Presidential Breakdowns in Latin America: Causes and Outcomes of Executive Instability in Developing Democracies

Presidential Breakdowns in Latin America

This page intentionally left blank

Presidential Breakdowns in Latin America Causes and Outcomes of Executive Instability in Developing Democracies Edited by Mariana Llanos and Leiv Marsteintredet

PRESIDENTIAL BREAKDOWNS IN LATIN AMERICA

Copyright © Mariana Llanos and Leiv Marsteintredet, 2010. All rights reserved. First published in 2010 by PALGRAVE MACMILLAN® in the United States—a division of St. Martin’s Press LLC, 175 Fifth Avenue, New York, NY 10010. Where this book is distributed in the UK, Europe and the rest of the world, this is by Palgrave Macmillan, a division of Macmillan Publishers Limited, registered in England, company number 785998, of Houndmill s, Basingstoke, Hampshire RG21 6XS. 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: 978–0–230–61819–0 Library of Congress Cataloging-in-Publication Data is available from the Library of Congress. A catalogue record of the book is available from the British Library. Design by Newgen Imaging Systems (P) Ltd., Chennai, India. First edition: March 2010 10 9 8 7 6 5 4 3 2 1 Printed in the United States of America.

To Clara and Valentin M.Ll. To Dixie L.M.

This page intentionally left blank

Contents

List of Tables

ix

Acknowledgments

xi

List of Abbreviations

xiii

Introduction: Presidentialism and Presidential Breakdowns in Latin America Mariana Llanos and Leiv Marsteintredet

1

Part I Comparative Perspectives 1. Presidentialism and Early Exits: The Role of Congress Ana María Mustapic 2. Presidential and Democratic Breakdowns in Latin America: Similar Causes, Different Outcomes Michael E. Álvarez and Leiv Marsteintredet

Part II

17

33

Cases of Repeated Breakdown

3. Presidential Breakdowns in Argentina Mariana Llanos

55

4. Parliamentary Solutions to Presidential Crises in Ecuador Andrés Mejía Acosta and John Polga-Hecimovich

73

5. Civil Society, Social Protest, and Presidential Breakdowns in Bolivia Miguel A. Buitrago

91

Part III The Cases of Impeachment 6. The Collor Impeachment and Presidential Government in Brazil Argelina Cheibub Figueiredo 7. The Impeachment of Carlos Andrés Pérez and the Collapse of Venezuelan Partyarchy Rickard Lalander

111

129

viii

Contents

8. Paraguay: The President in his General’s Labyrinth Detlef Nolte

147

Part IV Presidential Breakdowns Restoring Democratic Order 9. Guarding Privileges and Saving the Day: Guatemalan Elites and the Settlement of the Serranazo Maren Christensen Bjune and Stina Petersen

165

10. The Dominican Republic and the Fall of Balaguer 1994–1996: Presidential Breakdown or Democratic Transition? Leiv Marsteintredet

181

11. Peru and the Fujimori Presidential Breakdown in 2000: Continuismo Gone Bad Einar Berntzen and Tor-Einar Holvik Skinlo

197

Conclusions: Presidential Breakdowns Revisited Mariana Llanos and Leiv Marsteintredet

213

Epilogue Mariana Llanos and Leiv Marsteintredet

229

References

239

Contributors

255

Index

257

Tables

Table I.1

Presidential breakdowns in Latin America

Table 1.1

Early presidential Salidas, 1980–2006

20

Table 2.1

Determinants of presidential and democratic breakdowns

40

Determinants of presidential breakdowns and democratic breakdowns

47

Coalition duration and proxies by time period, 1979–2006

79

Table 2.2 Table 4.1

4

Table 5.1

Causes of presidential instability

102

Table 6.1

Government coalitions, parliamentary support, and executive law making strategy in Brazil, 1988–2007

117

Table 10.1 Challenges to presidents in the Dominican Republic

185

Table C.1 Table C.2

Presidential breakdowns: A country-based summary of causes and outcomes

214

Models of presidential breakdown

220

This page intentionally left blank

Acknowledgments

The first seeds for this project were sowed during the Joint Sessions of Workshops organized by ECPR in Cyprus in April of 2006. Five of the contributors to this volume first met and presented papers in the workshop on “Parliamentary Practices in Presidential Systems. (European) Perspectives on Parliamentary Power in Latin America,” organized by Detlef Nolte and Agustín Ferraro. Over a glass of white wine and some canapés under the warm sun in Nicosia, Detlef Nolte and Leiv Marsteintredet had the idea of establishing a Hanseatic project on Latin America, involving the universities of Bergen and Hamburg. That fall, our institutions applied for funding from the Norwegian Research Council and the German Academic Exchange Service (DAAD) to develop a cooperative project based on a series of workshops bringing together senior and young researchers on the topic of presidential breakdowns in Latin America. To our surprise, considering that we competed against geophysicists, chemists, and others working in the “hard” sciences, we got funding for two years. Without this funding, this book would never have materialized. Two meetings were held in Hamburg and two in Bergen, for which papers were prepared on several case studies: Argentina, Bolivia, the Dominican Republic, Guatemala, and Peru, as well as comparative overviews. As the project progressed, we realized that the workshop participants had enough material to consider publishing a book. We wanted to edit the first publication that compared all cases of presidential breakdowns, but our initial idea was to ignore cases of impeachment, as these had already been covered in journals and by other books (especially the cases of Brazil and Venezuela), and more recently in Pérez-Liñán’s book of 2007. Thanks to the insightful comments of an anonymous reviewer at Palgrave, we realized that these cases would enrich our book and make our volume more complete. Fortunately for us, the authors of the chapters on Venezuela (Rickard Lalander), Brazil (Argelina Cheibub Figueiredo), and Paraguay (Detlef Nolte) were sporting enough to accept our relatively late invitation to join the project. A special thanks to them, and also to Andrés Mejía-Acosta and John Polga-Hecimovich who also received a late invitation to write about Ecuador and, despite many other obligations, met the editors’ deadlines without complaint. We wish to thank all authors for their patience throughout the various rounds of revisions and sometimes strict demands to comply with deadlines and word count requirements.

xii

Acknowledgments

Alexandra Barahona de Brito went beyond her duties in reading, correcting, and pointing out errors in our chapters. Her help has been invaluable, and has helped tremendously to improve the final product. We also want to thank Anabel Leandro ([email protected]), a photo-artist living and working in Buenos Aires, who we came into contact with through a photo-sharing webpage, Flickr, while searching for a photo for the front cover. Her photograph of the streets of Buenos Aires symbolizes for us much of what we believe to be central to presidential breakdowns. Most of the funding for this project was kindly provided by the Norwegian Research Council and the DAAD. However, our institutions, the Department of Comparative Politics at the University of Bergen, and the GIGA Institute of Latin American Studies in Hamburg, were generous enough to provide funding for the final editing, translation, and proofreading of this book. This turned out to be extremely important since few of us are native English speakers. Leiv Marsteintredet wishes to thank the Meltzer foundation at the University of Bergen, the U.S.-Norway Fulbright-Hayes Foundation, and the Norwegian Research Council’s Leiv Eriksson Scholarship for funding a year at University of California, Berkeley, and other research-trips. Finally, we extend our thanks to our families who were patient and understanding when, due to a nine-hour time difference between Berkeley and Hamburg, we exchanged e-mails well beyond office hours.

Abbreviations

ACS AD ADN AGD ALBA AP ARENA CACIF

CAP CC CFP COB CONAIE COPEI

COPRE CPI CSUTCB

Civil Society Assembly (Asamblea de la Sociedad Civil)— Guatemala Democratic Action (Acción Democrática)—Venezuela Nationalist Democratic Action (Acción Democrática Nacionalista)—Bolivia Agency of Deposit Guarantees (Agencia de Garantía de Depósito)—Ecuador Bolivarian Alternative for the Americas (Alternativa Bolivariana para las Américas) Country Alliance (Alianza País)—Ecuador National Renovation Alliance (Aliança Renovadora Nacional)— Brazil Coordinating Committee of Agricultural, Commercial, Industrial, and Financial Associations (Comité Coordinador de Asociaciones Agrícolas, Comerciales, Industriales y Financieras)—Guatemala Carlos Andrés Pérez—Venezuela Constitutional Court (Corte Constitucional)—Guatemala Concentration of Popular Forces (Concentración de Fuerzas Populares)—Ecuador Bolivian Workers’ Central (Central Obrera Boliviana) Confederation of Indigenous Nationalities of Ecuador (Confederación de Nacionalidades Indígenas del Ecuador) Christian Democratic Committee of Independent Electoral Politics (Com ité de Organi zación Política Elec toral Independiente)—Venezuela Presidential Commission for the Reform of the State (Comisión Presidencial para la Reforma del Estado)—Venezuela Joint Parliamentary Commission of Inquiry (Comissão Mista Parlamentar de Inquérito)—Brazil Single Union Confederation of Peasant Workers of Bolivia (Confederación Sindical Única de Trabajadores Campesinos de Bolivia)

