, to 7.3 percent; and their combined total was the lowest since 1975. The implications of this dissolution for Portugal’s semi-presidentialism are substantial, and considerably widen de facto presidential power. As Shugart points out, a president’s dissolution power allows him to refer disagreements with (or within) the assembly to voters (Shugart 2005: 15–16). As outlined above, the practice of Portuguese semi-presidentialism delimited this notion of a disagreement to instances of a breakdown of agreement within the assembly; and broader semi-presidential practice only included one further situation, namely dissolution when the presidential mandate is more recent. The dissolution of December 2004 extends this notion of disagreement: dissolution occurred despite no breakdown of parliamentary support for the government – in other words, no intra-assembly disagreement over policy – and no “clearer” presidential mandate. As such, it leaves room for future presidents in interpreting the notion of disagreement that justifies a parliamentary dissolution, and opens a window of opportunity in terms of a greater role of the president within the executive.

9780230_242920_10_cha09.indd 167

11/11/2010 5:18:12 PM

168 Carlos Jalali

This window of opportunity is not limited to actual dissolutions of parliament. As Shugart (2005: 18) points out, the threat of dissolution “can be a powerful bargaining chip for a president when the majority expects the president to prevail after a dissolution.” As such, the “non-legislative” power of dissolution can also be used for “legislative” purposes, with presidents able to use the threat of dissolution as leverage against government decisions they disagree with. Such “legislative” use of the power of dissolution occurred in 2003 over the Iraq war, in what was by all accounts a relatively novel use of presidential powers in Portugal. Then prime minister (later European Commission president) Barroso intended to send troops to Iraq following “Operation Iraqi Freedom.” This was a predictable position given Barroso’s support for intervention in Iraq, reflected in his hosting of the “Azores summit” between Bush, Blair, and Aznar in March 2003. However, Barroso was forced to back down when president Sampaio told the prime minister that, as supreme commander of the armed forces, he would not support sending troops to Iraq without an EU or UN mandate. By all accounts, Sampaio did not explicitly threaten Barroso with dissolution should the prime minister persist with his plans for troops in Iraq, but it was made clear that he would use his full range of constitutional powers. It did not exercise Barroso’s political instinct to realize that a fight with the president over the deeply unpopular Iraqi issue could never be won (especially when his government was unpopular enough due to a growing economic crisis). The 2004 dissolution also widens the possibilities for a “legislative” use of “non-legislative” powers by the president in the future, by making the threat of dissolution credible in a greater variety of situations. This tends to confirm our overall assessment of Portugal’s semi-presidentialism. Political leadership rests primarily with the prime minister, and no president has contested this since 1979. However, this does not make Portugal’s presidents mere “figureheads,” even when facing an opposing parliamentary majority (contra Bahro et al. 1998), and presidents have sought (and succeeded) to use their powers in order to influence legislative outcomes.

Using legislative powers: Binding governments with Lilliputian threads Analyses of semi-presidentialism tend to concentrate on the non-legislative prerogatives of presidents, particularly their impact on government formation and cabinet nominations (Schleiter and Morgan-Jones 2005; Neto and Strøm 2006). However, as Magalhães (2001) points out, this bias underestimates presidential capacity to influence policy outcomes, in the Portuguese case at least. Portuguese presidents have two “legislative” resources at their disposal: referring legislation for constitutional review (both a priori and a posteriori);

9780230_242920_10_cha09.indd 168

11/11/2010 5:18:12 PM

Portugal: The President is Not a Passenger 169

and the political veto. In addition, the president can return decree-laws to the government, forcing their redrafting or submission to parliament. These powers are characterized in Portuguese political parlance as “conventional weapons” (in opposition to the “atomic bomb” of dissolution), and their use has not only had considerable impact on legislation, it has also constrained governments’ capacity for political leadership. This pattern was first evident during the presidency of Eanes, generating the concept of “institutional guerrilla warfare” between president and prime minister. Interestingly, this “guerrilla warfare” predated the constitutional revision of 1982, coming in the aftermath of the failure of the “presidential governments” of 1978–1979 (Barroso 1986), albeit becoming considerably more evident after the president’s reelection in December 1980 (Campinos 1986: 214). The impact of a president’s legislative powers on the government is reflected by the reaction of Cavaco Silva to the “institutional guerrilla warfare” of Soares during the latter’s second term. Cavaco Silva enjoyed an unparalleled mandate, both in parliament and politically, having led his party to a second single-party majority, with over 50 percent of the vote. Yet Soares’ use of constitutional litigation and political vetoes unnerved the then prime minister, with his remark that the government faced “blockading forces” an ill- disguised reference to Soares. The “guerrilla” nature of this “warfare” makes it difficult for governments to break free from the “Lilliputian threads” of the president’s legislative interventionism. Dramatizing the conflict risks generating “open warfare,” and the potential gains from this option do not seem to compensate its risks. Presidents generally engage in this “guerrilla warfare” over issues that are popular (namely, Eanes’ 1981 veto of a law increasing the remuneration of public office-holders), and are careful to maintain an “above party” stance. Moreover, they not only seek to preserve their popular backing, they also engage in this type of conflict more frequently when governments are facing a popularity downturn (Campinos 1986; Magalhães 2001). As such, the gains for the government of dramatizing the conflict into “open warfare” are at best relatively limited: the president’s use of his “legislative” powers, while not without impact, do not challenge the government’s leadership of the executive. But the costs of fighting the president can be considerable, especially when assessed through the “fog of war,” not only risking the government’s popularity but also its very survival, given the president’s power to dissolve parliament. The use of a president’s legislative powers – both in terms of constitutional litigation and veto power – has increased over time. Eanes referred no more than ten laws and decree-laws for judicial review during his second term, and issued a political veto on no more than five laws or decree-laws. This is considerably less than the legislative activism of both Soares and Sampaio in their second terms. The former vetoed or called for preventive

9780230_242920_10_cha09.indd 169

11/11/2010 5:18:12 PM

170 Carlos Jalali

review of 53 laws and decree-laws (three times more than Eanes); and the latter reached a total of 70 laws and decree-laws vetoed or sent for preventive review. The balance between preventive review and vetoes has shifted in the latter’s favor over time, pointing to an increasingly “political” use of presidential powers. While Eanes used the preventive review considerably more than the veto, Soares used them in relatively equal measure (30 vetoes and 23 preventive reviews) and Sampaio’s second term saw a predominance of the political veto (63 vetoes and 7 preventive reviews).12 The precise impact of these “legislative powers” is not easily quantified. However, it does appear to constrain governments. As Magalhães (2001) points out, three such effects can be discerned in the Portuguese case. First, the final formulation of a number of policies suggests a considerable degree of presidential influence. Second, it imposes costs on governments in terms of the policy-making process, by slowing down the legislative process and diverting attention from other issues. Third, as developed further below, the use of presidential “legislative powers” generates external costs for a government, with greater public scrutiny and debate over its decisions. Equally, a president’s capacity to impose such costs on governments gives him leverage over legislation, as governments may prefer to negotiate “behind-thescenes” rather than face the costs of public interventionism. This latter situation appears to be at work during the Sampaio presidency and during much of the first term of Cavaco Silva, who prefer backroom negotiations to open “guerrilla warfare,” and to some effect (namely, Sampaio’s role during the Santana premiership, or Cavaco Silva’s role in brokering a PS-PSD deal over the reform of the justice system in September 2006). Additionally, the considerably weaker activism in presidential first terms must be noted: Soares’ use of the veto or preventive review in his first term was less than half that of his second term; in Sampaio’s case, less than third. This shows the limits of legislative interventionism by presidents, not least motivated by the need to maintain a majority for re- election. Even second term presidents are careful to cultivate popular support and, while greater interventionism has had costs for presidents’ approval ratings, it did not stop them from remaining consistently the most popular political figures in Portugal. As will be seen next, presidents actively use this position to sway the “hearts and minds” of the electorate, and influence policy-making.

A president’s extra- constitutional powers: Setting the agenda The agenda-setting process is defined by Dearing and Rogers (1996: 1–2) as an “ongoing competition among issue proponents to gain the attention of media professionals, the public, and policy elites.” What the Portuguese experience has demonstrated is that presidents have a wealth of resources at their disposal in terms of media and public agenda-setting. To paraphrase Bernard Cohen, Portuguese presidents may have not been successful much

9780230_242920_10_cha09.indd 170

11/11/2010 5:18:13 PM

Portugal: The President is Not a Passenger 171

of the time in telling people directly what to do through policy, but have been very successful in telling people what to think about. Presidential pronouncements have considerable media and public repercussions, and presidents are well aware of this. This serves to reinforce the impact of a president’s “legislative” powers, generating considerable salience for issues favored by the president and, in some cases, accentuating opposition to government proposals. This agenda-setting capacity was first evidenced during the Eanes presidency. When his “hard power” to appoint governments was constrained – first by the failure of his governments in 1978–1979 and then by the constitutional revision of 1982 – Eanes used the “soft power” of agendasetting to considerable effect. Barroso (1983: 236–237), examining the conflicts between the president and the parliamentary majority from 1979 and 1983, characterizes his speeches and public pronouncements as “a new weapon of the president,” highlighting the endogenous capacity that a president has to increase his public visibility and speeches, and thus accentuate his agenda-setting role. This agenda-setting dimension has been a key feature of subsequent presidencies also. Thus, Soares’ institutional “guerrilla warfare” in his second term was also largely based on this dimension, which Soares defined as a president’s “magistratura de influência.” This occurred in two manners. First, through his public speeches, both at formal state events and at events “created” by the president, such as “open presidencies” (see Neto and Lobo 2009: 247, 249). Second, and as highlighted above, the president’s use of his “legislative” powers of veto or constitutional review also provide him with considerable agenda-setting power, even if the veto is overridden by the parliamentary majority. As Soares put it, explaining the impact of his presidential vetoes: Newspapers talked [about the veto], the opposition parties sought to debate the issue, some members of parliament of the PSD became irritated and, attempting to argue, manifestly lost reason. The ideas made their way into the popular consciousness. Soares in Avillez (1997: 274) Subsequent presidents have used this agenda power to good effect, even when the institutional climate has not been one of the “guerrilla warfare.” Thus, Sampaio’s comment that there was “more to life than the deficit” in 2003 considerably strengthened the opposition to the Barroso government’s economic policy, which was predominantly preoccupied with curbing Portugal’s public deficit to within the Eurozone’s stability and growth pact’s three percent norm. A more recent example is the debate over the location of a new international airport for Lisbon. The president’s public intervention in mid-2007 expressing doubts over the government’s choice of location reinforced opposition to this choice and

9780230_242920_10_cha09.indd 171

11/11/2010 5:18:13 PM

172 Carlos Jalali

helped keep the issue in the limelight. In the end, the sustained public pressure over this issue led the government to strike a deal with the president, with the location of the airport decided by a technical study by the National Civil Engineering Laboratory. This was to reverse the original government decision and the airport is now formally set for construction at a different location. The process helps illustrate the ability a president has to influence major policy decisions through his agenda-setting and constitutional powers.

Conclusion Portugal’s semi-presidential experience is frequently characterized as being composed of two distinct periods. The first is seen as a period of “clear” semi-presidentialism, until the constitutional revision of 1982. The second is described as a period of presidential effacement with the reduction in presidential powers after the 1982 revision. Indeed, this has led some authors to characterize the post-1982 president as being little more than a “passenger” within the executive, his role reduced to that of indirectly elected presidents. As this chapter demonstrates, this is far from constituting a complete and accurate portrayal of Portugal’s semi-presidential experience. While there was a shift in the use of presidential powers, this began in the aftermath of the 1978–1979 failed presidential governments rather than with the 1982 constitutional revision per se. At the same time, presidents have shown themselves able to adapt to this situation, using their “legislative” (veto and constitutional review) and “soft” agenda-setting powers in order to influence government decisions, while retaining their “hard” non-legislative powers – as the 2004 parliamentary dissolution was to rather unexpectedly demonstrate. Presidents are thus far from being powerless “passengers” in the executive – most evidently so during their second terms, unencumbered by the need to prepare their reelection. At the same time, this evolving pattern suggests that presidential patterns are not fully consolidated in the Portuguese case. As the precursor of modern semi-presidentialism put it, a constitution is “un esprit, des institutions, une pratique” (de Gaulle 1964). De Gaulle thus confirms Sartori’s (1997) distinction between the formal and material constitutions, as well as the prevalence of the latter in terms of actual practice. Material constitutions emerge out of a practice, and as such are based on the tacit understandings of those who work it. Yet, to paraphrase Low, these tacit understandings may not always be understood (Sidney Low 1904, quoted in Bogdanor 1996: 6). While the boundaries for the use of presidential power are largely defined in Portugal, the 2004 dissolution highlighted that they were not yet equally “understood” by all relevant actors. This partial indefinition means that we cannot exclude further shifts in Portugal’s semi-presidentialism. However, unless they are accompanied by changes to the party system, these shifts are

9780230_242920_10_cha09.indd 172

11/11/2010 5:18:13 PM

Portugal: The President is Not a Passenger 173

likely to result more in changes of degree than kind when it comes to the practice of semi-presidentialism.

Notes 1. For the full text of the first Pact, see Mello (1976: 235–241). 2. Viz. the president’s confirmation of the first civilian defense minister in 1980, thus helping defuse a potentially fraught relationship between the government and the armed forces. 3. Linz and Stepan (1996: 116–129) date the completion of both transition and consolidation to the constitutional revision of 1982, with the subsistence until then of the Council of Revolution as preventing full transition. 4. Although a greater party and party system influence is discernible in the 1980 elections with the PSD- CDS coalition’s aggressive anti-Eanes and anti-MFA stance. 5. See Soares in Avillez (1996: 23–29) for details. 6. The lapsing of the military clause is reflected in the nature of presidential candidates. Out of a total of 10 presidential candidates in the 1976 and 1980 presidential elections, eight were military men (and the two civilians were the least voted candidates in both elections). From 1986 on, none of 21 candidates that have contested presidential elections were military men. 7. The Communist Party and the Soares-led PS were bitter rivals during the revolutionary period. While the Communist Party ultimately supported Soares as the candidate of the left in the second ballot, it also made clear that this support did not make Soares any more palatable than before. 8. Misleadingly named “Party of the Democratic Social Centre.” 9. As they did his forays into foreign policy, construed as attempts at greater presidentialization. Thus, Soares confirmed the resignation of Medeiros Ferreira as his foreign minister against Eanes’ wishes (Soares in Avillez 1996: 64–66). 10. The Pintasilgo government was replaced on January 3, 1980, after the legislative elections held on December 2, 1979. 11. “Sampaio exigiu demissões,” Público online, December 16, 2000. 12. Data compiled from Campinos (1986), Magalhães (2001), and Freire and Pinto (2005), with author’s own calculations for Eanes’ second term.

9780230_242920_10_cha09.indd 173

11/11/2010 5:18:13 PM

10 Taiwan: Democratic Consolidation under President-Parliamentarism Yu-Shan Wu and Jung-Hsiang Tsai

The stability of a semi-presidential system is often put to the test when the president’s party does not hold a majority in parliament. Under these circumstances, the president-parliamentary version of semi-presidentialism where the government is accountable to both the president and the legislature is generally considered to be less stable than the premier-presidential version where the government is responsible only to the legislature. However, there may be other factors that militate against this generalization, which is the case in Taiwan. The constitution of the Republic of China (ROC) – Taiwan – was significantly amended in 1997. Since then the country has had a president-parliamentary version of semi-presidentialism. The ROC president is an all-powerful figure who holds the president of the Executive Yuan (premier Xingzheng yuanzhang) accountable to him, which is a typical case of president-parliamentarism. Since 1997 Taiwan has experienced both a united and a divided government. The true test of the system came when the country experienced its first divided government in 2000. The newly elected president, Chen Shui-bian, unilaterally appointed the prime minister without consulting the opposition that held the majority in the Legislative Yuan (Lifa yuan, the parliament). From 2000 to 2008, Chen insisted on the presidential prerogative of appointing the prime minister, and the parliament failed to challenge him. Divided government occurred in the context of dual accountability, presumably a recipe for disaster. However, Taiwan was able to consolidate its democracy during this protracted period of minority government, and satisfied the “two turnovers” test in 2008 when the KMT’s presidential candidate Ma Ying-jeou was elected president in March, restoring united government and removing the most serious threat to political stability.1 During its divided government, although the interbranch relationship between the executive and the legislature was extremely tense from time to time, Taiwan’s young democracy functioned reasonably well. What then accounts for the democratic consolidation of Taiwan, despite the institutional tendency toward instability of its president-parliamentary system? Here lies our primary curiosity. 174

9780230_242920_11_cha10.indd 174

11/11/2010 5:19:20 PM

Taiwan 175

The following discussion is separated into two sections. First, we trace the process of Taiwan’s entry into semi-presidentialism, the choice of a presidentparliamentary system, the advent of divided minority government, and the restoration of united government. Second, we explore the causes of Taiwan’s relative stability under such unfavorable circumstances.

Taiwan’s entry into semi-presidentialism Taiwan’s basic government structure was laid out in the 1947 Constitution of the Republic of China promulgated on the Chinese mainland. After the defeat of the ROC government by the communists in 1949, the Kuomintang regime fled to Taiwan and reestablished a state structure on the island. Although the Constitution’s legitimacy has been repeatedly challenged by the pro-independence political forces, it is still in force, making Taiwan one of the very few nascent democracies that have kept its pre-transition constitution (Wu 2006). One reason for its longevity is the symbolism of the 1947 Constitution. For all those who do not want to see a legal and permanent separation of Taiwan from the mainland, whether in Taipei, Beijing, or Washington, the Constitution has its intrinsic value of keeping a oneChina legal framework. The other reason is the Constitution was written by the framers for a nascent democracy, not for a party dictatorship. Its main architect, Chang Chün-mai (Carsun Chang), was arguably the most famous advocate of parliamentary democracy in China at the time (Chi 1992: 135–155). As such, the Constitution aimed at building liberal democratic institutions and progressive social rights. It was not an instrument for the dictatorship of the ruling Kuomintang (KMT). The Constitution established a parliamentary system with a strengthened position for the government vis-à-vis the parliament (Legislative Yuan) (see Table 10.1). Its most important impact was the creation of the position of prime minister (President of the Executive Yuan), who led a government that was responsible to the parliament. This cardinal principle has never been altered. As the original political structure laid out by the 1947 Constitution is parliamentary, the president of the Republic of China was elected by the National Assembly and not by popular vote. The president can nominate the prime minister to the parliament, but cannot directly appoint him. An appointment is possible only after parliamentary confirmation. There was no stipulation about the dismissal of the premier by the president. Only a few substantive powers were granted to the president, one of which was emergency decree power. However, presidential decrees were for national emergencies and were exercisable only when the parliament could not act. They then needed to be approved by the parliament within one month of promulgation. The way in which the premier was accountable to the parliament showed signs of the framers’ intention to strengthen the administration. Parliament could not pass a motion of no-confidence to topple the

9780230_242920_11_cha10.indd 175

11/11/2010 5:19:20 PM

176 Yu-Shan Wu and Jung-Hsiang Tsai Table 10.1 The ROC’s political development and constitutional regimes Stage

Constitutional regime President

First stage (1948) Second stage (1948–1991)

Parliamentarism

Chiang Kai-shek

Presidential dictatorship

Third stage (1991–1997) Fourth stage (1997–2000) Fifth stage (2000–2008) Sixth stage (2008–)

Emerging semipresidentialism Semi-presidentialism (united) Semi-presidentialism (divided) Semi-presidentialism (united)

Chiang Kai-shek Yen Chia-kan* Chiang Ching-kuo Lee Teng-hui Lee Teng-hui Lee Teng-hui Chen Shui-bian Ma Ying-jeou

Note: *The interregnum of Yen Chia-kan was unique in that he assumed presidency after the death of Chiang Kai-shek in 1975. He served until Chiang Kai-shek’s term ended in 1978. During Yen’s presidency, Chiang Ching-kuo was premier, and the young Chiang wielded the ultimate power. That aberration from presidential dictatorship proved transitory.

cabinet. The premier could only be forced to resign when his administration and the parliament disagreed on important government policies, bills, or the budget, and when he requested the parliament reconsider its position he was turned down by a two-thirds majority. Even under these circumstances the premier had the option of bowing to the parliament and implementing its resolution. In theory, the premier could hold on to his position during his tenure no matter what the parliament did. This was obviously an imperfect method of accountability to the parliament. The ROC’s experiment with democracy came to a halt in May 1948 when the Temporary Provisions Effective during the Period of Communist Rebellion (Dongyuan kanluan shiqi linshi tiaokuan) was promulgated. The communists launched a full-scale revolt and the KMT forces retreated. One year later martial law was imposed on Taiwan where the KMT Government had fled. From 1948 to 1991 Taiwan went through a protracted period of authoritarian rule under the KMT. Although vehemently anticommunist, President Chiang Kai-shek and his son, Chiang Ching-kuo, ruled the country through a Leninist party-state. The parliamentary structure of the Constitution was suspended. Instead, there was a presidential dictatorship (see Table 10.1). The legacy of the rule of the two Chiangs is a heavy dose of strongman politics and a popular belief that the country should be led by the president who is ultimately responsible for the performance of the

9780230_242920_11_cha10.indd 176

11/11/2010 5:19:20 PM

Taiwan

177

government. People in Taiwan are also used to seeing the prime minister act as the chief lieutenant of the president, which is a natural result of combining the prime minister as head of government (from the Constitution), and the president as the ultimate ruler of the country (from the experience of authoritarian rule). In 1991, Taiwan responded to the global trend of democratization by annulling the Temporary Provisions.2 The 1947 system was formally restored and the extraordinary presidential powers annulled legally. However, no sooner had the Temporary Provisions been confined to history than Lee Teng-hui, Chiang Ching-kuo’s KMT successor and the new ROC president, began strengthening the presidency. He accomplished this by commanding the KMT’s super majority in the National Assembly to amend the Constitution by passing additional articles suffixed to the main text (xianfa zengxiu tiaowen).3 This was a period of emerging semi-presidentialism (see Table 10.1). Lee’s first major move was to institute the popular election of the president. He also restored many presidential powers that the two Chiangs enjoyed during the authoritarian years. During this process, Lee faced resistance by a minority faction in the KMT headed by Premier Hau Pei-tsun (1990–1993). The conflict between Lee and Hau was not just about power, but also about the appropriate constitutional structure of the country, with Lee favoring a strong president and Hau advocating parliamentarism embodied in the 1947 Constitution. The independence-minded Democratic Progressive Party (DPP) went along with Lee’s grand scheme, because changing the way the president was elected had a great symbolic meaning: it showed the world that Taiwan was an independent state, separate from China. In 1996 Lee was reelected by popular vote as the ninth President of the Republic of China. In 1997 President Lee made a grand swap with the DPP in the fourth amendment of the Constitution, granting the parliament power to vote noconfidence on the prime minister and the cabinet in exchange for the presidential power to appoint prime minister without parliamentary consent. According to the 1947 Constitution, the president could not appoint the prime minister without parliamentary consent (Article 55), while the parliament had no power to remove the prime minister (Article 57). Such basic rules of investiture and dismissal were not changed by the constitutional amendments of 1991, 1992, and 1994. Now the parliament lost the power to approve the presidential nomination of the prime minister, but gained the power to remove him. This swap was made possible because Lee’s handpicked vice president/premier, Lien Chan, was boycotted by the DPP, which prompted Lee to launch the fourth constitutional amendment that made the president the sole decider of prime minister. In return, the parliament gained the power of dismissal. It turned out to be a bad deal for the parliament because the president gained more than the parliament. In one stroke, dual accountability was established. The post-1997 government found itself

9780230_242920_11_cha10.indd 177

11/11/2010 5:19:20 PM

178 Yu-Shan Wu and Jung-Hsiang Tsai

accountable to both the parliament and the president, the former for the parliament’s vote of no- confidence, and the latter for the president’s power of investiture (and by extension the power of dismissal). In short, through the constitutional amendments in the 1990s, Lee was able to consolidate the president’s position, and turned Taiwan into a semi-presidential system (direct presidential election with a government responsible to parliament), of the president-parliamentary variant (the president can, in practice, remove the prime minister unilaterally) (see Table 10.1). Taiwan’s entry into semi-presidentialism was a natural result of democratizing a bicephalous system that had its roots in the 1947 Constitution and the authoritarian interlude. Prior to democratization, Taiwan had had an all-powerful president and a prime minister who was theoretically responsible to the parliament, but there were no free competitive elections for either. Democratization brought about an assertive Legislative Yuan that wanted to hold the government responsible. On the other hand, a popular election provided the president with the greatest legitimacy to rule the country. Although there were different voices even in the ranks of the KMT against Lee’s moves to aggrandize his own power, the fact that the KMT held a dominant position and Lee commanded wide social support made it almost inevitable that he would move Taiwan into semipresidentialism, just like many similarly situated post-Leninist regimes. Institutional legacy and the existence of a dominant political force with a popular presidential candidate pretty much determined the constitutional structure of the country.4

From united to divided government At Taiwan’s full entry into semi-presidentialism in 1997, it was not noticed that dual accountability was inherent in the new system. This was because a united government existed. After the December 1998 parliamentary elections, the KMT’s majority in the Legislative Yuan was beefed up from 51.8 percent of the seats to 54.7 percent (see Table 10.2). President Lee was in full control of the political scene as the ROC president/KMT chairman. The congruent relation between the president and the parliament hid the vulnerability of the semi-presidential system. The realization that Taiwan might be heading for political turbulence was limited to a few political observers and scholars. Prior to 2000, the DPP’s chance of beating the KMT in the presidential election looked bleak, as the former had never won any national race against the latter. However, the split of the KMT in the 2000 presidential election resulted in two tickets. Lien Chan, Lee’s handpicked successor, competed with James Soong, former Taiwan Province Governor and a political star, for the KMT votes. With the KMT camp divided, the DPP’s candidate, Chen Shui-bian, secured a plurality and won the March 2000 presidential election.

9780230_242920_11_cha10.indd 178

11/11/2010 5:19:20 PM

Taiwan

179

Table 10.2 Major parties’ seats in the legislative Yuan KMT Dec 1986 Dec 1989 Dec 1992 Aug 1993c Dec 1995 Dec 1998 May 2000 d Dec 2001 Dec 2004 Jan 2008

PFP

NP

59 n/a n/a 80.8% 72 n/a n/a 71.3% 102 n/a n/a 63.4% 94 n/a 7 58.4% 4.3% 85 n/a 21 51.8% 12.8% 123 n/a 11 54.7% 4.9% 114 17 8 50.7% 7.5% 3.6% 68 46 1 30.2% 20.4% 0.4% 79 34 1 35.1% 15.1% 0.4% 81 1 n/a 71.7% 0.9%

PanBluea

DPP

101 62.7% 106 64.6% 134 59.6% 139 61.8% 115 51.1% 114 50.7% 82 72.6%

12 16.4% 21 20.8% 50 31.1% 51 31.7% 54 32.9% 70 31.1% 64 28.4% 87 38.7% 89 39.6% 27 23.9%

PanTSU Greena Others n/a n/a n/a n/a n/a n/a n/a 13 100 5.8% 44.5% 12 101 5.3% 44.9% n/a 27 23.9%

2 2.8% 8 8.0% 9 5.6% 9 5.6% 4 2.4% 21 9.3% 17 7.6% 10 4.4% 10 4.4% 10 23.9%

Total 73b 100% 101b 100% 161 100% 161 100% 164 100% 225 100% 225 100% 225 100% 225 100% 113 100%

Notes: a We record Pan-Blue and Pan- Green seats and shares only when there is more than one party in the camp. b Only elected legislators are counted. c In August 1993 the New KMT Alliance split from the KMT and formed the New Party, thus changing the distribution of seats among political parties in the Legislative Yuan. d Although there were no parliamentary elections held in 2000, the presidential election of that year brought about the People First Party which claimed 17 converting legislators in the parliament, and changed the political landscape.

Suddenly Taiwan shifted from a united to a divided government (see Table 10.1). This is the Achilles heel of semi-presidentialism. All the inherent contradictions of the system were exposed and Chen faced a difficult situation. He was a minority president who won only 39.3 percent of the popular vote. Furthermore, he now faced a parliament where his own party held only 28.4 percent of the seats, compared with the KMT’s 50.7 percent and the Pan-Blue’s 61.8 percent.5 Finally, he was closely watched by Beijing and Washington. Both were extremely wary of his inclination toward independence. What Chen did was not untypical of a semi-presidential president. Unable to appoint a DPP politician to be the prime minister for fear of the negative response from the opposition and from abroad, while unwilling to surrender power to the KMT after just winning the presidential election, Chen chose Tang Fei, the defense minister of the outgoing KMT cabinet, to be his first prime minister.6 This surprise move was not taken after consulting

9780230_242920_11_cha10.indd 179

11/11/2010 5:19:21 PM

180 Yu-Shan Wu and Jung-Hsiang Tsai

with the KMT. Tang was invited to form a new government as an individual, unrelated to the KMT. However, Tang proved to be simply a transitional figure. Once Chen felt he was in full control of the situation, and when Tang’s policy concerning the fourth nuclear power plant was in variance with his, Tang was swiftly replaced by a DPP politician, Chang Chun-hsiung, in October 2000. From that time on, Chen changed his prime minister four times. All his choices, from Chang Chun-hsiung (2000–2002), Yu Shyi-kun (2002–2005), Frank Hsieh (2005–2006), Su Tseng- Chang (2006–2007), and Chang Chunhsiung again (2007–2008), were veteran DPP politicians and powerful faction leaders in the ruling party (see Table 10.3). When Chen insisted on his prerogative of unilaterally appointing the prime minister, the Pan-Blue controlled parliament failed to vote no- confidence on the government. This was the case not only when the president had the latest popular mandate while the parliament carried an older, and weaker mandate (from May 2000 through December 2001, and from March 2004 through December 2004), but also when the parliament was elected after the president, thus holding a more recent mandate than the president (from December 2001 through March 2004 and from December 2004 through May 2008). Whenever the parliament had a more recent mandate from the people right after parliamentary elections, Chen appointed a new prime minister (Yu Shyi-kun replaced Chang Chun-hsiung after the December 2001 LY elections, and Frank Hsieh replaced Yu Shyi-kun after the December 2004 LY elections). However, what Chen did was to replace a DPP politician with another one, disregarding the fact the DPP was always outnumbered by its Pan-Blue rivals in the parliament. In short, Chen appointed and dismissed prime ministers at will. The president held the premier accountable to him. For eight consecutive years, Taiwan experienced a divided government and a minority government.7 This is a clear indication that Taiwan has a president-parliamentary regime (see Table 10.3).

Reversion to congruence In 2000, the DPP’s candidate, Chen Shui-bian, won the presidential election mainly by the division of the Pan-Blue ticket and by toning down his pro-independence rhetoric. After his inauguration Chen made several accommodating moves. He appointed former KMT defense minister Tang Fei as his first premier, and extended an olive branch to the Chinese mainland. However, as he faced mounting pressure from the Blue camp, and the economic performance of his government left much to be desired, Chen shifted back to a radical line, mobilizing Taiwanese identity as a campaign instrument against the pro-unification KMT. Roughly at the same time, former president Lee split with the KMT and founded the Taiwan Solidarity Union. Lee began describing the country as “abnormal,” and demanding

9780230_242920_11_cha10.indd 180

11/11/2010 5:19:22 PM

Taiwan

181

Table 10.3 Semi-presidentalism in the ROC

President Lee Teng-hui (KMT, 1996– 2000)

Parliamentary majority Pan-Blue (1996–1998)

Presidentparliamentary relation Prime minister United United

Pan-Blue (1999–2001) Chen Shui-bian (DPP, 2000–2004)

United Divided Divided

Pan-Blue (2002–2004) Chen Shui-bian (DPP, 2004–)

Pan-Blue (2005–2007)

Divided Divided Divided Divided Divided

Ma Ying-jeou (KMT, 2009–)

Pan-Blue (2008–) United

United

Cabinetparliament relations

Lien Chan (KMT, Majority 1993–1997) Vincent Siew Majority (KMT, 1997–2000) Majority Tang Fei (KMT*, Minority 2000–2000) Chang ChunMinority hsiung (DPP, 2000–2002) Yu Shyi-kun Minority (DPP, 2002–2005) Minority Frank Hsieh (DPP, Minority 2005–2006) Su Tseng- Chang Minority (DPP, 2006–2007) Chang ChunMinority hsiung (DPP, 2007–2008) Liu Chao-shiuan Majority (KMT, 2008– 2009) Wu Den-yih Majority (KMT, 2008–)

Note: *Although Tang Fei was a leftover from the outgoing KMT cabinet, he joined the new cabinet on an individual basis.

a “rectification of the country’s name,” that is, replacing the Republic of China with the Republic of Taiwan. Lee’s quantum jump from “light Blue” to “dark Green” emboldened and actually pressured Chen to race to fundamentalism for fear of losing influence to the newly founded TSU. The 2004 presidential election thus witnessed a grand shift from the moderating, converging pattern of the 1990s to competitive radicalization. The Pan Blue’s Lien-Soong electoral coalition was forced to follow Chen in repeated demonstrations of their ultimate commitment to Taiwan. On the eve of the election, two bullets were fired at Chen and his running mate, Vice President Anette Lu, when they were canvassing in Tainan, Chen’s home county. As it turned out, Chen and Lu won the presidential election by a mere 0.2 percent of the popular vote.

9780230_242920_11_cha10.indd 181

11/11/2010 5:19:22 PM

182 Yu-Shan Wu and Jung-Hsiang Tsai

In his second term in office, Chen found himself on acrimonious terms with the Pan-Blue opposition, vehemently hated by the Communist leaders in Beijing, and disliked by most American officials for his blatant manipulation of anti- China feelings in Taiwan and provocative actions taken without prior consultation with Washington. The economy remained weak, lagging behind Taiwan’s East Asian neighbors. In 2005, the first family and Chen’s immediate entourage in the presidential office became involved in a series of corruption scandals, leading to first lady Wu Shu- chen’s indictment. The president’s own integrity was also undermined. Besieged by mounting problems, Chen again appealed to fundamentalist support. The strategy backfired and in the 2005 local elections for county magistrates and city mayors, the DPP managed to keep only six posts, compared with the KMT’s 14 (17 if all Pan-Blue victories are counted). Even with this electoral fiasco, Chen did not tone down his rhetoric. He continued dominating the party well into his second term, equating support for him with loyalty to the party and its core values. The January 2008 parliamentary elections proved voters were more concerned with bread and butter issues than with identity and changing the country’s name. The DPP now lagged behind the KMT by 14.8 percent of the popular vote. This gap was further augmented by the new electoral system when it came to seat distribution in the Legislative Yuan. Whereas in the past Taiwan used the SNTV system that provided a nice match between the vote share and seat share, the new Mixed Member Majoritarian (MMM) system with two-thirds of the seats from single member districts and one-third from proportional representation (with a 5 percent threshold), strongly favors the winner. The results devastated the DPP. It captured only 27 seats (23.9 percent) in the LY, compared with the KMT’s 81 seats (71.7 percent) (Table 10.2). The March 2008 presidential election proved no less disastrous for the DPP than the preceding parliamentary elections. Reflecting the popular mood at the time, the KMT’s candidate, Ma Ying-jeou, defeated his DPP opponent Frank Hsieh by a huge margin of 17 percent of the popular vote. What the 2008 parliamentary and presidential elections meant for Taiwan’s constitutional system was its reversion to a united government. After eight years of bitter conflict between the president and the parliament, both are now firmly in the hands of the same party. Although political observers were quick to point out the latent competition between President Ma Ying-jeou, the KMT Chairman, Wu Poh-hsiung, and Legislative Yuan Speaker, Wang Jin-Pyng, the new KMT troika, no one seriously questioned the political dominance of President Ma, whose charisma proved highly instrumental in bringing about the KMT’s grand victories. Ma’s original idea of how to operate the ROC’s semi-presidential system was quite different from that of his predecessor. The KMT had long advocated that the constitutional structure of Taiwan is a dual- executive system (shuang shouzhang zhi), and it had been highly critical of the DPP’s

9780230_242920_11_cha10.indd 182

11/11/2010 5:19:22 PM

Taiwan

183

disrespect of the majority in the parliament. During the campaign for the parliamentary elections, Ma promised that if the Pan- Green were to capture the majority in the LY, and if he was elected president, then he would respect the parliamentary majority and appoint a Pan- Green prime minister. This showed that Ma did not insist on holding the prime minister accountable to him if the KMT and its allies failed to capture a parliamentary majority. Under those circumstances, it would be hard for Ma to dismiss the premier as long as the latter enjoyed the support of the parliament. On the surface this would look like the French system during cohabitation.8 However, appointing a Green premier under divided government did not mean Ma would leave the whole cabinet to the discretion of the DPP. As the additional article 2 to the ROC Constitution stipulates that the president shall determine major policies for national security, Ma interpreted that as empowering the president to have a say on the personnel of the cabinet members directly related to national security, such as ministers of defense and foreign affairs and director of mainland affairs council, the cabinet-level agency in charge of cross-Strait relations. This meant a portion of the cabinet would still be responsible to him, even under a divided government.9 Ma’s inclination to follow a “functional division of power” between him and the premier can be further explored by the president’s behavior after winning the presidential race. In May 2008 Ma invited Liu Chao-shiuan, a veteran technocrat, to be his prime minister and to form the new KMT government. Ma then worked out with Liu who should take what position in the cabinet. During the process Ma made it clear that he should determine the ministers of foreign affairs and defense, and the director of the mainland affairs council. Liu had more say on other ministerial positions, primarily in the economic and domestic realms. During the early stages of Ma’s presidency, he thought the “functional division of power” was the appropriate formula between the president and the premier. He thought this was a correct reading of the Constitution. Ma then carved out the bulk of state affairs for the prime minister to take charge, while constrained himself to security matters. The president thought he should stand on the “second line,” leaving the premier in full control of the government. This resulted in an uproar from the people who did not know why they had elected a president who constrained himself and left everything to a prime minister whom none had expected to be in that position when Ma was elected. The constitution-abiding, self- constraining role that Ma imposed on himself was in contradiction with popular expectations of an all-powerful president who could lead the nation through challenges. Since Ma’s inauguration, skyrocketing energy prices, the international financial crisis, and most critically the devastating impact of typhoon Morakot that struck Taiwan in early August 2009 all forced Ma to take the lead in directing the government. He began frequently holding press conferences, delivering speeches, visiting areas struck by natural disasters, and giving interviews

9780230_242920_11_cha10.indd 183

11/11/2010 5:19:23 PM

184 Yu-Shan Wu and Jung-Hsiang Tsai

to show his leadership in an attempt to restore public confidence in the government amidst unprecedented challenges. Ma’s partnership with Liu was finally broken after the government was faced with mounting criticism from both domestic and international media for lack of effective actions in the wake of typhoon Morakot that left 700 dead, the worst in Taiwan’s typhoon history. Ma’s popularity plummeted to unprecedented lows. One month after Morakot struck, Ma accepted the resignation from Liu and his cabinet. Wu Den-yih, the KMT’s secretary general, was swiftly appointed by Ma as the new premier and a new government was formed. Among the members in the cabinet, Ma’s favorites filled important posts. The erstwhile functional division of power between the president and premier simply disappeared. Ma took the lead in changing the premier and reshuffling the cabinet. In October 2009, Ma further strengthened his position when he resumed the KMT’s party chairmanship that he had relinquished to Wu Poh-hsiung when he got entangled in an embezzlement case in early 2007.10 Now he became the indisputable leader of the government (replacing the premier) and the ruling KMT (assuming chairmanship). The way in which Ma was forced to stand on the front line vividly showed how popular expectations limited the choices of political actors under semipresidentialism. People in Taiwan do not accept a titular head of state. When the president is elected by popular vote, so are his campaign promises, and it takes a strong presidency to fulfill those promises. The logic of direct presidential election and the lack of parliamentary tradition make Taiwan an unlikely place where one would find the premier and government solely responsible to the parliament.

Measuring presidential powers Taiwan’s president has considerable powers. Based on the schema developed by Shugart and Carey (1992) and Shugart (1996, 1999), two categories of presidential powers can be identified: presidential legislative powers include package veto power, partial veto power, decree power, and exclusive introduction of legislation. Nonlegislative powers include cabinet formation, cabinet dismissal, censure, and dissolution of assembly. Table 10.4 enumerates all the presidential powers and how they are graded. This schema makes it possible for us to gauge the variation of the presidential powers in Taiwan. Concerning the president’s legislative powers, he did not have a direct package veto, although he could endorse the premier’s request for the parliament to reconsider bills it had passed. A two-thirds majority in the parliament was needed to override such a “veto.” After 1997 the threshold was lowered to one-half, making it easier for the parliament to override the executive’s veto. There has been no presidential partial veto, either before or after 1997. In the 1947 Constitution, the president could issue emergency decrees when the parliament was not in session, but they needed to be approved by the

9780230_242920_11_cha10.indd 184

11/11/2010 5:19:23 PM

Taiwan

185

Table 10.4 Measuring Taiwan’s presidential powers Presidential powers Legislative powers Package veto Partial veto Decree Exclusive introduction of legislation Legislative powers total Nonlegislative powers Cabinet formation Cabinet dismissal Censure Dissolution of parliament Nonlegislative powers total Total

Pre-1997 Post-1997 2 0 2 2

1 0 2 2

6

5

1 2 1 0 4

3 2 2 1 8

10

13

parliament within one month of issuance. Additional article 2 made it possible for the president to issue emergency decrees anytime, subject to parliamentary approval within ten days. Finally, only the executive could propose a budget. The parliament could not amend the budget by increasing government expenditures. On the nonlegislative powers, the 1997 amendments significantly shifted the balance in favor of the president at the expense of the parliament. Previously the president’s candidate for premier would need parliamentary confirmation for final appointment. After 1997 the president could appoint the premier without consent from the parliament. Concerning dismissal of the cabinet, the president maintained the power to dismiss the premier, before and after 1997. For censure of government, the pre-1997 system allowed for a soft version of a no-confidence vote. When the parliament disagreed with the government over important policies, the parliament could pass a resolution demanding a change of policies, and the government with presidential endorsement could request parliamentary reconsideration, which the parliament could veto with a two-thirds majority. This was a “soft” censure because the premier could either abide by the parliament’s decision or resign. After 1997 a new mechanism of accountability was put in place whereby the parliament could vote no-confidence on the government and force it to resign if the motion is carried with a simple majority. This increase of parliamentary power is countered by the president’s discretion to dissolve the parliament when a censure motion is passed. In the Shugart-Carey schema the president’s power is raised both in censure and dissolution of assembly. As shown in Table 10.4, there was only a slight change in the president’s legislative powers in the 1997 amendment, from six to five. However, the nonlegislative powers of the president increased quite considerably, from

9780230_242920_11_cha10.indd 185

11/11/2010 5:19:23 PM

186 Yu-Shan Wu and Jung-Hsiang Tsai

four to eight. In all, the amendment brought about an increase of presidential powers by three points. Among the presidential powers, the ability to hold the premier and the cabinet accountable is of particular importance. This feature makes Taiwan a president-parliamentary system. After the 1997 amendment, the presidential powers further increased. As a president-parliamentary system with increasing presidential powers, Taiwan weathered eight years of divided government from 2000 to 2008. It seems that all the negative elements converged and Taiwan’s nascent democracy would be seriously threatened. However, Taiwan managed to pass the test. Its Freedom House rating remained stable qualifying it as a free country throughout the eight years with two turnovers of power.11 How is it that president-parliamentarism with dual accountability, high overall presidential powers, and divided government and minority government fail to destabilize Taiwan’s semi-presidential system? In the following section, we explore the causes.

Explaining the abnormality It has become a consensus in the study of constitutional engineering that high presidential powers in a semi-presidential country have ominous implications for political stability and democratic survival (Shugart and Carey 1992; Shugart 1996, 2005; Parish 1998; Fish 2000; Schleiter 2003; Skach 2005; Elgie 2007b; Elgie and McMenamin 2008; Elgie 2008). In the literature presidential-parliamentarism is the institutional culprit, and dual accountability is the root cause of instability. This is particularly the case when there is a minority government appointed by the president against the opposition from the parliament. Under such circumstances the president and the parliament clash over control of the government, and the premier is torn between two supervising bodies that are in a tug of war. Policy paralysis and dire government performance ensue. In the process of overcoming this structural obstacle, the president may pursue a constitutional dictatorship. The serious conflict between the president and the parliament may result in frequent cabinet reshuffles and government collapses. Squeezed in between, the prime minister may antagonize the parliament, or displease the president.12 This type of divided government may lead to legislative paralysis, and even to democratic breakdown (Skach 2005). In nascent democracies that adopted semi-presidentialism strong presidents may thwart democratic consolidation compared with moderate or weak presidents, other things being equal (Shugart and Carey 1992; Colton 1995; Taras 1997). This consensus in the literature makes Taiwan’s democratic consolidation a particularly interesting case that requires explanation. The reason for Taiwan’s abnormality lies in the context of its semipresidentialism. Strong presidents are not a necessary or sufficient condition for democratic instability in semi-presidential countries, since they do not

9780230_242920_11_cha10.indd 186

11/11/2010 5:19:23 PM

Taiwan

187

govern in a vacuum. Presidential powers have to be considered with other factors such as parliamentary dynamics in the context of specific countries. At issue here is what kind of executive-legislative interactions in presidentparliamentary systems entail greater instability, and how do the particularities in Taiwan account for its democratic consolidation? The existing literature lacks specifics on the exact mechanisms and actual processes that lead to political instability in president-parliamentarism. The specific conditions under which a particular president-parliamentary regime operates need to be spelled out. In a president-parliamentary system, the threat to democracy may arise from three conditions. They are: the overwhelming powers of the president, excessive competition between the legislative and the executive (interbranch competition), and excessive competition between the president and the premier (intra- executive competition). In scenario one, the presidential powers may be too strong for the parliament to check and the country may slide into presidential authoritarian rule. In intra- executive competition (scenario two) there is cohabitation, with the cabinet siding with the parliamentary majority against the president. Excessive competition of this kind may cause paralysis of the administration, with the president and the premier clashing over state power. A military coup may ensue and the system may plunge into democratic breakdown. In interbranch competition (scenario three) one finds a divided government with the president and the cabinet on one side, and an opposing parliamentary majority on the other side. Excessive competition of this kind may prompt either side to resort to extra- constitutional measures to win the fight. The system may also plunge into democratic breakdown. Between 2000 and 2008 Taiwan found itself in scenario three. This is the case because the Legislative Yuan was controlled by the Pan-Blue camp, but President Chen insisted on appointing his fellow DPP members to the premier’s position. As the president was checked by an opposition- dominated parliament, the danger of overwhelming presidential powers was reduced. The country could not possibly slide into presidential authoritarian rule under the parliament’s watch. One need not be worried that overwhelming presidential powers would threaten democracy, a phenomenon typically associated with scenario one. Excessive intra- executive competition is also excluded in this case, for the president always handpicked the premiers, despite opposition from the parliamentary majority. The relationship between Chen and his premiers (a total of six) was one of hierarchical subordination. The rapid change of the premier meant that no holder of this office could be sure of his tenure, hence the lack of opportunity to build up his own power base around premiership. There was a vice president who might be the president’s successor, so the premier was by no means heir apparent to the president. This further reduced the prestige of the premier. In short, during 2000–2008 no signs of intra- executive competition were

9780230_242920_11_cha10.indd 187

11/11/2010 5:19:24 PM

188 Yu-Shan Wu and Jung-Hsiang Tsai

visible, even though infighting among the DPP’s various factions went on behind closed doors. Given that Taiwan had a divided government and minority government between 2000 and 2008, excessive interbranch competition was the most serious danger to the country’s nascent democracy. However, there was a balance between the president and the Pan-Blue- dominated parliament that kept the system from collapse. The president and parliament bickered over numerous issues. However, neither the president nor the parliament prevailed over the other. The president prevailed over the parliament on the formation of government, while the parliament had the final say on legislation. Even though both were totally dissatisfied with the situation (the president wanted his policies passed by parliament and the parliament wanted a Pan-Blue premier appointed), they settled on this equilibrium without attempting to change it for a more advantageous (and thus less desirable for the other side) solution. First of all, the president in Taiwan can appoint the prime minister without the consent of parliament according to the 1997 constitutional amendments. President Chen Shui-bian appointed six prime ministers without consulting the parliament from 2000 to 2008 during the period of minority governments. All but one of the premiers were veteran DPP politicians. Chen’s ability to have his way in handpicking premiers was directly related to the weaknesses of the Pan-Blue bloc. This presidential ascendancy was attributed to low credibility of a successful vote of no- confidence on Chen’s prime ministers. That low credibility was in turn the result of dismal electoral prospects and lack of will to fight by the opposition leadership, and high coordination hurdles for the opposition parties (Wu 2005). Specifically the KMT and the People First Party (PFP) found their popular support and seat share in the parliament locked in a zero-sum relation because of a fixed share of Pan-Blue support in the society. The two developed different attitudes toward the ruling DPP. When the PFP was strong (from 2000 to 2004), it wanted to challenge President Chen but found the KMT a reluctant partner, bruised in the previous election’s defeat and fearful of a showdown with the president. When the KMT and the PFP traded their electoral lots in 2004 (see Table 10.2), the KMT became the hardliner, while the PFP played a pivotal role, equidistant between the other two political parties. The zerosum relation between the KMT and the PFP made it possible for President Chen to drive a wedge between them and maintain presidential supremacy in appointing the premier and controlling the government. Although the president had his way in forming the government, he could not stop the parliament from retaliation in other battlegrounds. The PanBlue bloc in the Legislative Yuan resisted Chen’s policy initiatives, and threatened the impeachment of the president. Impeachment turned out to be empty words and mere political gesturing. Legislative boycotts, however, were concrete and serious. The Pan-Blue opposition in the parliament

9780230_242920_11_cha10.indd 188

11/11/2010 5:19:24 PM

Taiwan

189

enjoyed procedural advantages with its majority status. Because members of the all-powerful procedure committee that set agenda in the Legislative Yuan were distributed according to seat shares of the parties, the opposition naturally controlled the legislative agenda. Government bills were easily blocked. For those bills that sifted through the procedure committee, the opposition held the advantage when it came to vote. In the LY the ruling DPP and the Pan-Blue opposition were almost always at each other’s throat, voting down bills initiated by the opposing camp. In roll- call votes, crossparty/cross- camp collaboration was quite rare, at around 12 percent. More significantly, the cooperation between the DPP and the KMT was extremely rare, at a mere 3 percent. This pattern of mutual denial obviously worked to the disadvantage of the DPP government. In the first parliamentary session under the DPP, only 38.5 percent of government bills were passed, compared with 72.7 percent in the previous session under majority government (Sheng 2003). In the Fourth LY that straddled power transfer, the KMT majority government enjoyed a success rate of 98.3 percent in roll-call votes, compared with a mere 34.4 percent for the DPP minority government. Even when a government bill managed to get through the legislative process, the opposition could change its content and turn it into a bill that the opposition preferred. This pattern was consistent throughout the years of the DPP’s minority government. Overall, the opposition- controlled parliament in Taiwan during the period of minority government took the helm in the parliament, significantly checking the president’s power in the legislative area.13 In short, the president and the parliament during the period of minority government dominated different territories. President Chen could decide the premier and the cabinet portfolios, while the parliament dominated the legislative realm. The competition between the executive and the legislative was tense, but it gradually equalized. Neither won or lost. Compared with other cases of divided government and minority government, the lack of incentive for the opposition to challenge the president’s power in forming the government, and the compensation that the opposition could take in the legislative realm proved crucial in constraining its actions. Had there been greater incentives for the opposition to directly challenge the president over the formation of government, we could expect more violent confrontations between the executive and the legislature. The stability of Taiwan’s president-parliamentary system might then be thrown into serious doubt.

President-parliamentary systems and democratic consolidation Semi-presidentialism is a constitutional system that is easy to choose, but difficult to operate (Wu 2007b). Dual accountability in its presidentparliamentary version is a unique problem for semi-presidentialism.

9780230_242920_11_cha10.indd 189

11/11/2010 5:19:24 PM

190 Yu-Shan Wu and Jung-Hsiang Tsai

Conventional wisdom holds a pessimistic view on president-parliamentarism, linking it to government paralysis, systemic conflict, and political instability. Here we find the Achilles heel of semi-presidentialism. Minority governments make such systems even more perilous for democratic consolidation. Taiwan adopted the president-parliamentary version of semi-presidentialism in 1997, and moved into a divided government in 2000. Against the empirical and theoretical backdrop of “the perils of president-parliamentarism,” one would expect great political instability and even threat to democratic survival in this nascent democracy. However, Taiwan’s case in 2000–2008 proved otherwise, as the democratic system stabilized and consolidated. During this period of time the presidential powers were checked by the opposition- dominated parliament, the intra-executive friction was minimized by the president handpicking the premiers, and the conflict inherent in the divided government was contained by the lack of willingness of the opposition to directly challenge the president in government formation. This phenomenon is attributable to the low credibility of a successful vote of no confidence on President Chen’s prime ministers. Different electoral prospects and high coordination hurdles for the opposition parties were the main reason behind it. Even though the opposition- dominated parliament did not directly challenge the president over government formation, it successfully defeated the president’s policy initiatives by dominating the legislative realm. The executive and the legislature settled on an equilibrium, one in which neither the president nor the parliament was the total loser, or total winner. Under those circumstances, Taiwan’s presidentparliamentarism proved more resilient in sustaining a nascent democracy than the literature would expect. Taiwan became a rare case of democratic consolidation under president-parliamentarism.

Notes 1. In the chapter the term divided government means the president and the parliamentary majority are not from the same political party, and minority government means there is no majority support for it in the legislature. From 2000 to 2008 Taiwan experienced a period of divided government and minority government in which the president picked prime ministers mainly from his own party, in defiance of the majority held by the opposition in the parliament. However, this is not “divided minority government,” which, à la Skach, would indicate the situation where the president and the prime minister are from opposing parties and where neither enjoys a majority in the legislature. 2. Prior to that, martial law was lifted in July 1987, clearing the way for further measures of liberalization and democratization. 3. Those amendments are not permanent ones, as shown by the foreword “To meet the requisites of the nation prior to national unification, the following articles are added or amended to the ROC Constitution.” 4. See Chapter 2.

9780230_242920_11_cha10.indd 190

11/11/2010 5:19:24 PM

Taiwan

191

5. The Pan-Blue camp is composed of the KMT and those that split from it. On Taiwan’s political spectrum the Pan-Blue are pro-unification, while the PanGreen led by the DPP are pro-independence. 6. This reminds one of François Mitterrand’s attempt to invite Jacques ChabanDelmas to form a government in 1986 when the president knew that he would have to cohabit with a right-wing politician, but loathed the possibility that Jacques Chirac would become the prime minister. Chaban-Delmas was a Gaullist whom Mitterrand could accept. A similar case happened in Niger. When President Mahamane Ousmane faced a parliament controlled by the opposition, he refused to appoint the opposition’s nominee for prime minister, Hama Amadou, and chose another opposition politician, Amadou Boubacar Cissé, a member of MNSD-Nassara, without that party’s consent. Cissé was blocked by a vote of no- confidence by the National Assembly, and expelled by his party. In all three cases, France in 1986, Niger in 1995, and Taiwan in 2000, one finds a reluctant president trying to find a lesser evil to become the prime minister when facing a parliament controlled by the opposition. 7. For a comparison with cohabitation in France, see Cabestan (1997), and Wu (2000b). 8. For the French institution of cohabitation, see Poulard (1990), Pierce (1991), Elgie (2002). 9. A similar situation can be found in Poland from 1990 to 1997. According to the 1992 Little Constitution the president was given the power to determine (or to be consulted) over the appointment of the ministers of internal affairs, defense, and foreign affairs. It translated to “presidential ministers” in the cabinet led by a prime minister whom the president was forced to cohabit with. The 1997 Constitution removed that presidential power. 10. Institutionally the KMT’s chairman wields the greatest power in the party. It has always been the case throughout the eras of Chiang Kai-shek, Chiang Chingkuo, Lee Teng-hui, and Lien Chan. Ma succeeded Lien after the first popular vote for the KMT chairman held in 2005. However, because Ma was indicted by the prosecutor in February 2007 in an embezzlement case, he swiftly resigned from his party post and at the same time declared candidacy for the KMT’s presidential nomination. Ma’s political ally Wu Poh-hsiung then took over to become the new KMT chairman. After winning the 2008 presidential election, Ma was not in direct control of the ruling party, breaking the tradition of the KMT. Without the liaison of the party, the relationship between the president and the parliamentary majority was kept distant, adding to the difficulty of coordination. 11. Taiwan’s Freedom House average rating for Political Rights and Civil Liberties was 2 in 1996, qualifying the country as “free” for the first time. The rating then fluctuated between 1 and 2 up until 2010, keeping Taiwan in the “free” territory (the cutoff point between “free” and “partly free” is 3). 12. When the prime minister received more support than the president, the president’s jealousy would justify him to sack the prime minister. In 1998 President Boris Yeltsin of Russia discharged Prime Minister Viktor Chernomyrdin with widespread popularity to show his supreme authority. See Wu (2000a: 93). 13. In several landmark cases, such as working-hour bill, fourth nuclear power plant dispute, and referendum law, the Pan-Blue opposition scored victory and forced the government to accept the opposition’s version of the bill.

9780230_242920_11_cha10.indd 191

11/11/2010 5:19:24 PM

11 Disintegrated Semi-Presidentialism and Parliamentary Oligarchy in Post-Orange Ukraine Kimitaka Matsuzato

This chapter focuses on what can be termed “disintegrated semipresidentialism,” a system in which coordination between the president and parliament is almost absent, based on a case study of Ukraine. In contrast to the other extreme of semi-presidentialism, which is highly presidentialized semi-presidentialism (see Chapter 11),1 “disintegrated semi-presidentialism” has not attracted the scholarly attention it deserves. A reason for this lack of study is that usually this is no more than a transitional, uncertain conjuncture of power, rather than a viable regime type, and, therefore, there cannot be many examples. Indeed, a number of post- communist countries underwent this uncertainty in their departure from a Soviet pseudoparliamentary regime, but chose between president-parliamentary and premier-presidential regimes relatively quickly, thus raising the integrity of the regime (Figure 11.1). Ukraine’s peculiarity is that it consolidated “disintegrated semi-presidentialism” with its 1996 constitution, and the constitutional amendment in December 2004, in the midst of the Orange Revolution, barely improved this situation. A major reason for Ukraine choosing “disintegrated semi-presidentialism” was the compromise characteristics of Ukrainian politics. For this nascent nation state, Yeltsin’s scenario in 1993, as a result of which a presidentparliamentary constitution was adopted in the smoldering ashes of the shelled parliament (the Supreme Soviet), was inconceivable. Conscious of their inexperience in state building, the Ukrainian elite are more willing to compromise than their Russian counterparts. Because of the natural time lag in political development in these countries, Ukraine could learn from Russia’s failures. However problematic he was, Kuchma neither shelled the Supreme Rada (parliament), nor invaded Crimea with armed forces. A result of this admirable avoidance of bloodshed was, according to Yulia Tymoshenko’s brilliant analogy, the “IKEA catalogue” syndrome (solutions satisfying all, but lacking any institutional logic to sustain efficiency) (Tymoshenko 2007). 192

9780230_242920_12_cha11.indd 192

11/11/2010 5:20:04 PM

Ukraine: Disintegrated Semi-Presidentialism 193 Integrity Presidentparliamentary regime

Premierpresidential regime

Strong legislature

Strong president Russia 1991–93 Armenia 1991–95 Kazkhstan 1991–95 Lithuania 1991–98 Poland 1990–97 Ukraine 1991–96

Disintegrated premierpresidental regime

Constitutional Amendment 2004

Constitution 1996 Disintegrated presidentparliamentary regime

Absence of integrity Figure 11.1

President and legislatures under post-communism

Since 2000, a number of Commonwealth of Independent States (CIS) have attempted to make their political regimes more stable by remodeling the dynamic, but unpredictable “patronal presidentialism” (Hale 2006) that operated in the 1990s.2 This occurred more often in Ukraine than in other CIS countries because its constitutional amendment in 2004 was nothing but a preemptive measure to limit the prerogative of the expected winner, Viktor Yushchenko, and neutralize the impact of the Orange Revolution. However, as Ukraine’s political history from 2005 to 2009 demonstrated, President Yushchenko never became a passive conveyer of the parliamentary will. This paradox seems to have universal significance; the transition to a premierpresidential regime from a president-parliamentary regime, as a rule, tries to minimize the president’s role in premier appointment and cabinet formation, so that this process depends on electoral results and parliamentarians’

9780230_242920_12_cha11.indd 193

11/11/2010 5:20:04 PM

194

Kimitaka Matsuzato

initiative.3 Nevertheless, the president continues to play an independent role in the appointment of the prime minister simply because parliamentary elections do not necessarily produce a definite “will” and majorities take shape in various ways. The activation of party politics as a result of the transition to a premier-presidential regime from a president-parliamentary regime increases the opportunities for the president to maneuver. I have described this paradox of the premier-presidential regime, focusing on Lithuania (Matsuzato and Gudžiskas 2006) and this study is a continuation of this idea. This chapter analyzes why the 1996 constitution consolidated “disintegrated semi-presidentialism;” describes the tendency toward parliamentary oligarchy in Ukraine on the eve of the Orange Revolution; and examines why this revolution failed to create an effective system of checks and balances. The final section argues that Ukraine’s shift toward a premier-presidential regime did not limit the president’s political independence and active commitment to cabinet formation.

The emergence of disintegrated semi-presidentialism in Ukraine In the 1990s, the CIS countries, with only two exceptions,4 adopted semipresidential constitutions with a strong presidency, which was a natural evolution of the Soviet executive diarchy between the Central Committee of the Communist Party and the government (Matsuzato 2005; Huskey 2007).5 Semi-presidentialism at the union level6 was automatically transplanted into Ukraine. On June 19, 1991, the Supreme Rada confirmed the document titled “Concept of the New Constitution of Ukraine,” drafted by the constitutional commission, which had started its work in August 1990, immediately after Ukraine’s declaration of sovereignty. Unless we count the Constitutional Accord of June 10, 1995, this was the most “presidentialized” document ever proposed in Ukraine before 1996; the prerogative of the prime minister was minimal and the president almost directly bore responsibility for the work of the cabinet ministers, while he did not have the right to dissolve parliament (Matsuzato 2005: 49).7 On December 1, 1991, presidential elections were held simultaneously with a national referendum to request Ukraine’s independence; Leonid Kravchuk became the first Ukrainian president, gaining 61.6 percent of the vote. In February 1991, Kravchuk persuaded the Supreme Rada to amend the existing constitution so the president could propose to parliament a candidate for prime minister and seven leading ministers for confirmation (Wilson 1997: 75). At that time, this “right to propose” was thought of as the president’s newly obtained prerogative, not an obligation, so the president’s right to dissolve parliament in case of its disobedience was out of the question. By 1995, Ukraine remained as the only post-Soviet country that had not adopted a new constitution, which made the viability of this country

9780230_242920_12_cha11.indd 194

11/11/2010 5:20:05 PM

Ukraine: Disintegrated Semi-Presidentialism 195

questionable. The deep political and economic crises that Ukraine faced forced the Supreme Rada to sign the Constitutional Accord with President Kuchma on June 10, 1995, which minimized parliament’s commitment to the appointment of a prime minister and the formation of his cabinet (Birch 2008: 223). This document set a time limit (“a year,” namely by June 1996) for the adoption of the Ukrainian Constitution. Kuchma tried in vain to have a constitution modeled after this Constitutional Accord. Even during the year in which this accord was effective, Ukrainian prime ministers Evhen Marchuk and Pavlo Lazarenko were politicians independent from Kuchma and enjoyed parliamentary support by their own authority. Moreover, the highly presidentialized semi-presidentialism that the Constitutional Accord requested contradicted the existing loose semi-presidential practice that had been consolidated during from 1992 to 1995. A major reason that Ukrainian judicial specialists resisted Kuchma’s attempt to make the Constitutional Accord the basis of the new constitution was that such fundamental changes in state administration would be unbearably costly (Kozyubra 1997: 12). The time limit that the Constitutional Accord set for the adoption of the constitution activated its preparation. Kuchma and Parliamentary Chair Oleksandr Moroz from the Socialist Party (SP) agreed to introduce a working group based on professional (nonpartisan) principles. The working group published a constitutional draft on November 15, 1995. Both Kuchma and Prime Minister Lazarenko were dissatisfied with the draft and wanted more prerogatives than outlined in the draft. So were the leftist parties, who traditionally advocated a parliamentary system. However, it was obvious that any attempt to revise the draft would cause an even stronger counteraction from other actors. Under this peculiar political situation, a small parliamentary group guided by Mykhailo Syrota took the initiative to revise the draft. Syrota’s group actively exploited “IKEA catalogue” tactics; strengthening the president’s prerogative, particularly in cadre appointment, Syrota practically deprived the president of the right to dissolve the Supreme Rada.8 This principle of “mutual nonintervention” made a compromise between the president and parliament possible, and the Supreme Rada adopted the constitution on June 28, 1996. The lack of any presidential prerogative to dissolve parliament even when parliament rejected the presidential candidate for prime minister was quite exceptional for CIS countries and undoubtedly benefited the development of parliamentarism in Ukraine, while the president enjoyed more freedom in his cadre policies in comparison with the Russian president. Above all, the Ukrainian president continued to appoint regional governors, while Russia’s governors were popularly elected from 1996 to 2004.9 In 1996, the president and parliament of Ukraine compromised on the condition that they would not affect each other’s status and competences. The lack of integrity of the Ukrainian semi-presidentialism became even more extreme because of the negligence of the president in exercising his constitutional obligations. For

9780230_242920_12_cha11.indd 195

11/11/2010 5:20:05 PM

196

Kimitaka Matsuzato

example, Kuchma neglected to sign laws that he disliked even when parliament overrode Kuchma’s veto. Because of this behavior, many important constitutional laws (necessary for the concretization of constitutional provisions), for example, the Law on the Cabinet of Ministers, were shelved. The lack of constitutional laws intensified the institutional disintegration.

The developing role of parliamentary oligarchs in the late Kuchma period Specific concepts, such as “parliamentary majority” and “imperative mandate,” laid as a foundation of the 2004 constitutional amendments, were a product of the painful development of Ukrainian parliamentarism. Even as a result of the 1998 parliamentary elections when the Communist Party (CP) was at its zenith, leftists were far from the majority. The total of the Communist, Socialist, and Progressive Socialist deputies was only 171 out of 450 deputies. Even by allying with Hromada, an oligarchic opposition guided by Pavlo Lazarenko (24 deputies), they could not gain a majority. According to the electoral law at that time, half of the deputies were elected from single-mandate districts and, as a natural result, nonaffiliated deputies composed “the largest party” (145 deputies).10 Under this situation, President Kuchma easily formed temporary majorities for different bills and issues. For the development of parliamentarism in this unfavorable institutional environment, it was necessary to release parliament from its leftist leadership and prevent the president’s intervention in majority formation. The first example of this attempt was the so- called Velvet Revolution on January 24, 2000 (Zerkalo nedeli 2000). After his own extremely dirty reelection in 1999, Kuchma had no alternative but to appoint the president of the National Bank, “Mr. Hryvna,” Viktor Yushchenko, as prime minister of Ukraine. “Having lived for nine years amorally” (Yushchenko 2000), Ukraine was on the brink of becoming a failed state. The astronomical foreign debt, on which Ukraine existed, needed to be liquidated in the next few years. To restructure the budget, Yushchenko needed a parliamentary majority that would support the retrenchment policy, obviously unpopular among the voters. At the end of 1999, 11 right and centrist parties with about 240 deputies (more than half of all deputies) agreed to create a pro-government majority. On January 24, 2000, this majority split parliament to cast out its chairman, Oleksandr Tkachenko, and other leftist leaders and eventually forced the left-wing deputies to support Yushchenko’s budget. Although the glorification of this event as the “Ukrainian Velvet Revolution” is questionable, this event injected the term “parliamentary majority” into the political jargon of Ukraine. During the same months, Kuchma was obsessed with organizing a referendum to transform Ukraine’s loose semi-presidential regime into a more presidentialized one. This referendum, conducted in April 2000, was excused as consultative but remained

9780230_242920_12_cha11.indd 196

11/11/2010 5:20:06 PM

Ukraine: Disintegrated Semi-Presidentialism 197

illegitimate for European institutions. Foreign observers interpreted and continue to interpret the thrust for the parliamentary majority to support Yushchenko and Kuchma’s referendum as two sides of the same coin (Birch 2008: 232–233). In fact, Kuchma resisted the formation of a parliamentary majority for fear of making Yushchenko more autonomous,11 while many members of this majority were critical of Kuchma’s referendum. Although somewhat paradoxical in hindsight, the next example of the formation of a nascent parliamentary oligarchy restricting presidential prerogative was Speaker Volodymyr Lytvyn’s resistance to Kuchma’s attempt to change the constitution. In 2002, Kuchma began to think that the victory of Yushchenko, who had shifted to the opposition after being discharged from the post of premier in May 2001, in the coming presidential election was inevitable. Kuchma reversed his earlier super-presidential position and proposed a constitutional reform to strengthen the prime minister’s power at the cost of the president (Protsyk 2003: 1087). This proposal was unacceptable not only for Yushchenko’s Our Ukraine camp, which counted on its leader’s victory, but also for Kuchma’s successor, Viktor Yanukovych, who did not share his chief’s pessimism and did not wish to curtail his own future competence. Nevertheless, Kuchma urged pro-presidential parties to prepare a radical constitutional amendment. The draft proposed that the prime minister and cabinet of ministries be nominated by parliament, not the president. Second, parliament would be obliged to form a parliamentary majority, otherwise it would be dissolved. In other words, the formation of parliamentary coalitions shifted from the sphere of politics to constitutional law. Third, Ukrainian elections would shift to an entirely proportional system combined with a system of so- called imperative mandate. According to this system, deputies who moved from one faction to another, or who just lost membership to their original faction, were deprived of their MP status.12 Thus, instead of the president, the parliamentary oligarchs would appoint the prime minister, form the cabinet of ministers, and control parliamentarians. At that time, many Ukrainian parliamentarians and jurists were critical of Kuchma’s attempt. Mykola Kozyubra, an architect of the 1996 constitution, expressed his misgivings that the fundamental amendments to the constitution adopted only eight years ago might discredit constitutionalism in general in the eyes of the Ukrainian public. He added that Moldova did not become more democratic after it returned to a parliamentary regime in 2000.13 Mykhailo Syrota remarked that “only between 14 and 19 out of 49 constitutional laws” had been adopted during the 8 years since the adoption of the 1996 constitution. Fundamental amendments to the constitution in such a situation would create additional legal chaos.14 Many jurists and political scientists remarked on the danger of changing the constitution in haste on the eve of the presidential elections. There were supporters of the reform, too. For example, Petro Martynenko, a former judge of the

9780230_242920_12_cha11.indd 197

11/11/2010 5:20:06 PM

198

Kimitaka Matsuzato

Constitutional Court, argued that the reform was “a step forward” since it would transform the present incomplete semi-presidential system into a fuller semi-presidential one.15 Understandably, Yushchenko’s Our Ukraine and the Bloc of Yulia Tymoshenko (BYuT) protested against this reform proposal, while Kuchma succeeded in attracting Socialist and Communist deputies by the adoption of a purely proportional electoral system (which is, as a rule, advantageous for disciplined parties; in the Ukrainian context, left-wing parties and BYuT). On the other hand, the introduction of the purely proportional system disappointed a number of deputies of the pro-presidential Party of Regions (Yanukovych’s party), which was composed of local bosses for whom the existing single-member constituency system was advantageous. On April 8, 2004, this proposal failed to gain the necessary support of two-thirds of the deputies because of absenteeism or abstention of six deputies from the Party of Regions and similarly five deputies from the pro-presidential faction, Labor Ukraine (Laboratoriya 2004). Kuchma assumed that Parliamentary Speaker Volodymyr Lytvyn had organized this betrayal among pro-presidential parties, and relations between Kuchma and Lytvyn became tense. On September 3, 2004, Lytvyn proposed that parliament play an active role in making the coming presidential elections democratic and clean. Newspapers reported that this sudden shift of pro-presidential factions to a more democratic, proYushchenko position was motivated by discontent with the economic and property distribution policies in the late Kuchma period, the beneficiaries of which were becoming all the smaller circle of people (Ternavs’kyi 2004).

What changed? The constitutional amendments on December 8, 2004 On December 8, 2004, in the parliamentary building, surrounded by thousands of demonstrators and picketers who were oblivious to the minus-twenty- degree- centigrade temperature of the Kyivan winter, parliamentarians reached a compromise whereby presidential elections would be repeated under “fairer conditions.” As an equivalent for this concession, the former pro-Kuchma deputies requested realization of the constitutional amendment described above. Parliament adopted an amendment law and Kuchma signed it immediately. Personally, Yushchenko was against this compromise, but leaders of his Our Ukraine party persuaded him (Sylina et al. 2004), while Tymoshenko claimed that this amendment was Kuchma’s attempt to stay in power (Tymoshenko 2004). The situation was similar to that of June 1996. No actor enjoyed influence strong enough to force his will on others, or block other actors’ initiatives alone, but the parliamentarians were not able to postpone the decision; they needed to adopt a compromise decision urgently to calm down the demonstrators surrounding the parliament. This is why the “IKEA catalogue” approach appeared again.

9780230_242920_12_cha11.indd 198

11/11/2010 5:20:06 PM

Ukraine: Disintegrated Semi-Presidentialism 199

From a legal point of view, however, the amendment on December 8 doubly violated the procedure requested by the Ukrainian Constitution. First, according to Article 158 of the constitution, once parliament has voted against a constitutional amendment, the same amendment may not again be proposed to parliament for a year, but the amendment law adopted on December 8 was very similar to the bill rejected in April. Second, bills on constitutional amendments can be proposed to parliament only with the “conclusion” of the Constitutional Court on its constitutionality (Article 159). There was no “conclusion” because the bill on December 8 was a result of a quick deal among parliamentary factions. The amendment, which was scheduled to go into effect after the parliamentary elections in March 2006, introduced a peculiar form of premierpresidentialism. First, it drastically changed the procedure of appointing the prime minister. The 1996 constitution authorized the president to appoint the prime minister, who needed to gain parliament’s support to be appointed. According to the 2004 amendments, parliamentary factions, within a month of the start of a new parliament or after the demise of the previous majority coalition, are to form a majority coalition, to which more than 226 deputies should belong. Otherwise, the president will dissolve parliament. This coalition proposes to the president a candidate for prime minister (Article 83) and within fifteen days of receiving the coalition’s proposal, the president proposes to parliament a candidate for prime minister (Article 106–9) and parliament appoints the prime minister.16 This system of premier appointment is perhaps the most arguable element of the 2004 amendment. Informed observers and political scientists are aware that a significant portion of coalition governments in Europe have been minority coalitions. The most significant merit of coalition politics is its capability to provide multiple options in difficult situations and thus stabilize democracy. In Ukraine, contrarily, the constitution obliges parliamentarians to form a majority coalition and to artificially close off various options.17 Moreover, this provision eventually obliges the voters (supposedly the source of sovereignty) to vote in a certain way; unless they vote to enable the formation of one or another majority, they will face an unstable parliament and political crisis. It is not difficult to identify the parliamentarians’ purpose bestowed on this problematic constitutional requirement. Even in semi-presidential systems with a strong parliament, the president may continue to influence cabinet formation simply because parliamentary elections often result in a lack of an obvious majority. In this case, there will be room for the president’s personal judgment. Given this, a way to minimize the president’s influence on premier appointment and cabinet formation is to oblige parliamentarians to form a majority. The 2004 amendment introduced the imperative mandate (Article 81), which the Venice Commission criticized as a violation of “European standards.” In a liberal democracy, parliamentarians are responsible to the electorate, not their party leaders, and the proportional electoral system does

9780230_242920_12_cha11.indd 199

11/11/2010 5:20:06 PM

200

Kimitaka Matsuzato

not change this principle. However, similar to the concept of parliamentary majority, the imperative mandate was a device to develop party politics artificially in the unfavorable environment of the late Kuchma period, even at the cost of the free expression of popular will.18 The amendment strengthened parliament’s position by prolonging the term of parliament from four to five years, thus identical to that of the president, and by subordinating the term of the prime minister and cabinet of ministers to that of parliament, whilst the 1996 constitution had subordinated the term of these offices to the president. In addition, Kuchma often declined to sign bills that he disliked, even if parliament had overridden his veto by a two-thirds majority. The 2004 amendment, in such cases, authorizes the parliamentary chairman to sign the bill to make it effective (Art. 94). To compensate the loss of presidential prerogative in cabinet formation, the amendment supplemented the president’s prerogatives in other spheres of checks and balances with parliament. 1. According to the 1996 constitution, the president appointed ministers, considering the prime minister’s proposals (Kuchma often ignored this provision, appointing ministers without consulting with the prime minister). According to the 2004 amendment, parliament appoints the defense and foreign affairs ministers according to the president’s proposal, and appoints other members of the cabinet according to the prime minister’s proposal (Art. 114). The existence of the two “presidential ministers” makes it possible for the president to use these posts to influence or upset the work of the cabinet of ministers, as happened in Poland under President Walesa. The Venice Commission criticized this provision, too. 2. The 2004 amendments significantly increased the possibility for the president to dissolve parliament. In addition to the power to dissolve parliament if the plenary meeting of the legislature does not start work within thirty days after the beginning of the regular season, the president may dissolve parliament if a majority coalition does not take shape according to Article 83 or if parliament cannot form a new cabinet of ministers within 60 days of the end of the previous cabinet. 3. Because the president lost the prerogative to appoint or discharge the prime minister and ministers, he instead may propose to parliament a vote of no- confidence against the cabinet of ministers. As noted above, the Venice Commission criticized the provisions on the parliamentary majority, imperative mandate, “presidential ministers,” and eventually the fundamental ideas of this constitutional reform. Venice Commission member Sergio Bartole, a professor at Trieste University, stated that this reform aimed to create a “terrible monster” of an intra-parliament

9780230_242920_12_cha11.indd 200

11/11/2010 5:20:07 PM

Ukraine: Disintegrated Semi-Presidentialism 201

“corporative majority” aloof from the public (Bartole 2007). The vice secretary of the commission, Thomas Markert, remarked that parliament was able to remove any minister at any time, while the president continued to exert executive functions; sandwiched between these organs, it was difficult for the government to maintain stability and work effectively. Markert also recommended abolishing the pure proportional electoral system (Markert 2008). EU high representative for the Common Foreign and Security Policy, Javier Solana, remarked that the 2004 amendments “did not spell out sufficiently clearly the division of competences between various leaders” (Solana 2007). During the preparation of the 1996 constitution, Ukrainian jurists and parliamentarians vigorously studied foreign constitutions and constitutional practices. This constitution enjoyed a positive reputation among foreign specialists and European institutions for several years (until the exposure of negative phenomena in the late Kuchma period), since it was comparatively democratic among the constitutions of CIS countries. In contrast, the 2004 amendment assumed an air of provincialism. The terminology used in this amendment, such as parliamentary majority and imperative mandate, had evolved in a purely Ukrainian context. Although Ukrainian specialists had become much more experienced and knowledgeable than they had been ten years ago, they were not mobilized this time to survey foreign experiences.19

Parliamentary oligarchs and the president after the Orange Revolution The 2004 constitutional amendment drastically raised the status of the prime minister both judicially and politically. This is in part because in post- Orange Ukraine the most plausible pretenders to the presidential post are the most influential parliamentary oligarchs or party leaders: Viktor Yushchenko (incumbent president) as leader of Our Ukraine (OU), Yuliya Tymoshenko as leader of the party named after herself (Bloc of Yuliya Tymoshenko: BYuT), and Viktor Yanukovych, leader of the Party of the Regions. In the premier-presidential regime, it is not difficult for Tymoshenko and Yanukovych to become the prime minister even as a result of their modest electoral success. Out of the 57 months from the Orange Revolution to October 2009, these two figures occupied the premiership for 46 months (see Table 11.1), leaving only 11 months to a technocrat obedient to the president, Yurii Yekhanurov. By comparison, in Ukraine after independence until the Orange Revolution, potential pretenders to the presidential post (Kuchma, Marchuk, Lazarenko, Yushchenko, and Yanukovych) occupied the premier post for 76 months, while grey figures obedient to the president (Fokin, Zvyahil’s’kyi, Masol, Pustovoitenko, and Kinakh) served as premier for 77 months. It was almost

9780230_242920_12_cha11.indd 201

11/11/2010 5:20:07 PM

9780230_242920_12_cha11.indd 202

11/11/2010 5:20:07 PM

Maneuver for formation Intra-cabinet conflict (Tymoshenko vs. Petro Poroshenko)

Reasons for collapse

OU declared its move to oppose Yanukovych.