xiv

Abbreviations

CTV DGC DP EMP FARC FEDECAMARAS

FENASTEG

FRA FREPASO FRG FSTMB

FUN GAM GDP GSP ICC ID IFES IA IMF IPEA JCE JD JLCA

Confederation of Workers of Venezuela (Confederación de Trabajadores de Venezuela) Guatemalan Christian Democracy (Democracia Cristiana Guatemalteca) Popular Democracy (Democracia Popular)—Ecuador Presidential Military Guard (Estado Mayor Presidencial)—Guatemala Revolutionary Armed Forces of Colombia (Fuerzas Armadas Revolucionarias de Colombia)—Colombia. Federation of Chambers and Associations of Commerce and Production (Federación de Cámaras y Asociaciones de Comercio y Producción)—Venezuela National Federation of Workers Unions of Guatemala (Federación Nacional de Sindicatos de Trabajadores de Guatemala) Alfarista Radical Front (Frente Radical Alfarista)— Ecuador Country in Solidarity Front (Frente País Solidario)— Argentina Guatemalan Republican Front (Frente Republicano Guatemalteco) Federation of Unions of Mine Workers of Bolivia (Federación Sindical de Trabajadores Mineros de Bolivia) National Unity Front (Frente Unidad Nacional)— Bolivia Mutual Support Group (Grupo de Apoyo Mutuo)— Guatemala Gross domestic product Generalized System of Preferences Tax on Capital Circulation (Impuesto a la Circulación de Capitales)—Ecuador Democratic Left (Izquierda Democrática)—Ecuador International Foundation for Election Systems Irrelevant alternative International Monetary Fund Institute of Research on Applied Economics (Instituto de Pesquisa Econômica Aplicada)—Brazil Junta Central Electoral (Junta Central Electoral)— Dominican Republic Youths for Democracy (Jóvenes por la Democracia)— Paraguay Court for Contentious Administrative Proceedings (Juzgado de Letras de lo Contencioso Administrativo)— Honduras

Abbreviations JNE MAS MAS MBR-2000 MEP MERCOSUR MIN MIP MIR MNR MNRH MNRI MP MPD MUPP NDI NFR OAS ONPE PAC PAN PAP PC PCB PCdoB PCE PCV

xv

National Election Board (Junta Nacional Electoral)—Peru Movement to Socialism (Movimiento al Socialismo)—Bolivia, Venezuela Solidarity Action Movement (Movimiento de Acción Solidaria)—Guatemala Revolutionary Bolivarian Movement 200 (Movimiento Bolivariano Revolucionario)—Venezuela People’s Electoral Movement (Movimiento Electoral del Pueblo)—Venezuela Southern Common Market (Mercado Común del Sur)— Argentina, Brazil, Paraguay, and Uruguay Movement of National Integration (Movimiento de Integración Nacional)—Ecuador Pachakuti Indigenous Movement (Movimiento Indigena Pachakuti)—Bolivia Revolutionary Left Movement (Movimiento de Izquierda Revolucionario)—Bolivia, Venezuela Nationalist Revolutionary Movement (Movimiento Nacionalista Revolucionario)—Bolivia Historical Nationalist Revolutionary Movement (Movimiento Nacionalista Revolucionario Historico)—Bolivia Left Nationalist Revolutionary Movement (Movimiento Nacionalista Revolucionario de Izquierda)—Bolivia Provisional measures (medidas provisórias)—Brazil Popular Democratic Movement (Movimiento Popular Democrático)—Ecuador Pachakutik Plurinational Unity Movement (Movimiento Unidad Plurinacional Pachakutik)—Ecuador National Democratic Institute—USA New Republican Force (Nueva Fuerza Republicana)—Bolivia Organization of American States National Office of Electoral Procedures (Oficina Nacional de Procedimientos Electorales)—Peru Voluntary Civilian Self-Defence Committees (Patrullas de Autodefensa Civil, PAC)—Guatemala National Progress Party (Partido de Avanzada Nacional)— Guatemala Popular Action Party (Partido de Acción Popular)—Peru Colorado Party (Partido Colorado)—Paraguay Bolivian Communist Party (Partido Comunista Boliviano) Brazilian Communist Party (Partido Comunista do Brasil) Ecuadorian Conservative Party (Partido Conservador Ecuatoriano) Venezuelan Communist Party (Partido Comunista Venezolano)

xvi

Abbreviations

PDC PDH PDN PDS PFL PJ PL PLD PLRA PMDB PN PP PRD PRE PRIAN PRN PRSC PS-1 PSC PSD PSD PSP PT PUNACE PUR RECADI SIN STF

Christian Democratic Party (Partido Demócrata Cristiano)— Bolivia Human Rights Ombudsman (Procuraduría de Derechos Humanos)—Guatemala National Democratic Party (Partido Democrático Nacional)— Venezuela Social Democratic Party (Partido Democrático Social)—Brazil Liberal Front Party (Partido da Frente Liberal)—Brazil Justicialist or Peronist Party (Partido Justicialista) Liberal Party (Partido Liberal)—Honduras Dominican Liberation Party (Partido de la Liberación Dominicana) Authentic Radical Liberal Party (Partido Liberal Radical Auténtico)—Paraguay Brazilian Democratic Movement Party (Partido do Movimento Democrático Brasileiro) National Party (Partido Nacional)—Honduras Progress Plan (Plan Progreso)—Bolivia Dominican Revolutionary Party (Partido Revolucionario Dominicano) Ecuadorian Roldosista Party (Partido Roldosista Ecuatoriano) Institutional Renewal Party of National Action (Partido Renovador Institucional de Acción Nacional)—Ecuador National Reconstruction Party (Partido da Reconstrução Nacional)—Brazil Social Christian Reformist Party (Partido Reformista Social Cristiano)—Dominican Republic Socialist Party One (Partido Socialista Uno)—Bolivia Social Christian Party (Partido Social Cristiano)—Ecuador Democratic Socialist Party (Partido Socialista Democrático)— Argentina Social Democratic Party (Partido Social Demócrata)— Guatemala 21 January Patriotic Society Party (Partido Sociedad Patriótica 21 de Enero)—Ecuador Workers’ Party (Partido dos Trabalhadores)—Brazil National Union Party of Ethical Citizens (Partido Unión Nacional de Ciudadanos Éticos)—Paraguay Republican Union Party (Partido Unión Republicana)—Ecuador Differential Exchange Rates Regime (Régimen Cambiales Diferenciales)—Venezuela National Intelligence Service (Servicio de Inteligencia Nacional)— Peru Federal Supreme Court (Supremo Tribunal Federal)—Brazil

Abbreviations TSE UCN UCR UCS UD UDP UNACE UNE URD URNG USAD VAT WB WDI

xvii

Supreme Electoral Tribunal (Tribunal Supremo Electoral)— Guatemala National Centrist Union (Unión del Centro Nacional)— Guatemala Radical Civic Union (Unión Cívica Radical)—Argentina Union of Civic Solidarity (Unión Civica Solidaridad)—Bolivia Democratic Unification (Unificación Democrática)—Honduras Democratic Popular Union (Unión Democrática y Popular)— Bolivia National Unity of Ethical Colorados (Unidad Nacional de Colorados Éticos)—Paraguay National Unity of Hope (Unidad Nacional de la Esperanza)— Guatemala Republican Democratic Union (Unión Republicana Democrática)— Venezuela Guatemalan National Revolutionary Union (Unidad Revolucionaria Nacional Guatemalteca) Santo Domingo Autonomous University (Universidad Autónoma de Santo Domingo)—Venezuela value added tax World Bank World Development Indicators

This page intentionally left blank

Introduction: Presidentialism and Presidential Breakdowns in Latin America Mariana Llanos and Leiv Marsteintredet To change the government in a presidential regime when the president is unwilling . . . requires a break with the rules of democratic election of the chief executive: government crises almost by definition become regime crises. —(Linz 1978, 74)

When Juan Linz published his book on the breakdown of democratic regimes, the third wave of democratization was just beginning in Latin America. At the time, military regimes and regime instability had been the rule for decades, and in this context Linz argued that, among other factors, the characteristics of presidentialism, the prevailing regime type in the region, could explain Latin America’s endemic instability and authoritarian tendencies. Thirty years on, however, the third wave seems to have marked the end of democratic breakdown and long-lived authoritarian regimes. Latin America is still dominated by a variety of presidential regimes and, with few exceptions, democracy, albeit imperfect, holds sway from the Río Bravo to Tierra del Fuego. Even though democratic regime breakdowns are less of a peril now than they were thirty years ago, government crises have persisted into the democratic era. These crises, which in contemporary Latin America occur separately from regime crises, are the object of study of this book. In effect, over the past decades, Latin America has witnessed a new pattern of political instability characterized by the extraordinary exit from power of presidents before their constitutionally fixed mandate has ended, without the ensuing government crisis leading to regime breakdown. This phenomenon has been characterized, variously, as presidential removal (Pérez-Liñán 2007), presidential failure (Hochstetler and Edwards 2009), presidential fall (Hochstetler 2006), or interrupted presidency (Kim and Bahry 2008; Marsteintredet and Berntzen 2008; Negretto 2006; Valenzuela 2004). In