Confrontations between the president and prime minister

BYuT, OU-PSD, Parliamentary Chair BL Lytvyn’s initiative

The BYuT, CP, and PR voted for the Tymoshenko did not accuse Russia after presidential impeachment law and the South Ossetian War. for increasing the prime minister’s prerogative at the cost of president. Confrontation around the pre-term parliamentary elections

Legend: OU: Our Ukraine; BYuT: Bloc of Yuliya Tymoshenko; SP: Socialist Party of Ukraine; PR: Party of Regions; CP: Communist Party of Ukraine; PSD: People’s Self-Defense; BL: Bloc of Lytvyn.

12/08 -

9/08-12/08 BYuT

Tymoshenko, 12/07-9/08 BYuT, OU-PSD Yuliya

The PR and BYuT voted for the Law on the Cabinet of Ministers (Jan. 2007). On April 2, 2007, Yushchenko declared his intention to dissolve parliament. Pre-term parliamentary elections in September 2007 (PR 175, BYuT 156, OU 72, CP 27, BL 20) 9/07-12/07 PR, CP The SP failed to pass the threshold in the parliamentary elections.

Yushchenko signed Yanukovych, 8/06-10/06 Anti-Crisis Viktor Coalition (PR, the “Universal of National Unity.” SP, CP) + OU 10/06-9/07 PR, SP, CP

Yushchenko’s deal with the PR Parliamentary elections in March 2006 (PR 186, BYuT 129, OU 81, SP 33, CP 21) 3/06-8/06 Tymoshenko could not become the prime minister because of the SP’s “betrayal.”

OU, SP, PR

9/05-3/06

Yekhanurov, Yurii

Ruling coalition

OU, BYuT, SP

Office term

Tymoshenko, 2/05 -9/05 Yulia

Prime minister

Table 11.1 Cabinets of ministers in Ukraine, 2004–2009

Ukraine: Disintegrated Semi-Presidentialism 203

an established tradition before the Orange Revolution that the Ukrainian president, both Kravchuk and Kuchma, appointed strong figures (his own potential rivals) as prime minister when he faced serious economic crises and replaced them with obedient practitioners once the crisis had been overcome to some extent (Protsyk 2003: 1080, table 1). The 2004 constitutional amendment politicized the premiership by making it ex officio the president’s rival. Raising the prime minister’s status significantly, the 2004 amendment did not provide mechanisms for coordination between the president and the prime minister. The president lost the “first-move advantage” (Protsyk 2005: 724–726) in cabinet formation, but continues to exercise enormous prerogatives in the appointment of high state officials, above all, regional governors. Since the 2004 amendment gave the president the power to nominate a candidate for foreign minister, Yushchenko believed that diplomacy continued to be the president’s sphere of competence, whilst during his term as prime minister, Yanukovych delivered active speeches regarding the NATO accession. The lack of a constitutional mechanism for the division of labor between the president and the prime minister resulted in yearly political crises in Ukraine. In 2006, it took more than four months for Yanukovych to be appointed as prime minister after his victory in the parliamentary elections in March. On April 2, 2007, President Yushchenko issued a decree to dissolve parliament on the pretext that a number of deputies, elected from OU and the BYuT, had in practice moved to the anti- Orange coalition. Yushchenko regarded this as usurping the people’s will expressed in the 2006 elections, as well as violating the “imperative mandate” provision of the constitution, although the constitution does not empower the president to dissolve parliament for such reasons. The following month, Yushchenko unconstitutionally tried to discharge the procurator general, who stood on parliament’s side. This issue almost resulted in armed conflict between the Ministry of Internal Affairs’ troops and the presidential guards on May 25 (Kievskie vedomosti 2007). In August 2008, Prime Minister Tymoshenko did not accuse Russia over its intervention in the second South Ossetian War, and Yushchenko’s group criticized Tymoshenko for being a traitor. Tymoshenko exposed the presidential secretariat as arranging exports of weapons to prewar Georgia, and that Georgia’s payments for these weapons had “disappeared.” The lack of a coordinating mechanism for the executive dyarchy has become a pretext for political speculation. A striking example was Tymoshenko’s position vis-à-vis this issue, changing almost every year. In January 2007, she allied with then-prime minister Yanukovych to adopt the Law on the Cabinet of Ministers, which aimed to strengthen the prime minister at the cost of the president even further than proposed in the 2004 amendments.20 After the preterm parliamentary elections in September

9780230_242920_12_cha11.indd 203

11/11/2010 5:20:08 PM

204

Kimitaka Matsuzato

2007, the BYuT and Yushchenko’s OU-PSD (People’s Self-Defense) formed a winning coalition to make Tymoshenko the prime minister. This coalition’s first work was to amend the law on the Cabinet of Ministers to the advantage of the president and eventually replaced it with a new law in May 2008. Four months later, after coming into conflict with Yushchenko over the Georgian issue, Tymoshenko allied with the Party of the Regions and the CP and voted for increasing the prime minister’s prerogative at the cost of the president’s. Both the Orange and anti- Orange blocs dissolved quickly and all cabinets of ministers collapsed, or attempts to form a cabinet failed due to “betrayals” within these blocs. President Yushchenko found plenty of opportunity to use his power to nominate the prime minister because of the absence of a stable coalition. For example, Tymoshenko’s first postOrange government collapsed because of intra- cabinet conflicts between Tymoshenko and Petro Poroshenko (then-secretary of the Security and Defence Council). Yurii Yekhanurov’s next cabinet was confirmed by parliament only because Yushchenko made a deal with the Party of the Regions. In addition, after the parliamentary elections in March 2006, Tymoshenko tried to create a coalition rallying the BYuT, OU, and SP (the leftist flank of the Orange bloc). This attempt failed because the SP chose to ally with the Party of the Regions and the CP, which granted their leader, Oleksandr Moroz, the position of parliamentary speaker. Yushchenko cooperated with this anti- Orange coalition by signing the Decree (“Universal”) of National Unity and proposed Yanukovych as the candidate for prime minister. Thus, the 2004 constitutional amendment did not make the Ukrainian president a passive conveyer of parliament’s will. An institutional expression of this independence was the presidential secretariat (previously known as the presidential administration), which began to play an independent, party-like role. In September 2006, immediately after Yanukovych formed his government, Yushchenko appointed his old comrade from the National Bank period, the former Transcarpathian governor and one of the founders of OU, Viktor Baloha,21 head of the presidential secretariat. Baloha purged Yushchenko’s dilettante entourage, the remnants of the Orange Revolution, and launched a hard line against the anti-presidential forces. The quick decline of Orange and anti- Orange coalitions was at least partly for personal reasons. At least until 2008, Yushchenko and Yanukovych had a strong tendency toward grand coalitions, regarding Tymoshenko, not each other, as the most formidable rival in the coming presidential election. A most striking example was Yanukovych’s “life-saving” help delivered to Yushchenko in the midst of the constitutional crisis of 2007. Yanukovych and Speaker Moroz salvaged Yushchenko from this crisis that he caused by “dissolving” parliament; the three leaders agreeing to hold preterm parliamentary elections in September 2007.

9780230_242920_12_cha11.indd 204

11/11/2010 5:20:08 PM

Ukraine: Disintegrated Semi-Presidentialism 205

The bipolar constellation of the Orange and anti- Orange coalitions in 2004 proved to be ephemeral because this was not supported by any social cleavage, while observers tend to exaggerate the renowned West/East split of Ukraine. Sociological surveys show that Ukrainian voters care about issues such as employment, welfare, education, and honest politicians, as is the case with homogeneous countries, but do not necessarily count language, religion, NATO accession, and other West/East issues as important when choosing who to vote for.22 In the electoral politics of Ukraine, the West/ East axis emerged only when socioeconomic or other axes, for some reason, were obscured (Matsuzato 2001). During the Orange Revolution Ukrainian elites relied upon the West/East axis for popular mobilization, but different tactics can be observed in the 2009–2010 presidential campaign. In TV debates and other campaigns Yanukovych’s camp often put forward brilliant Ukrainian-speaking female politicians, while Tymoshenko’s camp put forward male Russian-speaking politicians. The desire to penetrate the opponent’s electorate is a normal democratic practice and it was obvious that if the candidates mobilized only their own electorate, then the stalemate in 2004 would be repeated. Lijphart has argued that the religious or religious/secular cleavage is effective for mobilizing the vote, but much less relevant than the socioeconomic cleavage in coalition making (Lijphart 1982). Here, Ukraine’s West/East axis may function in the same way as the religious axis. In short, it is questionable as to whether Ukraine is a country divided by world views and cultural/linguistic preferences. Rather, this cleavage would seem to be constructed by the elite for electoral convenience. Yet the same elite do not rely upon this cleavage in coalition formation, counting more on clientelistic and personal advantages. As a result, the Orange parties have lost their ideational appeal, thus making Yanukovych the most popular politician on the eve of the 2010 presidential elections.

Conclusion Having overcome the political crisis in 2007, President Yushchenko declared 2008 the year of constitutional reform. A pro-presidential jurist, Volodymyr M. Shapoval, published a draft of the proposed presidential constitution in February 2008 (Proekt 2008) and influential parties responded by publicizing their own draft constitutions. In terms of procedure, Yushchenko proposed adopting a new constitution, rather than amending the existing one, for the purpose of enacting it by national referendum (as was the case with the Russian 1993 constitution and other CIS constitutions), thus evading the strict requirements for constitutional amendment (two-thirds of parliamentarians’ support). This attempt, however, was harshly criticized by the Venice Commission (European Commission 2008). Domestically, Yushchenko’s “constitutional blitzkrieg” was doomed to fail, because in

9780230_242920_12_cha11.indd 205

11/11/2010 5:20:08 PM

206

Kimitaka Matsuzato

post- Orange Ukraine, few people, except for the president, have an interest in a strong presidency. Parliamentarians have been accustomed to the oligarchic practice by which the country cannot be run efficiently but is much better than their humiliating position under Kuchma. Big business is irritated by the huge donations requested at each presidential election (often by multiple candidates). They remember well the years when Kuchma used the tax police to repress opposition businessmen and witnessed closely what happened to their colleagues in Russia under Putin. Regional leaders do not want a strong presidency, because during the years since the Orange Revolution Yushchenko demonstrated that he would continue Kuchma’s politicized cadre policies in regard to Ukrainian regions. Ukraine’s endeavor to readjust the president-parliamentary regime to the epoch of parliamentary oligarchy was indigenous. Parliamentarians devised concepts, such as the parliamentary majority and imperative mandate, to fortify the position of legislature in the unfavorable environment of the late Kuchma period. As a result, the 2004 amendment suffered from a provincialism that the Venice Commission did not endorse. It was paradoxical that the Orange Revolution, which is believed to have opened Ukraine’s road to Europe, distanced Ukraine from European constitutional practices. However, the concepts of parliamentary majority and imperative mandate are deeply rooted in the minds of Ukrainian politicians and both Yushchenko and the Party of the Regions included them in their draft constitutions published in 2008. Where can we find an exit from the present situation? Oleksandr Syniookyi, leader of an NGO Laboratory of Legislative Initiative, does not think that parliamentary systems are a priori better than pure presidential or semi-presidential systems. However, once the malfunctioning of a political regime has reached this level, he believes that a shift to a full parliamentary regime is the only solution. In this case, Yushchenko, or his successor, will have no other option but to participate in parliamentary elections seriously and comply with the result.23 However, it is unlikely that a president who is accustomed to the sweet taste of semi-presidentialism, where maneuvering in cabinet formation can replace the painful struggle for parliamentary majority, will reject it. What we witnessed in Ukraine after the Orange Revolution was a symbiosis of oligarchic politics and semi-presidentialism. The lack of a definitive cleavage in Ukraine, be it socioeconomic or cultural, makes stable coalition politics difficult, but it also makes unconceivable compromises possible. If the events of April to May 2007 had happened in Russia, it would have ended in bloodshed. The oligarchic characteristics of Ukrainian politics widen the room for the president to act as an independent broker in coalition formation. As the omnipotent presidency in the late Kuchma period raised the then-nascent parliamentary oligarchy, parliamentary oligarchs now bestow on the exhausted president valuable opportunities for maneuvering.

9780230_242920_12_cha11.indd 206

11/11/2010 5:20:08 PM

Ukraine: Disintegrated Semi-Presidentialism 207

Notes 1. See Elgie and Moestrup 2007: 245–246. Matsuzato’s concept of “almost full presidential semi-presidentialism” (Matsuzato 2006) and Siaroff’s “Category 6” (Siaroff 2003) implies the same regime type. Along with Taiwan, we find its examples in the Transcaucasus. 2. As Hale also argues, the CIS elites’ preference changed around the turn of the century. In the 1990s, presidential-parliamentary regimes with strong, patronal presidency took root in CIS countries as a device for distributing Socialist state properties through patron-client networks and as a way of struggling against leftist forces by mobilizing administrative resources through dirty campaigns. Since the turn of the century, however, CIS elites became dissatisfied with this regime type because of its unpredictability, exemplified by the political uncertainty in Russia in 1999 and various colored revolutions. They tried to build a more stable political regime by depersonalizing (de-presidentializing) politics and, for this purpose, created or developed a ruling ideational party, such as United Russia and the New Azerbaijan Party. These parties are tangibly different from pro-presidential parties in these countries in the 1990s, which were no more than façades for clientelistic networks (Hale 2008; Colton and Hale 2009). This is why the succession from Putin to Medvedev was incomparably smoother than that from Yeltsin to Putin. Another way to “de-presidentialize” politics was to strengthen the parliamentary oligarchy. It is symptomatic that Russia, Ukraine, and Armenia, renowned for their clientelism, shifted to pure or predominantly proportional systems of election to parliament, while Lithuania with a developed party system remains under a mixed (half proportional, half single-mandate) system of parliamentary elections. These institutional changes, understandably, affected the operation of semi-presidentialism in CIS countries. 3. For this purpose, the constitutional amendment of Ukraine in 2004 even prohibited minority coalitions. 4. Turkmenistan and the unrecognized state of Transnistria, both of which chose full presidential systems. 5. In contrast to this traditionalist choice, actors might choose semi-presidentialism consciously in a critical political situation in order to neutralize the opposition by sharing power with it. See the examples of Poland, Lithuania, and Mozambique in Matsuzato (2006), Matsuzato & Gudžiskas (2006), and Manning (2007). 6. This was not fully-fledged semi-presidentialism because Mikhail Gorbachev was elected as the president of the USSR by the Congress of People’s Deputies (parliament), not by popular vote, on March 15, 1990. 7. A former judge of the Ukrainian Constitutional Court, Petro Martinenko, provides an interesting interpretation that the Ukrainian Constitution of 1996 could not “reach” a standard semi-presidentialism with a fully-fledged system of checks and balances because Ukraine’s constitutional process started from the highly presidentialized ideas described in the “Concept.” Interview with me on March 9, 2004 in Kyiv. 8. To be more precise, the president was empowered to dissolve parliament if the plenary meeting did not start work within thirty days after the beginning of the regular sessional season (Art. 90). However, this was a very improbable scenario to take place. 9. Oleh Protsyk noted that despite the smaller scale of the country, President Kuchma issued more numerous appointment decrees than Yeltsin did. Protsyk attributes this to “a higher degree of politicization of bureaucratic appointment and a higher level

9780230_242920_12_cha11.indd 207

11/11/2010 5:20:09 PM

208 Kimitaka Matsuzato

10.

11. 12. 13.

14. 15.

16.

17.

18.

19.

20.

21.

22.

of cabinet instability in Ukraine” (Protsyk 2004: 650). The situation was the same at the local level. In Russia, even during the period when the president appointed governors before 1996, Yeltsin only appointed governors and regional capital mayors, while the Ukrainian president appointed and continues to appoint governors and the whole legion of district chiefs (after the consultation with the governor). During the 1990s, Kuchma even tried to appoint vice governors. See the results of the parliamentary elections of 1998 at the official site of the Central Electoral Committee of Ukraine (www.cvk.gov.ua/pls/vd2002/ webproc0v?kodvib=1&rejim=0, accessed January 11, 2010). Interview with Mykhailo Syrota, parliamentarian in 1994–2002 and after 2007, March 5, 2004, Kyiv. Kovryzhenko and Zamnius (2003); and interview with Yurii Klyuchkovs’kyi, parliamentarian from Our Ukraine on March 11, 2004 in Kyiv. Interview with Mykola Kozyubra, member of the constitutional commission in 1992–1996, judge of the Ukrainian Constitutional Court in 1996–2002, on March 2004 in Kyiv. The aforementioned interview with Syrota. Interview with Petro Martynenko, dean of the Law Faculty, International Solomon University, former judge of the Constitutional Court, on March 9, 2004, in Kyiv. This interval of “fifteen days” gave Yushchenko a good pretext not to confirm the parliamentary coalition’s candidate automatically, but retain the possibility of nominating other candidates. If the president’s role was to confirm the candidate proposed by the coalition, Yushchenko argued, the amendment should not have given him 15 days to consider. There have been examples among constitutions of the world where parliament is obliged to form a government within a certain time limit but it seems a rare case that the constitution obliges parliament to form a government based on a majority coalition. Before long, however, it proved easy for deputies to bypass this provision. Deputies who, in fact, began to act and vote with another faction never submitted official documents to resign from the party to which they belonged. No deputies elected in 2006 and 2007 have lost their deputy’s mandate, based on Article 81, but frequent turncoats continue to characterize Ukraine’s party politics. For example, parliament did not consider the Polish experience of adopting the German system of a constructive vote of no- confidence in her 1997 constitution, which was effective in overcoming the political instability and frequent government changes under Walesa’s presidency. As a reward for Tymoshenko’s support for this law, the anti- Orange coalition agreed to spread the imperative mandate to regional deputy elections – an advantageous step for Tymoshenko’s party run by her own charisma and strict party discipline. The yea vote for the cabinet law, 366 out of the 450 deputies, was sufficient in number to impeach the president and, unsurprisingly, frightened him. On the friendship between Yushchenko and Baloha, as well as Baloha’s contribution to OU in the 2002 parliamentary elections, see Matsuzato (2002). A number of those in Yushchenko’s entourage believe that Baloha’s behavior has created unnecessary enemies of Yushchenko (Mustafin 2008). In some circumstances, the West/East axis might possibly become a convenient device for mobilizing the vote. To propose convincing policies in education,

9780230_242920_12_cha11.indd 208

11/11/2010 5:20:09 PM

Ukraine: Disintegrated Semi-Presidentialism 209 welfare, and other material issues, deputy candidates need to have solid expertise and a plausible budget perspective, whilst it is more or less possible to attract the audience’s attention with empty hands, if they speak out on such issues as language and NATO accession. Taras Kuzio describes how tactically Yanukovych used the NATO issue (Kuzio 2006). 23. Interview with Oleksandr Syniookyi, vice head of the board of the NGO, Laboratory of Legislative Initiative, on August 18, 2007 in Kyiv. On this NGO see: www.parlament.org.ua, accessed January 11, 2010.

9780230_242920_12_cha11.indd 209

11/11/2010 5:20:09 PM

12 Semi-Presidentialism and Moldova’s Flawed Transition to Democracy William Crowther

Scholarly debate regarding the virtues and vices of presidential versus parliamentary regimes has been a fixed part the intellectual landscape of political science over the past half century. The study of semi-presidential variants emerged as a subsidiary line of research in the wake of Maurice Duverger’s seminal work in the 1970s on the French Fifth Republic (Duverger 1980). While a number of prominent scholars (Liijphart 1992; Shugart and Carey 1992; Sartori 1997) ventured into the field, interest remained relatively modest, in all probability due to the limited number of cases to which the model pertained. As the 1990s progressed, however, studies of semi-presidentialism have burgeoned, probably in part due to the proliferation of these political systems in the former Soviet space (See, for example Baylis 1996; Pasquino 1997; Elgie 1999; Roper 2002; Siaroff 2003). In the post- communist context a significant amount of work has been done on the importance of institutional choice for political outcomes (Birch 2003; Olson and Crowther 2003). Fish (2006) and supports the case for the importance of strong legislatures for democratic development. Easter (1997), among others, has argued that presidentialism introduced an antidemocratic bias into the political process. Roper’s (2008) examination of Moldova leads him to conclude that semi-presidential arrangements act to destabilize transition political systems. Colton and Skach (2005: 124–125) conclude that semi-presidential arrangements are inherently unstable, and urge constitutional reform in the direction of parliamentary dominance. Based on a broader comparative study that incorporates many of the post- communist cases, Elgie (2008) argues that semi-presidentialism may be perilous but not for the reasons that its critics usually suggest. Moldova’s experience with semi-presidentialism (from 1994 through 2000) was far from good. Examining that experience may shed light on the weaknesses of the semi-presidential model. The case is also of some intrinsic interest because of the country’s 2000 decision to revise its constitution, strengthening the role of parliament. While recognizing the power of institutional arguments, a close examination of the Moldovan experience will 210

9780230_242920_13_cha12.indd 210

11/16/2010 7:14:57 PM

Semi-Presidentialism and Moldova’s Flawed Transition 211

alert analysts to the weaknesses of explanations weighted too heavily in the direction of unicausality. This chapter examines the development of the Moldovan semi-presidential system, its functioning, and the factors that led to its abandonment. It also assesses the “parliamentary” model introduced in 2001. It will be argued that the configuration of party politics and extra-institutional factors, predominantly elite conflict, are fundamental to understanding how Moldovan semi-presidentialism functioned in practice.

1990 and the transition parliament Moldova’s democratic experience began with a “transition parliament” formed during the closing phase of the Soviet period. The 1990 election process produced a legislative body that was heavily managed by Communist authorities and can be considered “partly free.” During the lead-up to the elections, the Soviet Republic’s Communist Party was sharply divided between a reformist wing headed by First Secretary Petru Lucinschi, and pro-Moscow hard-liners. Increased cooperation between the newly emerged Popular Front of Moldova and reform- oriented Communists paved the way for an opposition victory. A total of 90 (24.4 percent) Supreme Soviet legislators were selected from the list supported by the Popular Front. With the support of many Communist deputies, the Popular Front was able to command more than half of the votes in the new legislature and gained control of the leadership, electing Mircea Snegur, a prominent leader among the reform- oriented Communists, Speaker (Crowther 1997: 294). The 1990 election, though a significant step toward independence, failed to produce a stable post- communist government. First, in the weakly constitutionalized and politically charged transition environment, difficulties almost immediately arose concerning relations between the new legislature and the executive branch. Under the prevailing constitutional arrangements, the President of the Republic was elected by the legislature. In December 1990 Snegur was elected and assumed this office. As factional conflict within the elite intensified, Snegur sought to strengthen his position by standing for direct popular election under new provisions introduced in 1991. Members of the government were to be proposed by the president but required confirmation by the legislature before taking office. Under these conditions, individual level conflict between Snegur and his political opponents in the Parliament rapidly resolved into institutional conflict over control of the government. Simultaneously, parliament became the primary arena for the increasingly critical conflict that spawned ethnic mobilization. This pitted the Romanian-speaking majority against russophone minorities concentrated in the east and south of the country. The breakdown of political consensus inside the legislature was precipitated by ideologically motivated activists immediately on the opening of the 12th parliament. Legislators polarized,

9780230_242920_13_cha12.indd 211

11/16/2010 7:14:57 PM

212 William Crowther

and many minority representatives abandoned the body in May 1990, as a precursor to the formation of a separatist state on the left bank of the Dnistr River. The outbreak of ethnic conflict, along with increasing economic adversity, undermined support for the Popular Front government, which as a result was unable to consolidate its hold on power. In August 1992 the Democratic Agrarian Party banded together with Popular Front defectors and with the Conciliation party group (later the Socialist Party) to name Andrei Sangheli, a ranking member of the previous regime, prime minister, in direct opposition to President Snegur. Sangheli’s government promised more efficient economic management and a more moderate approach to the nationality question. Prime Minister Sangheli’s efforts to govern were hampered by unabated conflict inside parliament between the Popular Front and moderate Deputies who comprised his core supporters. Seeing their support wane and their policies reversed, the Popular Front (rechristened the Christian Democratic Popular Front in February 1993) pursued a course of obstructionism, which led to further fragmentation among the deputies. The conflict in parliament coupled with a growing dispute between legislative leaders and the president, obstructed efforts to enact badly needed reforms. Work on a new constitution languished. Mircea Snegur, having shifted his power base from the legislative branch to the presidency was unable to engineer a constitutional reform that would serve his needs and was thus uninterested in the undertaking (Matsuzato 2006: 329). Through his party allies, however, Snegur retained sufficient influence to block alternate proposals. Work on the constitution came to a halt, and leaders continued to work within the patchwork institutional structure that had been constructed on the framework inherited from the Soviet period. By early 1994 Moldova’s disastrous experience with transition induced a second stage of leadership change. Popular reaction against the separatist conflict, continued economic decline, and public elite infighting led to widespread popular disenchantment with the country’s post- communist government. In an effort to check declining confidence, the pro-Popular Front president of the legislature, Alexandru Moşanu, was replaced by former Communist Party First Secretary Petru Lucinschi. When this too failed to produce progress, Moldovan leaders accepted that the existing legislative arrangement was no longer viable. Deputies voted to dissolve the transition legislature and hold early elections on February 27, 1994.

Constitutional reform, semi-presidential institutions, and failed elite politics New rules introduced for the 1994 legislative election were one element of a more comprehensive restructuring of Moldovan political institutions. The

9780230_242920_13_cha12.indd 212

11/16/2010 7:14:58 PM

Semi-Presidentialism and Moldova’s Flawed Transition 213

previous legislature, considered too large to be functional, with 380 members, was replaced by a dramatically smaller body of 104 deputies. A closed list proportional representation electoral system with a four percent threshold employing a single nationwide electoral district was introduced in the hope of improving party discipline. The election results essentially reversed the political configuration of the early transition. Pro-Romanian parties were marginalized in favor of those supporting Moldovan identity and ethnic accommodation. The Agrarian Democrats won 43.2 percent of the vote and 56 of the 104 seats in the new legislature. The Socialist Bloc (the Socialist and Edinstvo parties) captured 22 percent of the vote and 28 seats. The Bloc of Peasants and Intellectuals captured 9.2 percent of the vote and 11 seats. Finally, the formerly dominant Popular Front managed to retain only 7.5 percent of the vote and nine members in parliament (International Foundation for Electoral Systems 1994: 3–5). Top leadership positions in the legislature changed little as a consequence of the 1994 election. The increasingly influential Petru Lucinschi retained his position as President of the Parliament, but the underlying situation changed dramatically. Moderates gained a working majority, and were thus positioned to act on their legislative programme. Previously deadlocked items were reentered on the legislative agenda and quickly enacted. Under Lucinschi’s leadership work on the new constitution was resumed, with strong input from the EU’s Venice Commission. The new constitution, which was ratified by parliament on July 29, 1994, established Moldova as a democratic republic with a semi-presidential system, separation of powers, judicial review, and guarantees for basic human rights (Constituţia Republicii Moldova 1994). It was modeled on Romania’s 1991 document, and confirmed Moldova as premier-presidential regime, in which the powers of the president, not markedly strong under the previous rules, were further circumscribed. According to Kimitaka Matsuzato’s scale of “closeness to full presidentialism,” Moldova dropped from a three to a one on a scale ranging from one to ten (Matsuzato 2006: 323). Roper (2002: 259–260) ranked Moldova at or near the middle of the pack in scales of legislative and nonlegislative powers of presidents in semi-presidential systems. Under the new constitution, the president was conferred with the right to nominate candidates for prime minister on the advice of the parliamentary majority. Parliament confirmed the prime minister in office through a vote of confidence (Article 82). While Article 83 permitted the president to attend and preside over cabinet meetings, this in fact occurred only rarely. The president was empowered to dissolve parliament in cases of deadlock over government formation for 45 days, or deadlock over legislation for three months (Article 85). Article 88 gave the president the right to submit issues directly to the population as referenda. Finally, while the president was given decree power under Article 94 of the constitution, this was

9780230_242920_13_cha12.indd 213

11/16/2010 7:14:58 PM

214 William Crowther

constrained, with decrees in the critical areas of foreign policy and national defense requiring countersignature by the prime minister. The legal authority of the Moldovan president was thus clearly relatively limited in comparison with the more executive- dominant post-Soviet regimes. This outcome can best be understood as a “conjunctural” result of the immediate demands of inter- elite conflict rather than the consequence of either more profound structural variables or a deep deliberative process. At the time of the constitution’s adoption President Snegur was engaged in a (failing) rearguard struggle against a politically resurgent Petru Lucinschi. Snegur, taking his cue from the Russian constitutional direction, argued for enhanced presidential authority as an anchor for stability in an increasingly insecure political environment. Lucinschi, with his power base in the legislature, championed a constitutional design that enhanced his ability to limit the president’s political reach. Lucinschi’s course was clearly eased by support for the premier-presidential model by western constitutional experts, by support from the leaders of other legislative factions (who did not want their authority infringed upon by the president), and by support for the “Romanian” model from a significant pro-Romanian constituency within the general population. That the president’s constitutional position was not eroded further (transition to a straightforward parliamentary model that one might have expected) is most likely accounted for by a combination of factors. President Snegur retained residual influence in the Parliament and pushed for presidential powers. Petru Lucinschi may well have been uncertain regarding his own long-term interests and therefore unwilling to press for a clearer parliamentary model. As Frye has argued on the basis of an analysis of post- communist choices regarding presidential institutions, “Under conditions of uncertainty, powerful actors tend to hedge their bets and create institutions that are less in their favor than if they were operating under perfect certainty” (Frye 1997: 546–547). The division of power established in the 1994 Constitution, per se, need not have resulted in instability. One could argue that the country was well placed to take advantage of the semi-presidential model, given the presence of a majority party in parliament, the Agrarian Democrats, who were aligned with President Snegur following his break with more radical elements in the Popular Front. Unfortunately a number of factors came together in the wake of the 1994 elections to undermine political stability. Primary among these was the unremitting contest for dominance among political elites. In the context of the transition a raw struggle developed not just over political power, but for control over assets and opportunities in the post- communist economy. Lucan A. Way has focused on this characteristic of both Ukrainian and Moldovan politics, terming it “rapacious individualism” and arguing that it is central to the country’s post- communist development (Way 2003; Way 2005). Second, Moldova lacked highly institutionalized political parties.

9780230_242920_13_cha12.indd 214

11/16/2010 7:14:58 PM

Semi-Presidentialism and Moldova’s Flawed Transition 215

Arguably the single really institutionalized party in Moldova, the Party of Communists of the Republic of Moldova (PCRM), was not yet present in the legislature. The Communists had been banned from political activity in 1991 but were re-legalized by a decision of the parliament in April 1994. Other parties functioned largely as the patronage parties, the factional vehicles of particular leaders, and did little to structure decision-making or constrain leaders. Third, with no established history of independent statehood prior to the Soviet period, Moldova lacked a strong constitutional tradition (or for that matter deep commitment to the rule of law). In this context the Moldovan variant of semi-presidentialism resulted in 1) legislative factionalism; 2) efforts at “presidential expansionism”; and 3) the emergence of competing institutional centers of power. These factors blurred lines of authority, and the policy stalemate of the preelection period reappeared. As Petru Lucinschi worked to undercut President Snegur’s support within the parliament, dissent within the Agrarian party group quickly became at least as intense as that between it and the other parliamentary parties. Agrarian Democratic Party deputies were enticed into contending factions that grouped around the rival leaders. With reform once again stalled, the economic collapse unabated, and the sovereignty issue unresolved, an open struggle for power broke out during the run-up to the 1996 presidential election. President Mircea Snegur initiated his fight for a new term early in the year by moving to a more pro-Romanian position, sharpening his attacks on legislative leaders, and attempting to assert more control over the government. In March, Snegur attempted to dismiss Defense Minister Pavel Creangă over the objection of Prime Minister Sangheli. The president’s decision was appealed successfully to the Constitutional Court, which found in April that ministers could only be removed by the prime minister or through a parliamentary vote. Snegur, who surely must have anticipated this outcome, made confrontation with the Parliament the centerpiece of his campaign. He argued that his primary opponents, Prime Minister Sangheli and President of the Parliament Petru Lucinschi, were using their control over the legislature to bloc reform. He also argued that the constitutional design of 1994 was fatally flawed, and proposed that Moldova be transformed into a “Presidential Republic.” The outcome of the 1996 presidential elections indicated popular hesitation concerning this course. President Snegur came in first out of nine candidates in the first round of the November elections, with 38.7 percent of the vote in comparison to 27.7 percent for second-placed Petru Lucinschi. Vladimir Voronin, leader of the PCRM, took 10.2 percent of the vote. The Agrarian candidate, Andrei Sangheli, was relegated to a fourth place finish (9.5 percent) While President Snegur attracted only an additional 6 percent in the second round, bringing his total to 46 percent, Petru Lucinschi drew support from the entire the left-wing grouping, capturing 54 percent. The former Communist Party First Secretary thus returned to the peak of

9780230_242920_13_cha12.indd 215

11/16/2010 7:14:58 PM

216 William Crowther

political power, while shifting his institutional base from the parliament to the presidency (Crowther and Josanu 2004). By 1996, however, the conditions that plagued Moldovan political life had become too deeply entrenched to be easily reversed. Despite the hopes of many Moldovans, Petru Lucinschi was unable to restore order to the country’s political processes. A new technocratic government backed by President Lucinschi was formed under (independent) Prime Minister Ion Ciubuc. The Ciubuc government was composed in large part by holdovers from the previous administration, but now included two ministers from the PCRM (a probable consequence of Communist support for Lucinschi in the presidential race). While officially pro-reform, Ciubuc’s government proved indecisive, especially with regard to privatization progress was also hampered by the parliament. The continued presence of many antireform deputies, lack of party discipline, and continued political fragmentation all worked against comprehensive reform. It should be noted, though, that freezing the economy between state ownership and comprehensive privatization and marketization did not work to everyone’s disadvantage. Substantial wealth continued to be siphoned off into the hands of political elites and their private sector allies through the sort of insider-trading practices that emerged across much of post- Soviet space. In the absence of effective leadership from the top, the Moldovan economy spiraled downward. The impact of this continuous decline on the population was disastrous. By 1997, Moldova was poorer than any country in Central Europe, and poorer than any former Soviet republic except Tajikistan and Uzbekistan with a per capita of GDP $527 in 1997 (World Bank 2002: 4). Out-migration, particularly among members of the upper social strata reached unprecedented levels as Moldovans both lost hope for improvement in their own country and were forced to seek income elsewhere. Not surprisingly under these circumstances, popular confidence in the country’s political leadership plummeted in the run-up to 1998 parliamentary elections. Only one of 15 parties and electoral blocs could claim as much as 15 percent of popular support in the months leading up to the election. Just as telling, that party was the PCRM, the primary opponent of reform. The PCRM’s unexpected resurgence under Vladimir Voronin clearly presented a significant challenge to President Lucinschi, who had positioned himself as a moderate “post- communist socialist” and benefitted from the support of otherwise disenfranchised left-leaning voters. Now, forced to compete for their support, he established the Movement for a Democratic and Prosperous Moldova (MDPM) as an unofficial presidential party. The MDPM described itself as social democratic and also made strengthening the position of the president a central tenant of its campaign, arguing that this would resolve the republic’s policy deadlock and allow reform to go forward.