2

Mariana Llanos and Leiv Marsteintredet

this book, this new form of government or presidential crisis is referred to as presidential breakdown. It is not our purpose here to add to the proliferation of new concepts or neologisms, but by looking at the cases in detail and from a longer-term perspective as we do here, the concept of breakdown seemed to us more embracing and neutral than others, which narrowly refer to a particular way of leaving power (such as through resignation, removal, deposal, or ouster), make implicit normative judgments (“failure”), or refer both to the fall and rapid succession of presidents (“interruption”). In our view, presidential breakdown denotes the very specific phenomenon of early termination of an elected presidency, which entails a broad variety of manifestations, causes, and implications, all of which have motivated the organization of this book. Presidential breakdowns are interesting for a host of reasons, one of which is that they challenge a previously accepted understanding of how presidential regimes function. When contrasted with other regime types, presidential regimes have generally been characterized as involving the direct election of the executive and legislative branches, and by the fact that both are elected for a fixed term (Linz 1994; Shugart and Carey 1992). The phenomenon of presidential breakdowns clearly challenges the assumptions that a president cannot be removed prior to the expiration of his or her fixed term of office, and, very likely to Linz’s surprise, that there is a link between the premature departure from power of presidents and the breakdown of democratic regimes.1 However, the fact that government crises do not result in regime crises and that some of the essential features of presidential regimes have been altered does not mean that Linz’s and others’ institutional analyses of presidential systems are entirely wrong. As Pérez-Liñán (2003, 157) states, because of the democratic transformation that took place in Latin America in the 1980s, struggles between presidents and congresses are less likely to destabilize regimes, although they can still destabilize governments, and even lead to their removal. From this perspective, this book is part of the body of work intended to improve our understanding of presidentialism as a regime type. The central aim of this book is to provide the first comprehensive account of all presidential breakdowns that have occurred so far during the third democratization wave in Latin America. 2 Since the scholarly literature on presidential breakdowns first emerged in 2003 and 2004, there have been few published case studies that apply the lessons we have learned from those experiences. This book therefore aims to fill this gap. The bulk of the work consists of various chapters with country studies, in which presidential breakdowns are analyzed by reference to existing theories on the topic. Presidential breakdowns are the result of a complex causality and a chain of events that are difficult to fully capture with statistical models. The topic therefore lends itself to case study analysis, comparative and process-tracing methods, which are especially appropriate to deal with complex causality, and complement statistical analyses. Some chapters in the book offer the first

Introduction

3

comprehensive analysis of the presidential breakdown in particular countries (Ecuador and Bolivia), while others reassess cases already studied in recent literature on the topic (Argentina and Peru). Two additional chapters, in the first section of the book, aim to shed new light on the study of presidential crises and breakdowns in Latin America from a comparative perspective. In the remainder of this introduction we discuss Latin American presidential breakdowns, and then the literature that attempts to pinpoint the causes of these breakdowns. The final section outlines how the book is organized and provides a brief summary of the contents of the chapters, stressing their author’s position vis-à-vis the comparative literature.

Presidential Breakdowns: A Latin American Phenomenon Even though presidential breakdown is not a solely Latin American phenomenon, the majority of the cases that have occurred during the current wave of democratization can be found in this region, more specifically in South America (Hochstetler 2006; Hochstetler and Edwards 2009). Table I.1 below presents a list of the cases of presidential breakdowns in Latin America, all of which are discussed in this volume. If one considers all third-wave presidential democracies, one finds that only one president left office prematurely in Asia (President Estrada in the Philippines in 2001, see Kasuya 2003), while in Africa presidents have usually been removed through coups (Kim and Bahry 2008, 813). In Latin America, by contrast, only two of the fourteen cases of presidential breakdowns dealt with here were (failed) coup attempts, and even in these cases civilian rulers either maintained power throughout the whole process or quickly regained it. 3 Clearly, then, presidential breakdown as we define the phenomenon seems to occur more frequently in Latin America than in other regions. This is not that surprising: not only are the world’s great majority of pure presidential regimes to be found in Latin America (Siaroff 2003), but U.S. policy toward the region since the end of the cold war has discouraged crisis solutions that do not comply with democratic rules (Marsteintredet, this volume; Valenzuela 1997; Villagrán de León 1993). It seems evident that regions of the world share some political tendencies and countries within them exert mutual influences on one another because of their proximity. However, country specificities must be considered along with regional trends and influences. Within the Latin American region, the rate of presidential breakdowns is higher in South America, where twelve out of sixty-four elected presidents between 1978 and 2005 were removed, compared with only two out of thirty-eight in Central America and the Caribbean. Three countries have experienced more than one presidential breakdown (Argentina, Bolivia, and Ecuador), five have experienced one, and the remaining ten have not experienced any breakdowns during this last

4

Mariana Llanos and Leiv Marsteintredet

Table I.1

Presidential breakdowns in Latin America

Country and Year of Election*

President

Breakdown

Raúl Alfonsín Fernando de la Rúa

June 1989 December 2001

Hernán Siles Zuazo Gonzalo Sánchez de Lozada Carlos Mesa

July 1985 October 2003 June 2005

Fernando Collor de Melo

November 1992

Joaquín Balaguer

Summer 1994/96

Abdalá Bucaram Jamil Mahuad Lucio Gutiérrez

February 1997 January 2000 April 2005

Jorge Serrano Elías

May–June 1993

Raúl Cubas

March 1999

Alberto Fujimori

September 2000

Carlos Andrés Pérez

May 1993

Argentina 1983 1999 Bolivia 1982 2002 2003 Brazil 1990 Dominican Republic (1986) 1990 Ecuador 1996 1998 2002 Guatemala 1989/90 Paraguay 1998 Peru (1990) 2000 Venezuela 1988

* The years in brackets refer to the first time reelected presidents were elected to the post. The breakdown of elected vice president Mesa is included, but the breakdown of caretaker presidents, resignations due to illness, or the accidental deaths of presidents are not considered presidential breakdowns and therefore excluded from the table.

democratic spell. Since cross-country differences are broad in Latin America, country analyses can show how seemingly similar factors have different effects across countries and time (causal complexity). The chapters in this book attempt to capture this causal complexity and provide a more in-depth analysis of the cases than has been offered in previous medium- to large-N studies. Country studies also allow us to focus more easily on differences across presidential breakdowns and evaluate the extent to which it is possible

Introduction

5

to draw general conclusions or to speak of a similar phenomenon throughout the region.

Explaining Presidential Breakdowns Linz (1978; 1990; 1994) initiated a debate on presidential institutions that is still ongoing, at least with respect to Latin America. His original argument was that due to the two essential features of presidentialism (Linz 1994, 6)—namely, the fixed term of office and the direct election of the executive and legislative—this regime type suffered from both rigidity and pervasive deadlock problems. The fixed term makes crisis and conflict resolution more difficult, in contrast with the flexible solutions available to parliamentary regimes, including the dissolution of parliament, the vote of confidence, and the possibility of calling new elections. The direct election of the executive and legislative powers gives both president and Congress direct, and dual, democratic legitimacy, thus inducing interinstitutional struggles and making it unclear which will prevail when there is a conflict between the two. Presidentialism shapes the political process or “way of ruling” both directly and indirectly (Linz 1990, 56), causing other problems related to the regime type, such as lack of incentives to build coalitions, polarization during elections due to the personal election of the chief executive, weak and undisciplined parties, and the myth of strong leadership (see also Linz 1994). Linz further argued that the absence of majorities in Congress, which increases the likelihood of deadlock, was a predictable consequence of these essential features of presidentialism. For Linz, then, it was no accident that in situations of interinstitutional conflict caused by the factors related to the presidential system, the armed forces were often tempted to intervene as a mediating power. Linz’s sweeping generalizations about presidential regimes have been widely criticized as well as refined.4 There have been varying levels of empirical support for Linz’s claims (Álvarez 1998; Przeworski et al. 2000; Cheibub 2002; Cheibub and Limongi 2002; Cheibub, Przeworski, and Saiegh 2004; Stepan and Skach 1994), but the most recent findings suggest that the problems Linz attributed to presidentialism are either nonexistent or are much less pervasive than was previously thought. Notably, the problems that presidentialism allegedly causes for democracy are actually caused by other factors, such as the military legacy of presidential regimes (Cheibub 2007). However, the potential risks posed by certain institutional constellations within presidentialism, particularly minority governments, which were pinpointed in the early 1990s in the institutional literature (Jones 1995; Mainwaring 1993; Shugart and Carey 1992; Shugart and Mainwaring 1997) still have explanatory value for the presidential breakdown phenomenon. Indeed, in their

6

Mariana Llanos and Leiv Marsteintredet

attempt to explain this new form of political instability in the region, several scholars have resorted to the institutional arguments inspired by Linz, changed the dependent variable from democratic breakdown to presidential breakdown, and thereby reignited the debate.