9780230_242920_13_cha12.indd 216

11/16/2010 7:14:58 PM

Semi-Presidentialism and Moldova’s Flawed Transition 217

Right-wing forces remained deeply divided. Following his break with the Agrarian Democrats in August 1995, former President Snegur had also formed a political party to advance his political agenda, the Party for Revival and Conciliation in Moldova (PRCM). Common opposition to Lucinschi led to the construction of an uneasy alliance between the PRCM, and the rightwing Christian Democratic Peoples’ Party (CDPP). Emulating the Romanian opposition to Ion Iliescu, these parties campaigned as an electoral alliance, the Democratic Convention of Moldova. A third right-wing party, Party of Democratic Forces (PFD), headed by Valeriu Matei, decided to contest the elections separately. Matei took an extremely critical stance on the political establishment, and argued for eventual (but not immediate) reunification with Romania. The 1998 election did little to resolve either the leadership dispute or the fragmentation inside parliament. Four parties surpassed the four percent threshold, with none approaching a majority. In fact, the political situation was further complicated by the Communists entry into the legislature with 30 percent of the popular vote and 40 seats. The Democratic Convention (19.4 percent 26 seats) and Lucinschi’s Bloc for a Democratic and Prosperous Moldova (18.2 percent and 24 seats) were nearly equally matched. Finally the Party of Democratic Forces entered parliament with 8.8 percent of the vote and 11 seats. This outcome in essence gave the main left-wing parties 48.2 percent of the vote and a potentially strong majority of 64 seats in the 101-seat parliament, as opposed to 28.2 percent and a potential 37 seats for the political right. A centre-left/left vs. centre-right/right division would thus have resulted in a decisive governing majority for the left. It would also have in fact reflected accurately the public’s ideological preferences at the time. But the fratricidal power struggle within the post- communist elite dictated otherwise. Despite their obvious policy differences, the leaders of all of the nonCommunist parties united against the PCRM. Collectively they formed the Alliance for Democracy and Reform (ADR), in order to keep Voronin and the Communists out of government. The ADR was comprised of 61 members of parliament representing the CDM, the PDF, and MDPM. Their coalition agreement called for development of a joint program and the allocation of offices proportionally to all three party groups. The MDPM’s legislative leader Dumitru Diacov was voted in as Parliamentary Chairman while Valeriu Matei and Iurie Roşca became his deputies. Mircea Snegur, while not in the parliamentary leadership was named President of the Alliance for Democracy and Reform. This deeply fragmented legislature and a constitutional framework that did not provide the president with sufficient authority to assert control over the political agenda set the stage for a new round of institutional deadlock. Comprised of pro-reform and antireform, and pro-nationalist and antinationalist factions, this legislative “alliance” proved to be anything but

9780230_242920_13_cha12.indd 217

11/16/2010 7:14:58 PM

218 William Crowther

functional. While Dumitru Diacov and the MDPM were at least initially a pro-presidential, the leaders of each of the other coalition parties were longstanding opponents of Lucinschi with aspirations for capturing control of the executive branch themselves. They thus had no interest in seeing the president’s agenda succeed. Further complicating the situation, following months of wrangling, Diacov broke with President Lucinschi and formed his own party, the Democratic Party of Moldova (PDM), aggravating divisions within parliament and further reducing Lucinschi’s influence within that body. The course of legislative- executive relations and parliamentary politics after the 1998 election was a testament Moldova’s systemic disfunctionality. Blocking an effort by Mircea Snegur’s supporters to name one of their own as prime minister, President Lucinschi engineered the retention of Ion Ciubuc. Prime Minister Ciubuc, however, confronted nearly insurmountable obstacles. His government’s survival was dependent upon support from reform-averse PCRM deputies who would, if pushed too far, abandon the coalition. Thus his ability to press home meaningful change was limited. Furthermore, charges of political corruption bitterly divided the elite. Accusations concerning financial fraud by Lucinschi appointees were met with counter charges against their initiators. Legal action was followed by charges of judicial prejudice (“telephone justice”), the selective leaking of information by the authorities, and charges of a cover-up by President Lucinschi and his allies (Quinlan 2002). Overburdened, Prime Minister Ciubuc resigned in November 1998. A successor government formed under Ion Sturza fared little better. Sturza, a moderate reformer, had served as Deputy Prime Minister under Ciubuc. As prime minister he retained all but four members of the Ciubuc cabinet. His government was voted into office with great difficulty. Communist deputies voted against Sturza on the first two ballots. The CDPP opposed the new government as well, on the grounds that it included corrupt elements from the preceding administration. Given this marked lack of confidence, the new government had little prospect for success. Indeed, even as Prime Minister Sturza struggled to introduce much needed reforms and govern the country, his presumed backer, President Lucinschi, ever more openly questioned the viability of the existing constitutional structure. In affirmation of Graham Allison’s aphorism “where you stand depends on where you sit,” in the wake of the 1998 election President Lucinschi took up the argument that his predecessor, Mircea Snegur, had earlier espoused, calling for a presidential form of government in order to break the political deadlock. A May 23, 1999 referendum initiated by Lucinschi indicated that the majority of the population favored this solution. The vote was not considered valid, however, because the constitution required three-fifths of the population to participate, and in this case only 58 percent did so. Despite this fact Lucinschi pressed forward, taking the vote as a mandate

9780230_242920_13_cha12.indd 218

11/16/2010 7:14:58 PM

Semi-Presidentialism and Moldova’s Flawed Transition 219

for action. On July 1, he formed a special commission on constitutional reform. The commission proposed that the president would have the right to name and dismiss the prime minister and cabinet. The presidential term was extended from four to five years. The number of deputies in parliament was to be reduced from 101 to 70, and a single-member district plurality electoral system would be introduced for legislative elections (East European Constitutional Review 1999). This reform initiative was taken by Lucinschi’s rivals as a personally motivated power grab and a fundamental threat to their interests. Outside Lucinschi shrinking camp, Moldovan political leaders universally opposed increased presidential powers. The PCRM contended that Lucinschi’s initiative was contrary to democratic principles and warned that it could lead to a totalitarian regime. The Democratic Convention and Party of Democratic Forces both argued that the president’s proposal was an effort to emulate the authoritarian “super-presidency” that had emerged in Russia. Even the normally compliant MDPM deputies questioned the necessity of the move. Eleven deputies (supporters of Lucinschi) withdrew from the MDPM, forming a party of independents. President Lucinschi certainly recognized that his proposal would not be supported by parliament. Rather, it appears that he was positioning himself for presidential elections in the fall of 2000, intending to campaign against parliament’s obstructionism. Once having gained a second term, he could force the parliament’s dissolution, work with supportive elements to achieve a manageable legislative majority, and revise the government as he desired. By mid-August legislators had drafted a counter proposal and submitted it to the constitutional court for review. The deputies’ proposal was designed to sharply restrict the president’s power and transform Moldova into a parliamentary republic. It would have shifted to a system under where the president would be elected by the parliament, and would have prohibited the expansion of presidential powers, except in cases of war or emergency. Over the course of the next several weeks the two sides sparred over their competing reform plans, with each attempting to push its proposal forward while blocking the alternative. In the midst of this controversy, the Sturza government was brought down, on November 9, 2000, through a vote of no confidence engineered by the Communists and the right-wing Christian Democratic Peoples’ Party. Sturza’s dismissal initiated the final scene in a bleak decade-long chronicle of self- destructive political infighting. President Lucinschi first proposed Vladimir Voronin as prime minister, hoping that the 40-member Communist legislative group would at least provide the core for an alternative parliamentary coalition on the left. Voronin’s candidacy was predictably blocked by the centre-right. A second, “technocratic” candidate, Valeriu Bobuţac, also failed to accumulate the necessary number of votes, leading Lucinschi to threaten the dissolution of parliament. Under conditions prevailing at

9780230_242920_13_cha12.indd 219

11/16/2010 7:14:58 PM

220 William Crowther

the time, a snap election would almost certainly have resulted in large gains for the PCRM. To avoid this outcome, on December 22, legislators accepted the president’s nomination of a compromise candidate, Dumitru Braghiş, a moderate former Communist, as prime minister. Once again, however, the underlying pattern of factionalism reasserted itself. The Braghiş cabinet made some progress on privatization and reform issues, but was ultimately as untenable as its predecessors. Prime Minister Braghiş was dependent on the unstable alliance between the Communists and the far right CDPP, and was further hindered by the ongoing power struggle between President Lucinschi and parliament. With presidential elections looming ever closer (November 2000) and Lucinschi polling very well, legislative leaders overcame their internal rivalries struck preemptively. On July 5, 2000 parliament overwhelmingly approved (92 of the total 101 votes) constitutional amendments transforming Moldova into a parliamentary republic (Lege Nr. 1115 din 05.07.2000 cu privire la modificarea şi completarea Constituţiei Republicii Moldova Publicat: 29.07.2000 în Monitorul Oficial Nr. 088). The basic premise of the legislators’ constitutional reform was simple: reduce the role of the president in domestic affairs and have the president indirectly elected by the parliament (Article 78). Article 83, which allowed the president to preside over or participate in cabinet meetings and to consult with the government on urgent issues was eliminated (Constitution of the Republic of Moldova, July 29, 1994). Article 89 of the Constitution was amended to allow parliament to dismiss the president through a two-thirds vote, eliminating the necessity of a national referendum on dismissal that existed under the 1994 provisions. The president was, under Article 80, also limited to two terms in office. The president retained the right to nominate a candidate for the office of prime minister, following consultation with the parliamentary party factions (Article 98). Despite discussion of eliminating this power, the president also retained, under Article 73, the right to introduce legislation, along with the government and deputies. The president also retains his role as commander in chief of the armed forces, and under Article 86, the right to take part in international negotiations on behalf of Moldova. Protsyk (2005: 155) has argued on the basis of an examination of several post- communist cases, including Moldova, that parliamentary fragmentation created ambiguity regarding lines of authority and has legitimated presidents’ efforts to exercise expanded executive authority. This argument would certainly appear to be borne out by the Moldovan case throughout the 1990s. Divisions inside the legislature loomed large in the causes for the crisis of the semi-presidential system. Many other factors were at work as well. Lack of party institutionalization certainly played a role. Clearly, so did the formation of highly articulated clientelist networks centering on highly placed political elites, and those elites’ ongoing struggle for power

9780230_242920_13_cha12.indd 220

11/16/2010 7:14:58 PM

Semi-Presidentialism and Moldova’s Flawed Transition 221

and resources. The obvious difficulties and shortcomings of the 1991–2000 institutional model were therefore if anything, in Marxist parlance “over determined.”

Parliamentary regime, presidential outcome A first serious test of the new institutional alignment was provoked by the initial presidential election held under its provisions. Under the new system the president is elected through a secret ballot of deputies at a special session of the parliament. The candidate is elected by a three-fifths vote of the deputies. If no candidate receives the required number in the initial vote, a second ballot is taken between the top two candidates. If this fails to produce a victor with the required three-fifths majority repeat elections must be conducted within 15 days. If the repeat balloting again fails to elect the president of the republic, then the incumbent president dissolves the parliament and establishes the date for legislative elections (Law of the Republic of Moldova on the Procedure of Election the President of the Republic of Moldova Nr.1234-XIV din 22.09.2000, Monitorul Oficial al R.Moldova nr.139– 140/996 din 02.11.2000). The immediate political intent of the electoral reform was clearly 1) to prevent Petru Lucinschi from winning a second in term in office through direct elections, and 2) to limit the political reach of his successor and place the government more firmly under parliamentary control. President Lucinschi was successfully thwarted. However, far from achieving this second goal, the new electoral procedure precipitated the dissolution of parliament and produced exactly the executive- dominant outcome that it intended (at least by many of its proponents) to evade. In preparation for the December 2000 presidential election, PCRM deputies nominated their leader, Vladimir Voronin, while the centreright Democratic Convention and the Party of Democratic Forces jointly put forward Pavel Barbalat. Neither candidate gained the necessary threefifths majority in first round balloting. The same result was repeated in the mandated second round. In essence, with 40 Deputies the PCRM faction was effectively able to exercise a veto over presidential candidates. On December 21, the centre-right factions boycotted the repeat election in an attempt to avoid Voronin’s election. Only the forty PCRM deputies and eight independents appeared in parliament. After obtaining a ruling from the Constitutional Court that under these circumstances the election could be considered a failure, Lucinschi announced dissolution of parliament in mid-January and set early parliamentary elections for February 25, 2001 (East European Constitutional Review 2001). Holding early elections at this juncture worked very much in favor of the Communists. Public opinion was firmly set against the “democratic parties” which had in the estimation of many misgoverned the country since

9780230_242920_13_cha12.indd 221

11/16/2010 7:14:59 PM

222 William Crowther

independence. Campaigning against the status quo, the PCRM took 50.7 percent of the vote and 71 of 101 seats in parliament. The Braghiş Alliance came in a distant second, with 13.4 percent of votes and 19 parliamentary seats. Only one other party, the CDPP, surpassed the 6 percent threshold, gaining 8.2 percent of the vote and 11 seats. The combination of this stunning electoral turnaround and the newly modified constitution completely redefined the character of Moldovan governance. The PCRM was transformed into an utterly dominant political force in the parliament. Its 71 seats not only gave it the numbers required to choose the president unilaterally, but also provided more than the twothirds majority required to amend the country’s constitution. On April 4 Vladimir Voronin was elected President of Moldova with the support of all 71 Communist deputies. The 15 CDPP deputies, not wanting to lend legitimacy to the election of a Communist by their presence, boycotted the proceedings. On April 20 a new Communist backed government was formed under Prime Minister Vasile Tarlev. While the appointment of Tarlev (a moderate with a background as a successful business manager) and retention of 11 members of the previous cabinet was seen as an indication of flexibility, the PCRM’s legislative agenda, especially in its early years in power, was uncompromising. Communist legislators set out to reverse many of their predecessors’ initiatives (Tomiuc 2001). Parliament once again became the scene of intense political conflict, but with very different consequences. The Communists’ initiatives were taken by the opposition as a direct assault on national independence. However, in sharp contrast to previous sessions, a disciplined PCRM majority was able to squelch legislative resistance, allowing President Voronin and Prime Minister Tarlev to govern virtually unimpeded. In late 2001, after vainly opposing President Voronin and his parliamentary supporters for months, the Christian Democratic Peoples’ Party turned to extra-parliamentary resistance. CDPP deputies organized a series of street demonstrations and became the focal point of a growing popular protest movement. Arguing that the countries’ institutions had been “hijacked” by totalitarian forces, they called for mass action to save the democracy. As support for the demonstrators spread, the government employed increasingly heavy-handed tactics to marginalize its critics. On January 22, Justice Minister Ion Morei banned the CDPP for a period of 30 days and threatened harsher sanctions should the party’s leaders continue their protests. Communist deputies threatened to withdraw CDPP legislators’ immunity in order to clear the way for judicial proceedings against them. Finally, a growing perception that the protests threatened to spiral out of control, as well as concern regarding the response of the Western diplomatic community, led the PCRM leadership to moderate its approach. In order to avoid diplomatic isolation and the loss of external financial support upon which it depended, the Communist leadership agreed to participate in a

9780230_242920_13_cha12.indd 222

11/16/2010 7:14:59 PM

Semi-Presidentialism and Moldova’s Flawed Transition 223

Council of Europe mediation effort in April 2002. This marked an unmistakable turning point and brought an end to the period of mass mobilization in Chişinău. The Moldovan Communists clearly recognized the constraints imposed by the international environment and pulled back from their most provocative initiatives. But beneath the veneer of civility President Voronin continued to dominate the government and the PCRM continued to exercise nearly absolute control over the legislative process. Even as partisan conflict escalated in the months following the 2001 legislative elections, Moldova’s economy surprisingly began to show steady improvement. The PCRM benefited both from this recovery and from the improvement in governance that resulted from its control over both parliament and the presidency. As public perception of conditions improved, support for the Communist Party and its leader increased as well (Moldovan Eurobarometers, January/February 2001, April 2002). Furthermore, recognizing the electoral salience of the “Europe issue,” and its own vulnerability, the PCRM conspicuously turned to the West, calling for a policy of “Europeanization,” and improved relations with Romania. In February 2005 Moldova ratified the EU-Moldova Action plan, which considerably strengthened the government’s relationship with the European Union, while reassuring Western oriented members of the opposition. Despite these advantages, as the 2005 parliamentary contest approached the PCRM faced an increasingly coherent electoral opposition. Recognizing their inability to counter the ruling party individually, disparate opposition parties ranging from the centre-left to the right joined together to form the “Democratic Moldova Bloc” electoral alliance. This included the Social Liberal Party, Dumitru Diacov’s Democratic Party, and the “Our Moldova” Alliance (made up of four political parties, and headed by Chişinău’s popular Mayor, Serafim Urechean). These opposition parties, which clearly enjoyed encouragement from Western governments, along with the Christian Democrats, painted the PCRM controlled government as old style Soviet, pro-Russian, and antidemocratic. Opposition arguments proved less than entirely persuasive. The March 6, 2005 parliamentary election produced a second decisive victory for the PCRM. Though less decisive than their first victory, the Communists won a highly credible 46 percent of the popular vote, and took 56 seats in parliament. Their strongest opponents, the Democratic Moldova Bloc, captured 28.5 percent of the vote and 34 seats, while the CDPP followed with 9 percent of the vote and 11 legislative mandates. The PRCM thus became the first party since independence to retain control of Moldova’s government for two consecutive terms. Their reduced majority left the party with strong working majority in parliament, but fewer votes than required to elect the president or amend the constitution. When the XVI Parliament held its first session on March 24, 2005 the Communists’ lacked the required three-fifths majority to elect a president.

9780230_242920_13_cha12.indd 223

11/16/2010 7:14:59 PM

224 William Crowther

If they remained united, opposition parties were in a strong position to block Voronin’s reelection as president. Deadlock over election of a president and parliamentary dissolution was therefore once again possible. But on the conclusion of the parliamentary campaign, consensus within the opposition coalition rapidly broke down. In a startling turnaround, the PCRM’s harshest parliamentary critic during the preceding four years, Christian Democratic leader Iuri Roşca, along with the Democratic Party and the Social-Liberals, broke ranks with their allies and supported Voronin’s reelection on April 6, 2005. President Voronin was thus positioned to retain his dominant role in Moldovan politics and clear signs suggest that Voronin was firmly committed to retaining a firm hold on the leavers of power. Authority was concentrated on the institution of the presidency, to a certain extent undermining the central role played by the PCRM in the early period following the Communists’ return to power. Voronin promoted a system of governance that Moldovans characterize as “vertical power,” where institutional autonomy is eroded in favor of control from above. The PCRM employed an imbedded clientelist network built on the framework of local administration to maintain its political hegemony. Despite nearly continuous calls for more decisive action, reform of local administration was sluggish. The Communist Party’s dominance on the local level has been crucial to its continued success in national electoral contests. The state’s role in the economy also continued to be substantial, and maintaining good relations with leading politicians remained critical to carrying on high-level business activity. Financial transparency is limited, and corruption persisted as a massive and unresolved problem. Substantial privatization occurred under the Communists, but as in many post-Soviet countries, the issue was not simply “privatization,” but “privatization to whom.” Substantial assets have ended up in the hands of the leaders of political clans. President Voronin’s son, Oleg, played a prominent role in Moldova business, throughout the Communists’ tenure in office, as did party stalwart, Vadim Mişin, to provide only two examples associated with the current ruling clique. The converse of President Voronin’s consolidation was a long period of opposition disarray. Despite the hopes of many that a more coherent liberal democratic bloc would take shape in reaction to the PCRM hegemony, leaders of the non- communist parties have shown little ability to cooperate. CDPP leader Iurie Roşca’s decision to cooperate with the PCRM and accept a leadership position in the parliament severely undermined his credibility, even among core supporters. Former Chişinau Mayor Serafim Urechean’s reputation suffered when he emerged as the Kremlin’s preferred candidate in 2005 despite his previously touted “nationalist” stance. The Liberal Party’s candidate, Dorin Chirtoaca, was able to capture the important position of Chişinău mayor in 2007. But the democratic parties proved unable to

9780230_242920_13_cha12.indd 224

11/16/2010 7:14:59 PM

Semi-Presidentialism and Moldova’s Flawed Transition 225

capitalize on the victory, due to internal divisions and the national government’s hostility to the municipal administration. PCRM dominance was clearly reflected in Moldovan public opinion in the post-2001 period. Since the beginning of his first term in office, Vladimir Voronin enjoyed more public support than any other political leader (Figure 12.1). The only leaders whose support was close to that the president were former Prime Minister Vasile Tarlev, and President of the Parliament, Marian Lupu, both Voronin loyalists for most of the period. Support for the most prominent of the opposition politicians (Urechean, Roşca, and Diacov), on the other hand, weakened following the 2005 elections. Even the most popular figure among them, Serafim Urechean, had less than half the level of positive evaluation as that enjoyed by President Voronin in 2008 (Moldovan Eurobarometer 2006). Attitudes toward political parties were equally stark. While support for the Communists as measured by respondents’ stated voting intentions has eroded over time, they continue to dominate the political scene, with support varying from a high of 73 percent to about 45 percent between 2000 and 2008 (Figure 12.2). Fragmented, and often directing their attacks against one another, even the most successful of opposition parties generally ranged around the 10 percent support level. 80 70 60 (%)

50 40 30 20 10

Au g07

-0 6 Au g

-0 5 Au g

04 Au g-

03 Au g-

-0 2 Au g

-0 1 Au g

Au g

-0 0

0

Vladimir Voronin

Vasile Tarlev

Marian Lupu

Sarafim Urechean

Eugnia Ostapciuc

Dumitru Braghis

Dumitru Diacov

Iurie Rosca

Oleg Serebrian

Dorin Chirtoaca Figure 12.1

Support for leaders 2000–2008

9780230_242920_13_cha12.indd 225

11/16/2010 7:14:59 PM

226 William Crowther 80 70 60

(%)

50 40 30 20 10

PCRM Figure 12.2

PD

PPCD

AMN

PSL

PL

Feb-08

Aug-07

Feb-07

Aug-06

Feb-06

Aug-05

Feb-05

Aug-04

Feb-04

Aug-03

Feb-03

Aug-02

Feb-02

Aug-01

Feb-01

Aug-00

0

PSD

For which party would you vote 2000–2008

The PCRM’s political hegemony was finally brought to an end through two highly contested legislative elections held in April and July of 2009. In the first of these the Communists 49 percent of the popular vote and 60 legislative seats, was well ahead of any competitors, but one seat short of the number required to elect a new president. Popular perception that the election had been rigged by the ruling party led to rioting by regime opponents, who burned the lower floors of the parliament building. The three opposition parties that surpassed the electoral threshold (Liberals, Liberal Democrats, and Our Moldova Alliance) boycotted the presidential vote scheduled by parliament, forcing dissolution of the legislature and repeat elections. This contest, held on July 29, gave the opposition a slim legislative majority (52 seats to the PCRM’s 48), allowing them to form a coalition government, ending eight years of Communist rule.

Conclusion Moldova’s decade-long experience with semi-presidentialism was far from positive. Policy-making was difficult, and sometimes impossible. Government formation was difficult, and the rate of turnover was high. The power struggle between the president and his competitors in the legislature were among the most prominent characteristics of national politics. This picture conforms to the negative assessments of the consequences of semi-presidentialism by the model’s critics. One must question,

9780230_242920_13_cha12.indd 226

11/16/2010 7:14:59 PM

Semi-Presidentialism and Moldova’s Flawed Transition 227

however, how much of Moldova’s political instability is attributable to semipresidentialism, and how much is attributable to other factors. Moldova has been plagued since its independence by weak party institutionalization and high levels of fragmentation. As Protsyk (2005) points out, this characteristic has been a critical factor in intra- executive conflict. Elite conflict is endemic, rule of law has been slow to develop, and in fact a substantial (but hardly quantifiable) part of the country’s politics for the first decade following independence went on altogether outside the public sphere. To what extent would a different institutional structure have altered the situation? In this case it seems unlikely in the extreme that a parliamentary model would have produced a more democratic outcome as its proponents suggest. Perhaps more probable is the possibility that a stronger presidential system would have allowed the executive (Lucinschi, for example) to induce a higher degree of order. This solution, however, would obviously have come with its own set of risks. In large part, the pattern of governmental stability reflected changes in the party environment within the parliament. In the extremely volatile period of the early transition, Prime Minister Mircea Druc and his successor, Valeriu Muravschi, were removed in rapid succession and the radical nationalist faction was displaced from leadership in the parliament. Prime Minister Sangheli served without interruption from 1992 through 1996, despite the intervening 1994 election and his own growing leadership aspirations. In 1996, newly elected President Lucinschi successfully proposed a technocratic government under Ion Ciubuc that lasted two years, but was never really able to develop a positive relationship with the legislative leadership and was unable to effectively formulate or implement policy. Legislative elections in 1998 further destabilized the situation, bringing to power a parliamentary coalition fundamentally opposed to the president (Roper 2002: 267). Two succeeding governments under Ion Sturza and Dimitri Braghis equally failed to bridge the rift between president and parliament. Only after the decisive elections of 2001 was stability restored. With the PCRM firmly in control, the decade-long tussle for control over government evaporated. The Tarlev government, supported by President Voronin and the Communist legislative majority that he commanded, faced no significant challenge. It remained in place from its inception in 2001 until March 2008. Tarlev’s replacement by Zinaida Greceanîi, a long time PCRM member who had served as Minister of Finance and First Deputy Prime Minister since 2002, underlined the stability that characterized the period of PCRM dominance. The course of events since 2000 underscores the limited reach of purely institutional approaches. There is absolutely no argument as to whether government was more stable in the post-reform period, nor that, to a lesser extent, governance has improved since semi-presidentialism was abandoned. But despite quite substantial changes to the constitution there was

9780230_242920_13_cha12.indd 227

11/16/2010 7:15:00 PM

228 William Crowther

been no transition to a parliamentary system, if we take that to mean that the parliament plays the dominant role in the political life of the country by exercising control over the government. In fact, if not in law, Moldova moved from a semi-presidential model to a fully presidential system due to the Communists control of both branches of government and President Voronin’s control of PCRM legislative faction. As in the earlier period, this counterintuitive outcome resulted in the first instance from nonconstitutional factors. Party leaders made use of their control over the state to engineer a favorable electoral climate and construct a comprehensive clientelist network. Opposition leaders were largely unable or unwilling to sustain cooperation, in general focusing instead on maximizing their individual outcomes. Finally, all of this occurred in a variant of post- communism that was not highly reformed. Much of the PCRM leadership was socialized in the Soviet period, and comfortably reverted to the executive- dominant “vertical power” model preferred by President Voronin. Similarly, substantial numbers of voters responded positively to the restoration of central control under the Communists. A realistic assessment of the impact of the parliamentary system will therefore have to wait until further systemic progress occurs in the direction of democratic governance. With power in the hands of a coalition of democratic opposition parties, parliamentary institutions may well have the desired effects of promoting both governmental responsibility and stability. What we can conclude at this point is that semi-presidentialism did not prove helpful, and that the introduction of a parliamentary constitutional model did not in itself generate democratic progress in Moldova.

9780230_242920_13_cha12.indd 228

11/16/2010 7:15:00 PM

13 Semi-Presidentialism in the Weimar Republic: A Failed Attempt at Democracy Yu- Chung Shen

The Weimar Republic is an important case for the study of semipresidentialism for two reasons: Weimar was one of the first semi-presidential countries, and it was an attempt at democracy that failed. In this chapter, I define Weimar’s semi-presidentialism as a “vertical dual executive system,” meaning that the president and/or the premier could be the real political leader at the same time. The designers of the constitution intended the president to play a moderating role during normal times and to serve as a political leader in reserve in the event of an emergency. However, because of the fragmented party system, parliament remained powerless. The actual operation of the state transgressed the essence of the constitution after 1930 and finally caused the breakdown of Weimar’s democracy. This chapter examines the Weimar Republic in three parts. First, why and how did the Weimar Republic adopt a semi-presidential constitution? Second, what were the features of Weimar’s semi-presidential constitution? Third, how did the constitution operate in practice and in what way did its semi-presidential nature cause or contribute to the Weimar Republic’s demise?

The choice of semi-presidentialism in 1919 This section discusses the making of the semi-presidential constitution from two perspectives: the influence of Germany’s domestic and international context after World War I, and the role of the designer of this new constitution, Hugo Preuß. The domestic and international context The collapse of the monarchical system in Germany in 1918 was sudden. During a military council in Spa on September 28 and 29, Generals Paul von Hindenburg and Erich Ludendorff came to the conclusion that Germany could not possibly win the war. Accordingly, they decided to accept the 229

9780230_242920_14_cha13.indd 229

11/11/2010 5:22:02 PM

230 Yu- Chung Shen

conditions for armistice proposed by US President Woodrow Wilson and to be the delegates of Germany at the peace negotiations (Kolb 2002: 3). The Allies did not accept the transformation of Germany into a constitutional monarchy and, having lost support from the military, Kaiser Wilhelm II went into exile in Holland. The domestic power vacuum following the breakdown of the monarchy and international pressure from the Allies hastened the birth of a new republic. There were at least three possible forms the new republic could have taken at that time: a constitutional monarchy (supported by the conservative parties, such as the Deutschnationale Volkspartei, DVP), a democratic republic (supported by the Sozialdemokratische Partei, SPD), and a socialist republic (supported by the radical Left, such as the Unabhängige Sozialdemokratische Partei, USDP and the Spartakus alliance). Ultimately, the military allied itself with the SPD to support the establishment of a democratic republic. The military opposed the radical Left’s plans to dismantle the armed forces and turn Germany into a socialist republic like Russia. Moreover, the military believed that its defeat in World War I was due to the radical Left being a drag on the military (the so-called Dolchstoß legend). For these reasons, the military cooperated with the SPD to suppress the revolution, which was led by the Spartakus alliance, and to build a democratic republic.1 The new republic of Germany in 1918 resulted from the pressures of international and domestic social movements. This is clear from Hugo Preuß’s statement in 1919 about the new republic being born not with luck, but with lucklessness, namely through a military defeat (Preuß 1926: 421). In this context, Preuß believed that a powerful president was necessary as a replacement for the Kaiser in the new republic. The president needed not only to be a symbol for a united state but also to have the power to manage the dangerous predicaments in domestic and international affairs of the new republic (Mauersberg 1991: 73–76). Hugo Preuß’s vision for the new republic On November 15, 1918, Friedrich Ebert, the chairman of the Council of the People’s Deputies (Der Rat der Volksbeauftragten),2 appointed Hugo Preuß as the State Secretary in the Ministry of the Interior with the task of drafting a new constitution. He was appointed for two reasons. First, Preuß was at that time the Left’s most learned law scholar. Second, by appointing him, Ebert sought to bridge the divide between his Social Democrats and the middle class (Weber 2001: 35). According to Preuß’s thinking, the new republic should be constructed with a Reich President, who represented all of society. Preuß wanted to see the Reich President play an independent political role as the head of state, alongside a traditional government by elected representatives and civil servants (Stirk 2002: 502). However, the new republic would still be based on the principles of a parliamentary system. The goal of the new constitution was

9780230_242920_14_cha13.indd 230

11/11/2010 5:22:03 PM

Semi-Presidentialism in the Weimar Republic

231

to develop and consolidate a parliamentary democracy (Preuß 1926: 387). The new republic was to have both a directly elected head of state and an appointed head of government accountable to the legislature. Influenced by the failed experience of parliamentary democracy in France, Preuß made provisions for a strong Reich President who had equal status with the Reichstag. He thought that a Reich President popularly elected and experienced in political tactics would be better able to gauge political situations and to judge and act courageously. Not only Preuß, but also other important scholars of the time, such as Max Weber, supported having a strong Reich President head the new republic. Weber’s main reasons were, first, that it was essential for the people to have a head of state whose legitimacy rested unquestionably on the will of the population as a whole, without the intervention of intermediaries. Second, only a Reich President with the votes of millions of people behind him could have the authority to make the necessary political and economic reforms in post WWI- Germany. Third, only the election of the Reich President by the people would provide an opportunity to select leaders and create an outlet there. Fourth, particularism cries out for a bearer of the principle of the unity of the Reich government.3 Any indirectly elected president of the republic appointed by parliament under the banner of a particular party constellation, or coalition, would be politically finished as soon as there was a shift in coalitions. Therefore, in Preuss’ view, the republic required a directly elected powerful Reich President, elected for a fixed term. There needed to be a Reich President who would be a worthy representative of the nation and who would adequately personify, with all the authority needed, a state as considerable as the Reich. The Reich President also needed to be powerful in order to act as a counterweight to the Reichstag, which otherwise might become omnipotent (Brunet 1922: 154). Even though the Weimar constitution endowed the Reich President with many important powers, the essence of the constitution was still founded on parliamentarism. The Reich President was intended to play the role as a political arbiter between the Reichstag and the government and only to act as a political leader when the parliamentary system did not work. The constitution granted him the right to dissolve the Reichstag or to call a referendum. These powers were to be used to appeal to the judgment by all people in the event of a deadlock between the Reichstag and the cabinet (Stirk 2002: 504). The Weimar Constitution was designed with a strong Reich President. It was a reflection of the postwar context of Germany and the subjective beliefs of Hugo Preuß. The collapse of the monarchy in 1918 was an abrupt event. The new republic was built without preparation as an “improvising democracy” (Kolb 2002: 1). In the context of a power vacuum, a strong Reich President replacing the Kaiser as the head of state was the final result.

9780230_242920_14_cha13.indd 231

11/11/2010 5:22:03 PM

232 Yu- Chung Shen

Furthermore, important powers were also bestowed on the Reich President to avoid the disorder of a multiparty parliamentary system like in France. Thus, the new republic was designed as a parliamentary system with a directly elected president; that is to say as a semi-presidential system.

The framework of the Weimar Constitution Generally speaking, a new constitution is usually adopted either at independence or during a time of crisis. The Weimar Constitution is no exception, it was developed at a time of internal revolt and foreign invasion. The victory of the SPD was most clearly apparent in the drafting of a new constitution. The election of the National Assembly,4 which convened to create the new constitution, was held on January 19, 1919, and the SPD was the biggest winner in this election. The results are shown in Table 13.1. Although the SPD was disappointed not to have secured an absolute majority in the constituent body, it was able to ally itself with other centrist parties, the DDP (Deutsche Demokratische Partei, DDP) and the Ztr (Zentrum, Ztr), to form a coalition named the Weimar Coalition. This coalition controlled 78 percent, more than a two-thirds majority, of the National Assembly and led the drafting of the new constitution that was finally adopted on July 31, 1919 by a vote of 262 to 75. The next section will discuss the structure of the government and the relationship between the president, the parliament, and the chancellor. The Reich President According to Articles 41 and 43 of the Weimar Constitution, the Reich President was directly elected by the people every seven years.5 The intention of Preuß was for the Reich President to be the delegate of all the people, and to prevent parliamentary absolutism thanks to the president’s countervailing powers. Table 13.1 The result of the 1919 parliamentary election in Weimar Germany Party

Vote (%)

SPD Z DDP DVP DNVP USPD Others Sum

37.9 19.7 18.5 4.4 10.3 7.6 1.6 100

Seats (no.) Seats (%) 165 91 75 19 44 22 7 423

39 21.5 17.7 4.5 10.4 5.2 1.7 100

Source: Kolb 2002: 308–309.

9780230_242920_14_cha13.indd 232

11/11/2010 5:22:03 PM

Semi-Presidentialism in the Weimar Republic

233

As indicated, the Reich President possessed many important constitutional powers. One of the most disputed powers was enshrined in Article 48, which gave the Reich President sole authority to rule for a limited period without recourse from parliament in the event of major disruption or threat to public safety and order. Some saw this clause as legitimizing the dictatorial power of the Reich President (Friedrich 1933: 197). However, the intent behind Article 48 was for the president to “restore” the constitutional order in times of national crisis. Only a clear parliamentary majority could challenge the president’s invoking of his emergency powers under Article 48. Article 48 came to be a reflection of the stability of the republic. In the troubled beginning of the Weimar republic, from 1919 to 1925, Article 48 was invoked 136 times. During the more stable days of the republic, from 1925 to 1930, the Article was exercised only 9 times. Finally, during the last years of the republic, from 1930 to 1932, it was exercised 109 times (Gusy 1991: 50–51). Reflecting the frequency with which Article 48 was invoked during the period from 1930 to 1932, the number of days that parliament was in session decreased because of the Reich President’s use of his emergency powers. The laws passed by the Reich President under Article 48 also increased. See Table 13.2. That said, Article 48 could be exercised only in the event of a major disruption or a threat to public safety and order. Moreover, the Reich President could not make new laws under this article to change the essence of the constitution such as the framework of government or the system of the state (Gusy 1991: 70). Furthermore, the Reichstag could order the Reich President to suspend the measures that had been taken. All of this meant that, if it so wished, the Reichstag had the power to maintain the equilibrium between legislative and executive powers. Even so, Article 48 subsequently came to be regarded as the origin of dictatorship. According to Article 25 of the constitution, the Reich President could dissolve the Reichstag without being requested to do so by the Reich Chancellor Preuß included this proviso to allow the Reich President to act as a moderator between the cabinet and the legislature, underpinned by the legitimacy of direct elections.6 Table 13.2 The number of sitting days and the laws legislated by the Reichstag or by the Reich President

Time 1930 1931 1932

Sitting Laws days of the passed by Reichstag Reichstag 94 41 13

98 34 5

Laws passed by the Reich President under Article 48 5 44 60

Source: Möller 2004: 204.

9780230_242920_14_cha13.indd 233

11/11/2010 5:22:04 PM

234 Yu- Chung Shen

In addition to emergency powers and the ability to act as mediator between the Reichstag and the Reich Chancellor, the Reich President had another important constitutional power: calling a referendum. It was Preuß’s original intention that the Reich President be able to call a referendum to resolve a conflict or deadlock between the cabinet and the Reichstag (Preuß 1926: 388–389). As discussed above, I define the Weimar Constitution as being semipresidential with a vertical dual executive system. In normal times the Reich Chancellor would govern with the backing of the majority of the parliament, while the president would live in the shadows. That is why Preuß emphasized parliamentarianism as the basic principle behind the constitution (Preuß 1926: 426). However, during a national crisis, the Reich President was enabled to assume all powers, in order to restore normality to the country, thereby sidestepping the Reich Chancellor. The Reichstag and the Reich Chancellor The Reich Chancellor was the chief of the cabinet and was appointed by the Reich President. However, the cabinet was responsible to the Reichstag under Article 54. Thus, the party-political situation in the Reichstag should guide the Reich President in his appointment of the Reich Chancellor. Max Weber thought that the government should work on the basis of a compromise between parliament and public opinion (the directly elected Reich President) (Haungs 1968: 32). Despite the president’s considerable powers, the population still served as an adjudicator when a deadlock or conflict occurred (Finger 1923: 311). In the event of the dissolution of parliament or the call for a referendum, the result depended on the majority of the popular vote. Furthermore, the Weimar Constitution also included provisions for countersignatures on presidential decrees. To go into effect, the decrees were to be countersigned by the Reich Chancellor or the responsible cabinet ministers. The ambit of countersignature even covered the armed forces. The responsibility of policies was therefore shared. The relationships of the president and the cabinet were changeable. In terms of the essence of the Weimar Constitution, the normal constitutional order was meant to work as a parliamentary system. The Reich President played a critical role only when the cabinet and parliament were gridlocked or when the country was in a state of emergency. The essence of the constitution was that the people were meant to act as the ultimate arbiter between the Reich President and the Reichstag but not between the cabinet and the Reichstag (Finger 1923: 316; Haungs 1968: 30). Carl Schmitt viewed the Reich President as the defender of the constitutional order with mediating and reserved authority powers (Schmitt 1931: 137). Because of Germany’s domestic and international conditions, it was impossible for the country to remain unified without a political leader

9780230_242920_14_cha13.indd 234

11/11/2010 5:22:04 PM

Semi-Presidentialism in the Weimar Republic

235

as strong as the Reich President.7 More important, in Schmitt’s view the Reich President needed to be a moderator between the government and the Reichstag when the Reichstag had no clear majority (Schmitt 1931: 138– 139). Schmitt’s vision of the Weimar Constitution conformed to modern constitutional doctrine. The constitution displayed the characteristics of a system of “checks and balances.” Constitutional power was separated into legislative and executive power, and balanced the powers of both (Schmitt 1928: 186). For the constitutional framework see Figure 13.1. We can conclude that the structure of the Weimar Constitution had three important characteristics. First, the Reich President represented an authority on equal footing with the Reichstag. They both were legitimized through direct elections. Second, the Reich President and the Reich Chancellor both had executive powers defined in the constitution. But the essence of the constitution was that the Reich Chancellor led the government, and the Reich President played the role of moderator. He could be a political leader in moments of crisis, in abnormal political times. The Reich Chancellor was only the chief of government in normal times, but the Reich President could appoint and dismiss ministers and even the Reich Chancellor personally. Third, the equilibrium of legislative and executive powers played out in the cabinet. The cabinet was to act as a juncture between the Reich President

Legislative

Executive Proposal of bills from E (Art. 68).

Enacted of bills by L (Art. 68).

Call for a referendum from E (Art. 73).

Nonconfidence Vote by L (Art. 54).

Dissolution from E (Art. 25). Deposed by L (Art. 43). Impeached by L (Art. 59).

The whole people Figure 13.1 legislative

Structure of the Weimar Constitution between the executive and the

Note: Arrows mean the target of the power enforcement. Source: Carl Schmitt 1928(1993): 197.