Presidential Weakness, Congress, and the Streets For Valenzuela, presidential breakdowns confirm the perils associated with the presidential regime type (Valenzuela 1993; 2004). The personalization of authority in the figure of the president, together with serious economic problems (including high inflation, economic stagnation, state financial crises, foreign and domestic public indebtedness, unemployment, and increased inequality) and memories of former powerful and nondemocratic presidents, increased the expectations of citizens regarding the performance of thirdwave presidents (Valenzuela 2004, 12–13). The failure to fix the country’s problems, which under presidentialism are not viewed as party or movement failures but rather as failures of the chief executive, generated mounting unrest and put presidential stability into question. This argument calls to mind the dangers that O’Donnell (1994) associated with delegative democracy. Delegative democracies are strongly majoritarian as they empower an individual to be the embodiment and interpreter of the high interests of the nation for a specific amount of time, in isolation from political institutions and organized interests. Needless to say, in these systems, the shift from broad popularity to general discontent, from omnipotence to impotence, can be as rapid as they are dramatic. In the absence of military threats, a widely reviled president must hang on until the end of his term (O’Donnell 1994, 67), or face his fate under pressure from the streets and Congress (Valenzuela 2004, 12). It is precisely these two latter factors, which Valenzuela stresses, that have been dealt with and considered to be crucial explanatory variables in the comparative literature dealing with the breakdown of presidencies. In effect, Valenzuela (2004, 14) found that, with the exception of Carlos Andrés Pérez in Venezuela and Raúl Cubas in Paraguay, 5 weak minority presidents are a common denominator of presidential breakdowns; even Pérez and Cubas faced political revolts in Congress, which Valenzuela sees as confirming the prevalence of the logic of confrontation in presidential systems. Thus far, different comparative analyses have corroborated that insufficient party support in Congress is an important explanatory variable of presidential breakdowns (Álvarez and Marsteintredet, this volume; Hochstetler 2006; Kim and Bahry 2008). Similarly, Pérez-Liñán (2007) demonstrates that presidents need a legislative shield to protect themselves from impeachment attempts. The formation of a shield depends on the size of the president’s party, its cohesion, and the capacity of the president to gain support among opposition legislators (Pérez-Liñán 2007, 246). Despite these arguments, there is a growing literature on coalition presidentialism that shows that coalitions are not necessarily short-lived and ad hoc under presidential

Introduction

7

rule (Altman 2000; Amorim Neto 2006; Cheibub, Przeworski, and Saiegh 2002; Dehesa 1998), and there are successful examples of coalition politics (such as in Brazil and Chile) that contradict the Linzian argument about the anti-cooperative logic of presidential institutions. For Negretto (2006), minority presidents should not necessarily experience problematic interbranch relations and face the risk of breakdown, particularly if they control the median and veto voter in Congress. More specifically, it is the minority government that is not supported by non-median voter in Congress that runs the risk of not completing the fixed term. Some of the above-mentioned analyses acknowledge that institutions are clearly not the only factor causing presidential breakdowns. They also highlight the explanatory value of event variables, particularly street protests, and, in some cases of presidential breakdowns, of political scandals. High levels of protest on the streets (often in reaction to scandals) were found to be a noninstitutional factor causing or triggering presidential breakdown (Hochstetler 2006; Pérez-Liñán 2003; 2007). Other works also point to the convergence of institutional and noninstitutional factors in presidential breakdowns (Morgenstern, Negri, and Pérez-Liñán 2008; Pérez-Liñán 2008). Hochstetler (2006) and Pérez-Liñán (2007) both argue that street protest is more important than congressional politics for causing breakdowns, a claim strongly supported in this volume by Buitrago’s analysis of the Bolivian case. Hochstetler (2006) argues that when legislative action did not provoke, or was not accompanied by popular reaction, presidential challenges failed. Thus, the “street” plays the role of poder moderador, a role previously corresponding to the armed forces. Likewise, Pérez-Liñán remarks that “popular uprisings, rather than legislative action, are the main force behind the collapse of elected administrations” (Pérez-Liñán 2007, 206). Finally, Hochstetler and Edwards (2009) show that presidents who repressed protesters, particularly those who used deadly force (presumably with the intention of protecting their position in office), set in motion their own demise.

What Triggers Demands for a President’s Ouster? If anything is clear from the above, it is that there are some necessary preconditions for setting in motion institutional or street challenges. Yet, relatively less attention has been paid to the factors that motivate Congress and/or the public to seek to remove a president from office. Two factors, however, stand out: scandals involving the president; and economic performance. We also briefly discuss two other factors that are examined thoroughly in other chapters in this book and in its conclusion: namely, regime-related rather than personal scandals that imperil the survival of democracy; and intra-governmental factors including, but not exclusive to, presidential leadership. The importance of scandals in explaining presidential breakdowns depends on the level of maturity of democracy in two ways. First, there is

8

Mariana Llanos and Leiv Marsteintredet

clearly a nexus between a scandal and the role of the media in exposing a scandal to the public (Pérez-Liñán 2007, chs 4–5). A precondition for the media to play a watchdog role is that freedom of expression should be guaranteed, and we can only assume that the quality of this guarantee correlates positively with the level or maturity of democracy. However, only some of the presidential breakdowns resulted directly from personal scandals implicating the president and leading to impeachment procedures (the cases of presidents Collor de Melo, Cubas, and Andrés Pérez). Thus, in comparative terms, the explanatory value of the variable is limited (Morgenstern, Negri, and Pérez-Liñán 2008, 185). The level of democratic maturity is also linked to which type of scandal is most likely to affect presidential survival. In hybrid regimes, such as those of Guatemala, Peru, and the Dominican Republic, it was not personal scandals but scandals tout court, which imperiled precarious democratic regimes, which condemned presidencies to failure. These cases of presidential breakdown seem to follow a different logic that is not so easily explained by the theories referred to above. As regards structural variables, bad economic and socioeconomic performance can cause presidential breakdown as it triggers a series of protests and weakens presidential popularity. Hochstetler (2006) has argued that it may be the generally disliked neoliberal reforms rather than economic performance that affect the survival of presidents in Latin America. However, as the examples of Paz Estenssoro in Bolivia and Menem in Argentina suggest, broad programmes of neoliberal reform do not necessarily stand in the way of presidents completing their mandates. Likewise, as Valenzuela (2004, 11–12) points out, presidents who have avoided making decisive policy choices because they feared a public outcry—Valenzuela gives as examples Collor de Melo, Siles Zuazo, Alfonsín, and Bucaram—have also paid a price for their relative inaction, as national currencies collapsed and inflation spun out of control. So it appears that it is not necessarily neoliberal reforms per se that make the difference for presidential survival, but rather how an economy is handled. Álvarez’s and Marsteintredet’s (this volume) findings support this claim. They argue that economic performance affected both democratic and presidential breakdowns in Latin America, a conclusion that is also confirmed in global analyses (Hochstetler and Edwards 2009). As Linz initially argued (1990, 60), it appears that presidents are at least partly responsible for their early demise if we consider that presidential instability is causally related with economic performance, scandals involving the president, and presidential dealings with Congress and the “street.” To date, however, the comparative literature has not dealt with presidential leadership and intra-government politics in an explicit and systematic manner, as the focus has been on presidential interaction with Congress or other political actors. Several chapters of this book concentrate on this and other related issues. We briefly present the main arguments below, stressing how the standpoint adopted by the authors of the chapters in this book fits within the analyses and explanations provided by existing comparative analyses.

Introduction

9

Organization The chapters of this book are divided into four sections. The two chapters in the first section provide new comparative insights on the ongoing debate on presidential breakdowns. The second section focuses on countries that have experienced more than one presidential breakdown, namely, Argentina, Bolivia, and Ecuador. The third section includes the cases involving personal presidential scandals and impeachment (Brazil, Venezuela, and Paraguay). The fourth and final section includes cases of presidential breakdown caused by attempts to rupture the precarious democratic order in hybrid regimes (Guatemala, the Dominican Republic, and Peru). In her chapter, Mustapic argues that presidential breakdowns challenge the postulate that it is only in parliamentary regimes that legislatures participate in forming and dismissing governments. For this author, Latin American presidential breakdowns evince two kinds of institutional solutions to crisis in which the composition of Congress plays a relevant role: there is the “presidential solution,” when the outgoing president is in charge of dealing with his own succession, a situation that typically occurs when there are no possible alternative majorities in Congress; and the “parliamentary solution,” which occurs when an alternative majority forms in congress, which withdraws its support for the sitting president, hastens his resignation from office, and chooses a successor. Thus, the analysis of the presidential breakdowns shows that the president’s fixed mandate is no longer an essential feature of presidentialism, but that the Congress’s fixed mandate is: indeed, it is only the first that can be altered without provoking a regime breakdown. Álvarez and Marsteintredet use statistical analyses to compare the causes of democratic and presidential breakdowns in Latin America in the postSecond World War period in order to find out whether presidential breakdowns are merely the modern equivalent of democratic breakdowns in an era when military takeovers are considered beyond the pale. They find that the two phenomena share some causal features, namely, economic hardship and societal mobilization, but that there are structural differences between the two. While democratic breakdowns seem to occur in historically unstable regimes that score relatively lower in terms of democratic performance, presidential breakdowns are not affected by past democratic experience. On the other hand, institutional factors such as party fragmentation and a support for a president in Congress seem to be more relevant when explaining presidential breakdowns than when addressing democratic breakdowns. Argentina is one of the cases of repeated presidential breakdown. In her chapter, Llanos analyses the breakdown of elected presidents Raúl Alfonsín and Fernando de la Rúa and, briefly, that of the two caretakers that followed de la Rúa, namely, the presidencies of Adolfo Rodríguez Saá and Eduardo Duhalde. The author argues that the first two cases confirm the importance of the variables of minority status, adverse economic context, and public disorder on the streets that have been highlighted in the literature, but that

10

Mariana Llanos and Leiv Marsteintredet

a closer look at these cases highlights the relative importance of both institutional and president-centered explanations. Llanos shows how defeats in the midterm elections of 1987 and 2001 activated the perils inherent in minority governments (interinstitutional conflicts and confrontation) and enhanced presidential vulnerability. Meanwhile, the different ways in which the presidents approached the exercise of power explain why their exits from power followed such different paths and were based on different institutional solutions in each case, an observation that also applies to the caretakers’ departure from power. Mejía Acosta and Polga-Hecimovich analyze the ousting of presidents Abdalá Bucaram, Jamil Mahuad, and Lucio Gutiérrez in Ecuador. The authors find that the factors usually associated with these presidential crises, such as the role of the military, indigenous and popular mobilization, and economic crisis, contributed to but not decisive for the demise of these presidents. Instead, the Ecuadorian presidential crises were linked to the rapid erosion of legislative coalitions, which began to unravel as a consequence of a series of institutional reforms approved in the mid-1990s. In principle, these were meant to curb corruption and strengthen presidential authority, but in practice they blocked the formation of policy coalitions and created the conditions for presidential instability. In his chapter on the breakdown of the presidencies of Hernán Siles Zuazo, Gonzalo Sánchez de Lozada, and Carlos Mesa in Bolivia, Buitrago claims that the difficult relations between these minority presidents and Congress, intra-government conflicts related to coalition politics, and social protests against certain policy measures in particular contributed to weakening the position of the president in office and became a potential source of presidential instability. However, it was only when generalized and high-intensity protests put pressure on an already weakened president that the outcome was presidential breakdown. The author argues that the important role of unions and new social movements is rooted in the historically politically active role played by civil society, and the fact that social protests have been the main means for these civil society groups to interact with the state. The section dealing with presidential breakdowns involving impeachments begins with Cheibub Figueiredo’s analysis of the Brazilian case. The author argues that Collor de Melo’s inability to obtain and sustain legislative support was a crucial factor in explaining the success of his impeachment, and that Congress, rather than popular demonstrations, was the force propelling the impeachment of the president. The impeachment followed the classical procedure, whereby a constitutional mechanism designed to deal with exceptional, albeit anticipated, situations in presidential systems was activated. The author also stresses that the congressional opposition did not act irresponsibly but rather independently, placing limits on Collor de Melo’s extremist policy preferences, compelling the president to broaden his coalition, and finally impeaching him on the basis of evidence gathered by the commission established for the purpose of investigating the president’s alleged corruption. Contrary to claims that presidential breakdowns are a

Introduction

11

one-off exercise in “checks and balances,” and that they signal a crisis of presidential democracy, the author argues that the effects of the impeachment of Collor de Melo were mainly positive for democratic governance in Brazil. Lalander argues that in order to understand why Carlos Andrés Pérez of Venezuela was impeached it is necessary to consider deeper, underlying political factors and not just the level of support the president had in Congress and the scandal that finally triggered the impeachment process. The author describes the political actors involved in the process, particularly emphasizing the collapse of the partyarchic model involving the Democratic Action (Acción Democrática, AD) and Christian Democratic Committee of Independent Electoral Politics (Comité de Organización Política Electoral Independiente, COPEI) parties. If partyarchy once guaranteed the relative stability of the Venezuelan political system, it also proved to undermine it by hindering the institutionalization of political accountability of political leaders. In that context, President Andrés Pérez became a scapegoat and his impeachment crystallized the opposition against the failed parties. The breakdown of President Andrés Pérez is thus also a story of the breakdown of a party system. Nolte’s chapter on Paraguay stresses that this case constitutes a typical case because the causes of presidential breakdowns usually cited in the comparative literature—a fragile legislative shield, strategic presidential errors, scandals, and violent mass demonstrations—all played a role in the demise of Cubas. However, the author stresses the many interests that converged to oust President Cubas from power (notably those of his possible successors), and the power struggle within the Colorado Party, which became particularly dangerous when it evolved into an institutional conflict between three state powers: the presidency, Congress, and the Supreme Court. In line with Lalander’s analysis, Nolte argues that the impeachment was only the last in a chain of events that had increased the vulnerability of the Cubas presidency to breaking point. Christensen and Petersen’s chapter deals with the failed self-coup of President Serrano of Guatemala, and it is the first in the last section dealing with attempts to stamp out the already precarious democratic features of hybrid regimes. The authors argue that the president’s minority situation in Congress and his bold confrontational attitude may explain the president’s decision to opt for a self-coup. However, they also show that the subsequent rapid restoration of constitutional order was mainly the result of an elite settlement, rather than of pressure from the street or Congress—in which the participation of de facto power structures, namely, of the business elite and the armed forces, was crucial. In retrospect, the restoration of democracy as the result of a settlement among traditional forces outside formal institutions did not help to make this democracy more accountable and representative. Finally, the epilogue discusses the most recent presidential crisis in Honduras, and examines its implications for theories of presidential breakdown.

12

Mariana Llanos and Leiv Marsteintredet

Marsteintredet discusses the difficulties of classifying the fall of President Balaguer of the Dominican Republic as a presidential breakdown: although his fall from power was an extraordinary and forced shortening of a presidential term of office, it is actually the only Latin American case in which the shortening of the term was formalized through a constitutional reform. At the same time, because democracy in the Dominican Republic was then at the “lower” end of the democracy scale, the author finds that existing theories lose explanatory value when it comes to the breakdown of Balaguer’s presidency. Indeed, theories of transition to democracy and of hybrid regimes are much more helpful when explaining this case, since elite pacts, and international pressure from the United States were the key factors behind the demise of President Balaguer. In a similar vein, in the last country chapter, Berntzen and Skinlo argue that the decision of President Fujimori of Peru to shorten his third presidential term from five years to one year, and his abrupt resignation before ending that abbreviated one-year term, clearly qualify this as a case of presidential breakdown, as has been highlighted by the comparative literature. However, the literature cannot explain the outcome of Fujimori’s fall, which led not only to a change of government but also to a change of regime, from electoral authoritarianism to democracy. The literature on gray-zone regimes and transitions from (electoral) authoritarian regimes accounts more fully for the outcome of Fujimori’s abrupt resignation. The concluding chapter by Llanos and Marsteintredet sums up the arguments presented by the authors of the various chapters in this book, particularly highlighting the intra-governmental dimensions of presidential breakdown, and suggests future lines of research on the topic.

Notes 1. In fact, the only successful democratic breakdown since the beginning of the third wave of democratization in Latin America did not lead to a change of the chief executive. Fujimori’s self-coup (autogolpe) allowed him to remain in power for a further eight more years. 2. Just as this book was going into production, on June 28, 2009, the Honduran President Manuel Zelaya was deposed and sent into exile. At the time of writing (August 2009), no solution had been found for this presidential crisis, which has been condemned as an undemocratic military coup by all international actors. The case resembles other instances of presidential breakdown treated in this book, however. Although proximity does not allow us to reach any definitive conclusions about this case as we do regarding other presidential crises in this book, we provide an epilogue recounting the events so far and analyze them in comparative perspective. 3. Namely, President Serrano’s failed autogolpe in Guatemala in 1993, and the ousting of President Mahuad in Ecuador in 2000 by a civil-military junta.

Introduction

13

4. The literature based on Linz’s analysis of presidentialism is extensive and we do not provide a detailed account of its defenders and detractors. For an overview, see Elgie (2005). 5. In 2005, after the publication of Valenzuela’s essay, the breakdown of the presidency of Gonzalo Sánchez de Lozada in Bolivia was the third one that involved a president with a majority in congress.

This page intentionally left blank

I

Comparative Perspectives

This page intentionally left blank

1 Presidentialism and Early Exits: The Role of Congress Ana María Mustapic*

Introduction Conventional wisdom has it that democratic legislatures have three basic functions: to represent, legislate, and oversight. In parliamentary regimes, they have one other function: forming government. It is recognized, then, that it makes a difference whether a regime is presidential or parliamentary. But while forming government is not one of the responsibilities of presidential legislatives, there are some exceptions. The Argentine constitution before the 1994 reform, for instance, and the current Bolivian constitution, both establish that if no candidate obtains an absolute majority of votes, Congress must elect the president.1 In any case, the fact that a president can be designated by Congress in such specific circumstances does not entail the abandonment of a basic principle governing presidentialism, namely, that presidents are not accountable to Congress. The crucial difference between both regimes, then, is the different nature of relations between the executive and legislative branches in each system. In parliamentary governments, the fusion of both powers happens not only because the chief executive is nominated by parliament but also because it depends on parliament whether the former remains in power. By contrast, the separation of powers that is typical of presidential regimes means that access to and permanence in the presidency does not depend on Congress. But is this really the case? It is relevant to reexamine this statement in light of the many early presidential departures from power from the 1980s onward in Latin America’s democracies. First, it is not strictly true that it is not up to Congress whether a president remains in power, as the possibility of impeachment proves. As it happens, this was not taken into account in the literature because