9780230_242920_14_cha13.indd 235

11/11/2010 5:22:04 PM

236 Yu- Chung Shen

and the Reichstag. The cabinet, not the Reich President, represented the executive system and was responsible to the Reichstag.

The instability of operations under the Weimar Constitution Politics in the Weimar Republic were extremely unstable due to the radically fragmented party system. This section discusses the fragmented party system and the constitutional structure before and after 1930. The constitution at work before 1930 The constitutional functioning of the Weimar Republic can be separated into three stages. The first stage represents the difficult postwar years of public crisis, which lasted until 1923. During this stage the constitutional structure was influenced by peacemaking after World War I. The first cabinet was formed in 1919 with the backing of more than 75 percent of the seats in Reichstag. However, this coalition was formed through a compromise between the bourgeois groups and the socialists with widely differing political opinions (Winkler 2005: 102). The cabinet collapsed after only 128 days over the Treaty of Versailles, and as a result the distrust between the Left and the Right also worsened. The right-wing parties believed that the leftists and the Bolsheviks were traitors. They believed that the leftist parties were stabbing a knife in the backs of the army and the empire (the so- called Dolchstoß-Legende).8 The first Reichstag election under the new constitution was held on June 6, 1920 and the result highlighted a fragmented party system and produced the first minority government. From that time on, minority government was the most common type of government. Weimar’s fragmented party system was a result of the country’s historical sociopolitical divisions and was worsened by a purely proportional representative electoral system. The electoral formulae used over the course of the Weimar Republic did not work to encourage majorities, but rather exacerbated an already fragmented, poorly institutionalized, and quite polarized system (Skach 2005: 41). The radically diverging views of the main parties led to great difficulties in forming viable coalition governments. While agreements could be reached on domestic politics between the Ztr, the liberal DVP, and the conservative DNVP (Deutschnationale Volkspartei, DNVP), these parties could not agree on foreign policy; conversely, while a grand coalition encompassing the SPD through the Ztr to the DVP (excluding the DNVP) could agree on foreign policy, they could not agree on domestic issues (Fulbrook 1990: 171–172). After the parliamentary election in 1920, the government was formed by a bourgeois coalition of the DDP, the DVP, and the Ztr. This government was tolerated by the SPD. This meant that the SPD did not join the coalition, but neither did it work against it (Kolb 2002: 43; Winkler 2005: 139). Even so, the bourgeois coalition collapsed after 315 days because of the reparations

9780230_242920_14_cha13.indd 236

11/11/2010 5:22:05 PM

Semi-Presidentialism in the Weimar Republic

237

and the controversial issue of Upper Silesia (Oberschlesien).9 These issues were all related to the Treaty of Versailles. After the first bourgeois coalition, only a government led by the SPD could and would handle diplomatic policy with the Allies. However, the disagreement between the left-wing and the right-wing parties about diplomatic policy made them as incompatible as fire and ice. On January 11, 1923, French and Belgian troops marched into the Ruhr to compel Germany to pay war damages. Chancellor Wilhelm Cuno inaugurated a passive policy and the government bore financial responsibility for the strike.10 This had destructive consequences. Unemployment was rising and inflation was out of control. The Ruhr occupation caused Cuno’s government to collapse but also created a united front that enlisted the support of all political forces. The SPD could not hold its ground against the DVP. On August 13, 1923, Ebert appointed Gustav Stresemann (DVP) as Reich Chancellor and formed the first grand coalition including the SPD, the DDP, the Ztr, and the DVP. Stresemann endeavored to improve relations with the Allies, and created a better international environment for Germany from 1923 to 1930. On April 9, 1924, the Allies submitted a new reparations agreement known as the Dawes Plan. The compulsory cooperation between the SPD and the DVP, under the pressure of the Ruhr occupation, was devoid of mutual trust and the two parties were at loggerheads over policy issues. As a result, the first grand coalition’s lifespan was short. Stresemann wanted to lengthen working hours to mitigate inflation, but this policy was not accepted by the SPD. On November 22, 1923, the Reichstag failed to support Stresemann’s vote of confidence by 231 to 156. The SPD, the KPD, the BVP (Bayerische Volkspartei), and the DNVP all cast dissenting votes. The grand coalition survived for less than 100 days. After the collapse of the grand coalition, President Ebert appointed Wilhelm Marx (Ztr) as the Reich Chancellor at the head of a coalition government formed only by the Reich President and without negotiations between the main parties. On March 13, 1924, Marx dissolved the Reichstag and called for new elections. This was the first time the Reichstag had been dissolved. The radical parties were the big winners in the new elections. The KPD, an anti-republican party, increased its parliamentary representation from 16 to 62 seats. Another anti-republican party, the NSDAP (Nationalsozialistische Deutsche Arbeiterpartei, NSDAP), led by Adolf Hitler and participating in the election for the first time, also won 32 seats. These two anti-republican parties won 94 seats in total, accounting for about 20 percent of the Reichstag. The effective number of parties increased from 6.4 to 7.1. Reich Chancellor Marx decided to dissolve the Reichstag once again. Thus, the Reichstag was dissolved twice in 1924. The result of the second parliamentary election did not change anything. Finally, under the DVP’s insistence and the rightleaning position of the new Reich Chancellor, Hans Luther, a new government was formed by the Ztr, the BVP, the DVP and the DNVP. It was the

9780230_242920_14_cha13.indd 237

11/11/2010 5:22:05 PM

238 Yu- Chung Shen

first time that the DNVP had participated in a coalition government. It was also the first time that the DDP had not participated. The DDP and the SPD decided to stay in opposition while tolerating this right-leaning bourgeois coalition. On February 28, 1925, the Reich President, Friedrich Ebert, died and a new presidential election was held on March 29. The members of the Weimar Coalition, the SPD, the DDP and the Ztr, built a Peoples’ Bloc (Volksblock) and nominated Wilhelm Marx (Ztr). On the other side was a Reich Bloc (Reichsblock), which consisted of the DVP and the DNVP. They nominated a military hero, Paul von Hindenburg. The election symbolized a parliamentary republic versus a monarchical system. The key minorities were the KPD, the BVP, and the NSDAP. In the end the BVP and the NSDAP decided to stand for the Reich Coalition, and the KPD decided to nominate its own candidate. This led to the victory of the Reich Coalition. Paul von Hindenburg won the presidency in the second round by 48.3 percent to 45.3 percent of the vote for Marx. The weakness of parliamentary politics from 1919 to 1925, and the victory of the right-leaning camp became a key indicator that the president had a large influence on the formation of a government. The presidential election was therefore seen as a constitutionally silent transition (ein stiller Verfassungswandel) (Kolb 2002: 84–85). After 1925, the mood in politics was right-leaning, because of Hindenburg’s victory and because the DNVP joined the coalition government led by Luther, a right-leaning and conservative Reich Chancellor. Three months later, the SPD initiated a successful vote of no- confidence against Luther’s government. President Hindenburg appointed ex- chancellor Marx as the new Reich Chancellor. But the government was still without a majority. On December 17, 1926, the SPD again initiated a vote of no-confidence to oppose military cooperation with Russia. Marx’s government fell again, this time by a vote of 249 to 171. It violated Hindenburg’s will when the SPD initiated the overthrow of the cabinet twice. President Hindenburg also felt that the government should be led by the president through the use of Article 48 (Winkler 2005: 320). The influence of the president in forming a government became ever greater hereafter (Haungs 1968: 131–134; Kolb 2002: 86). A right-leaning coalition government, including the Ztr, the DVP, and the DNVP, was again formed on January 29,1927 under Hindenburg’s intervention. The bourgeois coalition was stable at first because of the Dawes Plan and the Locarno Treaty, but the coalition soon broke up due to the proposal for a Reich School of Law.11 On March 31, 1928, Hindenburg dissolved the Reichstag to resolve gridlock within the coalition. The SPD won reelection and formed a coalition government again. A new grand coalition government, led by Hermann Müller (SPD), was formed including the SPD, the Ztr, the DDP, and the DVP. The issue of economic reform and the policy of unemployment insurance had become increasingly serious problems

9780230_242920_14_cha13.indd 238

11/11/2010 5:22:06 PM

Semi-Presidentialism in the Weimar Republic

239

following the international financial meltdown. The conflict inside the grand coalition was made worse because of the Young Plan.12 On October 3, 1929, Gustav Stresemann died and the DVP moved to the right again under Alfred Hugenberg’s lead. On October 24, 1929, the financial meltdown destroyed Germany’s fragile economy. The right-wing parties formed a new coalition without the SPD and gained Hindenburg’s support. At that time, Hindenburg decided to form a right-wing coalition government and operated under Article 48 (Brüning 1970: 156–161). Thus, the grand coalition fell on March 27, 1930. After the breakdown of this coalition, the president took over responsibility for cabinet formation until the fall of the Weimar Republic. Dictatorship under the executive and the breakdown of the Weimar Republic after 1930 After the resignation of Müller’s government, Hindenburg did not look for a new cabinet led by the largest party in the Reichstag, the SPD. Instead, he appointed Heinrich Brüning (Ztr) to be the successor, which signified the beginning of the “Reich President- cabinet” (Präsidialkabinett). From then on, the Reich Chancellor knew that the basis of the government was not in the unstable Reichstag, but rather in the Reich President, under Article 48.13 The Reich President- cabinet was a compromise. The declared motive was not to build a government to confront the Reichstag, but rather to build a government in an unstable situation; not to build a dictatorship under the Reich President, but to build a government which worked together with him; and not to exclude the possibility of building a government under the Reichstag, but to resolve the predicament that the government work with a “tolerating majority” (Tolerierungsmehrheit) (Bracher 1984: 274).14 Even so, Brüning expressed strong opinions on the statement of the new government’s policy on April 1, 1930, making it clear that it would be the last attempt to work and resolve the chasm within the Reichstag. The SPD called for a vote of no confidence in the Brüning government but lost the vote 253 to 187 (Winkler 2005: 377). This signified that a majority within the Reichstag supported the Reich President- cabinet. Brüning’s main assignment was to overcome the financial meltdown. His financial policy was to reduce social expenditure, thereby sparing business and industry. It was hard for the SPD to accept such a plan. On July 18, the SPD again called for a vote of no-confidence that was passed by a majority of 236 to 222 and demanded that the government annul this plan. Facing this situation, Reich Chancellor Brüning proposed to Reich President Hindenburg that the latter dissolve the Reichstag and call a new election (Winkler 2005: 379–381; Nolte 2006: 174). In his memoirs, Brüning wrote that there were two reasons for the dissolution. It was not only intended to let the people decide on whether the financial bills should be adopted or not, it was also a vote of confidence on the parliamentary system, which he

9780230_242920_14_cha13.indd 239

11/11/2010 5:22:06 PM

240 Yu- Chung Shen

felt was useless, disorderly, and a democracy only in form without essence (Brüning 1970: 182). The vote of no confidence in July had another important consequence. The SPD, which was devoted to building a republic, cooperated with the NSDAP and the KPD, which were against the republic, and tried to overrule the decisions of the Brüning government with an absolute majority. The SPD opposed the Reich President- cabinet and the president’s use of emergency powers, but the NSDAP and the KPD opposed the republic itself. The success of the no- confidence vote and the following dissolution opened the door for these two anti-republican parties to gain further influence. There were only two winners in the restaged election: the NSDAP, led by Adolf Hitler, and the KPD. These two anti-parliamentary democracy parties gained 187 out of the 577 seats (31.89 percent) in the Reichstag. All other parties received fewer votes than before. Thus, it could be said that the election results were a great victory for the anti-parliamentary democracy movement. This election was a turning point. More than one-third of the people supported the NSDAP and the KPD against the parliamentary democracy establishment. After the “catastrophic election” (Katastrophenwahlen; Nolte 2006: 169), Hindenburg and Brüning understood that it was impossible to build a government on the basis of a majority of the Reichstag. Therefore, until 1933, when Hitler was appointed Reich Chancellor, three chancellors and their cabinets were formed upon the principle of the Reich Presidentcabinet.15 Parliamentary democracy existed in name only. Paul von Hindenburg then changed the role of the Reich President from defender to reformer of the constitution (Bracher 1984: 44–46). He used his constitutional powers to strengthen executive power and to settle predicaments without the express majority support of the Reichstag (Bracher 1987: 133). The end result was a dictatorship of the executive, while party politics and parliamentary democracy were further crippled as the president made ample use of his emergency powers under Article 48 (Bracher 1987: 136– 138). The Reich President went from playing the role of a defender, as per the expectation of Hugo Preuß, to being the breaker of the constitution. According to the essential aspects of the constitution, the Reich President could not use emergency powers to change the principle that the constitution included the essence of parliamentary democracy. However, after 1930, the Reich President appointed the Reich Chancellor and formed the “Reich President- cabinet.” Moreover, the operation and the legitimacy of the Reich President- cabinet relied on Article 48 and not the Reichstag. Even though the measures should have been repealed by the demands of the Reichstag, and the Reichstag could call for a no- confidence vote, history shows that the noconfidence vote called by the SPD against the first Reich President- cabinet in 1930 was not passed. Even though the SPD, the NSDAP and the KPD could have formed an absolute majority to demand that Reich Chancellor Brüning annul the measures, which were passed into law using emergency

9780230_242920_14_cha13.indd 240

11/11/2010 5:22:06 PM

Semi-Presidentialism in the Weimar Republic

241

powers, parliament was dissolved by Reich President Hindenburg, which led to the victory of the NSDAP and the KPD, the anti-parliamentary democracy parties, in the new parliamentary election. In hindsight, we can conclude that the Reich President broke with the principle of parliamentary democracy and brought about a dictatorship. Since 1919, cabinets and Reichstags had been weak and their weakness had encouraged Reich presidential independence. The presidential powers, under Article 48, seemed at first the only hope of saving the new constitution but were eventually the cause of its downfall (Kennedy 2004: 119). The fragmentation of Weimar’s party politics and constitutional design There were structural and institutional reasons for the fall of the Weimar Republic. First, a polarized party system occurred due to the historical background and the different positions of Germany’s foreign and constitutional policies. The competition between the main parties occurred in many different policy areas and were perceived as a zero-sum game. The party system was fragmented because of the diversified social base. There were left-wing socialists (in the SPD, the USPD, and the KPD), conservative nationalists (in the DNVP, the NSDAP, and the DVP), parties interested in Catholicism as a political base (Ztr and the BVP), a liberal party (the DDP), and others. Some of the parties could cooperate on specific issues, but competed against each other on others. For example, the grand coalition formed by the SPD, the DDP, the Ztr, and the DVP could work together on domestic issues but could not reach agreement on foreign policy. The bourgeois coalition, formed by the DDP, the Ztr, the DVP, and the DNVP, cooperated on foreign policy but competed internally on domestic politics. These parties had their respective social bases, and no party could win a majority alone. This made for a special situation in that the basis of many of the coalition governments was personalities not parties. Gustav Stolper, a member of the Reichstag, indeed stated that the coalition government was formed by personal ministers, not by parties. Also, there were only opposition parties and no governing parties in the Reichstag (Boldt 1997: 45). Both the right-wing and the left-wing coalitions faced a critical challenge, not only from the other side but also from the anti-republican parties (Gusy 1997: 386). We can say that the party system of the Weimar Republic was fragmented, zerosum, polarized, and lacked the capability of forming a majority coalition (Boldt 1997: 46; Bracher 1984: 84–85; Huber 1984: 260). Before the twentieth century, anti-partiness (Antiparteienaffekt) and anti-pluralism (Antipluralismus) were both characteristics of Germany’s traditional political culture (Lösche 1997: 149–150) and political individualism was harmful to party politics in the Weimar Republic. After Hindenburg’s victory in 1925, this problem became more acute (Gusy 1997: 388, 399). Briefly, the Weimar Republic’s power fragmentation occurred

9780230_242920_14_cha13.indd 241

11/11/2010 5:22:06 PM

242 Yu- Chung Shen

not only because of the large number of active parties but also because of the zero-sum nature of the system and the polarized competition as well as the existence of anti-republican parties. Furthermore, besides political parties, there were also many other political actors, like the German army, the bureaucracy, and judicial officials, who had significant influence on everyday politics. These actors further contributed to the power fragmentation. At the level of institutions, the superiority of the executive, especially the role of the Reich President, held a potential for executive dictatorship. The Reich President played an important role in this vertical dual executive constitution. He could act independently of an unstable Reichstag and he could use constitutional powers, such as dissolution (Article 25), referendum (Article 73), and the emergency powers (Article 48) to affect politics and to defend the constitution (Schmitt 1931: 158–159). According to the spirit of the constitution, the Reich President was to be a political mediator and use these powers only in abnormal times. But the de facto situation was that the Reich President played a much larger role than originally envisioned and even overruled the principles of the parliamentary essence of the constitution. The Reich President could not be the defender of the constitution for the following reasons. First of all, even though the Reich President was elected directly by the people, it was still difficult for him to claim that he was acting as a delegate of the whole population. Because of the multiplicity of political parties, it was almost impossible for a candidate to receive a majority of votes. In Hans Kelsen’s view, the Reich President could not play a neutral role presiding over the parties due to his lack of strong legitimacy (Kelsen 1931: 44). In fact, the first Reich President was elected by the assembly and nobody received a majority of the votes in the first ballot of the other two elections in 1925 and 1932. Reich President Hindenburg was elected by just 48.3 percent in the second ballot in 1925. The second controversy over the role of the Reich President is with regard to his constitutional powers. The Reich President could call a referendum and dissolve the Reichstag as a delegate of all people. But these powers were designed only as a way of arbitrating a conflict between the executive and the legislature. In other words, the Reich President could not use these powers to rule or to decide between the executive and the legislature. Even though the Reich President possessed emergency powers (Article 48) and could apply the emergency powers to maintain or restore public security or public order, he could not use them to change the principle of the constitution. In a critical moment in 1930, Hindenburg used the emergency powers and dissolved the Reichstag. His intention was not to call a judgment by the people, but rather he wanted to hold on to the reins of government without the Reichstag. These measures transgressed the essence of the constitution that the Reich President should preside neutrally over all parties. Hugo

9780230_242920_14_cha13.indd 242

11/11/2010 5:22:07 PM

Semi-Presidentialism in the Weimar Republic

243

Preuß, the designer of the Weimar Constitution, probably never expected this situation to occur. Thus, at the institutional level, we can say that the power arrangement of the Weimar Republic’s constitution benefited the executive, not the legislature. This is illustrated by the president’s power of dissolution, his freedom in appointing the chancellor, his unlimited emergency powers due to the fragmented party system, and even his power to call a referendum. Moreover, the Reichstag was weak because of the polarized multiparty system. This made the Reich President even more of a fountainhead of the executive dictatorship after 1930.

Conclusion This chapter examines the fall of the Weimar Republic from the concept of “power fragmentation” in party politics and constitutional design. A constitutional arrangement is a rule used to restrict or even induce political behavior. In the experience of the Weimar Republic, the polarized and fragmented party system and a fragmented power arrangement in the constitution made the political system unstable even before its collapse in 1933. The specific conditions of the party system and the semi-presidential constitution worked together to make the Weimar Republic’s transitional tasks even more complex than they already were, thereby facilitating constitutional dictatorship (Skach 2005: 70). Nicholls proposed three general explanations for the failure of the Weimar Republic as follows. First, the German revolution in 1918 was not sweeping enough and did not create a stable and functional institution for the new regime. Second, the Versailles Treaty saddled the Republic with impossible burdens. Third, the constitution was impracticable (Nicholls 2000: 172). In fact, Nicholls’ first two points are the reasons that made the Weimar’s party system fragmented and polarized. His third viewpoint emphasized that the essence of the Weimar’s constitution was parliamentary politics, but the Republic was never able to build a stable block of parliamentary supporters, which made it difficult to practice a parliamentary democracy. Moreover, a powerful president became the source of the executive dictatorship due to a lack of checks on his power by parliament. The experience of the Weimar Republic was a failure since the constitution attempted to let the Reich President be a defender and a moderator in a fragmented society. The Reich President could never be a neutral delegate in a fragmented and polarized party system. It proved counterproductive that the Reich President could use his powers to affect politics, especially when the Reichstag was fragmented and could not check executive power because of the multiparty system. This chapter discussed the background of the constitutional design, the constitutional powers, and the constitutional work in

9780230_242920_14_cha13.indd 243

11/11/2010 5:22:07 PM

244 Yu- Chung Shen

the Weimar Republic. As one of the first examples of semi-presidentialism, it would be worth further research.

Notes 1. On November 10 General Groener, the quartermaster-general of the army, telephoned Ebert, the chairman of the SPD. Groener promised to support the new government led by the SPD in exchange for retaining the military’s position in the new government. This was the so- called “Ebert- Groener-Bündnis” (alliance). 2. There was a transitional government after the November Revolution. The Council of the People’s Deputies was composed of six committee members, three of which were from SPD, while the other three were from the USPD. 3. In this text, Weber put forward seven views. The full text can be seen in the text “Der Reichspraesident.” Berliner Boersenzeitung, February 25, 1919. 4. This was the most democratic election in Germany’s history. Not only could women vote, but the voting age was reduced to 20. See Kolb, 2002: 15. 5. The election law stipulates that anyone who receives more than one-half of the votes is elected. When no absolute majority is secured, a second ballot is held, at which the candidate receiving the greatest number of votes is elected. For details please see: Anschütz 1960: 247–248. 6. The Reich President could also call a referendum to resolve the deadlock. These two constitutional powers were to be treated as the Reich President’s instruments to resolve conflict or deadlock between the executive and the legislative power. see Haungs 1968: 28, or Preuß 1926: 388–389. 7. Hugo Preuß also argued that the president can play the role as a symbol of the unification of Germany. See Preuß 1926: 428. 8. Even Paul von Hindenburg, the Reich President after 1925, agreed with this viewpoint. See Winkler 2005: 121; Kolb 2002: 39; Huber 1984: 37. 9. The Allies decided Germany’s obligations and asked the German government to accept. If the government refused, the Allies would occupy the industrial cities in the east. The people who lived there voted to belong to Germany and not to Poland. The DVP did not accept the obligation and withdrew from the government. Thus the bourgeois coalition fell. 10. It refers to the “Ruhr Occupation.” French and Belgian invaded Ruhr, an important industrial area, to compel Germany to pay the damages. The economy was therefore close to collapsing. 11. One of the motives for the Ztr and the DNVP in forming the coalition in 1927 had been to create a Reich School of Law, which reflected the interests of the Catholic and Protestant Churches in denominational schools. But the DVP, true to its national liberal tradition, was suspicious of clerical influence in education, and would not accept a change in the status of interdenominational schools. See Huber 1984: 622; Winkler 2005: 331. 12. The Young Plan was a new reparations agreement replacing the Dawes Plan. According to the Young Plan, the total amount and a fixed number of years would be decided. In addition, economic sovereignty would also be returned to Germany. Although the Young Plan was advantageous to Germany, it still caused an acute debate inside the coalition government and parliament. For details regarding the Young Plan see: Winkler 2005: 347–348; Nolte 2006: 163; and Huber 1984: 692–696.

9780230_242920_14_cha13.indd 244

11/11/2010 5:22:07 PM

Semi-Presidentialism in the Weimar Republic

245

13. As Brüning was to be appointed the new Reich Chancellor, he told Reich President Hindenburg that the new government did not depend on the Reichstag, but on the Reich President and on Article 48. See: Brüning 1970: 161. 14. Tolerating majority means that although the government was formed only with the backing of a parliamentary minority, more than one-half of members of parliament were tolerant of it. In this situation, some parties did not join the coalition, but they also did not work against the government. 15. They were: Brüning (for a second time from 9 October 1931 to 1 June 1932), Franz von Papen (from June 1, 1932 to December 3,1932) and Kurt von Schleicher (from December 3, 1932 to January 30, 1933).

9780230_242920_14_cha13.indd 245

11/11/2010 5:22:07 PM

14 Haiti: Semi-Presidentialism in a Fractured society Sophia Moestrup

Haiti illustrates the difficulties of making democracy work when constitutional powers are overly skewed in favor of parliament. Haiti has a semipresidential regime1 that provides for extensive limits on presidential powers and numerous checks and balances at different administrative levels. Despite the best intentions of the constituent assembly, democracy in Haiti has suffered repeated setbacks since the 1987 constitution was adopted. This chapter argues that weaknesses of the constitutional design combined with a highly fractured party system have contributed to these setbacks. An often quoted Haitian saying states that, “Konstitisyon sé papié, bayonet sé fè,” meaning “the constitution is made of paper, bayonets are of steel” (Bellegarde-Smith 1990: 134). Given this cynical popular opinion of the relevance of constitutional rule in Haiti, why focus on the role of Haiti’s current semi-presidential constitution in the face of more recent democratic troubles of the country? Haiti is frequently referred to as the political and economic basket case of the Americas. Despite improvements in the security situation and political stability over the past few years, the deadly earthquake in January 2010 and election-related violence in 2009 have rekindled concerns that Haiti is faced with nearly intractable structural problems. How can constitutions matter in such a context? The origins of Haiti’s woes have been attributed to root causes such as cultural traits (Harrison 1993), material constraints (Fatton 2007), and economic dependency (Dupuy 2006) that perpetuate a pattern of authoritarian politics and mass poverty. Such arguments offer little hope of Haiti ever having a consolidated democracy. They also leave us with many unexplained cases of “Third Wave” (Huntington 1991) democratic successes in other post-authoritarian, postcolonial, poor countries such as Benin,2 Ghana, Mali, and Cape Verde, to name a few. Instead, I maintain that the specific provisions of the Haitian 1987 constitution – ironically the fact that in an attempt at curbing presidential excesses, powers are unbalanced in favor of parliament – contributed to the democratic troubles of the country during the past two decades. This is not to say that economic and social 246

9780230_242920_15_cha14.indd 246

11/11/2010 5:22:53 PM

Haiti 247

problems have not negatively impacted the survival of democracy in Haiti – they have. However, I would argue that there was nothing preordained in the democratic woes of the country since the fall of the Duvalier regime in 1986. The argument presented here is that after the fall of Duvalier Haiti had a window of opportunity to change its course following decades of political mismanagement and human rights violations. However, in the context of a very tense and difficult social situation, the weaknesses of the otherwise highly democratic constitution paved the way for a military coup in September 1991, and have since contributed to recurring instances of politically-motivated violence and democratic overturn. There are important parallels between the post-Duvalier period and the post-1946 period3 in Haitian history (Smith 2009: 192–195). In both instances, the promise of democratic change floundered thanks in large part to internecine struggles between former allies within the popular movement that had brought down the previous dictatorship. In the post-Duvalier period, the 1987 constitution contributed to the fracturing of the pro- democracy alliance. It is notable that the 1987 constitution was adopted in a context of rapid institutional change, where the constituent assembly responsible for drafting the constitution was dominated by technocrats and operated relatively autonomously. This minimized the problems of endogeneity – the situation where “[i]nstitutions are at once the dependent variable, the outcome of a set of decisions by political actors, and an independent variable, a factor that affects the decisions made by political actors” (Elgie and Moestrup 2007b: 241). The 1987 constitution was not written by or for a dominant set of political actors. In other words, the distribution of powers enshrined in the constitution was not designed to cater to the strategic interests of any particular group. This exogenous imposition of a semi-presidential constitution – a first in Haitian history – on the main political actors provides an ideal setting for analyzing the impact of introducing this constitutional form in a highly fractured nascent democracy. In particular, the Haitian experience allows us to learn more about the power-sharing effects of premier-presidentialism4 in a polarized society – the type of society that seemingly would have the most to benefit from the purported advantages of this regime type. Unfortunately, in the Haitian context of a highly fractured polity, the premier-presidential system did not shape incentives for political actors as expected. Instead of facilitating power-sharing and democratic consolidation, the constitution contributed to further divisions, while causing a stalemate and hindering effective governance. This chapter outlines the social context for the current Haitian constitution and the transition to democracy in the country. In addition, it presents salient constitutional provisions, focusing on problematical elements, and

9780230_242920_15_cha14.indd 247

11/11/2010 5:22:53 PM

248

Sophia Moestrup

a discussion of how these provisions contributed to political conflict and democratic breakdown in the country. This discussion is particularly timely as Haitian political actors consider ways of rendering the constitutional framework more conducive to democratic consolidation through a series of constitutional amendments.

Social context and democratic transition Haiti’s history since its independence from France in 1804 has been a bloody and troubled one, with a succession of generally unaccountable leaders. This unfortunate pattern culminated with the dictatorial regime of François Duvalier (1957–1971) under whom as many as 50,000 people may have been killed (Laguerre 1994).5 Reflecting the political instability in the country, the constitutional history of Haiti has seen more than 20 constitutions come and go, the majority of which provided for a strong executive and limited countervailing powers for parliament (HippolyteManigat 2000). Political infighting between black and mulatto elites has historically been an important ingredient in the country’s political instability. The black elite had its main origins in the leaders of the war for independence against France that started as a slave revolt in 1791. The mulattos constituted the majority of freedmen, former slaves who had been granted freedom by their masters and could own property and slaves themselves (Paquin 1983). Overall, irrespective of color, the Haitian elite has had little interest in the mainly subsistence agricultural sector, and the state has invested only minimally in services and infrastructure that could have contributed to development and poverty reduction for the majority of the population. Haiti is the poorest country in the Western Hemisphere, with a GDP per capita of $332 in 2003, down nearly 50 percent from 1980. Haiti also has one of the world’s highest income inequalities, with a Gini coefficient of 0.66 in 2001. At the same time, Haiti has been identified as the most remittance- dependent country in the world, with financial transfers from the large Haitian diaspora constituting about 25 percent of GDP (Verner 2007). When François Duvalier came to power in 1957, he presented himself as a social revolutionary and promoter of the interests of an emerging black middle class. However, his regime rapidly became one of the worst dictatorships Haiti has known. At his death in 1971, his son Jean- Claude took over. Jean- Claude initiated some political liberalization and removed a number of regime stalwarts from leading positions, in an attempt to cater to the wishes of international development partners for greater respect of human rights. Most changes were only cosmetic, however. As popular dissatisfaction mounted and found ways of expressing itself thanks to the limited liberalization, the military withdrew its support for the regime. Duvalier

9780230_242920_15_cha14.indd 248

11/11/2010 5:22:53 PM

Haiti 249

and his family ultimately fled to France on February 7, 1986 (Heinl and Heinl 1995). A military junta took over and organized elections for a constitutional assembly in October 1986, in an attempt to appear legitimate. Of the constituent assembly’s 59 members, 41 were elected, while the remaining 18 were appointed by the military government; more than half (34) of the members of the assembly were law professionals. The constituent assembly deliberated freely from November 1986 to March 1987, working on a draft transmitted by the government, which was freely amended (Michel 1992). The new constitution was approved by over 99 percent of the voters in March 1987 (Bellegarde-Smith 1990: 134). The military let the document pass, most probably with the expectation that bayonets would, once again, prevail over paper. And so they did for the next couple of years, until the military fractured into competing factions as coups were followed by countercoups. The economic situation deteriorated and the country became increasingly ungovernable. Popular demonstrations, strikes, and pressure from the US finally led to the installation of an interim civilian government who was in charge of the organization of the first free and democratic elections in Haiti (Gros 1997). This was a hopeful political moment, harboring the possibility of a positive change. This interim government acquitted itself of its major task, the organization of general elections in December 1990, under the leadership of Ertha Pascal-Trouillot as interim president. However, no serious attempt was made at national reconciliation, and political tensions remained high. The clear winner of the presidential elections in December 1990 was the priest and liberation theologian Jean-Bertrand Aristide, who won in the first round with 67.5 percent of the votes. International observers judged the elections to have been free and fair, despite some administrative irregularities. Aristide’s candidacy was supported by a coalition of leftist parties and movements, the FNCD (National Front for Democratic Convergence). Aristide had called his electoral campaign lavalas, a Creole word that means cleansing flood. He was sworn into office on February 7, 1991. The first round of legislative and local elections coincided with the presidential elections; a majority of the seats had to be decided in runoff elections held in January 1991. No single party obtained an absolute majority in either house of the National Assembly (NDI 1991). The initial willingness of the military to support a change of political course is exemplified by the fate of an attempted coup d’état on January 7, 1991, even before the newly elected president and parliament had taken office. The coup was led by a well-known Duvalierist, Roger Lafontant. The army leadership remained loyal, refused to support the coup, and Lafontant was arrested and imprisoned. Massive popular demonstrations opposing the overthrow of the government probably contributed to the military’s

9780230_242920_15_cha14.indd 249

11/11/2010 5:22:53 PM

250 Sophia Moestrup

decision of not supporting Lafontant. The attempted coup was, however, symptomatic of the continued existence of strong opposition to democracy in the country. The new constitution, despite the best intentions of the constituent assembly, did not facilitate a change from the authoritarian practices of the past, as we shall see below. Instead of furthering power-sharing and negotiation, it contributed to a political stalemate and to strengthening the zero-sum nature of politics in Haiti.

Constitutional provisions The 1987 constitution, which governs Haiti today, provides for a semipresidential regime, a first in Haitian history. Though a revised constitutional text adopted by the regime of Jean- Claude Duvalier in 1985 made provisions for a prime minister, none was appointed (Hippolyte-Manigat 2000: 511). The primary concern of the constituents had been to provide checks on the executive powers of the president and to impede the rise of personal rule, given the traumatic experience of the Duvalier years. “Nous avons ficelé le Président de la République comme un saucisson. Je crois que nous l’avons mis hors d’état de nuire (We have tied up the president like a sausage. I think we have neutralized him)” declared one of the constituents as the document was signed (Michel 1992: 133). Past constitutions in Haiti have reflected fluctuating balances in power between president and parliament, but the 1987 constitution contains by far the strongest provisions for parliamentary powers, compared to any of the country’s previous constitutions. In particular, for the first time, cabinet formation was made dependent on the legislature’s approval (Moïse 1999: 127). The constitution provides for a popularly elected president, elected for five years, who can be reelected for a nonconsecutive term. The legislature consists of two houses, a Chamber of Deputies and a Senate that together form the National Assembly. Ninety-nine deputies serve four-year terms,6 while 30 senators are elected for six years,7 with one-third renewed every two years. The constitution makes the prime minister solely dependent on parliamentary confidence as the president does not have the authority to dismiss the prime minister on his or her own initiative, thereby placing Haiti squarely in the premier-presidential camp. The exact division of powers between the two executives is unclear, as is often the case in semi-presidential regimes. While the president chairs cabinet meetings (Article 154), the prime minister is formally the head of government, responsible for the design and conduct of public policies (Articles 155, 156). The two share responsibility for the national defense policy (Article 116). Presidential powers as provided for in the constitution are discussed in more detail below, using the categorization and coding of powers developed by Shugart and Carey (1992).

9780230_242920_15_cha14.indd 250

11/11/2010 5:22:53 PM

Haiti 251

Nonlegislative powers Cabinet formation The president shares executive power with a prime minister who must be approved by the legislature. The president must choose the prime minister candidate among the members of the party having a majority in the National Assembly. If no such majority exists, he or she must consult with the leaders of the two houses of the assembly before appointing a prime minister. In either case, both houses have to ratify the choice of prime minister (Article 137). The prime minister selects cabinet members in agreement with the president, and must present the government’s program for approval by both houses (Article 158). Cabinet dismissal and censure The president cannot on his or her own initiative dismiss the prime minister. Either of the two houses of the National Assembly, on the other hand, can adopt a vote of no-confidence against the government, thus dismissing it, but only once in a 12 month period (Articles 129–4 to 129–6). Either house can also pass a no- confidence vote against individual ministers. Dissolution of assembly As we can already see, the legislature’s powers of control over the executive are quite considerable; they are not matched by countervailing presidential powers. The president does not have the power to dissolve the National Assembly (Article 111–8). The legislature, on the other hand can impeach the president by a two-thirds Senate majority vote for high treason or any other crime or offence committed during his appointment (Articles 185, 186). Legislative powers Veto powers The president does not have veto powers: presidential package or partial vetoes can be overridden by a simple majority of the members of both houses of the legislature present at the time of the vote (Articles 117, 121–5). Decree powers The president cannot make or suspend laws. If the president imposes a state of emergency, the emergency declaration must be countersigned by the prime minister and the entire cabinet and include an immediate convocation of the National Assembly to vote on the measure (Article 278–1). The suspension of any constitutional provisions during a state of emergency must be approved by the legislature (Article 278–2). Even regulatory presidential decrees have to be countersigned by the prime minister or other cabinet members (Article 163). The president’s appointments to high-ranking

9780230_242920_15_cha14.indd 251

11/11/2010 5:22:54 PM

252 Sophia Moestrup

posts in the public administration have to be approved at cabinet meetings; and some must be referred to the Senate for approval (Articles 141, 142). In short, the only remaining discretionary power of the president is that of amnesty (Roy 1992: 86). Reserved policy areas and budgetary powers There are no provisions in the constitution for the exclusive introduction of legislation by the president. With regards to the national budget, the National Assembly can make amendments to the budget proposed by the government. However, the legislature has to accompany any proposed increase in expenditures by the designation of corresponding additional revenues (Article 228–1). Referenda The president does not have the right to initiate popular referenda. In particular, the constitution cannot be amended through a popular referendum; constitutional amendments can only be approved by a two-thirds majority of parliament through a complex procedure described in more detail below. Any amendments approved will only enter into force after the presidential elections. Discussion and additional constitutional provisions As shown in Table 14.1, the Haitian president has very limited constitutional powers, when compared with other semi-presidential regimes (see Chapters 6 and 9 for an overview of presidential powers in Western European and African semi-presidential electoral democracies, respectively, and Elgie and Moestrup (2008) for a summary of presidential powers in Eastern and Central Europe).8 The constitution in essence assigns the ideal role of arbiter to the president, in the event of cohabitation with a prime minister appointed by a legislative majority in opposition to the president (Moïse 1994: 28). In the social context of Haiti from 1990 to 1991 that was probably unrealistic. The popular election of the president focused popular expectations and hope on him, while the letter of the constitution gave him very little leeway for action. The president was elected with a two-thirds majority in the first round of elections with a massive voter turnout. The Senate, on the other hand, was elected with a voter participation of less than 20 percent in the second round of voting (ibid.: 63). Moreover, many legislators were elected more on the coattails of the president than on their own merit. The popular legitimacy of the legislature did not match its ability to block or delay government action. The 1987 constitution reflects a concerted effort at dispersing power and ensuring checks and balances, in an attempt at breaking with past traditions of power centralization and unaccountable governments. In addition

9780230_242920_15_cha14.indd 252

11/11/2010 5:22:54 PM

Haiti 253 Table 14.1 Presidential powers in Haiti’s 1987 constitution Nonlegislative powers Cabinet formation Cabinet dismissal Censure Dissolution of Assembly Total

1 0 0 0 1

Legislative powers Package veto/override Partial veto/override Decree Reserved policy areas Budgetary powers Proposal of referenda Total

0 1 0 0 1 0 2

to providing for a dual executive, the constitution divides legislative power between two houses, the Chamber of Deputies (lower house) and the Senate (upper house), the concurrence of both being needed for cabinet formation and dismissal. As we shall see in the following, this has contributed to long delays in cabinet appointments and slow progress in policy reforms. Furthermore, Articles 61 to 87 in the constitution provide for a complex decentralized governing structure composed of layers of elected and appointed representatives at five administrative levels.9 The powers of local assemblies include the right to nominate candidates for the judiciary as well as for the independent electoral commission. The constitution vests extensive powers in the independent electoral commission: it is responsible both for organizing elections and for resolving electoral disputes (Articles 191, 197) – provisions that clearly make conflicts of interest possible. The nine members of the electoral commission are selected by the executive, parliament, and the Supreme Court (three each) from a list of 27 candidates proposed by the nine departmental assemblies. The departmental assemblies have yet to become functional. As a result, in lieu of the permanent independent electoral commission provided for by the constitution, the country is still operating with an interim commission that has been the object of fierce political competition, and a tool for the manipulation of elections (Erikson et al. 2007: 73). Another important institution provided for by the constitution that was never put in place, is the conciliatory commission responsible for solving conflicts between the executive and the legislature (Article 206). Without this institution, the regime has been left without an arbiter of conflicts between the executive and the legislature (Jean-François 1995).

9780230_242920_15_cha14.indd 253

11/11/2010 5:22:54 PM

254 Sophia Moestrup

Finally, a controversial, though temporary, provision of the constitution was that it banned Duvalierists from assuming public office for a period of 10 years (Article 291). This provision by itself guaranteed the constitution widespread popularity and legitimacy. Though the adoption of such a provision was understandable, given the political context, it did not serve to ensure the loyalty of former supporters of the Duvalier regime to the new constitution, nor their support for the fragile democratic experiment that was underway, as the attempted coup by Lafontant in January 1991 demonstrated.

Political outcomes Aristide was the clear winner of the presidential elections in December 1990. The FNCD, the coalition that had supported his candidacy, won 13 out of 27 seats in the Senate and 27 out of 81 seats in the Chamber of Deputies. With the support of a small party with two elected members, the FNCD realized a slight majority in the Senate (Moïse 1994: 48). The unity of the coalition of forces having supported Aristide’s election was, however, soon to fracture. Aristide’s choice for prime minister, René Préval, was ratified by the National Assembly shortly after Aristide was sworn in as president. Buoyed by his widespread popularity, the president took direct charge of executive matters. The vague provisions in the constitution for the distribution of power between the president and prime minister left room for an active president, in the event that the two executives were of the same political orientation. By not taking a distance apart from ordinary politics, Aristide lost the possibility of using the prime minister as a safety valve in the event of political conflict. While Aristide initially was backed by a majority of legislators, the institutional logic of the constitution helped to create divisions between former allies. The Senate quickly proceeded to stake out its claims on power. Conflict over reforms in the public administration and over nominations for the Supreme Court, the ambassadorship to France and Haiti’s comptroller agency (Cour supérieure des comptes et du contentieux administratif ) ensued (Moïse 1994: 73–79). In August 1991, the prime minister was asked to defend his policies in the National Assembly, the first step toward a vote of no- confidence against the government. The political situation became extremely tense, as Aristide mobilized supporters into mass protests, his only available measure of political pressure against the parliament. After meeting with Aristide, the legislature finally decided to postpone a vote on the Prime Minister’s performance until January 1992 (Jean- François 1995: 250–257). The spectre of mob justice and neck-lacing incited by Aristide’s fiery talks, combined with controversial policy measures undertaken by his government aimed at economic redistribution and military reforms, provided the

9780230_242920_15_cha14.indd 254

11/11/2010 5:22:54 PM

Haiti 255

impetus for yet another coup, on September 30, 1991, only seven months after the first transition to democracy. This time, the coup was successful. Ironically, it was initiated by the military leadership that Aristide himself had appointed. The divisions among erstwhile supporters of the democratic transition, visible in the conflict between president and legislators, served to weaken the young democracy, creating an opportunity for an authoritarian reversal (Fatton 2002: 82). Aristide went into exile and returned only three years later, under US military protection. When Aristide returned to Haiti in 1994 to resume his interrupted presidency, the constitutional framework had not changed. What had changed, however, was the appreciation by the different political actors of the importance of controlling a National Assembly majority. The specific characteristics of Haiti’s semi-presidential regime, with no countervailing powers for the president, make control over the legislature extremely important. Contrary to what had been the case in 1990, the legislative elections of June 1995 were therefore fought with extreme determination. Aristide supporters, organized in the movement, Organisation Politique Lavalas (OPL), won an absolute majority in both chambers, amidst widespread fraud accusations. Senatorial elections held in April 1997 were likewise widely viewed as fraudulent (The Economist 1998: 36). Former Prime Minister René Préval, who succeeded Aristide as president in January 1996, quickly ran into difficulties getting his policies through parliament, following the fissuring of the OPL majority. In November 1996, Aristide withdrew from the OPL to form a new party, Fanmi Lavalas. An important contributing factor to this fracture had been the dispute over whether Aristide should serve three more years to make up for the three years he had spent in exile. The bungled April 1997 senatorial elections deepened the split between former allies. Préval faced stiff opposition from the OPL majority in the National Assembly, resulting in a severe political stalemate (Erikson et al. 2007: 70). The 1996–1997 budget was blocked in parliament for seven months. Important infrastructure projects were delayed and salaries went into arrears. Haiti was without a prime minister for 18 months as the legislature repeatedly rejected the candidates proposed by Préval. In January 1999, the president suspended the legislature, claiming that the legislators’ electoral mandate had expired, and proceeded to rule by decree. In March 1999, after a negotiated compromise with the opposition, the president finally appointed a new prime minister and government in charge of organizing new elections (The Economist 1999: 35–36). The 2000 elections did not solve the political impasse. The legislative elections were widely viewed as fraudulent. The independent electoral commission appears to have been far from independent. Aristide supporters swept up more than two-thirds of the seats, enough to approve constitutional amendments (Fatton 2002: 115, 120). Presidential elections in

9780230_242920_15_cha14.indd 255

11/11/2010 5:22:55 PM

256 Sophia Moestrup

December 2000 saw Aristide compete against three virtually unknown opponents, in an atmosphere of extreme political intimidation. In 2001, the well-respected democracy watchdog, Freedom House, declassified Haiti as an electoral democracy.10 Political radicalization led to armed clashes between Aristide supporters and opponents. The military no longer constituted an organized threat, as Aristide had dissolved the army in February 1995, ahead of the June 1995 legislative elections. However, the dissolution of the army had been accompanied by neither a disarmament program nor employment-generating measures, aimed at integrating former soldiers and officers into the workforce. An initially promising new police force trained with the help of the UN became politicized (Fatton 2002: 152). Several former officers left the force, crossing into the neighboring Dominican Republic where they joined up with former officers and soldiers of the disbanded army and initiated an armed takeover of Haiti in early 2004. The nonmilitarized opposition to the government also exerted pressure for Aristide’s departure through manifestations in the streets of Port-au-Prince (Maguire 2004). The political standoff came close to a civil war, entailing Aristide’s flight into exile on February 29, 2004. An interim government headed by Prime Minister Gérard Latortue organized new presidential elections in February 2006 that were won by René Préval. Despite some fraud accusations, international observers found the polls credible. Accordingly, Haiti regained its status as an electoral democracy, following Freedom House’s classification. The two rounds of legislative elections in February and April 2006 did not produce a majority party or coalition in the National Assembly, though Lespwa (Front for Hope), a coalition including a number of former members of Fanmi Lavalas, won a relative majority of seats in both houses. Government instability has continued under Préval’s second term. Lespwa supported Préval’s candidacy in the presidential elections and was instrumental in securing sufficient votes for legislative approval of Préval’s candidate for prime minister, Jacques-Édouard Alexis. Alexis lost the support of Lespwa, however, and was ousted in April 2008 by a no- confidence vote in the Senate, following price hikes and widespread riots. For five months the country was without a prime minister, as Préval was unable to have his candidate validated by the Chamber of Deputies, which twice rejected his choice. Not until September 2008 was Préval able to appoint his new prime minister, Michèle Pierre-Louis (ICG 2009). Pierre-Louis only lasted 14 months in the position, being ousted through a vote of no-confidence in October 2009 in the Senate. This time, the approval of Préval’s candidate to replace Pierre-Louis was swift: within a week, both houses of the legislature approved planning and cooperation minister Jean-Max Bellerive as the country’s new prime minister. Interestingly, two new members of Bellerive’s government are from Fanmi Lavalas, Aristide’s party that was not allowed to run in the 2009 Senate elections.11 For once, consensus appears to have

9780230_242920_15_cha14.indd 256

11/11/2010 5:22:55 PM

Haiti 257

prevailed in the appointment of the new prime minister and cabinet (Elgie 2009b). Préval’s governments have had to urgently deal with a host of pressing matters such as combating armed gangs,12 reforming the police, dealing with rising food prices, and the aftermath of natural disasters, while replenishing empty state coffers. Despite government instability, there have been some notable successes: security has improved markedly; Haiti’s international debt has been reduced by two-thirds by the cancelling 1.2 billion dollars of Haiti’s debt by the Bretton Woods Institutions; and notwithstanding the widespread destruction caused by four storms that hit the country last year, the Haitian economy was projected by the IMF to grow 2.4 percent in 2009, in the midst of the international financial crisis (Charles 2009). As presidential elections scheduled for November 2010 approach, there is concern despite progress on a number of fronts that a cycle of electionspurred violence could start anew. Haiti’s political scene is characterized by poorly structured political parties that lack a long-term program and a consistent membership basis. Parties tend to be conceived primarily as electoral vehicles. Instead of transforming popular movements into institutionalized political parties, political leaders such as Aristide and Préval have favored the maintenance of loser coalitions formed around their person. The danger of this approach is, however, “the tendency for rival factions to emerge within [the unstructured political organization] and to operate autonomously from the presumed central authority” (Dupuy 2003: 4). Once in parliament, parties splinter into factions as deputies form new allegiances and groupings across party lines, thereby sidelining the leadership of the political parties in negotiations with the executive.13 The government finds itself obliged to negotiate separate agreements with a number of these factions to get anything through the legislature, from the national budget to the appointment of ambassadors (ICG 2009: 5 fn. 30). The widespread splintering of political parties and lack of effective means for aggregating interests and channelling them through the political system is particularly pernicious in a conflict-prone society such as Haiti. Instead of accommodating different demands within the party system, divergent positions are brought to the political system “unfiltered” (Catón and Soldevilla 2008: 135). As a result, “popular mobilization outside regular institutional channels has become central to wielding power in Haiti” (Erikson et al. 2007: 70). This mobilization is facilitated by the continued abysmal living conditions of a large proportion of Haiti’s population.

Constitutional amendments Préval has characterized the 1987 constitution as “the single greatest threat to Haiti’s long-term stability” (Perito and Jocic 2008), due to its complex and cumbersome provisions and the unbalanced distribution of powers.

9780230_242920_15_cha14.indd 257

11/11/2010 5:22:55 PM

258 Sophia Moestrup

The constitutional provisions governing the passage of amendments to the constitution are themselves very laborious. First, a two-thirds majority of both houses in the National Assembly have to approve the need for constitutional amendments; this vote has to take place during the last regular session of the sitting legislature (Article 282–1). The vote on the proposed amendments has to occur during the first session of the subsequent legislature; and the amendments have to be adopted by a two-thirds majority in the presence of a quorum of at least two-thirds of both houses (Articles 284, 284–1). If approved, the amendments will not go into force until after the next presidential elections: specifically, the president under whom the amendments were adopted cannot benefit from these constitutional changes (Article 284–2). Following up on preliminary findings from a small working group, President Préval in February 2009 formally instituted a commission with the mandate to conduct wide consultations and make recommendations for constitutional amendments. The composition of the commission was remarkably balanced, despite rumors that the president’s real intent was to manipulate the constitutional amendment process to lift the limits on presidential terms. The 17 members represented a broad swath of Haitian society, from a representative of a rural community group, to a former minister under Jean- Claude Duvalier, to members of Haiti’s mulatto bourgeoisie, to a number of lawyers and academics from the black middle class.14 The commission consulted with different interest groups across the country and among the Haitian diaspora and presented its recommendations to the president in July 2009. On the basis of the commission’s report, Préval’s office subsequently prepared and submitted a list of proposed amendments to the National Assembly. On September 14, 2009, following two weeks of consultations and revisions to the text proposed by the executive, both houses overwhelmingly approved the motion to move forward with a revision to the constitution as well as a list of amendments to be submitted for adoption by the next legislature to be elected in 2010 (Jérôme 2009).15 This was on the last day of the last regular session of the legislature. Had the motion not passed, Haiti would have had to wait another four years, until the end of the next legislature, for the constitutional amendment process to resume. The proposed revisions are fairly limited, as it appears from a close comparison of the revised and the original constitutional text. By far the most controversial changes are those providing for the reinstatement of the Haitian army (new Articles 263 and forward) and the acceptance of the double nationality for Haitians (old Articles 13–15 eliminated), including the right of Haitians of double nationality to run for public office. The changes aimed at improving the overall functioning of the institutional framework appear mostly as tinkering. The amendments focus on simplifying cabinet appointment procedures and rendering cabinet dismissal by the legislature slightly more difficult. The prime minister must

9780230_242920_15_cha14.indd 258

11/11/2010 5:22:55 PM

Haiti 259

no longer be ratified by both houses of the National Assembly before cabinet formation. Only once the cabinet is formed, does the prime minister have to submit the government’s program for ratification by the legislature (new Article 158).16 The revised procedure for approving a new government thereby consolidates the ratification of the new prime minister and the approval of the government’s program into one step. This should facilitate approval, by providing the executive and the National Assembly an incentive to negotiate adoption of the whole government “package,” while reducing tensions around the prime minister position. With regards to cabinet dismissal, the introduction of a request to question the prime minister or any member of the cabinet (the initial step in the passage of a no- confidence vote) must now be supported by a quarter of the members of the house in question instead of only by five members (new Article 129–3). The failure of a no- confidence vote to pass is considered equivalent to a confidence vote, and there can be no new vote of no- confidence within the following 12 months (new Article 129.6).17 Finally, to ease passage of legislation, the government can now invoke urgency measures that result in a section-by-section vote of the said legislation, instead of a vote articleby-article (new Article 119.1). In an acknowledgement of the need for more effective mediation between the executive and the legislature, the amendments provide for a Constitutional Council (new Article 190bis) to replace the former conciliatory commission that was never appointed. The Constitutional Council is to determine the constitutionality of laws (in lieu of the Supreme Court as provided for in the 1987 constitution) and to interpret the constitutional text in the event of conflict between the executive and the legislature or between the two houses of the National Assembly over their respective powers (190ter 7). The nine members of the new Constitutional Council are to be designated by the executive, the legislature, and the judiciary (three members each). To reduce the frequency and cost of elections, the proposed amendments include changes to the requirements for runoff elections and to the renewal of senatorial seats. In the case of presidential as well as legislative elections for both houses, if the front-runner does not win an absolute majority but has at least 25 percent more votes than the closest runner-up, he/she is declared the winner without the need for a second round of elections (new Articles 90.2, 94.4 and 134). There also appears to be an attempt at making senatorial and deputy mandates and elections coincide: members of both houses are now to be elected for five years (instead of four years for deputies and six years for senators as in the 1987 constitution), with all legislative mandates ending on the last Sunday of November in the fifth year. In particular, senators are no longer elected on a rolling basis with a renewal of a third of the Senate every two years (old Article 95–3 eliminated). With regards to the permanent electoral commission, the proposed amendments eliminate

9780230_242920_15_cha14.indd 259

11/11/2010 5:22:56 PM

260

Sophia Moestrup

the role of the departmental assemblies in the selection of its members (old Article 192). Instead, the nine members of the commission are to be selected directly by the executive, the Supreme Court, and the National Assembly (three members each). It is interesting to note some of the changes proposed by the presidency that did not make it into the text adopted by the legislature. The limit of two presidential terms is maintained and the two presidential terms still have to be nonconsecutive. The text proposed by the executive included an amendment allowing for two consecutive terms. The amendments proposed by the executive also gave the president the power to dissolve the legislature, a provision that – perhaps not surprisingly – was not maintained in the adopted text. Finally, the procedures for amending the constitution have remained unchanged (Radio Vision 2000, 2009). Perhaps the most promising aspect is that the revision process thus far has followed the constitutionally mandated modalities for amending the constitution, a sign that Haiti could be on a path toward a constitution with constitutionalism (contrast with Fatton 2000). However, there is still a last important step to complete – the review, debate, and possible amendments to the proposed text by the next legislature, which originally was scheduled to be elected in February and March 2010.18 This could be a possible flashpoint if, for example, Préval manages to secure a sufficient legislative majority in both houses in the upcoming elections to secure passage of amendments that significantly increase presidential powers or remove presidential term limits. If Préval were to seek to secure an extension of his mandate through a constitutional amendment, Haiti would be back on an oft-travelled path of government decay through constitutional review.19

Conclusion A tracing of political processes in post-1987 Haiti demonstrates that the subtype of semi-presidentialism thought to be the most conducive to democratic survival (premier-presidentialism) does not necessarily further power-sharing and the institutionalization of democratic practice in a highly fractured society. In their concern to rein in presidential authoritarian excesses, the constituents erred in the opposite direction. The total absence of presidential countervailing powers in the Haitian constitution has provided incentives for obstructionism on the part of the National Assembly, and for nonconstitutional behavior on the part of the president. The skewed balance of powers in favor of the National Assembly has entailed a strong temptation to try and control a legislative majority by illegal means to avoid political gridlock, resulting in tampering with vote results and appeals to mob support (Gros 1997). The reactions have been a military coup and armed confrontation coming close to a civil war.

9780230_242920_15_cha14.indd 260

11/11/2010 5:22:56 PM

Haiti 261

The unbalanced distribution of powers in favor of the legislature in a context of weakly developed parties made the country suffer under the weaknesses of parliamentary systems (unstable governments, inefficient policy making), as the National Assembly flexed its constitutional muscles. The situation was further complicated by the presence of a president intent on making his powers match his popular legitimacy, leading to gridlock and stalemate in lieu of power-sharing. All nascent democracies are vulnerable to democratic breakdown at the hands of powerful interests who may decide on the “cancellation of the emergent game” of democracy, rather than following the new institutional rules (O’Donnell and Schmitter 1986: 69). Haiti’s young democracy was further weakened by rifts among former allies in the democratic movement. Power-sharing and negotiations have historically not come easily to Haitian political actors, in a context of deep class and color divisions and limited material resources. Dialogue has been further hampered by Haitian political leaders’ tendency to jockey for US support rather than to find negotiated solutions to their differences (Maguire 2004). The 1987 constitution did not provide incentives for the negotiated settlement of political differences. Institutions do not change identities or interests in the short run.20 Institutions do, however, have a fairly immediate impact on the strategies pursued by political actors to reach their goals (Pontusson 1995). Institutional frameworks that further instability and gridlock, on the one hand, and provide strong incentives to subvert electoral outcomes, on the other, are not conducive to the long-term survival of democracy. In other words, a specific institutional framework can undermine the institutionalization of democracy.21 I would argue that this has been an important issue in Haiti. The constitutional amendments that are currently underway attempt to address some of the imbalances in the Haitian constitutional framework. While maintaining a premier-presidential system, the proposed amendments provide incentives for greater negotiation around cabinet formation. They also limit the frequency with which votes of no- confidence can be initiated by the legislature to once a year, making the government less vulnerable to external shocks and to changes in political alliances. The remaining step is now for the next elected legislature to pass the amendments. Given the cumbersome procedures for amending the 1987 constitution, it is remarkable that the process has come this far. It is particularly encouraging that the amendment process thus far has followed constitutional prescriptions. The executive has not tried to force through changes (including through mobilization of mass support), while the legislature has accepted some reduction (albeit fairly limited) of its freedom of action. By engaging in negotiated changes to the constitutional framework, Haitian political actors are hopefully setting down a new path that will lead to greater power-sharing in the future. Peaceful elections (once

9780230_242920_15_cha14.indd 261

11/11/2010 5:22:56 PM

262

Sophia Moestrup

the country has recovered from the devastation of the January 2010 earthquake) and the passage of the constitutional amendments will be important milestones to reach on this path.

Notes 1. See Chapter 1 for a definition of the concept of semi-presidentialism, as the constitutional regime-form where there is a directly elected president and a prime minister and cabinet who are responsible to the legislature. 2. Many of the slaves imported to Haiti originated from Benin (Fouchard 1988). The religious tradition of the kingdom of Dahomey (ancient Benin), vodoun, became and still is the dominant religion among Haitians along with Catholicism, as in Benin. 3. January 1946, often referred to as the Haitian Revolution of 1946, marked the culmination of a period of popular mobilization and radicalism following the end of the United States occupation in 1934, where nationalist, Marxist and populist groups joined forces to bring down the authoritarian regime of President Élie Lescot. See Smith (2009) for an analysis of this important period in Haitian history. 4. See Chapter 4 for a discussion of the different characteristics of the two subtypes of semi-presidentialism developed by Shugart and Carey (1992): premierpresidential and president-parliamentary systems. The key difference between the two subtypes is that in a president-parliamentary system the prime minister and cabinet are accountable to both parliament and president; in a premierpresidential system, the prime minister and cabinet are accountable only to parliament and cannot be dismissed unilaterally by the president. 5. For an overview of the history of Haiti, including the Duvalier era and its end, see Heinl and Heinl (1995). 6. Deputies are elected in single-member districts, through a two-round first past the post system. 7. Elections take place in multimember districts (three members each), through a two-round first past the post system. 8. My coding differs slightly from that of Shugart and Carey (1992) who assess the Haitian president to have zero constitutional powers. While the president’s powers of cabinet formation are indeed limited, parliamentary recommendation of ministers is not a requirement for their naming (this warrants a one instead of a zero for cabinet formation powers). Similarly, while the president’s ability to veto legislation is very limited given that a simple majority is sufficient to override a presidential veto, the president does have the power to veto part of a bill instead of the whole package. Also, while parliament has the power to amend government budgets, it cannot increase expenditures without designating corresponding revenues. These two corrections put the president’s legislative powers at two instead of zero. 9. See Egset and Kuttner 2007 for an overview of the Haitian administrative governance structure. 10. Country reports for Haiti can be accessed at www.freedomhouse.org. 11. The electoral commission did not approve the candidacies submitted by Fanmi Lavalas because they lacked the signature of the party’s leader, Aristide.

9780230_242920_15_cha14.indd 262

11/11/2010 5:22:56 PM

Haiti 263 12. The spread of armed gangs has been an unfortunate characteristic of the urban scene in Haiti since the mid-1990s. At the source of this phenomenon is the manipulation by “entrepreneurs of violence” of popular movements and associations, a manipulation facilitated by the presence of a large impoverished population of unemployed youth (Erikson et al. 2007, 71). Over time, “Haitian gang culture has morphed from a primarily community or politically-based affiliation where violence was mostly over ideology or turf conflicts, to one increasingly based on crime” (ICG 2006: 5). 13. For example, following the ouster of the Alexis government in 2008, the Coalition des Parlementaires Progressistes (CPP), a majority parliamentary caucus formed by deputies from a number of parties, including Lespwa, demanded that Préval negotiate with the CPP about the choice of the new prime minister instead of with the leadership of the parties represented in parliament (Radio Kiskeya 2008). 14. The text of the presidential decree providing the mandate and composition of the commission as well as the report submitted by the commission to the president is available online at http://www.haiti-reference.com/histoire/constitutions/ rapport_president_const87_intro.php. 15. The text of the proposed amendments, as approved by the sitting legislature and to be validated by the legislature elected in 2010, can be accessed via the website of the Haitian embassy in Washington, DC at www.haiti.org/images/stories/pdf/ rsolution_amendement.pdf. 16. While the new Article 158 states that the prime minister presents his or her program for approval by the National Assembly (both houses at the same time), a no- confidence vote by either house still suffices to oblige the president to appoint a new prime minister candidate. 17. Compared to the old Articles 129–4 to 129–6, the difference is that a failed vote of no- confidence now also counts (as a vote of confidence), limiting the frequency with which no- confidence votes can be initiated to once a year. 18. Due to the devastating earthquake in January 2010, the elections were postponed to a later date yet to be specified at the time of writing. 19. “Haitian political history has shown it clearly: government decay starts with the drive for ‘constitutional review’ when the latter aims to amend the clause that forbids immediate reelection” (Moïse 1990: 298) quoted in Fatton (2007: 180). 20. Roland (2004) makes a useful distinction between “slow-moving” and “fastmoving” institutions. A prime example of the former is “[w]hat is often called ‘culture,’ including values, beliefs, and social norms,” while political institutions are classified as “fast-moving” (pp. 110–111). While Roland focuses on the effect of slow-moving on fast-moving institutions, the emphasis here is on the opposite causal relation. 21. Institutionalization is here defined as a situation where “all the participants in a political process understand and accept the rules of that process, and [where] struggles over the framework within which politics takes place have been settled” (Gorges 2001: 137).

9780230_242920_15_cha14.indd 263

11/11/2010 5:22:57 PM

15 Semi-Presidentialism: What Have We Learned? Robert Elgie, Sophia Moestrup, and Yu-Shan Wu

This book takes stock of what we have learned to date about the impact of semi-presidentialism on democracy from a cross-regional perspective. We do so by discussing institutional choice, operational subtypes, system performance, and the evolution of semi-presidentialism, following a logical sequence. In so doing, the present volume covers more ground than any previous work on semi-presidentialism, both in terms of comprehensive coverage of regions where the system concentrates, and in terms of a full institutional approach that looks at the upstream, midstream, downstream, and evolution of the system. Starting with a reliable definition of the concept of semi-presidentialism that is based on core constitutional features and is adopted throughout the volume, contributors investigate the presence of general patterns of systematic variation across the very diverse universe of cases of semi-presidentialism and their performance. The authors illustrate the actual dynamics of this widely adopted constitutional form through chapters that cover all regions in the world where semi-presidentialism is found; they review the performance of semi-presidentialism in successful democracies, in countries where democracy has broken down, and in countries where democracy has thus far survived but with many challenges. The book thus has a unique breadth and depth that allows us to tease out a number of well-founded conclusions and to suggest new avenues for research for the study of semi-presidentialism. In these concluding comments, we focus on the impact of the founding context and institutional variation, the dynamics of political processes under semi-presidentialism, and the prognosis for the evolution of semi-presidential regimes over time. We conclude by discussing whether semi-presidentialism appears as a useful constitutional model, taking varying contexts and institutional subtypes into consideration.

The founding context and institutional variation As outlined by Elgie in Chapter 1, there are a variety of reasons why a country may wish to choose semi-presidentialism. For newly democratizing 264

9780230_242920_16_cha15.indd 264

11/11/2010 5:23:31 PM

Semi-Presidentialism: What Have We Learned?

265

countries, semi-presidentialism typically appears as a convenient constitutional compromise between proponents of a presidential system, on the one hand, and supporters of a parliamentary type of government, on the other. Semi-presidentialism may be chosen to increase the legitimacy of a directly elected president (as in Portugal), or to implement a compromise between president and parliament by making the prime minister accountable to parliament (as in Taiwan). The choice of semi-presidentialism may also be part of a post- conflict settlement to provide elements of power-sharing, as in Kenya in 2008. Semi-presidential regimes are highly heterogeneous. Adopting semipresidentialism also entails a choice of the calibration of powers between the two elements of the executive, and between the executive and the legislature. Should the prime minister be accountable to the president, meaning that the president can dismiss the prime minister at will, or only to parliament? Should the president have the power to dissolve parliament actively? What should the veto powers of the president be? How extensive should the president’s emergency powers be? As Wu points out in Chapter 2, one can find three main clusters of semi-presidential countries in the world: West European, post-Leninist, and postcolonial. The varied sociopolitical context of these clusters into which semi-presidentialism is introduced is likely to influence the choices of power distribution. West European countries tend to endow the parliament with great powers, while post-Leninist and postcolonial countries tend to have powerful presidents. Parliamentary tradition is the strongest in Western Europe, and weakest in the postcolonial world. Wu argues that a starting point of the study of semi-presidential countries is to recognize how vastly different they are across clusters, and how similar they could be within clusters. Clustering provides a first cut in appreciating both the differences and similarities in the semi-presidential world. Two major subtypes of semi-presidentialism stand out based on a review of the president’s constitutional powers (following Shugart and Carey 1992) – premier-presidential systems, where the prime minister is only accountable to parliament, and president-parliamentary systems, where the prime minister is accountable to both the president and the parliament. Presidentparliamentary systems are in general characterized by greater legislative powers of the president as well, in addition to powers over government formation and dismissal. Not surprisingly, we find a prevalence of president-parliamentary systems in Africa and in the former Soviet Union, as illustrated by Moestrup and Protsyk. These are countries with a history of strong centralization of powers, a weak parliamentary tradition, and limited checks and balances. In the process of democratization, many of these countries adopted a subtype of semi-presidentialism where powers are skewed in favor of the president. Elgie and Schleiter’s large-n study of the impact of constitutional variation within semi-presidential regimes on democratic survival finds that,

9780230_242920_16_cha15.indd 265

11/11/2010 5:23:32 PM

266 Robert Elgie, Sophia Moestrup, and Yu-Shan Wu

controlling for other factors such as the economy and social fragmentation, president-parliamentary systems have a significant negative impact on democratic survival, entailing more than a fivefold increase of the risk of democratic collapse. On the other hand, cohabitation (when a president and a prime minister from opposing political parties have to share power) does not appear to have an independent effect. This is an interesting finding, given that, in the literature, cohabitation has been considered as the main Achilles heel of semi-presidentialism, as it can entail tensions between president and prime minister, produce gridlock and ultimately open the door for an authoritarian reversal. The findings presented in Chapter 4 indicate, however, that this potential complication does not have an independent impact on democratic survival, once the president’s constitutional position is taken into account. The purported reason for the comparatively poorer survival rate of president-parliamentary democracies is that incentives for power-sharing are undermined by the president’s and parliament’s ability to dismiss the government at will. There is no obligation to compromise in the event of conflict between the executive and the legislature. Overall, checks and balances tend to be weaker in president-parliamentary countries, as the president’s legislative powers enshrined in the constitution also are liable to be higher than in premier-presidential systems. A potential problem with any study of institutions, including semipresidentialism, is the issue of endogeneity. Is the effect of institutions endogenous to the choice of the institutions in the first instance? For example, president-parliamentary systems are particularly prevalent in the regions of the world that appear to be the most inhospitable to democracy. Do constitutional incentives to power-share really matter, compared to structural impediments to democratization such as a weak parliamentary tradition or a political culture inimical to competitive political processes that lead countries to adopt a president-parliamentary system in the first place? Protsyk’s review of semi-presidentialism under post- communism helps us address this question by illustrating how president-parliamentary regimes have performed worse in terms of Freedom House scores in all subregions within the former communist bloc where both premier-presidential and president-parliamentary systems are found (the Balkans, the western part of the Commonwealth of Independent States, Central Asia, and the Caucasus). Similarly, in Africa, as Moestrup shows, we find a much higher degree of democratic breakdown among president-parliamentary than premier-presidential systems: only 20 percent of president-parliamentary democracies have survived, compared to 60 percent of premier-presidential democracies. Thus, within regions where structural impediments to democracy presumably do not vary greatly and where we might assume that institutions were chosen under relatively similar historical and cultural circumstances, the choice of subtype of semi-presidential regime does seem to

9780230_242920_16_cha15.indd 266

11/11/2010 5:23:32 PM

Semi-Presidentialism: What Have We Learned?

267

matter for democratic performance. It is true that political traditions, the strength of the party system, and other institutional legacies may influence citizens’ expectations and presidential behavior. However, it would appear that in general, how the constitution structures incentives to power-share is of overriding importance for the probability of long-term survival of a semipresidential democracy. We will now turn to an investigation of the actual political dynamics behind these general trends. The individual country cases in the book allow us to pursue the study and the specification of the political processes and mechanisms that have led to the consolidation of democratic gains or, alternatively, to the breakdown of democracy in six very diverse countries, located in four different regions of the world. Three of these countries opted for a president-parliamentary system, and three for a premier-presidential one. The country cases are complemented by four regional studies of Western Europe, Asia, the post- communist region, and Africa, respectively.

Institutional variation and political dynamics So, what have the actual political dynamics been in our country cases and regions? Can we tease out some general patterns that help illustrate the mechanics of semi-presidentialism at work? How do the different constitutional incentive structures play themselves out in practice, in premierpresidential versus president-parliamentary systems? Moldova and Haiti are examples of countries having adopted a premierpresidential subtype of semi-presidentialism. Portugal changed from a president-parliamentary to a premier-presidential system in 1982 and Ukraine made a similar change in 2006. Taiwan, the Weimar Republic, Portugal pre-1982 and Ukraine from 1996 till 2004 are, on the other hand, examples of the president-parliamentary subtype of semi-presidentialism. We can see that there is no evident pattern: one premier-presidential regime has survived (Portugal), one broke down repeatedly (Haiti), and two have struggled by (Ukraine since 2004, and Moldova until 2000 when the country changed to a parliamentary constitution); similarly, one presidentparliamentary regime has survived (Taiwan), one broke down (Weimar), and one muddled through (Ukraine until 2004). Before we discuss the role of semi-presidential institutions in these outcomes, we need to place these countries in their appropriate economic context. Following Przeworski et al. (2000), we would not expect rich countries to collapse, but democracy in poor countries may be fragile. In terms of the countries discussed in this book, Taiwan is, despite financial crises, a relatively rich country. We would not expect democracy to breakdown in such a context, but institutions may still generate political crises and President Chen Shui-bian’s decision to avoid cohabitation and continue with a minority government from 2000– 2008 was politically traumatic. For its part, Portugal is now a developed

9780230_242920_16_cha15.indd 267

11/11/2010 5:23:32 PM

268 Robert Elgie, Sophia Moestrup, and Yu-Shan Wu

country, though in the late 1970s and early 1980s, before EU membership, it was relatively poor. Therefore, at that time, arguably democracy was not economically determined. The other countries discussed here were or are economically underdeveloped.1 In these countries, democracy collapsed in two (Haiti and Weimar) and survived, though only just, in the other two (Moldova and Ukraine). Controlling for economic performance, what was the observable impact of semi-presidentialism in all of these countries? The two stable democracies among our country cases, Portugal and Taiwan, both adopted semi-presidentialism in the process of democratization. In Portugal, the direct election of the president was introduced to increase the legitimacy of the office, as mentioned above. In Taiwan, accountability of the prime minister to parliament came about as a result of negotiated constitutional amendments between president and parliament. The functioning of semi-presidential institutions in Portugal and Taiwan has been eased by institutional incentives that have favored accommodation and adjustments on either the president’s or parliament’s side, while avoiding either parliamentary or presidential excesses. In Portugal, as Jalali shows, party system dynamics have largely precluded presidential leadership within the executive. The president has typically not been the leader of a political party and has lacked sufficient party support to count on in the event of open conflict with parliament. Instead, the president has skilfully used his limited legislative powers as well as his nonlegislative powers (the threat of dissolution of parliament) to influence policy making. Far from being powerless, presidents have used the media and their public agenda-setting powers to influence political outcomes, counting on the government’s unwillingness to fight the president head- on on controversial issues. Mutual checks on power between president and parliament have worked to the advantage of Portugal’s consolidating democracy. In Taiwan, the president, though lacking the backing of a majority in parliament from 2000–2008, was able to impose his choice of prime minister and avoid cohabitation, thanks to the low credibility of a successful vote of no- confidence, as Wu and Tsai point out. High coordination hurdles between opposition parties and the significant personal costs of running an election campaign dissuaded deputies from toppling minority governments appointed by President Chen, which would almost certainly have led to the dissolution of parliament. On the other hand, the majority opposition was able to successfully defeat presidential policy initiatives in parliament. The executive and the legislature settled on equilibrium where neither party was the total loser, nor the total winner. Ukraine and Moldova have suffered under disjointed constitutional arrangements inherited from piecemeal constitutional reform during the transition from Soviet rule, as Matsuzato and Crowther show, and both countries have experienced significant constitutional changes. Ukraine started out as a president-parliamentary system, with a strong presidency,

9780230_242920_16_cha15.indd 268

11/11/2010 5:23:32 PM

Semi-Presidentialism: What Have We Learned?

269

while Moldova adopted a premier-presidential system. Both countries were characterized by weak party systems, dominated by competing factions of post-Soviet elites struggling for control over assets and opportunities in the post- communist economy. Factionalism within the legislature contributed to presidential expansionism in both countries. In Ukraine, popular pressure for increased democratization following fraud in the 2004 presidential elections led to a compromise between parliament and president Kuchma for the introduction of a premier-presidential system (calculated on Kuchma’s side to reduce the powers of his successor). In Moldova, in the lead-up to the 2000 presidential elections, ongoing power struggles between president and parliament culminated with the adoption of constitutional amendments by parliament that transformed the country into a parliamentary republic. Even though Freedom House scores have improved in Ukraine following the transition to a premier-presidential system, as Protsyk mentions, persistent conflicts between the president and the parliamentary majority could easily lead to further constitutional changes. The Moldovan case illustrates how, paradoxically, a change to a parliamentary regime can actually serve to resolve tension between president and parliament over control over the executive in favor of the president – in the process inhibiting rather than helping democratic progress. As Crowther points out, in 2001 a new consolidated communist majority in the Moldovan parliament elected a communist president and backed a new communist prime minister. This new, coherent executive supported by the communist parliamentary majority enabled the president to control the political agenda to an extent that had previously been impossible. Authority effectively became concentrated in the presidency, supported by the surviving clientelistic networks of the communist party. Consequently, Moldova’s Freedom House score worsened slightly. Finally we have the very disparate cases of the Weimar Republic and Haiti, Weimar counting among the earliest examples of semi-presidential regimes, and Haiti among the latest. The Weimar Republic saw the introduction of a president-parliamentary regime with the president intended to play the role of a regulator while managing Germany’s difficult postwar issues, as Shen shows. In Haiti, in contrast, the constituent assembly opted for a premier-presidential system in an effort to break a historical pattern of presidential authoritarianism, as Moestrup discusses. While the president’s constitutional powers in the Weimar Republic were quite considerable, including significant emergency powers, they were very limited in the Haitian constitution. Common for the two countries was an extremely fractured party system and fragmented parliamentary representation, hindering the formation of stable governing majorities. The downfall of the Weimar Republic came about in a context of government instability as the president overstepped his intended role, dissolving parliament, appointing a minority government and ruling through his emergency powers. In Haiti,

9780230_242920_16_cha15.indd 269

11/11/2010 5:23:33 PM

270

Robert Elgie, Sophia Moestrup, and Yu-Shan Wu

on the other hand, the skewed balance of powers in favor of parliament entailed a strong temptation to seek control of the legislative majority in order to avoid political gridlock, resulting in tampering with vote results and appeals to mob support. In both instances, insufficient checks and balances (on the president in the Weimar Republic and on parliament in Haiti) in the context of a highly fractured society entailed a lack of incentives for power-sharing and negotiation. This overview of our six country cases, complemented with the regional analyses in the book, tends to confirm the broader findings outlined above. President-parliamentary systems can indeed easily lead presidents to overstep the boundaries of their powers, through a lack of incentives for negotiation with the legislature in the event that presidential and parliamentary majorities do not coincide. The Weimar case is the prime example of this dynamic. However, as Wu and Tsai’s analysis of Taiwan illustrates, the electoral and party systems can provide separate incentives for accommodation through mutual checks on power between the executive and the legislature, in the process facilitating the consolidation of democracy. Portugal epitomizes a well-functioning premier-presidential system, where the president serves effectively as a mediator through his power to influence the political agenda. Haiti, on the other hand, illustrates how skewing powers excessively in favor of parliament can be just as detrimental for power sharing and negotiated political solutions as powers overly tilted in favor of the president. In both cases, the lack of sufficient checks on power (of the parliament in the case Haiti) undermines the incentives for power-sharing and negotiation otherwise provided by the existence of a dual executive. In Haiti, a factionalized party system has served to further inhibit progress toward democratic consolidation, as the president has been faced with a legislature dominated by shifting alliances of factions representing particularistic interests. Negotiated settlements in such a context are akin to buy- offs and do not provide incentives for greater institutionalization of political parties. The lack of well-functioning parties inhibits the emergence of a practice of power-sharing between different societal interests, as has also been the case in Ukraine and Moldova. Interestingly, we note from our cases that the issue at hand has generally not been divisions within the dual executive, or cohabitation, the major purported danger expounded by critics of semi-presidentialism. This confirms the findings from Elgie’s and Schleiter’s study. Instead, we find conflict between the president and the legislature at the root of democratic troubles in the sample of countries discussed above. In some cases, the existence of the prime minister post contributes to these conflicts, as the president imposes his choice of prime minister on a fractured parliamentary majority or otherwise seeks to avoid cohabitation. In sum, the political dynamics discussed above illustrate the importance of countervailing powers between parliament and president to secure

9780230_242920_16_cha15.indd 270

11/11/2010 5:23:33 PM

Semi-Presidentialism: What Have We Learned?