18

Ana María Mustapic

impeachments were seen as an exceptional mechanism that was costly to implement in practise. However, since the “third wave” of democratization, there has emerged an abundance of material for the analysis of presidential impeachments in Latin America, as the book by Aníbal Pérez-Liñán (2007) illustrates. Second, the various existing works on the early departure of presidents from power—not only those included in this volume but also those published earlier—highlight how important parliamentary backing is in shaping the fate of presidents (Carey 2002; Pérez-Liñán 2005; Negretto 2006). From this, one can conclude that presidents depend on two accounts on Congress to remain in power: on the one hand, there is a de jure dependence, determined by the constitutional mechanisms available to depose presidents, on the other hand, there is a de facto dependence, as presidents need parliamentary support to remain in power. 2 With this framework in mind, I pursue two goals in this chapter. First, I describe the role of Congress in the process of solving presidential crises in Latin America from the 1980s onward. Second, I evaluate the implications that the foregoing analysis has for the conventional characterization of presidential regimes. In order to develop the first point, I attempt to show that the performance of Congress during crises may be more or less active, depending on the level of support it gives to a president. For this, I rely on a previous work (Mustapic 2005) that focuses on Argentina, and in which I refer to two “exit modalities” from crises. These modalities are conditioned by the nature of congressional support for presidents: a presidential and a parliamentary exit strategy. As regards the second goal, in light of the above analysis, I argue that it is pertinent to revise the way we have characterized presidential regimes to date. Conventionally, presidential regimes have two distinguishing features: the independent election of the executive and legislative, and the fixed mandate. In my view, the second feature should be revised by explicitly taking into account the fact that presidents lack the capacity to dissolve Congress. I consider that it is this institutional component of presidentialism that best captures the essence of the inner workings or political dynamic of that system. Put differently, the only fixed mandate that matters is that of Congress. Before proceeding, let me clarify my point of departure. This chapter takes it for granted that military takeovers are no longer viable in Latin America. The progressive demise of the military option is almost self-evident: coups no longer take place, even though the military may participate in some crisis episodes. Indeed, Mainwaring and Pérez-Liñán (2005) point out that before 1978 the odds of military coups in Latin America were twenty times higher. Their analysis shows that a key factor for democratic survival is the presence of a regional context that is more favorable to democracy. Obviously, the changes in the foreign policy of the United States and the collapse of the Soviet Union are relevant in this context. One of the effects of this shift is that international actors are more willing to impose high costs on openly authoritarian regimes (Mainwaring and Pérez-Liñán 2005). Given this background,

Presidentialism and Early Exits

19

when Latin America’s third- wave democracies have faced crises, they have been forced to choose mechanisms for survival from the menu of options provided by their own rules and institutions. My chapter proceeds as follows: in the first section, I introduce the universe of cases that will be examined here, present my hypothesis regarding the two paths for presidents to leave power in a crisis context, and then examine cases of presidential exits from power. In the following section, I focus on cases of parliamentary exits. I then assess the impact of early presidential exits on the working and characterization of presidential regimes. In the conclusions, I comment on the role that Congress has played in early presidential exits from power.

Presidential Exits from Crisis: Presidential Resignations My universe is composed of the presidential exits or salidas from power that took place after 1980. More specifically, I focus only on those cases that occurred under presidential regimes in those countries that, following the Mainwaring, Brinks, and Pérez-Liñán (2001) classification, are democracies or semi-democracies.3 This classification currently includes eighteen countries; as regards early exits from power, I exclude the resignation of Fujimori in Peru in 2000, since according to the classification of the above mentioned authors, this resignation occurred under an authoritarian regime. All the cases analyzed appear in table 1.1 below, which also includes additional data on the origin of mandates (elected or not), the year of the start of the mandate, the year in which the president left power, and the means whereby presidents were removed. Of the eighteen Latin American presidential regimes that qualify as democracies or semi-democracies, eight (or 44 percent) experienced early presidential salidas, be it through resignation or the ouster of the president. As regards the total of fifteen cases that appear in table 1.1 below, 66 percent (10 cases) took place in only three countries: Argentina, Bolivia, and Ecuador. In Mustapic (2005), I focused on Argentina, the country that has experienced no less than four presidential resignations since the return to democracy in 1983: the resignation of Alfonsín in 1989, of de la Rúa in 2001, of Rodríguez Sáa, also in 2001, and finally of Duhalde in 2003. On the basis of the Argentine case study, I formulated a hypothesis regarding the “types of crisis resolution.” I argued that there were two paths to resolve crises, both depending on the distribution of power among parties in Congress: presidential salidas and parliamentary exits. The first kind of salida happens when it is no longer possible to form alternative parliamentary majorities in Congress and when, with the onset of crisis, the president becomes involved in the management of a more or less orderly succession process. By contrast, the second kind of salida happens when the formation of a new opposition

Table 1.1

Early presidential Salidas, 1980–2006

Country

President

Origin of Mandate

Start of Year of Mechanism to Mandate Exit end Mandate

Argentina

Raul Alfonsín

Elected

1983

1989

Argentina

Fernando de la Rúa Adolfo Rodríguez Saa Eduardo Duhalde Hernán Siles Zuazo Gonzalo Sánchez de Lozada Carlos Mesa

Elected

1999

2001

Early Departure from Office/ Resignation Resignation

Congress

2001

2001

Resignation

Congress

2002

2003

Elected

1982

1985

Elected

2002

2003

Early elections/ Resignation Early elections/ Resignation Resignation

Line of 2003 succession/ Vice president Elected 1990

2005

Resignation

1992

Elected

1996

1997

Impeachment/ Resignation Congress declares “mental incapacity”/ Dismissal Congress declares “abandonment of office” (article 167 of the constitution)/ Dismissal Congress declares “abandonment of office” (article 167 of constitution)/ Dismissal Resignation

Argentina Argentina Bolivia Bolivia

Bolivia

Brazil Ecuador

Fernando Collor de Melo Abdalá Bucaram Ortiz

Ecuador

Jamil Mahuad

Elected

1998

2002

Ecuador

Lucio Gutiérrez Borbúa

Elected

2002

2005

Guatemala Jorge Serrano Elías Paraguay Raúl Cubas

Elected

1991

1993

Elected

1998

1999

Dominican Joaquín Republic Balaguer Venezuela Carlos Andrés Pérez

Elected

1994

1996

Elected

1989

1993

Impeachment/ Resignation Early elections/ Resignation Impeachment/ Dismissal

Presidentialism and Early Exits

21

bloc in Congress is feasible. In such cases, given the absence of parliamentary backing, the president is forced to resign. In the first instance, the impossibility of changing the status quo in Congress gives presidents some margin for maneuver. This allows them to engage in negotiations and have a decisive role in, say, forging agreements among political elites with a view to organizing a peaceful exit from power; in the second instance, the president lacks any such resources. The resignation of Raúl Alfonsín and Eduardo Duhalde in Argentina, of Joaquín Balaguer in the Dominican Republic, and of Hernán Siles Zuazo in Bolivia should be included in the category of presidential salidas. Let us consider the Argentine cases first. President Alfonsín, whose mandate should have ended in December 1989, decided to hand over power earlier in the midst of a hyperinflationary crisis, besieged by social conflict, and leading a government that lacked any authority whatsoever. It was this context that led him to call early elections in May rather than in October 1989, as had been the case until then.4 Once the new president, Carlos Menem, had been elected, Alfonsín negotiated his resignation before his full term was over. Thus, Menem took over the presidency in July 1989, five months before the date established by constitutional and legal norms. Deputies and senators concluded their mandates within the constitutionally stipulated time, however, and undertook to cooperate with the new president. For his part, Eduardo Duhalde assumed the presidency in January 2002 after being appointed by Congress. Indeed, the resignation of President Rodríguez Sáa meant that there was no head of government (or what in Argentina has been called a situation of “acephaly”), which set in motion the rules of succession for the second time in the space of a week. These rules charge Congress with the task of nominating a successor, who must complete the mandate of the outgoing president. 5 Clearly, the mode of accession to power—through congressional nomination rather than a popular vote—meant that President Duhalde was in a vulnerable position. To this were added the difficult circumstances that Argentina was experiencing at the time: the declaration of the debt default, the troubling consequences of the demise of convertibility, and the wave of street protests. Indeed, it was as a result of the political impact of the death of two protesters that Duhalde decided to opt for the early handover of power and to call early elections. This decision met no resistance, and Congress was willing to implement it immediately. As occurred under Alfonsín, only the executive was affected, since legislators remained in office until the end of their constitutionally established mandates. Given that the majority in both houses was of the same political color as the new president-elect, coexistence between the two powers did not present any difficulties. In the case of the Dominican Republic, President Joaquín Balaguer caved under pressure from the opposition and the United States (Espinal

22

Ana María Mustapic

2003; Valenzuela 2004), and ultimately agreed to the shortening of his mandate. The 1994 elections, in which Balaguer was declared the “winner,” triggered a serious crisis because of the various charges against proven cases of electoral fraud. An agreement, the so-called Pact for Democracy, was reached and ratified by Congress as part of the search for a solution to the crisis. In addition to stipulating a series of electoral and constitutional reforms, the Pact stipulated that Balaguer’s presidential mandate was to be shortened to two years, and that the mandates of legislative representatives should be upheld. Finally, in Bolivia, in 1985 Hernán Siles Zuazo accepted that elections should be held a year earlier; even though he had been elected in 1980, he had only been able to take office in 1982. He was forced to take this decision because he was cornered by the opposition in parliament, and because of a grave hyperinflationary crisis. In this case, too, the decision was implemented by means of an agreement reached through the so-called Dialogue for Democracy coordinated by the Catholic Church (Morales 1993). Unlike what happened in Argentina and the Dominican Republic, however, both the presidential and legislative posts were renewed. It could not be otherwise, since, as noted above, Bolivian constitutional rules delegate to Congress the election of the president if none of the candidates obtain an absolute majority of the vote. As in previous elections, nobody expected any candidate to win over 50 percent of the vote. For this reason, it was indispensable to call for legislative elections. The element that all these crisis resolutions have in common is the participation, to a greater or lesser degree, of the presidents themselves in the search for a solution. However, the most decisive element is that in none of these cases was it possible to form an alternative parliamentary majority to give renewed backing to the acting president. Under the Alfonsín presidency, the alternative to the two-party system constituted by the government party, the Radical Civic Union (Unión Cívica Radical, UCR), and the opposition Justicialist or Peronist Party (Partido Justicialista, PJ), would have been a coalition between both parties. But the PJ certainly had no interest in sharing the costs of the crisis, particularly in an electoral year. As for the Duhalde presidency, the latter had been elected by Congress with the support of both the PJ and the UCR, so there was no other majority that could replace the “grand coalition” that had backed his designation and administration. In the Dominican Republic, the political party arena was more fragmented, as there were three relevant political forces: President Balaguer’s Social Christian Reformist Party (Partido Reformista Social Cristiano, PRSC), the Dominican Revolutionary Party (Partido Revolucionario Dominicano, PRD), and the Dominican Liberation Party (Partido de la Liberación Dominicana, PLD). When Balaguer took office in 1994, his party, the PRSC, had 47 out of the 120 deputies, and 14 out of the 30 senators. In order to obtain a parliamentary majority, Balaguer had made an alliance with the PLD in Congress. The possibility of forming an alternative

Presidentialism and Early Exits

23

parliamentary majority was limited for the following reasons. An alliance between the PRSC and the PRD was not viable, since the PRD had been the main challenger of the PRSC in the contested elections of 1994: its candidate, Peña Gómez, had obtained 41.6 percent of the vote against Balaguer’s 42.3 percent. Neither was it possible to establish a coalition between the PRD and the PLD, given that these were antagonists: the PLD, led by Juan Bosch, was the product of a PRD split (Espinal, 1999). These differences were attenuated once the parties reached an agreement, which shortened Balaguer’s term in office from four to two years. The Bolivian party system was also a fragmented multiparty system. Siles Zuazo took over the presidency in 1982, heading a minority government with the backing of the Democratic Popular Unity (Unión Democrática y Popular, UDP) coalition, which included four parties but had won only 47 out of 130 congressional seats. Of the 47, only 24 deputies belonged to the president’s party, the Left Nationalist Revolutionary Movement (Movimiento Nacionalista Revolucionario de Izquierda, MNRI). In the senate, the UDP held only 10 of the 27 seats. The other main political forces represented in Congress were Banzer’s National Democratic Action (Acción Democrática Nacional, ADN) and Paz Zamora’s Revolutionary Left Movement (Movimiento de Izquierda Revolucionaria, MIR), which could use their veto power but which were not in a position to form alternative parliamentary majorities. To summarize, in a context of crisis, the difficulties of forming alternative majorities pave the way for a presidential salida. When a two-party system is in place, the alternative to the status quo is a government backed by both parties. However, it is very unlikely that the opposition party will be willing to share the costs of the crisis with the government party, its main competitor. In a multiparty system, there are more actors available to forge a possible alternative. But this may not be possible, as the resignation of Balaguer and Siles Zuazo show. In the first instance, where there were three political parties with quite similar electoral trajectories, the decision of a party to either form part of an alternative coalition or to support the government could have an impact on its future prospects, so a change in the status quo was problematic. In the second instance, in my view, one of the additional difficulties posed by a fragmented party system, such as that in Bolivia under Siles Zuazo, is that of coordination: who has the authority to articulate a formula for change? In both countries, the solution finally depended on the intervention of external agents, alien to the political class. In the case of the Dominican Republic, these were the United States and key international organizations; in the case of Bolivia, they included the Catholic Church as well as some international organizations. In any case, regardless of the configuration of the party system, the alternative that favors a presidential exit from crisis seems to be a “new beginning.” In a democracy, this means new elections. This explains why political elites agree to call early elections and thereby inaugurate a new presidential period.

24

Ana María Mustapic

Parliamentary Exits from Crisis Parliamentary exits from crises do not respond to agreed strategies; on the contrary, they suppose the persistence of variable degrees of interbranch disagreement between the president and the Congress. My hypothesis is that this type of early salida occurs when it is possible to form alternative parliamentary majorities in opposition to the president that are willing to abandon the executive. Presidential impeachment is the classic modality of parliamentary intervention, and the first I will address here. This is the typical instrument that presidential regimes grant Congress to get rid of a president. 6 In effect, the Congress always starts up the process; it is only after this, that national variations—such as the majority that is required to begin the process, or the way in which a trial is carried out—come into play (Pérez-Liñán 2001). The presidents of Brazil, Fernando Collor de Melo, Venezuela, Carlos Andrés Pérez, and Paraguay, Raúl Cubas, were impeached.7 The first one headed a coalition government in which his party held a mere 8 percent of the seats in the Chamber of Deputies. The other two had instead taken power with the support of majority forces: Democratic Action (Acción Democrática, AD) in the case Andrés Pérez, and the Colorado Party (Partido Colorado, PC) in the case of Cubas. But the PC was a strongly factious political force, while AD not only distanced itself increasingly from the president, in disagreement with his policies, but it was also divided. One could assume that it was party fragmentation, reflected in Congress, that opened the door to impeachment. But this is not a necessary condition: an impeachment moves forward when sectors within the forces supporting the president decide that it is better to abandon the chief executive than to carry on supporting him or her.8 That said, it is not my intention to focus here on impeachment, since it is not an innovation to presidential institutional practices; they form part of the conventional repertoire. I prefer, instead, to focus on presidential resignations in which no impeachment was involved. The peculiarities of such cases lie in their being alien to the original script of presidentialism. In this, newer, modality of presidential resignation, the intervention of Congress is facilitated because it is not necessary to muster up a broad majority to leave the president without support; all it takes is the formation of an alternative coalition. When one observes the distribution of seats in Congress, it is easy to prove that once a crisis was in motion, the partisan support for the resigning presidents—I refer here to de la Rúa, Sánchez de Lozada, Mesa, Bucaram, Mahuad, Gutiérrez, and Serrano Elías—was the minority.9 For this very reason, there existed a latent possibility of forming alternative parliamentary majorities, although this does not mean that it was always viable to actually constitute them. The loss of parliamentary support can, then, precipitate the removal of a president. Under certain conditions, legislators have at their disposal

Presidentialism and Early Exits

25

constitutional mechanisms to accelerate the process, even if they must broadly reinterpret the rules to achieve their goal. In Ecuador, for instance, Congress used constitutional rules to declare President Bucaram insane, and to accuse both presidents Mahuad and Gutiérrez of abandoning office.10 However, in other circumstances legislators resort to street mobilizations (see chapter 4, this volume), or, as in Guatemala, they count on the active participation of the military forces (see chapter 9, this volume). In any case, the resignation of a president is only the first chapter of the parliamentary exit strategy. The presidential salidas dealt with in the preceding section were solved by the anticipated call for elections, or the early handover of power. By contrast, parliamentary salidas include a second chapter to solve the succession problem. In general, when the president resigns, it is expected that the vice president will take his or her place. Thus, in Bolivia Mesa replaced Sánchez de Lozada, and in Ecuador Noboa Bejarano and Alfredo Palacio replaced Mahuad and Gutiérrez, respectively. But it might happen that there is no vice president, leading to what is known as a situation of acephaly, namely, the simultaneous absence of a president and a vice president. This is what happened when de la Rúa, Rodríguez Sáa, Mesa, Bucaram, and Serrano Elías left power. Let us look at this issue more closely. Taking into account all the countries included in table 1.1, we find there are four possible responses to situations of acephaly: the appointment of a successor by Congress; the calling of elections; the adoption of a mixed system, according to how many years of the mandate have been completed; and following a preestablished line of succession. In Argentina, Ecuador, Guatemala, the Dominican Republic, and Venezuela, it is legally established that Congress should appoint a successor.11 The difference tends to be the majorities that are legally required for a successor to be appointed. Thus, Guatemala’s single chamber must vote with an ample two-third majority, whereas in Bolivia, by contrast, there is a fixed line of succession rule, starting with the president of the chamber of deputies, followed by the president of the Senate, and the president of the Supreme Court. If the presidency is taken up by the latter, as occurred in the crisis of 2005, elections must be called immediately.12 Brazil and Paraguay have mixed systems: in Paraguay, if there is a situation of acephaly during the first three years of the constitutional period, elections are called to elect a successor to complete that period. If the situation arises during the last two years of the presidential mandate, Congress must nominate a successor to complete the period by absolute majority; in Brazil the situation is similar. With the exception of Bolivia, and to some extent Paraguay and Brazil, in the rest of the countries examined, the rules on acephaly delegate in Congress the decision of appointing a successor. In crisis situations, these regulations may well act as an incentive for legislators to play a more active role, thus giving rise to several strategies. Legislators may, for instance, hasten the crisis, as occurred with de la Rúa in Argentina (Mustapic, 2005); they may