271

sufficient checks and balances and incentives for power-sharing. The categorization of premier-presidential versus president-parliamentary systems provides a good first cut to determine whether such countervailing powers exist. In president-parliamentary systems the president’s power to dismiss the prime minister at will tends to be accompanied by significant legislative powers as well, overly tilting power in favor of the president. In premierpresidential systems, on the other hand, weaker presidential powers can provide incentives for the president to play more of a regulatory role, as in Portugal. Future comparative research can help further elucidate the specific conditions under which premier-presidentialism is most likely to help a nascent democracy take root. As we have seen, a weak party system and lack of incentives for accommodation on the parliament’s side are not conducive to the long-term survival of democratic institutions, even in premierpresidential systems.

The evolution of semi-presidentialism The first semi-presidential constitutions were adopted after the First World War. However, semi-presidentialism only became a common constitutional choice in the 1990s following the “third wave” of democratization. Since 2000 a number of countries have switched into the semi-presidential camp and for similar sorts of reasons that were witnessed a decade earlier. For example, in the Central African Republic and Mauritania there were coups and when democracy was reintroduced political leaders reverted back to carbon copies of former constitutions with a strong French influence. Equally, as Reilly notes in his chapter, East Timor, a former Portuguese colony, adopted a classic Lusophone semi-presidential constitution when it gained independence from Indonesia. Thus, the process of constitutional mimetism is ongoing. Most of the other countries that have chosen, or stumbled upon, semi-presidentialism since 2000 have tended to do so as a way of trying to signal that they have introduced a system with more checks and balances than was previously the case. For instance, in Egypt there was a desire to portray the system as being somewhat more democratic than the semi-authoritarian reputation that it had gathered. Thus, the direct election of the president was first introduced and then the government was made responsible to the legislature. Whether or not these changes have made a difference to the external perception of the Egyptian system is a moot point, but the motivation for introducing a semi-presidential constitution was clear and familiar to observers of other countries. The case of Georgia also fits this pattern, with perhaps a little more success in terms of how the democratization process has been perceived there. Interestingly, since 2000, Turkey is the only case where a parliamentary country has deliberately chosen semi-presidentialism.2 Even there, the choice of semi-presidentialism was motivated less by the classic French-like desire to introduce a strong

9780230_242920_16_cha15.indd 271

11/11/2010 5:23:33 PM

272

Robert Elgie, Sophia Moestrup, and Yu-Shan Wu

president and more to maximize the chances of the election of an AKP president whom parliament resisted. In other words, since 2000, no country has chosen semi-presidentialism because of a desire to strengthen the presidential leadership structures of the country. Moreover, no country has chosen semi-presidentialism because it was a natural compromise between equally balanced supporters of presidentialism and parliamentarism. Given institutions are “sticky” and with no sign of another “wave” of democratization on the horizon, perhaps we are now entering an era where only a limited number of countries are likely to adopt semi-presidentialism and where the most likely countries to do so are semi-authoritarian countries that want to put a veneer on their democratic credentials. Given those countries are the ones that are most likely to collapse or never to make the transition to democracy in the first place, then semi-presidentialism’s reputation as a poor governance choice is unlikely to be challenged. In terms of countries that have abandoned semi-presidentialism since 2000 and not (yet) returned, Moldova, as noted in Crowther’s chapter, did so in favor of parliamentarism. By contrast, both Angola and Niger did so in favor of presidentialism, only for the latter to experience a coup almost immediately. For its part, Madagasacar’s semi-presidential constitution was replaced by an interim regime with the coup in 2009. So, there has been no particular pattern in terms of the exit from semi-presidentialism in the last decade. While institutions may be sticky, it is clear that countries do change their constitutional structures (Elkins et al. 2009) and sometimes quite frequently. So, as part of a natural pattern of constitutional change, we should expect some countries to abandon semi-presidentialism in the years to come. What this means is that, overall, without a new “wave” of democratization that might again increase the number of semi-presidential constitutions, the number of semi-presidential regimes is likely to remain relatively constant for the time being. Since 2000, there seems to have been an incremental shift toward semi-presidentialism in some countries recently and an equivalent drift away. Thus, in terms of the number of countries with semipresidential constitutions, we should probably expect only a small change up or down in the foreseeable future. That said, we may well expect greater institutional evolution within the universe of semi-presidential countries. For example, there are tantalizing signs that countries are more willing to shift from one form of semipresidentialism to another. For example, only one country shifted directly from either premier-presidentialism to president-parliamentarism or vice versa in the both the 1980s and the 1990s (Portugal in the former decade and Madagascar in the latter). However, Armenia, Croatia, São Tomé e Príncipe, Senegal and Ukraine have all done so since 2000. If in general terms, as both Elgie and Schleiter and Moestrup argue in this volume, the subtype of semi-presidentialism makes a difference to democratic performance, then such shifts are potentially significant. In terms of the countries that have

9780230_242920_16_cha15.indd 272

11/11/2010 5:23:33 PM

Semi-Presidentialism: What Have We Learned?

273

made such a shift in the last decade, Senegal is the only one that has chosen to move from premier-presidentialism to president-parliamentarism and, in 2008, Freedom House recorded a shift in its status from Free to Partly Free. Many factors are likely to have caused such a downturn, but we should be sensitive to the possibility that the country’s institutional change may have had an effect in the apparent decline in the quality of governance there. At the time of writing, Georgia, where democracy is clinging on precariously, is considering a shift to premier-presidentialism. If the empirical findings in some of the chapters of this book are correct and bearing in mind that statistical associations are always general and do not necessarily tell us anything about individual countries, then Georgia may have embarked up a fruitful path. By contrast, Mali, West Africa’s democratic success story to date, has plans to shift to president-parliamentarism. This is a potentially worrying move. More generally still, many of the chapters in this book have emphasized the importance of “second-level” constitutional powers, such as the power of dissolution, decree power, emergency powers, and so on. In addition, in his chapter, Lin emphasized very forcefully and convincingly the importance of the interaction between electoral systems and institutions under semipresidentialism. In their recent book, Elkins et al. (ibid.) note that the vast majority of countries have not constitutionalized the precise form of their electoral system. If, as Lin suggests, the particular type of electoral system matters so much for the performance of semi-presidential countries and the reform of the electoral system is a potentially plausible political option in most semi-presidential countries, then the politics of semi-presidentialism may vary considerable in the future as governments change the electoral rules to their advantage and, in so doing, affect the conduct of semipresidentialism. In recent years, other types of reforms have also come into vogue. The abolition of term limits is one such example. Such a reform is almost always a proxy for the personalization of power and a sign that democracy is suffering or, if it is already in deep trouble, as in countries like Azerbaijan where presidential term limits were abolished in 2009, then it is a sign that any transition to democracy may still be a long way away. In his chapter, Protsyk records the constitutional changes that have taken place in the former USSR. As would-be autocrats try to maneuver their way into a lifetime of power, and sometimes a dynastical succession, then we might expect to see more changes that threaten the performance of democracy. Semi-presidentialism is not unique in this regard. Presidentialism suffers from this tendency too. However, what this point suggests is that the evolution of semi-presidential countries toward a balance of executive/legislative power, indeed perhaps toward a weaker president and more active government and parliament generally (Fish 2006), are the ones that, in the long-term, are more likely to be successful. As Moestrup argues, the Haitian example demonstrates that a strong parliament is not a good thing in and

9780230_242920_16_cha15.indd 273

11/11/2010 5:23:33 PM

274

Robert Elgie, Sophia Moestrup, and Yu-Shan Wu

of itself. However, as part of a move toward a diminution of presidential powers and a move to an overall balance of power and true limited government, then a shift away from highly presidentialized semi-presidentialism is perhaps the way in which countries with semi-presidential constitutions might be encouraged to evolve.

Conclusion This volume takes stock and expands our knowledge of semi-presidentialism. It breaks new ground by offering a truly global review of semi-presidential countries, and looking into each and every step of the institutional stream, starting with the choice of system, its operation, performance, impact, and evolution (both into and out of the system and among its subtypes). In this way, this study provides a breadth and depth that has been unavailable previously. We use the dichotomy of premier-presidentialism and presidentparliamentarism as our major classifying principle. We demonstrate the usefulness of the dichotomy, but also take note of other variables that may impact on democratic survival of a semi-presidential system. This volume thus moves forward the frontier of the study of semi-presidentialism, a constitutional system that has joined presidentialism and parliamentarism in a tripartite constitutional universe, and yet remained sorely understudied.

Notes 1. Using Angus Maddison’s data set, in 1933 Weimar Germany had a per capita GDP in 1990 Geary-Khamis $ that was very similar to countries such as Moldova in 2006. See http://www.ggdc.net/maddison/ 2. In both Serbia and Montenegro, semi-presidential constitutions were already in place prior to the breakup of the union. So, these countries are excluded from the list here.

9780230_242920_16_cha15.indd 274

11/11/2010 5:23:34 PM

Bibliography Ahn, K. W. 1997. The influence of American constitutionalism on South Korea. Southern Illinois University Law Review 22: 71–115. Alesina, A., A. Devleeschauwer, W. Easterly, S. Kurlat, and R. Wacziarg 2002. Fractionalisation. NBER Working Paper Series Working Paper 9411. Anschütz, G. 1960. Die Verfassung des deutschen Reichs. Darmstadt: Wissenschaftliche Buchgesellschaft Verlag. Araújo, A. de 2003. El Presidente de la República en la evolución del sistema político de Portugal in B. Gómez, A. Barreto, and P. Magalhães (eds.), El Sistema Político de Portugal, Madrid: Siglo XXI, 83–111. Arter, D. 1985. Government in Finland: A ‘semi-presidential system’? Parliamentary Affairs 38: 472–495. Arter, D. 1999. Finland in R. Elgie (ed.), Semi-Presidentialism in Europe. Oxford: Oxford University Press, 48–66. Avillez, M. J. 1996. Soares: Democracia. Lisbon: Público. Avillez, M. J. 1997. Soares: O Presidente. Lisbon: Público. Azevedo, E. and L. Nijzink 2007. Semi-presidentialism in Guinea-Bissau: The lesser of two evils? in R. Elgie and S. Moestrup (eds.) Semi-Presidentialism Outside Europe, London: Routledge, 137–160. Bahro, H., B. Bayerlein, and E. Veser 1998. Duverger’s concept: Semi-presidential government revisited. European Journal of Political Research 34(2): 201–224. Baker, B. 2006. Cape Verde: The most democratic nation in Africa? Journal of Modern African Studies 44(4): 493–511. Barroso, J. D. 1986. Les conflits entre le Président portugais et la majorité parlementaire de 1979 à 1983 in M. Duverger (ed.), Les Régimes Semi-Présidentiels, Paris: Presses Universitaires de France, 237–254. Bartol, S. 2007. Vneseni do konstytutsii zminy perebachyly stvorennya zhakhlyvoho monstra – “korporatsii bil’shosti”. Chaspys parlament, 1: 53. Bartolini, S. 1984. Sistema partitico ed elezione diretta del capo dello stato in Europa. Rivista Italiana di Scienza Politica 14(2): 223–243. Baylis, T. A. 1996. Presidents versus Prime Ministers: Shaping the executive authority in Eastern Europe. World Politics 48(3): 297–323. Beck, T., G. Clarke, A. Groff, P. Keefer, and P. Walsh 2001. New tools in comparative political economy: The database of political institutions. World Bank Economic Review 15(1): 165–176. Beissinger, M. R. and C. Young eds. 2002. Beyond State Crisis? Postcolonial Africa and Post-Soviet Eurasia in Comparative Perspective. Washington: Woodrow Wilson Center Press. Beliaev, M. 2006. Presidential Powers and Consolidation of New Postcommunist Democracies. Comparative Political Studies 39: 375–398. Bellegarde-Smith, P. 1990. Haiti, The Breached Citadel. Boulder, CO: Westview Press. Bernhard, M., T. Nordstrom and C. Reenock 2001. Economic Performance, Institutional Intermediation, and Democratic Survival. Journal of Politics 63(3): 775–803. Bernhard, M., C. Reenock and T. Nordstrom 2003. Economic performance and survival in new democracies: Is there a honeymoon effect? Comparative Political Studies 36(4): 404–431. 275

9780230_242920_17_bib.indd 275

11/16/2010 3:42:04 PM

276

Bibliography

Bernhard, M., C. Reenock and T. Nordstrom 2004. The legacy of western overseas colonialism on democratic survival. International Studies Quarterly 48: 225–250. Birch, S. 2003. Electoral Systems and Political Transformation in Post- Communist Europe (One Europe or Several?). London: Palgrave. Birch, S. 2008. Ukraine: Presidential power, veto strategies and democratization in R. Elgie and S. Moestrup (eds.), Semi-Presidentialism in Central and Eastern Europe. Manchester: Manchester University Press, 219–238. Blondel, J. 1992. Dual leadership in the contemporary world in A. Lijphart (ed.), Parliamentary versus Presidential Government. Oxford: Oxford University Press. Bogdanor, V. 1996. Politics and the Constitution: Essays on British Government. Aldershot: Dartmouth. Boix, C. and S. C. Stokes 2003. Endogenous Democratization. World Politics 55: 517–549. Boldt, H. 1997. Die Stellung von Parlament und Parteien in der Weimarer Rechsverfassung in E. Kolb and W. Mühlhausen (eds.), Demokratie in der Krise, München: R. Oldenbourg Verlag, 19–58. Box- Steffensmeier, J. M. and B. S. Jones 2004. Event History Modeling: A Guide for Social Scientists. Cambridge: Cambridge University Press. Bracher, K. D. 1984. Die Auflösung der Weimarer Republik. Düsseldorf: Droste Verlag. Bracher, K. D. 1987. Die Auflösung der Republik. Gründe und Fragen in G. Schulz (ed.), Weimarer Republik, Würzburg: Ploetz Verlag, 123–138. Bratton, M. and N. van de Walle 1997. Democratic Experiments in Africa: Regime Transition in Comparative Perspective. Cambridge: Cambridge University Press. Brown, N., F. Frison-Roche, and S. Moestrup. 2004. The Role of Constitutions in Political Transitions: Semi-Presidentialism in Palestine, Bulgaria, and Haiti. Seminar held by the Elliott School, The George Washington University, April 27. www.cersa.org/IMG/pdf/semi_presidentialism_reforme_ de_l_etat.pdf, accessed on November 3, 2007. Brunet, R. 1922. The New German Constitution. New York: Alfred. A. Knopf Press. Brüning, H. 1970. Heinrich Brüning Memoiren: 1918–1934. Stuttgart: Deutsche VerlagsAnstalt Verlag. Cabestan, J-P. 1997. The Constitutional Reform: Is Taiwan Moving towards a FrenchStyle Semi-presidentialism. China Perspectives 14: 40–44. Campinos, J. 1986. Le cas portugais de 1979 à 1983: le Président opposé à la majorité in M. Duverger (ed.), Les Régimes Semi-Présidentiels. Paris: Presses Universitaires de France, 209–236. Canepa, A. 2000. Il sistema semipresidenziale: Aspetti teorici e di diritto positivo. Torino: Giappichelli. Carey, J. M. and M. S. Shugart 1998. Executive Decree Authority. Cambridge: Cambridge University Press. Carlson, R. 1999. Presidentialism in Africa: Explaining Constitutional Choice. PhD dissertation in political science, University of Chicago. Carrilho, M. 1985. Forças Armadas e Mudança Política em Portugal no séc. XX: Para uma explicação sociológia do papel dos militares. Lisbon: Imprensa Nacional – Casa da Moeda. Catón, M. and F. Tuesta Soldevilla 2008. Political parties in conflict-prone societies in Latin America in Benjamin Reilly and Per Nordlund (eds.), Political Parties in Conflict-Prone Societies: Regulation, Engineering and Democratic Development, Tokyo: United Nations University Press, 126–157.

9780230_242920_17_bib.indd 276

11/16/2010 3:42:04 PM

Bibliography

277

Ceccanti, S., O. Massari, and G. Pasquino 1996. Semipresidenzialismo: Analisi delle esperienze europee. Bologna: Il Mulino. Charles, J. 2009. Growth seen for Haiti, Guyana: Storm-battered Haiti and Guyana are the only Caribbean nations to project economic growth this year. Miami Herald, November 10, 2009. Chehabi, H. E. and A. Stepan eds. 1995. Politics, Society, and Democracy, Comparative Studies. Boulder, CO: Westview Press. Cheibub, J. A. 2007. Presidentialism, Parliamentarism, and Democracy. Cambridge: Cambridge University Press. Cheibub, J. A. and S. Chernykh 2008. Constitutions and democratic performance in semi-presidential democracies. Japanese Journal of Political Science 9(3): 269–303. Chi, W- S. 1992. Ideological Conflicts in Modern China: Democracy and Authoritarianism. New Brunswick, NJ: Transaction Books. Clark, J. F. 2005. The collapse of the democratic experiment in the Republic of Congo in L. A. Villalón and P. VonDoepp (eds.), The Fate of Africa’s Democratic Experiments. Bloomington, IN: Indiana University Press, 96–125. Colton, T. J. 1995. Superpresidentialism and Russia’s backward state. Post-Soviet Affairs 11(2): 144–149. Colton, T. J. and H. Hale 2009. The Putin vote: Presidential electorates in a hybrid regime. Slavic Review 68 (3): 473–503. Colton, T. J. and C. Skach 2005. The Russian predicament. Journal of Democracy 16(3): 113–126. Conac, G. 2007. Semi-presidentialism in a Francophone context in R. Elgie and S. Moestrup (eds.), Semi-Presidentialism Outside Europe: A Comparative Study. London: Routledge, 78–91. Constitution de la République d’Haїti, 29 mars 1987. Port-au-Prince: Editions Henri Deschamps. Constituţia Republicii Moldova 1994. Chişinău: Direcţia de Stat pentru Asigurarea Informaţională, MOLDPRES. Copper, J. F. 2003. Taiwan: Consolidating its democracy? China: An International Journal, 1(2): 326–338. Cranenburgh, O. van 2006. Namibia: Consensus institutions and majoritarian politics. Democratization 13(4): 584–604. Cranenburgh, O. van 2007. “Big Men” rule: Presidential power in anglophone and francophone African countries. Paper presented for the Panel on ‘Presidentialism in Comparative Perspective’ at the General Conference of the ECPR, Pisa, 6–8 September, 2007. Crowther, W. 2007. Development of the Moldovan parliament one decade after independence: Slow going. Journal of Legislative Studies, 13(1): 99–120. Crowther, W. 2007. The politics of democratization in post- communist Moldova in K. Dawisha and B. Parrott (eds.), Democratic Changes and Authoritarian Reactions in Russia, Ukraine, Belarus, and Moldova, London: Cambridge University Press, 282–329. Crowther, W. and Y. Josanu 2004. Moldova in S. Bergland, F. Aarebrot, and J. Ekman, (eds.), Handbook of Political Change in Eastern Europe, 2nd edition, Cheltenham: Edward Elgar, 549–592. Cruz, M. B. da 1995. Instituições Políticas e Processos Sociais. Venda Nova: Bertrand. Dahl, R. 1971. Polyarchy: Participation and Opposition. New Haven, CT: Yale University Press.

9780230_242920_17_bib.indd 277

11/16/2010 3:42:04 PM

278

Bibliography

Dawisha, K. and B. Parrott eds. 1997a. Conflict, Cleavage and Change in Central Asia and the Caucasus. Cambridge: Cambridge University Press. Dawisha, K. and B. Parrott eds. 1997b. The Consolidation of Democracy in East- Central Europe. Cambridge: Cambridge University Press. Dawisha, K. and B. Parrott eds. 1997c. Politics, Power and the Struggle for Democracy in Southeast Europe. Cambridge: Cambridge University Press. De Gaulle, C. 1964. Conférence de presse du général de Gaulle, 31 janvier 1964 (extraits). Available at: www.ladocfrancaise.gouv.fr/dossier_actualite/cohabitation/conf64. shtml, accessed on June 29, 2004. Dearing, J. W. and E. M. Rogers 1996. Agenda-Setting. Thousand Oaks, CA: Sage Publications. Diamond, L., J. J. Linz, and S. M. Lipset 1995. Introduction, in L. Diamond, J. J. Linz and S. M. Lipset (eds.), Politics in Developing Countries. 2nd edition, Boulder, CO: Lynne Rienner, 1–36. Downs, A. 1957. An Economic Theory of Democracy. New York: Harper. du Bois de Gaudusson, J., G. Conac, and C. Desouches, Christine 1998. Les Constitutions africaines publiées en langue française, vol. 1 and 2. Paris: La Documentation française. Dupuy, A. 2003. Who is Afraid of Democracy in Haiti? A Critical Reflection. Haiti Papers, No. 7, June 2003 (a publication of The Haiti Program, a unit of Programs in International Affairs at Trinity College). Dupuy, A. 2006. The Prophet and Power: Jean-Bertrand Aristide, the International Community, and Haiti. New York: Rowman & Littlefield Publishers, Inc. Duverger, M. 1963. Political Parties: Their Organization and Activity in the Modern State. New York: Wiley. Duverger, M. 1978. Echec au roi. Paris: Albin Michel. Duverger, M. 1980. A new political system model: Semi-presidential government. European Journal of Political Research 8(2): 165–187. Duverger, M. 1986a. Le concept de régime semi-présidentiel in M. Duverger (ed.), Les Régimes Semi-Présidentiels, Paris: Presses Universitaires de France, 7–17. Duverger, M. ed. 1986b. Les régimes semi-présidentiels. Paris: Presses Universitaires de France. Duverger, M. 1996. Le système politique français. 21st edition. Paris: Presses Universitaires de France. Duverger, M. 1997. The political system of the European union. European Journal of Political Research 31(1): 137–146. East European Constitutional Review 1999. Constitution Watch: Moldova 8(4). East European Constitutional Review 2001. Constitution Watch: Moldova 10(2). Easter, Gereld, M. 1997. Preference for presidentialism: Postcommunist REGIME CHANGE in Russia and the NIS. World Politics 49(2): 184–211. The Economist 1998. Haiti, Help wanted: one government. July 25th–31st, p. 36. The Economist 1999. Haitians turn their backs on President Preval – and politics. Dec. 4th–10th, pp. 35–36. Egset, W. and S. Kuttner 2007. Governance and institutions in The World Bank, Social Resilience and State Fragility in Haiti, Washington, DC: The World Bank, 43–66. Elgie, R. 1998. The classification of democratic regime types: Conceptual ambiguity and contestable assumptions. European Journal of Political Research 33(2), 219–38. Elgie, R. 1999a. France in R. Elgie (ed.), Semi-Presidentialism in Europe. Oxford: Oxford University Press, 67–85. Elgie, R. 1999b. The politics of semi-presidentialism in R. Elgie (ed.), SemiPresidentialism in Europe. Oxford: Oxford University Press, 1–21.

9780230_242920_17_bib.indd 278

11/16/2010 3:42:04 PM

Bibliography

279

Elgie, R. 1999c. Semi-presidentialism and comparative institutional engineering in R. Elgie (ed.), Semi-Presidentialism in Europe. Oxford: Oxford University Press, 281–299. Elgie, R. 2002. La Cohabitation de longue durée: Studying the 1997–2002 Experience. Modern and Contemporary France 13(3): 297–311. Elgie, R. 2004. Semi-presidentialism: Concepts, consequences and contesting explanations. Political Studies Review 2(3), 314–330. Elgie, R. 2005. A fresh look at semipresidentialism: Variations on a theme. Journal of Democracy 16(3): 98–112. Elgie, R. 2007a. What is semi-presidentialism and where is it found? in R. Elgie and S. Moestrup (eds.), Semi-Presidentialism Outside Europe. London: Routledge, 1–13. Elgie, R. 2007b. Varieties of semi-presidentialism and their impact on nascent democracies. Taiwan Journal of Democracy 3(2): 53–71. Elgie, R. 2008. The perils of semi-presidentialism: Are they exaggerated? Democratization 15(1): 49–66. Elgie, R. 2009a. Cohabitation (15) – São Tomé (updated). www.semipresidentialism. com/The_Semi-presidential_One/, blog entry June 24, 2009, accessed on February 14, 2010. Elgie, R. 2009b. Haiti – Government appointed. www.semipresidentialism.com/The_ Semi-presidential_One/, blog entry November 11, 2009, accessed on November 12, 2009. Elgie, R. 2010. São Tomé – President withdraws ministers from government. www. semipresidentialism.com/The_Semi-presidential_One/, blog entry January 11, 2010, accessed on February 14, 2010. Elgie, R. and I. McMenamin 2008. Divided executives and democratisation. Japanese Journal of Political Science 9(3): 323–340. Elgie, R. and S. Moestrup 2007. The choice of semi-presidentialism and its consequences in R. Elgie and S. Moestrup (eds.), Semi-Presidentialism Outside Europe: A Comparative Study, London: Routledge, 237–248. Elgie, R. and S. Moestrup eds. 2007. Semi-Presidentialism Outside Europe. London: Routledge. Elgie, R. and S. Moestrup 2008. The impact of semi-presidentialism on the performance of democracy in Central and Eastern Europe in R. Elgie and S. Moestrup (eds.), Semi-Presidentialism in Central and Eastern Europe, Manchester: Manchester University Press, 239–258. Elgie, R. and S. Moestrup eds. 2008. Semi-Presidentialism in Central and Eastern Europe. Manchester: Manchester University Press. Elkins, Z., T. Ginsburg, and J. Melton 2009. The Endurance of National Constitutions. Cambridge: Cambridge University Press. Epstein, D. L., R. Bates, J. Goldstone, I. Kristensen, and S. O’Halloran 2006. Democratic transitions. American Journal of Political Science 50(3): 551–569. Erikson, D., E. Grant, F. Braun, G. Hall, K. Bain, M. Mattner, S. Kuttner, and W. Egset 2007. Social Resilience and State Fragility in Haiti. Washington, DC: The World Bank, 67–74. European Commission 2008. European commission for democracy through law (Venice Commission), “Opinion on the Draft Constitution of Ukraine” (Prepared by a Working Group Headed by Mr V.M. Shapoval). Adopted by the Venice Commission at its 75th Plenary Session (Venice, June 13–14, 2008). Fatton Jr., R. 2000. Constitution without constitutionalism: Haiti and the Vagaries of democratization. New West Indian Guide 74(1 & 2): 5–32.

9780230_242920_17_bib.indd 279

11/16/2010 3:42:05 PM

280 Bibliography Fatton Jr., R. 2002. Haiti’s Predatory Republic, The Unending Transition to Democracy. Boulder, CO: Lynne Rienner. Fatton Jr., R. 2007. The Roots of Haitian Despotism. Boulder, CO: Lynne Rienner. Finch, C. M. 2002. Mongolia in 2001: Political consolidation and continued economic reform. Asian Survey 42(1): 39–45. Finger, A. 1923. Das Staatsrecht des Deutschen Reichs: Der Verfassung vom 11. August 1919. Stuttgart: Ferdinand Enke Verlag. Fish, S. M. 1998a. Mongolia: Democracy without prerequisites. Journal of Democracy 9(3): 127–141. Fish, S. M. 1998b. The determinants of economic reform in the post- communist world. East European Politics and Societies 12(1): 31–78. Fish, S. M. 2001. The inner Asian anomaly: Mongolia’s democratization in comparative perspective. Communist and Post- Communist Studies 34(3): 323–338. Fish, S. M. 2006. Stronger legislatures, stronger democracies. Journal of Democracy 17(1): 5–20. Fouchard, J. 1988. Les marrons de la liberté. Port-au-Prince: Editions Henri Deschamps. Frain, M. 1995. Relações entre o Presidente e o primeiro-ministro em Portugal: 1985–1995. Análise Social 133: 653–678. Freedom House 2010. Freedom in the World Survey, 2010. www.freedomhouse.org. Freire, A. and A. C. Pinto 2005. O Poder dos Presidentes: A República Portuguesa em Debate. Lisbon: Campo da Comunicação. Friedrich, C. J. 1933. The development of executive power in Germany. American Political Science Review 27(2): 185–203. Frison-Roche, F. 2005. Le “modèle semi-présidentiel” comme instrument de la transition en Europe post- communiste, Bulgarie, Lituanie, Macédoine, Pologne, Roumanie et Slovénie. Bruxelles: Bruylant. Frison-Roche, F. 2007. Semi-presidentialism in a post- communist context in R. Elgie and S. Moestrup (eds.), Semi-Presidentialism Outside Europe: A Comparative Study. London: Routledge, 56–77. Frye, T. 1997. A politics of institutional change: Post- communist presidencies. Comparative Political Studies 30: 523–552. Fulbrook, M. 1990. A Concise History of Germany. Cambridge: Cambridge University Press. Fusaro, C. 1998. Le radici del semi-presidenzialismo. Viaggio alle origini di un modello cui si guarda in Italia. Soveria Mannelli: Rubbettino. Gallagher, M. 1999. Republic of Ireland in R. Elgie (ed.), Semi-Presidentialism in Europe. Oxford: Oxford University Press, 104–123. Galvan, D. 2001. Political turnover and social change in Senegal. Journal of Democracy 12(3): 51–62 Gasiorowski, M. J. 1995. Economic crisis and political regime change: An event history analysis. American Political Science Review 89(4): 882–897. Gasiorowski, M. J. and T. J. Power 1998. The structural determinants of democratic consolidation: Evidence from the Third World. Comparative Political Studies 31(6): 740–771. Giovannelli, A. ed. 1998. Il semipresidenzialismo: dall’arcipelago europeo al dibattito italiano. Torino: Giappichelli. Gorges, M. J. 2001. New institutionalist explanations for institutional change: A note of caution. Politics 21(2): 137–145. Grofman, B., S. Lee, E. A. Winckler, and B. Woodall eds. 1999. Elections in Japan, Korea, and Taiwan under the Single Non-Transferable Vote. Ann Arbor, MI: The University of Michigan Press. Gros, J- G. 1997. Haiti’s flagging transition. Journal of Democracy 8(4), 94–109.

9780230_242920_17_bib.indd 280

11/16/2010 3:42:05 PM

Bibliography

281

Gusy, C. 1991. Weimar - die Wehrlos Republik? Tübingen: Mohr Siebeck Verlag. Gusy, C. 1997. Die Weimarer Reichsverfassung. Tübingen: Mohr Siebeck Verlag. Hale, H. 2006. Democracy or autocracy on the march? The colored revolutions as normal dynamics of patronal presidentialism. Communist and Post- Communist Studies 39: 305–329. Hale, H. 2008. What Makes Dominant Parties Dominant in Hybrid Regimes? The Unlikely Importance of Ideas in the Case of United Russia. Paper presented at the 2008 Annual Meeting of the American Political Science Association, Boston, MA, August 28–31. Harrison, L. E. 1993. Voodoo politics. Atlantic Monthly, 271(6): 101–107. Haspel, M., T. Remington, and S. Smith 2006. Lawmaking and decree making in the Russian Federation: Time, space, and rules in Russian national policymaking. PostSoviet Affairs 22, 249–275. Haungs, P. 1968. Reichspräsident und Parlamentarische Kabinettsregierung. Köln: Westdeutscher Verlag. Heinl, R. D. and N. G. Heinl 1995. Written in Blood, The Story of the Haitian People 1492–1995. Lanham: University Press of America. Hippolyte-Manigat, M. 2000. Traité de droit constitutionnel haïtien. Port-au-Prince: Université Quisqueya. Howard, E. A. D. ed. 1993. Constitution Making in Eastern Europe. Washington, DC: Woodrow Wilson Center Press. Huber, E. R. 1984. Deutsche Verfassunsgeschichte seit 1789: Band VII: Ausbau, Schutz und Untergang der Weimarer Republik. Stuttgart: W. Kohlhammer Verlag. Huntington, S. P. 1991. The Third Wave: Democratization in the Late Twentieth Century. Norman: University of Oklahoma Press. Huskey, E. 2007. Eurasian semi-presidentialism: The development of Kyrgyzstan’s model of government in R. Elgie and S. Moestrup (eds.), Semi-Presidentialism Outside Europe: A Comparative Study. London: Routledge, 161–181. Hyden, G. 2006. African Politics in Comparative Perspective. Cambridge: Cambridge University Press. International Foundation for Electoral Systems 1994. Republic of Moldova: Parliamentary Elections. February Chişinău: TISH, Ltd. IRIN 2006. Senegal: Former Prime Minister freed after seven months behind bars. Irin News, February 8, 2006. Jamestown Institute Monitor 2002. Council of Europe Mediates Political Armistice in Moldova 8(77). Jean-François, H. 1995. Le Coup de Cedras. Pétionville: Mediatek. Jérôme, J. P. 2009. Haïti: L’amendement déclaré receivable. Le Nouvelliste, 15 September 2009 www.lenouvelliste.com/article.php?PubID=&ArticleID=74214, accessed on November 15, 2009. Jones, M. P. 1995. Electoral Laws and the Survival of Presidential Democracies. North Dame: University of North Dame Press. Katznelson, I. and H. V. Milner eds. 2002. Political Science: State of the Discipline. New York: W.W. Norton. Kaufman, D., A. Kraay, and M. Mastruzzi, 2007. Governance Matters VI: Aggregate and Individual Governance Indicators 1996–2006. Washington, DC: World Bank Policy Research Working Paper 4280. Kelsen, H. 1931. Wer soll der Hüter der Verfassung sein? Berlin: C. Schulze & Co., GMBH. Kennedy, E. 2004. Constitutional Failure: Carl Schmitt in Weimar. Durham and London: Duke University Press.

9780230_242920_17_bib.indd 281

11/16/2010 3:42:05 PM

282

Bibliography

Keogh, D. and A. McCarthy 2007. The Making of the Irish Constitution, 1937. Dublin: Mercier Press. Kievskie vedomosti 2007. 26 May, p. 4. Kim, Y. H. and D. Bahry 2008. Interrupted presidencies in third wave democracies. Journal of Politics 30(3), 1–38. Kirschke, L. 2007. Semipresidentialism and the perils of power-sharing in neopatrimonial states. Comparative Political Studies 40 (11): 1372–1394. Kolb, E. 1990. The Weimar Republic. London: Routledge. Kolb, E. 2002. Die Weimar Republik. München: Oldenbourg Verlag. Kontseptsiya 2008. Kontseptsiya systemnoho onovlennya konstytutsiinoho regulyuvannya suspil’nykh vidnosyn v Ukraini. Glavred, 24 April http://glavred.info/ archive/2008/04/23/190045-9.html, accessed on September 1, 2008. Kovryzhenko, D. and V. Zamnius 2003. Konstytutsiina reforma ta presydents’ki vybory. Chaspys parlament 7: 22–27. Kovryzhenko, D. 2007. Konstytutsiina reforma: otsinky, problem, shlyakhy vyrishennya. Chaspys parlament 7: 45–51. Kovryzhenko, D., O. Chebanenko, and V. Shramovych 2008. Konstytutsiina reforma ta konsolidovana demokratiya: Rol’ hromadyans’koho suspil’stva, vlady i opozytsii. Chaspys parlament 2: 2–25. Kozyubra, M. 1997. Proekt konstytutsii Ukrainy v redaktsii vid 15 lystopada 1995 roku. Konstytutsiya nezalezhnoi Ukrainy. Knyha druha. Dokumenty. Statti. Kyiv. Kuzio, T. 2006. Indecision and opportunism derail NATO in Ukraine. The Ukrainian Observer, 224 www.ukraine- observer.com/articles/224/933; accessed on September 1, 2008. Laboratoriya 2004. Laboratoriya zakonodavchykh initsiatyv. Zminy konstytutsii ne pryinyato. 8 April. Available at: www.parlament.org.ua/index.php?action=news&id=12&ar_ id=483&as=0 accessed on October 29, 2009. Laguerre, M. S. 1994. The Tontons Macoutes in J. Ridgeway (ed.), The Haiti Files, Decoding the Crisis. Washington, DC: Essential Books, 47–52. Lijphart, A. 1982. The relative salience of the socio- economic and religious issue dimensions: Coalition formations in ten western democracies, 1919–1979. European Journal of Political Research 10(3): 201–211. Lijphart, A. 1992. Introduction in A. Lijphart (ed.), Parliamentary versus Presidential Government. Oxford: Oxford University Press, 1–27. Lijphart, A. 2004. Constitutional design for divided societies. Journal of Democracy 15(2): 96–109. Lijphart, A. ed. 1992. Parliamentary versus Presidential Government. Oxford: Oxford University Press. Lindberg, S. I. 2006. Democracy and Elections in Africa. Baltimore: Johns Hopkins University Press. Linz, J. J. 1990. The perils of presidentialism. Journal of Democracy 1(1), 51–69. Linz, J. J. 1994. Presidential or parliamentary democracy: Does it make a difference? in J. J. Linz and A. Valenzuela (eds.), The Failure of Presidential Democracy: Comparative Perspectives. Baltimore: Johns Hopkins University Press, 3–87. Linz, J. J. 1997. Introduction: Some thoughts on presidentialism in postcommunist Europe in R. Taras (ed.), Postcommunist Presidents. Cambridge: Cambridge University Press, 1–14. Linz, J. J. and A. Stepan 1996. Problems of Transition and Consolidation: Southern Europe, South America, and Post- Communist Europe. Baltimore NJ: The Johns Hopkins University Press.

9780230_242920_17_bib.indd 282

11/16/2010 3:42:05 PM

Bibliography

283

Linz, J. J. and A. Valenzuela eds. 1994. The Failure of Presidential Democracy. Baltimore: Johns Hopkins University Press. Lobo, M. C. 2000. Governos partidários numa democracia recente: Portugal, 1976– 1995. Análise Social 35(154–5), 147–174. Lösche, P. 1997. Parteienstaat Bonn-Parteienstaat Weimar? E. Kolb (ed.), Demokratie in der Krise, München: R. Oldenbourg Verlag, 141–164. Magalhães, P. 2001. As armas dos fracos: o veto político e a litigância constitucional do Presidente da República in M. B. Cruz, (ed.), A Reforma do Estado em Portugal: Problemas e Perpectivas, Lisbon: Bizâncio, 489–506. Magalhães, P. C. 2007. What are (semi)presidential elections about? A case study of the Portuguese 2006 elections. Journal of Elections, Public Opinion and Parties 17(3): 263–291. Maguire, R. 2004. Statement, Hearing on Haiti, March 10, 2004. Committee on Foreign Relations, Subcommittee on the Western Hemisphere. Washington, DC: United States Senate. Mainwaring, S. 1993. Presidentialism, multipartism and democracy: The difficult combination. Comparative Political Studies 26(2): 198–228. Mainwaring, S. and M. S. Shugart 1997. Presidentialism and Democracy in Latin America. Cambridge: Cambridge University Press. Manning, C. 2007. Semi-presidentialism and the preservation of ambiguity in postwar Mozambique in R. Elgie and S. Moestrup (eds.), Semi-Presidentialism Outside Europe, London: Routledge, 121–136. Marcus, R. R. 2005. The fate of Madagascar’s democracy: Following the rules while eroding the substance in L. Villalón and P. VonDoepp (eds.), The Fate of Africa’s Democratic Experiments. Bloomington, IN: Indiana University Press, 153–174. Markert, T. 2007. Ukrains’ki polityky prahnut’ hraty z konstytutsiinymy pravylamy zamist’ toho, shchob hraty za tsymy pravylamy. Chaspys parlament, 1: 43. Markert, T. 2008. Ukrains’ki polityky prahnut’ hraty z konstytutsiinymy pravylamy zamist’ toho, shchob hraty za tsymy pravylamy. Chaspys parlament, 2: 43. Marshall, M. G., K. Jaggers, and T. R. Gurr (accessed April 2008). Polity IV Project:Political Regime Characteristics and Transitions, 1800–2004. http://www. systemicpeace.org/polity/polity4.htm. Martins, A. 2006. The Portuguese semi-presidential system: About law in the books and law in action. European Constitutional Law Review, 2: 81–100. Matos, L. S. de 1983. Significado e consequências da eleicão do presidente por sufrágio universal – o caso português. Análise Social 19(76), 235–260. Matos, L. S. de 1986. L’expérience portugaise in M. Duverger (ed.), Les Régimes SemiPrésidentiels. Paris: Presses Universitaires de France, 55–83. Mattila, M. 1997. From qualified majority to simple majority: The effects of the 1992 change in the Finnish constitution. Scandinavian Political Studies (20): 317–331. Matsuzato, K. 2001. All Kuchma’s men: The reshuffling of Ukrainian governors and the presidential election of 1999. Post-Soviet Geography and Economics 42(6): 416–439. Matsuzato, K. 2002. Elite and the party system of Zakarpattia oblast: Relations among the levels of party systems in Ukraine. Europe-Asia Studies, 54(8): 1287–1292. Matsuzato, K. 2005. Semi-presidentialism in Ukraine: Institutionalist centrism in Rampant clan politics. Demokratizatsiya 13(1): 45–58. Matsuzato, K. 2006. Differing dynamics of semipresidentialism across Euro/Eurasian borders: Ukraine, Lithuania, Poland, Moldova and Armenia. Demokratizatsiya 14(3): 317–345.

9780230_242920_17_bib.indd 283

11/16/2010 3:42:05 PM

284

Bibliography

Matsuzato, K. and L. Gudžiskas 2006. An eternally unfinished parliamentary regime? Semipresidentialism as a prism to view Lithuanian politics. Acta Slavica Iaponica, 23: 146–170. Mauersberg, J. 1991. Ideen und Konzeption Hugo Preuß’ für die Verfassung der deutschen Republik 1919 und ihre Durchsetzung im Verfassungswerk von Weimar. Frankfurt am Main: Peter Lang Verlag. McFaul, M., N. Petrov and A. Ryabov 2004. Between Dictatorship and Democracy: Russian Post- Communist Political Reform. Washington: Carnegie Endowment for International Peace. McMann, K. M. 2006. Economic Autonomy and Democracy: Hybrid Regimes in Russia and Kyrgyzstan. Cambridge: Cambridge University Press. McMenamin, I. 2008. Semi-presidentialism and democratization in Poland. Paper delivered at the conference on “Semi-Presidentialism and Democracy: Institutional Choice, Performance, and Evolution,” October 17–18, 2008, Academia Sinica, Taipei. Mehler, A. 2005. The shaky foundations, adverse circumstances, and limited achievements of democratic transition in the Central African Republic in L. Villalón and P. VonDoepp (eds.), The Fate of Africa’s Democratic Experiments. Bloomington, IN: Indiana University Press, 126–152. Mello, F. R. de ed. 1976. Dossier 2ª República – 1º volume, 25- 4-74 – 25/4/75. Lisbon: Edições Afrodite. Metcalf, L. K. 2000. Measuring presidential power. Comparative Political Studies 33, 660–685. Michel, G. 1992. La Constitution de 1987: Souvenirs d’un Constituant. Port-au-Prince: Imprimerie le Natal. Miranda, J. 1986. Le régime semi-présidentiel portugais entre 1976 et 1979 in M. Duverger (ed.), Les Régimes Semi-Présidentiels. Paris: Presses Universitaires de France, 133–153. Mitterrand, F. 1995. Discours 1981–1995. Paris: Europolis. Moestrup, S. 2007a. Semi-presidentialism in young democracies: Help or hindrance? in R. Elgie and S. Moestrup (eds.), Semi-Presidentialism Outside Europe. London: Routledge, 30–55. Moestrup, S. 2007b. Semi-presidentialism in Niger: Gridlock and democratic breakdown – learning from past mistakes in R. Elgie and S. Moestrup (eds.), SemiPresidentialism Outside Europe. London: Routledge, 105–120. Moestrup, S. and G. Ganzorig 2007. Semi-presidentialism in Mongolia: Tradeoffs between stability and governance in R. Elgie and S. Moestrup (eds.), SemiPresidentialism Outside Europe. London: Routledge, 182–200. Moïse, C. 1990. Constitutions et Luttes de Pouvoir en Haiti, vol. 2: De l’Occupation Etrangère à la Dictature Macoute (1915–1987). Montreal: CIDIHCA. Moïse, C. 1994. Une constitution dans la tourmente: le nouveau régime politique haïtien et la crise nationale, 1987–1993. Montréal: Editions Images. Moïse, C. 1999. Le pouvoir législatif dans le système politique haïtien: un aperçu historique. Montréal: Editions du CIDIHCA. Möller, H. 2004. Die Weimarer Republik: Eine unvollendete Demokratie. München: Deutscher Taschenbuch Verlag. Morais, I., J-M. Ferreida de Almeida, and R. L. Pinto 1984. O Sistema de Governo Semipresidencial: O Caso Português. Lisboa: Editorial Notícias. Moreira, A. 1989. O regime: Presidencialismo do Primeiro-Ministro in M. B. Coelho (ed.), Portugal: O Sistema Político e Constitucional, 1974–87. Lisbon: Instituto de Ciências Sociais, 31–37.

9780230_242920_17_bib.indd 284

11/16/2010 3:42:06 PM

Bibliography

285

Morgan-Jones, E. and P. Schleiter 2004. Government change in a presidentparliamentary regime. Post-Soviet Affairs 20(2): 132–164. Morlino, L. 1998. Democracy Between Consolidation and Crisis: Parties, Groups and Citizens in Southern Europe. Oxford: Oxford University Press. Müller, W. C. and K. Strøm eds. 2000. Coalition Governments in Western Europe. Oxford: Oxford University Press. Mustafin, O. 2008. Usi v sad? Dzerkalo tyzhnya, 8–14 March. National Democratic Institute for International Affairs 1991. The 1990 General Elections in Haiti. Washington, DC: NDI. Neto, O. A. and K. Strøm 2006. Breaking the parliamentary chain of delegation: Presidents and non-partisan cabinet members in European democracies. British Journal of Political Science 36(4): 619–643. Neto, O. A. and M. C. Lobo 2009. Portugal’s semi-presidentialism (re)considered: An assessment of the president’s role in the policy process, 1976–2006. European Journal of Political Research 48(2): 234–255. Nicholls, A. J. 2000. Weimar and the Rise of Hitler. New York: St. Martin’s Press. Nijzink, L., S. Mozaffar, and E. Azevedo 2006. Can Parliaments Enhance the Quality of Democracy on the African Continent? An Analysis of Institutional Capacity and Public Perceptions. CSSR Working Paper, no. 160, University of Cape Town. Nogueira Alcalá, H. 1986. El Regimen Semipresidencial: ¿Una Nueva Forma de Gobierno Democrático? Santiago: Andante. Nolte, E. 2006. Die Weimarer Republik. München: Herbig Verlag. Nousiainen, J. 1971. The Finnish Political System. Cambridge, MA: Harvard University Press. Nousiainen, J. 2000. Finland: The consolidation of parliamentary governance in W. Müller and K. Strom (eds.), Coalition Governments in Western Europe. Oxford: Oxford University Press, 264–299. Nousiainen, J. 2001. From semi-presidential to parliamentary government: Political and constitutional developments in Finland. Scandinavian Political Studies 24(2): 95–109. O’Donnell, G. and P. Schmitter 1986. Transitions from Authoritarian Rule, Tentative Conclusions about Uncertain Democracies. Baltimore: The Johns Hopkins University Press. Olson, David and William Crowther 2003. Committees in the New Democratic Parliaments of Central Europe. Columbus, Ohio: The Ohio State University Press. PACE 2002. Resolution 1280, adopted April 24. Paloheimo, H. 2001. Divided government in Finland: From a semi-presidential to a parliamentary democracy in R. Elgie (ed.), Divided Government in Comparative Perspective. Oxford: Oxford University Press, 86–105. Paloheimo, H. 2003. The rising power of the prime minister in Finland. Scandinavian Political Studies 26(3): 219–243. Paloheimo, H. 2005. Finland: Let the force be with the leader— But who is the leader? in T. Poguntke and P. Webb (eds.), The Presidentialization of Politics: A Comparative Study of Modern Democracies, Oxford: Oxford University Press, 246–268. Paquin, L. 1983. The Haitians, Class and Color Politics. Brooklyn: Multi-Type. Parish, S. 1998. Presidential decree authority in Russia in J. Carey and M. S. Shugart (eds.), Executive Decree Authority. Cambridge: Cambridge University Press, 62–103. Parliament of Moldova 1993. Parliamentary Staff Report, Informatie privind activitatea Parlamentullui si a comisiilor permannte, 28 September.

9780230_242920_17_bib.indd 285

11/16/2010 3:42:06 PM

286

Bibliography

Parliament of Moldova 1996. Parliamentary Staff Report, Bilantul activit|tii Paralamentului in anii 1994–1995: Date statistice. Parliament of Moldova 2004. Parliamentary Staff Report, Informaţia privind activitatea legisliativ| a Parlamentului Republicii Moldova anii 2001–2003 şi sesiunea de primavar. Parlamentul Republicii Moldova 1996. Chişinău: Studioul de Creaţia. Pasquino, G. 1996. ‘Duetti e duelli: l’adattabilità dei semipresidenzialismi in S. Ceccanti, O. Massari, G. Pasquino (eds.), Semipresidenzialismo: Analisi delle esperienze europee. Bologna: Il Mulino, 99–148. Pasquino, G. 1997. Semi-presidentialism: A political model at work. European Journal of Political Research 31(1): 128–137. Pasquino, G. 2005. Duverger and the study of semi-presidentialism. French Politics, 3(3): 310–322. Pasquino, G. 2007. The advantages and disadvantages of semi-presidentialism: A West European perspective in R. Elgie and S.Moestrup (eds.), Semi-Presidentialism Outside Europe. London: Routledge, 14–29. Pegoraro, L. and A. Rinella eds. 1997. Semipresidenzialismi. Padova: CEDAM. Pereira, A. G. 1984. O semipresidencialismo em Portugal. Lisbon: Ática. Perito, R. and J. Jocic 2008. Paper Versus Steel: Haiti’s Challenge of Constitutional Reform. Washington, DC: United States Institute of Peace. Pierce, R. 1991. The executive divided against itself: Cohabitation in France, 1986–88. Governance 4(3): 270–294. Pires, F. L. 1989. O sistema de governo: sua dinâmica in M. B. Coelho (ed.), Portugal: O Sistema Político e Constitucional, 1974–87. Lisboa: Instituto de Ciências Sociais, 291–319. Posner, D. N. and D. J. Young 2007. The institutionalization of political power in Africa. Journal of Democracy 18(3): 126–140. Pontusson, J. 1995. From comparative public policy to political economy: Putting political institutions in their place, and taking interests seriously. Comparative Political Studies 28: 117–147. Poulard, J. V. 1990. The french double executive and the experience of cohabitation. Political Science Quarterly 105(2): 243–267. Preuß, H. 1926. Staat Recht und Freiheit: Aus 40 Jahren Deutscher Politik und Geschichte. Tübingen: J. C. B. Mohr Verlag. Proekt 2008. Proekt: Konstytutsiya Ukrainy. Published in Glavred, 28 February http:// glavred.info/archive/2008/02/28/163946-5.html, accessed on October 2, 2010. Protsyk, O. 2003. Troubled semi-presidentialism: Stability of the constitutional system and cabinet in Ukraine. Europe-Asia Studies 55(7): 1077–1095. Protsyk, O. 2004. Ruling with decrees: Presidential decree making in Russia and Ukraine. Europe-Asia Studies 56, 637. Protsyk, O. 2005a. Prime Ministers’ identity in semi-presidential regimes: Constitutional norms and cabinet formation outcomes. European Journal of Political Research 44: 721–748. Protsyk, O. 2005b. Politics of intraexecutive conflict in semipresidential regimes in Eastern Europe. East European Politics and Societies 19(2): 135–160. Protsyk, O. 2008. Regime Type Effects on Presidential Use of Decree Making Authority in Post-Soviet States. Paper presented at American Political Science Association Meeting, Boston, MA, August 28-August 31, 2008. Przeworski, A., M. E. Alvarez, J. A. Cheibub, and F. Limongi 2000. Democracy and Development: Political Institutions and Well-being in the World, 1950–1990. Cambridge: Cambridge University Press.

9780230_242920_17_bib.indd 286

11/16/2010 3:42:06 PM

Bibliography

287

Przeworski, A. and F. Limongi 1997. Modernization: Theory and facts. World Politics 49: 155–183. Quinlan, P. D. 2002. Moldova under Lucinschi. Democratizatia (Winter). Raunio, T. and M. Wiberg 2003. Finland: Polarized pluralism in the shadow of a strong president in K. Strøm, W. C. Müller and T. Bergman (eds.), Delegation and Accountability in Parliamentary Democracies, Oxford: Oxford University Press, 301–324. Radio Kiskeya 2008. La Coalition des Parlementaires Progressistes (CPP) se définit comme «incontournable et garant de la majorité parlementaire» appelée à soutenir la politique «anti néo-libérale» du prochain gouvernement. April 28, 2008, http://radiokiskeya. com/spip.php?article4964, accessed on November 13, 2009. Radio Vision 2000 2009. Le Corps législatif approuve l’opportunité d’amender la Constitution. 15 September, 2009, www.radiovision2000haiti.net/?p=1008&wscr=1280x800, accessed on November 17, 2009. Raynal, J-J. 1993. Les institutions politiques du Niger. Saint-Maur: Sépia. Reilly, B. 2006. Democracy and Diversity: Political Engineering in the Asia-Pacific. Oxford: Oxford University Press. Reilly, B. 2007. Electoral systems and party systems in East Asia. Journal of East Asian Studies 7: 185–192. Reilly, B. 2008. Semi-Presidentialism and democratic development in East Asia. Paper delivered at the conference on “Semi-Presidentialism and Democracy: Institutional Choice, Performance, and Evolution,” October 17–18, 2008, Academia Sinica, Taipei. Reilly, B. and P. Nordlund eds. 2008. Political Parties in Conflict-Prone Societies. Tokyo: United Nations University Press. Reynolds, A. 2002. The Architecture of Democracy: Constitutional Design, Conflict Management and Democracy. Oxford: Oxford University Press. Reynolds, A., B. Reilly, and A. Ellis 2005. Electoral System Design: The New International IDEA Handbook. Stockholm: International Institute for Democracy and Electoral Assistance. Reyntjens, F. 1996. Constitution-making in situations of extreme crisis: The case of Rwanda and Burundi. Journal of African Law 40(2): 234–242. Ridgeway, J. ed. 1994. The Haiti Files, Decoding the Crisis. Washington, DC: Essential Books. Rinella, A. 1997. La forma di governo semi-presidenziale. Profili metodologici e “circolazione” del modello francese in Europa centro- orientale. Torino: Giappichelli. Robinson, J. A. 2006. Economic development and democracy. Annual Review of Political Science 9: 503–527. Roeder, P. 2005. Power dividing as an alternative to ethnic power sharing in P. Roeder and D. Rothchild (eds.), Sustainable Peace, Power and Democracy after Civil War. Ithaca, NY: Cornell University Press, 51–82. Roeder, P. G. and D. Rothchild eds. 2005. Sustainable Peace, Power and Democracy after Civil War. Ithaca, NY: Cornell University Press. Roland, G. 2004. Understanding institutional change: Fast-moving and slow-moving institutions. Studies in Comparative International Development 38(4): 109–131. Roper, S. D. 2002. Are all semi-presidential regimes the same? A comparison of premier-presidential regimes. Comparative Politics 34: 253–272. Roper, S. D. 2008. From semi-presidentialism to parliamentarism: Regime change and presidential power in Moldova. Europe-Asia Studies 60(1): 113–126. Roy, J. C. H. 1992. Entre la lettre et l’esprit de la constitution de 87. Port-au-Prince: Editions Henri Deschamps.

9780230_242920_17_bib.indd 287

11/16/2010 3:42:06 PM

288

Bibliography

Rubin, B. R. 2003. Presentation to Constitutional Commission of Afghanistan. http:// www.cic.nyu.edu/afghanistan/constitution.html#goverment, accessed on March 26, 2008. Samuels, D. and M. Shugart 2010. Presidents, Parties, Prime Ministers: A Framework for Analysis, Cambridge: Cambridge University Press – all chapters available at http:// dss.ucsd.edu/~mshugart/samuels-shugart.html, accessed on September 17, 2009. Sartori, G. 1994. Neither presidentialism nor parliamentarism in J. J. Linz and A. Valenzuela (eds.), The Failure of Presidential Democracy: Comparative Perspectives. Baltimore: Johns Hopkins University Press, 106–118. Sartori, G. 1995. Elogio del semi-presidenzialismo. Rivista Italiana di Scienza Politica, 25(1): 3–20. Sartori, G. 1997. Comparative Constitutional Engineering: An Inquiry into Structures, Incentives and Outcomes. 2nd edition. New York: New York University Press. Schedler, A. 2001. Measuring democratic consolidation. Studies in Comparative International Development 36(1): 66–92. Schedler, A. 2006. Electoral Authoritarianism: The Dynamics of Unfree Competition. Boulder, CO: Lynne Rienner. Schleiter, P. 2003. Mixed constitutions and political instability: Russia 1991–1993. Democratization 10(1): 1–26. Schleiter, P. and E. Morgan-Jones 2005. Semi-Presidential Regimes: Providing Flexibility or Generating Representation and Governance Problems? Working Paper Series, Department of Politics and International Relations, University of Oxford, Paper No. 001. Schleiter, P. and E. Morgan-Jones 2008. Russia: The benefits and perils of presidential leadership in R. Elgie and S. Moestrup (eds.), Semi-Presidentialism in Central and Eastern Europe, Manchester: Manchester University Press, 159–179. Schmitt, C. 1928. Verfassungslehre. Berlin: Duncker & Humblot. Schmitt, C. 1931. Der Hüter der Verfassung. Berlin: Duncker & Humblot. Schumpeter, J. A. 1947. Capitalism, Socialism and Democracy. New York: Harper. Seibert, G. 2004. Sao Tome and Principe Update. Situation Report. Institute for Security Studies, South Africa, November 9, 2004. Shen, Y- C. 2006. Cong diguo dao gonghe (From Empire to Republic). Issues and Studies 45(1): 29–51. Sheng, S-Y. 2003. The influences of the legislative branch and executive branch in the law-making process. Taiwanese Political Science Review 7(2): 51–105 (Written in Chinese). Shoesmith, D. 2003. Timor-Leste: Divided leadership in a semi-presidential system. Asian Survey 43(2), 231–252. Shoesmith, D. 2007. Semi-presidentialism and the democratic transition in a new, small state in Robert Elgie, and Sophia Moestrup (eds.), Semi-Presidentialism Outside Europe: A Comparative Study, London: Routledge, 219–236. Shugart, M. S. 1995. The electoral cycle and institutional sources of divided presidential government. American Political Science Review, 89(2): 327–343. Shugart, M. S. 1996. Executive-legislative relations in post- communist Europe. Transition 2(25): 6–11. Shugart, M. S. 2005a. Semi-presidential systems: Dual executive and mixed authority patterns. French Politics 3(3): 323–351. Shugart, M. S. 2005b. Comparative Executive–Legislative Relations: Hierarchies vs. Transactions in Constitutional Design. Center for the Study of Democracy, Paper 05- 07, available online at: http://repositories.cdlib.org/csd/05- 07.

9780230_242920_17_bib.indd 288

11/16/2010 3:42:06 PM

Bibliography

289

Shugart, M. S. 2006. Comparative executive-legislative relations’, in R.A.W. Rhodes, S. A. Binder and B. A. Rockman (eds.), The Oxford Handbook of Political Institutions. Oxford: Oxford University Press, 344–365. Shugart, M. S. and Carey, J. M. 1992. Presidents and Assemblies: Constitutional Design and Electoral Dynamics, Cambridge: Cambridge University Press. Siaroff, A. 2003. Comparative presidencies: The inadequacy of the presidential, semipresidential and parliamentary distinction. European Journal of Political Research 42(3): 287–312. Skach, C. 2005. Borrowing Constitutional Designs: Constitutional Law in Weimar Germany and the French Fifth Republic. Princeton, NJ: Princeton University Press. Skach, C. 2007. The “newest” separation of powers: Semipresidentialism. International Journal of Constitutional Law 5(1): 93–121. Smith, M. J. 2009. Red & Black in Haiti, Radicalism, Conflict, and Political Change, 1934– 1957. Chapel Hill, NC: The University of North Carolina Press. Soares, D., M. Maley, J. Fox, and A. Regan 2003. Elections and Constitution Making in East Timor. Canberra: State, Society and Governance in Melanesia Project, Australian National University. Solana, J. 2007. This is the time for political leaders to show their responsibility for the country’s future. Zerkalo nedeli, 15–21 September. Starr, H. and C. Lindborg 2003. Democratic dominoes revisited: The hazards of governmental transitions, 1974–1996. Journal of Conflict Resolution 47(4): 490–519. Stepan, A. and Skach, C. 1993. Constitutional frameworks and democratic consolidation: Parliamentarianism versus presidentialism. World Politics 46: 1–22. Stepan, A. and E. N. Suleiman 1995. The French fifth republic: A model for import? Reflections on Poland and Brazil in H. E. Chehabi and A. Stepan (eds.), Politics, Society, and Democracy. Comparative Studies, Boulder, CO: Westview Press, 393–414. Stirk, P. 2002. Hugo Preuss, German political thought and the Weimar constitution. History of Political Thought 23(3): 497–516. Suleiman, E. N. 1994. Presidentialism and political stability in France in J. J. Linz and A. Valenzuela (eds.), The Failure of Presidential Democracy. Baltimore: The Johns Hopkins University Press, 137–162. Sullivan, D. P. 2005. The missing pillars: A look at the failure of peace in Burundi through the lens of Arend Lijphart’s theory of consociational democracy. Journal of Modern African Studies, 43(1): 75–95. Svolik, M. 2008. Authoritarian reversals and democratic consolidation. American Political Science Review 102(2): 153–168. Sylina, T. et al. 2004. Anatomiya dushi maidanu. Dzerkalo tyzhnya, 11–17 December. Taras, R. ed. 1997. Postcommunist Presidents. Cambridge: Cambridge University Press. Tavits, M. 2009. Presidents with Prime Ministers: Do Direct Elections Matter? Oxford: Oxford University Press. Ternavs’kyi, O. 2004. Try nozhi u spynu reformy? Dzerkalo tyzhnya, 11–17 September. Thiébault, J-L. 2000. France: Forming and maintaining government coalitions in the fifth republic in W. C. Müller and K. Strom (eds.), Coalition Governments in Western Europe, Oxford: Oxford University Press, 498–528. Todosijevic, B. 2004. Serbia in S. Berglund, J. Ekman, and F. H. Aarebrot (eds.), The Handbook of Political Change in Eastern Europe, Cheltenham: Edward Elgar, 493–549. Tomiuc, E. 2001. Moldova: Moscow and Chisinau initial bilateral agreement. Radio Free Europe/Radio Liberty Weekly Magazine (November 6).

9780230_242920_17_bib.indd 289

11/16/2010 3:42:06 PM

290

Bibliography

Tsai, J-H. 2008. Sub-types of semi-presidentialism and political deadlock. French Politics 6(1): 63–84. Tsebelis, G. 2002. Veto Players: How Political Institutions Work. Princeton, NJ: Princeton University Press. Tymoshenko, Y. 2004. Kuchma, yak molodyi olen’, pryskakav pidpysaty politreformu. Dlya n’oho ta ioho otochennya – tse akt reinkarnatsii! Dzerkalo tyzhnya, 11–17 December. Tymoshenko, Y. 2007. Nova Konstytutsiya Ykrainy: vid parytetu syly do priorytetu prav, Dzerkalo tyzhnya, 30 June–6 July. UNDP 2008. Human Development Report 2008. Oxford: Oxford University Press. Wilson, A. J. 1980. The Gaullist System in Asia: The Constitution of Sri Lanka, 1978. London: Macmillian. Wilson, A. 1997. Ukraine: Two presidents and their powers in R. Taras (ed.), Postcommunist Presidencies, Cambridge: Cambridge University Press, 67–105. Verner, D. 2007. The Haitian people: Demographics, poverty and socioeconomic outcomes and risks, in World Bank, Social Resilience and State Fragility in Haiti. Washington, DC: The World Bank, 7–42. Villalón, L. A. and A. Idrissa 2005. The tribulations of a successful transition: Institutional dynamics and elite rivalry in Mali in L. Villalón and P. VonDoepp (eds.), The Fate of Africa’s Democratic Experiments. Bloomington, IN: Indiana University Press, 49–74. Villalón, L. A. and P. VonDoepp eds. 2005. The Fate of Africa’s Democratic Experiments. Bloomington, IN: Indiana University Press Way, L. A. 2003. Weak states and pluralism: The case of Moldova. East European Politics and Societies 17(3): 454–482. Way, L. A. 2005. Rapacious individualism and political competition in Ukraine, 1992–2004. Communist and Post- communist Studies 38: 191–205. Weber, K-D. 2001. Das Büro des Reichspräsidenten 1919–1934. Frankfurt am Main: Peter Lang Verlag. Wilson, A. 1997. Ukraine: Two presidents and their powers in R. Taras (ed.), Postcommunist Presidencies. Cambridge: Cambridge University Press, 67–105. Wilson, A. Jeyaratnam. 1980. The Gaullist System in Asia: The Constitution of Sri Lanka, 1978. London: Macmillian. Winkler, H. A. 2005. Weimar 1918–1933. München: C. H. Beck Verlag. World Bank 2002. Interim Poverty Reduction Strategy Paper. Prepared by the Moldovan authorities) (21 April), Washington, DC: The World Bank. World Bank 2003. Moldova: Public economic management review. Report No 25423MD, (20 February), Washington, DC: The World Bank, p. 5. World Bank 2005. Country Assistance Strategy for Sao Tome and Principe. Report no. 32078. Washington, DC: The World Bank. Wu, Y- S. 2000a. Russia’s Transition 1992–1999: A Political-Economic Analysis. Taipei: Wu-nan Press (Written in Chinese). Wu, Y- S. 2000b. The ROC’s Semi-presidentialism at Work: Unstable Compromise, Not Cohabitation. Issues and Studies 36(5): 1–40. Wu, Y- S. 2003. Triangular Interactions under Semi-presidentialism: A Typological Exposition. Paper presented at the Conference on Semi-presidentialism and Nascent Democracies, Academia Sinica, Taipei, Taiwan, ROC, October 24–25. Wu, Y- S. 2005. Appointing the prime minister under incongruence: Taiwan in comparison with France and Russia. Taiwan Journal of Democracy 1(1): 103–32.

9780230_242920_17_bib.indd 290

11/16/2010 3:42:06 PM

Bibliography

291

Wu, Y- S. 2006. Regime legitimacy and mode of constitutional reform: Comparing Taiwan and post- communist Nascent democracies in Europe. Issues and Studies 45(1): 1–28 (in Chinese). Wu, Y- S. 2007a. Taiwan’s developmental state after the economic and political turmoil. Asian Survey, 47(6): 977–1001. Wu, Y- S. 2007b. Semi-presidentialism in Taiwan: Easy to choose, difficult to operate in R. Elgie and S. Moestrup (eds.), Semi-Presidentialism Outside Europe: A Comparative Study, London: Routledge, 210–218. Wu, Y- S. 2008. Study of Semi-Presidentialism: A Holistic Approach. Paper delivered at the conference on “Semi-Presidentialism and Democracy: Institutional Choice, Performance, and Evolution,” October 17–18, 2008, Academia Sinica, Taipei. Yushchenko, V. 2000. My devyat’ let zhili amoral’no. Zerkalo nedeli, 15–21 January, p. 4. Zerkalo nedeli 2000. 29 January – 4 February 4, p. 1.

9780230_242920_17_bib.indd 291

11/16/2010 3:42:06 PM

9780230_242920_17_bib.indd 292

11/16/2010 3:42:07 PM

Index Africa Anglophone, 12 Francophone, 10–11, 12–13, 15, 22, 27 Lusophone, 11, 13, 15, 22, 27–28 semi-presidentialism in, 27–28, 47, 90, 134–155, 265, 266 Alkatiri, Mari, 35, 129–130, 132 Angola, 7, 11, 272 see also semi-presidentialism, in Africa Argentina, 4 Aristide, Jean-Bertrand, 249, 254–256 Armenia, 7, 42, 112, 114, 272 see also semi-presidentialism, under post-communism Asia Central, 43, 104, 111, 112, 266 East, 12, 22, 28, 32, 43, 117–133 Eurasia, 21, 24, 28, 35 semi-presidentialism in, 22, 28, 32, 117–133 Austria, 2, 7, 24, 49, 69, 82–89 see also semi-presidentialism, in Western Europe Azerbaijan, 7, 99, 273 see also semi-presidentialism, under post-communism Bagabandi, Natsagiyn, 35, 123, 126 Belarus, 5, 14, 106, 114–115, 142 see also semi-presidentialism, under post-communism Bosnia, 12, 99, 112 Brazil, 12 Bulgaria, 22, 99 see also semi-presidentialism, under post-communism Burkina Faso, 13–14, 22, 136, 142, 147 see also semi-presidentialism, in Africa Burundi, 10, 134, 136, 142, 145, 149, 150 see also semi-presidentialism, in Africa Cameroon, 14 Cape Verde, 22, 136, 142, 145–146, 147, 148

see also semi-presidentialism, in Africa Central African Republic, 22, 42, 142, 145, 147, 149, 271 see also semi-presidentialism, in Africa Chad, 13 Chen, Shi-bian, 77, 122, 125, 127–128, 132, 174, 178–182, 187–189, 190, 267–268 Chirac, Jacques, 76, 89, 95 cohabitation, 16, 29, 30, 35, 46–47, 49, 51, 55, 57, 58–59, 66, 76, 88, 89, 121, 138–140, 144–150, 161, 183, 187, 252, 266, 267, 268, 270 in Cape Verde, 145–146 in France, 94–95 in Mongolia, 119, 123, 125–127 in Niger, 146 in Sao Tome, 146, 150 Comoros, 10, 49, 134, 136, 142, 145, 150 see also semi-presidentialism, in Africa Congo, Republic of (Congo-Brazzaville), 10, 13, 134, 136, 142, 143, 145, 146, 147, 149 see also semi-presidentialism, in Africa Croatia, 22, 99, 112–113, 272 see also semi-presidentialism, under post-communism Cuba, 90 de Gaulle, Charles, 14, 94, 165, 172 de Valera, Éamon, 85 Duverger, Maurice, 1, 2, 15, 16, 29, 30, 44, 62–63, 82–86, 90–96, 163, 210 Eanes, António Ramalho, 159, 160–161, 165–166, 169–171 East Timor, see Timor Leste Ebert, Friedrich, 230, 237–238 Egypt, 12, 14, 271 electoral systems, 61–80, 107–110, 182, 198, 213, 236, 273

293

9780230_242920_18_ind.indd 293

11/16/2010 11:50:37 AM

294 Index Europe Central and Eastern, 12, 13, 22, 24, 26–27, 28, 32, 36, 43, 100, 102, 103, 111, 112, 123, 142, 252 Western, 12, 15, 21, 22, 24, 25, 27, 28, 32, 35, 36, 67, 72, 73, 81–96, 265 Finland, 1, 24, 32, 74–75, 77–78, 82–89, 92, 166–167 see also semi-presidentialism, in Western Europe France, 5, 6, 9, 10, 14, 17, 24, 27, 32, 74–76, 83–90, 93–95, 163–164, 167 see also semi-presidentialism, in Western Europe Gabon, 13, 14, 22 Georgia, 14, 22, 99, 112, 115, 271, 273 see also semi-presidentialism, under post-communism Germany, 1, 24, 42, 45, 48, 83, 85–89, 91, 120, 145, 229–244, 267–270 see also semi-presidentialism, in Western Europe Giscard d’Estaing, Valéry, 89 Guinea-Bissau, 22, 42, 136, 142, 144, 145, 147, 148, 149 see also semi-presidentialism, in Africa Gusmão, Xanana, 35, 124, 125, 129–130, 132 Haiti, 10, 246–262, 267, 269–270, 273–274 Hindenburg, Paul von, 229, 238–242 Hitler, Adolf, 237, 240 Iceland, 13, 17, 22, 24, 57, 69, 82, 83, 85, 86, 87, 89 see also semi-presidentialism, in Western Europe Ireland, 13, 17, 22, 24, 57, 82, 83, 85, 86, 89 see also semi-presidentialism, in Western Europe Italy, 91, 93, 94 Kazakhstan, 99, 109 see also semi-presidentialism, under post-communism Kenya, 1, 12, 14, 15, 132, 265

9780230_242920_18_ind.indd 294

Kuchma, Leonia, 192, 195–198, 200, 203, 206, 269 Kyrgyzstan, 22, 42, 99, 112 see also semi-presidentialism, under post-communism Lee, Teng-hui, 177–178, 180–181 Lijphart, Arend, 93, 94, 95, 134, 138, 140, 145, 150, 205 Linz, Juan, 16, 42, 44, 46, 47, 91, 94, 95, 98, 104, 105, 139 Lissouba, Pascal, 146, 147 Lithuania, 13, 22, 32, 99, 194 see also semi-presidentialism, under post-communism Lucinschi, Petru, 211–212, 213–220, 221, 227 Ma, Ying-Jeou, 122, 174, 182–184 Macedonia, 22, 99 see also semi-presidentialism, under post-communism Madagascar, 7, 142, 145, 149, 272 see also semi-presidentialism, in Africa Mali, 13, 22, 43, 136, 141, 142, 144, 145, 147, 149, 273 see also semi-presidentialism, in Africa Mauritania, 136, 142, 145, 271 Mexico, 12 Middle East, 12, 52 Mitterrand, François, 10, 76, 88, 95 Moldova, 99, 113, 197, 210–228, 267–270, 272 Mongolia, 12, 22, 32, 35, 36, 117–127, 131–132 see also semi-presidentialism, in East Asia Mozambique, 4, 17, 142, 145, 148 see also semi-presidentialism, in Africa Namibia, 22, 67, 136, 142, 143, 147, 148, 150 Niger, 7, 13, 35, 36, 47, 134, 141, 142, 144, 145–146, 147, 148, 149–150, 272 see also semi-presidentialism, in Africa Palestinian Authorities, 12 Patassé, Ange-Félix, 145

11/16/2010 11:50:37 AM

Index Peru, 12, 22, 45, 49, 67, 91 Philippines, 5, 121 Pires, Pedro Verona Rodrigues, 145–146 Poland, 14, 22, 99, 200 see also semi-presidentialism, under post-communism Portugal, 11, 22, 24, 27, 45, 85, 86, 88, 89, 156–173, 265, 267–271, 272 see also semi-presidentialism, in Western Europe premier-presidential, see semipresidentialism, sub-types of president-parliamentary, see semipresidentialism, sub-types of Préval, André, 255–258, 260 Ramos-Horta, Jose, 125, 130 Romania, 22, 99, 217 see also semi-presidentialism, under post-communism Russia, 2, 106, 107, 109, 142, 192, 195, 205, 206, 219 see also semi-presidentialism, under post-communism Sampaio, Jorge, 162, 166–170, 171 Sao Tome and Principe, 22, 35, 142, 144, 145, 146, 147, 149, 150, 272 see also semi-presidentialism, in Africa Sartori, Giovanni, 1, 2, 3, 16, 44, 91, 94–95, 125, 130, 162, 172 semi-presidentialism in Africa, 27–28, 47, 90, 134–155, 265, 266 in Asia, 22, 28, 32, 117–133 clustering of, 21–35 concept of, 2–5, 15–20 evolution of, 271–274 under post-communism, 11–12, 22, 98–116 spread of, 6–15, 22–28 sub-types of, 17–20, 21–22, 29–36, 45, 51, 57, 58–59, 61, 86–87, 100–105, 116, 138–139, 140–144, 146–150, 247, 265–271, 274 in Western Europe, 81–96 Senegal, 14, 22, 136, 141, 142, 143, 145, 147, 148, 150, 272, 273

9780230_242920_18_ind.indd 295

295

see also semi-presidentialism, in Africa Serbia and Montenegro, 115 see also semi-presidentialism, under post-communism Shugart, Matthew S., and John M. Carey, 1, 16, 17, 29–30, 45–46, 58, 61, 66, 72, 86, 91, 100, 105, 138, 141, 184, 186, 265 Silva, Anibal Cavaco, 166, 169, 170 Singapore, 14, 22, 117, 121 Slovakia, 14, 22, 99, 113, 115 see also semi-presidentialism, under post-communism Slovenia, 13, 22, 99 see also semi-presidentialism, under post-communism Snegur, Mircea, 211–212, 214–215, 217–218 Soares, Mário, 161, 166, 169–170, 171 South Korea, 5, 9, 12, 117, 121 Soviet Union, former, 7, 11, 12, 13, 22, 24, 103 see also semi-presidentialism, under post-communism Sri Lanka, 4, 10, 14, 22, 91 Taiwan, 12, 14, 22, 32, 74–77, 117–122, 124–129, 131–132, 174–190, 265, 267, 268, 270 see also semi-presidentialism Tandja, Mamadou, 148 Tanzania, 12, 14, 22 Timor Leste, 11, 22, 28, 35, 36, 117–126, 129–133, 271 see also semi-presidentialism Togo, 13, 22 Touré, Amani Toumain (ATT), 145 Tunisia, 14 Turkey, 1, 6, 14, 271 Tymoshenko, Yulia, 192, 198, 201, 203–205 Uganda, 5 Ukraine, 7, 32, 34–35, 36, 99, 106, 112–114, 192–209, 267–270, 272 see also semi-presidentialism, under post-communism

11/16/2010 11:50:37 AM

296 Index Vietnam, 12, 90 Voronin, Vladimir, 215–217, 219, 221–225, 227–228 Weimar Republic, see Germany Yanukovych, Viktor, 34, 197, 201–205

9780230_242920_18_ind.indd 296

Yemen, 12, 14, 22 Yugoslavia, former, 11, 12, 13, 99, 113, 115 see also semi-presidentialism, under post-communism Yushchenko, Viktor, 34, 113, 193, 196–198, 201–206

11/16/2010 11:50:38 AM