26

Ana María Mustapic

try to force the vice president to resign as well (this is what happened to Serrano Elías’ vice president, Gustavo Espina Salguero, and to Bucaram’s vice president, Rosalía Arteaga); or they may force the vice president to distance himself or herself from the president and to work more closely with Congress. Something of the sort occurred with Mesa, Sánchez de Lozada’s vice president. In conclusion, it is possible to argue the following when comparing presidential and parliamentary salidas. In the context of a presidential crisis, Congress is always relevant, be it by its action or by omission. First, this is so for the simple reason that it cannot be avoided. The fact that Congress cannot be dissolved will affect the calculations of the players involved in the crisis context. They may cooperate or opt for confrontation; what they cannot do is close Congress unless they decide they want to go down the path of democratic breakdown. Whatever the case, it is only through a constitutional convention that a new Congress can be brought to life, thanks to its sovereign power.13 Second, the political composition of Congress is relevant when predicting which type of salida is likely from the crisis. When that composition offers no scope to challenge the status quo, president and Congress are likely to converge; when there is room to constitute alternative alliances, it is more likely that the president will become isolated and be forced to resign.

Presidential Resignations and the Presidential System The other issue I wish to address here is the impact that early presidential exits have on how a presidential regime works. Given that presidents can leave office before they end their mandates, I must explain my view that the impossibility of dissolving Congress is relevant for an understanding of the “systemic logic” of presidentialism, to paraphrase Sartori (1994). With this aim in mind, in what follows I review some definitions and characterizations of presidentialism. Sartori has put forward one of the most recent definitions of presidentialism. He writes that “a political system is presidential if, and only if, the Chief of State (the president) is (a) popularly elected; (b) cannot be removed from office by a vote by Parliament or Congress during his preestablished mandate; and (c) in some way leads or directs the government he designates” (Sartori 1994, 99). Another author who cannot be ignored when it comes to the topic of presidentialism is Juan J. Linz. In a seminal text on the topic, Linz states the following: “The second main institutional characteristic of presidential systems is the fact that presidents are elected for a period of time that, under normal circumstances cannot be modified . . . The political process therefore becomes broken into discontinuous, rigidly determined periods without the possibility of continuous readjustments as political, social, and economic

Presidentialism and Early Exits

27

events might require” (Linz, 1994, 8). The presidency is at the heart of the reasoning of both these authors, and neither includes as a necessary feature of presidential regimes the impossibility of dissolving Congress. In the definition proposed by Shugart and Carey, which is more complete than the two above, this issue makes an indirect appearance when the authors introduce the second of their four criteria defining a presidential regime: “the terms of the chief executive and assembly are fixed, and are not contingent on mutual confidence” (1992, 19).14 Certainly, neither of the authors ignores that dissolution is not a proper tool of presidential regimes. Given that I wish to capture the “systemic logic” of presidentialism, the question that has to be addressed is the extent to which this omission is an obstacle to properly understanding that system. In order to respond to this question, it is worth taking a closer look at Linz’s argument. Linz’s emphasis on the fixed presidential mandate can be explained by his interest in highlighting the problems inherent in the dynamic of presidentialism and the risks it poses to democratic stability. He develops his argument in two steps. First, he calls our attention to the fact that there is a potential for conflict inherent in the popular and independent origin of the president and Congress. As both powers are equally legitimate because of their shared democratic origins, if there is no unity of purpose between them it is very likely that there will arise a situation of institutional paralysis, if not confrontation. Linz then adds that there is a natural tendency toward inertia in presidentialism, given the rigidity introduced by the fixed mandate. The president cannot be removed easily as there are no institutional mechanisms to resolve crisis situations when the president and Congress are in conflict. The absence of instruments available to presidential systems contrasts with the flexibility of parliamentary regimes. In the latter system, the motion of censure and the no-confidence motion make it possible to replace a prime minister and/or dissolve parliament, thus creating the conditions to deal with governance challenges. By contrast, every time that there is an important conflict within a presidential system, its resolution depends largely on the political skill of the president. If, for whatever reason, such skills are lacking, the only alternative may be to remove him or her from office. In this case, there are only two available mechanisms: resignation or a political trial. According to Linz, both courses are difficult to navigate. First, the fixed presidential mandate can serve as an institutional “trench” from which to resist any pressure to resign. Second, it is unlikely that legislators who are loyal to the president will support a measure as extreme as a political trial. Given the limited instruments available in presidential systems, the paths to resolve a crisis peacefully are closed off. For these reasons, Linz concludes, presidential crises tend to push the system to the edge of democratic breakdown. In his view, the history of the failure of Latin American presidentialism eloquently testifies to this fact. Linz wrote this in 1988 at the beginning of the third wave of democratization when the resignations examined in this chapter had yet to take place.

28

Ana María Mustapic

These new facts make it impossible to ignore presidential resignations as a novel alternative. This is all the more so when the possibility of military coups, which conditioned Linz’s reasoning, seems ever less likely. It is probable that a president threatened by a military coup will resist resignation, if for no other reason than to salvage his or her dignity. However, it is plausible to suppose that a president who is not threatened by a coup but rather by ungovernability—whether it is because he or she has no backing or because he or she has lost his or her authority—may prefer to resign. In my view, presidential resignation is an alternative that should be included in the menu of presidential regime dynamics, and it goes without saying that it cannot be included if one presupposes the dissolution of Congress. Thus, the presidential fixed mandate can become more flexible through resignation; but, in contrast, the absence of instruments to render the mandate of Congress more flexible is a feature inherent to presidentialism. For this reason, I consider that, in addition to the independent election to office, the impossibility of dissolving Congress is what most clearly distinguishes presidentialism from other systems. The Guatemalan crisis of 1993 illustrates this point very well. The conflict between President Serrano and legislators resulted in the presidential decision to close Congress, a measure that ruptured the regime. Predictably, this measure, which is the functional equivalent of dissolving parliament, deepened the crisis and ultimately led to the resignation of the president. Even after Serrano’s resignation and the appointment of a successor, the problem with Congress persisted, because of its high level of corruption and continued extortion of the executive. One of the alternatives proposed under the new presidency of Ramiro de León Carpio was to force the resignation of a particularly troublesome group of legislators known as the “purgables.” Despite the strong pressure exerted on them, this strategy did not work. Finally, after a popular consultation and constitutional reforms, new legislative elections were called. One of the issues that Guatemalan analysts have highlighted is precisely that there were no institutional means to remove these legislators from the playing field. It was necessary to resort to extraordinary measures—in this case a constitutional amendment—to bring about that outcome. As this shows, Congress becomes a central feature of the problems of presidentialism. This affirmation is supported by other evidence. The constitutions of Paraguay and Chile during the Stroessner and Pinochet dictatorships, for example, provided for the dissolution of parliament by the executive.15 This clearly demonstrates that authoritarian rulers find it expedient to have a mechanism to discipline Congress and avoid unexpected challenges to presidential power. Another means of limiting the power of Congress is apparent in the constitutional reforms undertaken in various Latin American countries to increase the legislative powers of presidents to the detriment of Congress. The constitutionally established “necessity and urgency” decree powers in Argentina, and “provisional measures” in Brazil, are a good example of this (Gargarella 1997).

Presidentialism and Early Exits

29

Thus, it turns out that Congress is key to government stability, independently of how it fulfils its duties of representation, legislation, and oversight. The capacity to give or withdraw support from presidents puts Congress in a position analogous to that of parliamentary legislatures with their capacity to censure executives. Some analysts consider that the resolution of presidential crises illustrates how Congress has been strengthened in Latin America (Carey 2002; Pérez-Liñán 2003). From the point of view put forward here, the resolution of these crises shows that Congress plays a hitherto ignored role. A president cannot get rid of Congress unless he is willing to resort to a coup; but Congress can get rid of a president just by ceasing to support him or her.

Conclusions Everything seems to indicate that the authoritarian alternative to solve crisis situations is increasingly unavailable on the immediate horizon of Latin American countries. This situation, by limiting certain strategic choices, particularly the recourse to military coups, has generated new opportunities for the treatment of presidential crises. It did so by directing the search for solutions to the utilization of the toolkit offered by the existing institutional rules. In other words, when democracy is the only game in town, unforeseen alternatives within the game appear on the table. One of them is presidential resignation. The analysis of presidential resignations in Latin America has allowed us to distinguish between two types of responses to crisis situations: one presidential, the other parliamentary. The difference between them lies in whether there is scope to constitute alternative congressional majorities. Where the conditions to form them are absent, the exit strategy is presidential and involves, other than a more passive role for Congress, calling early elections or early resignation. By contrast, when it is possible to form alternative congressional majorities, the exit strategy tends to be parliamentary. In such cases, even if forced—as when the vice president comes under pressure to resign as well—the line of succession has been followed. The analysis also highlights the peculiar role that Congress plays, in particular its contribution during the crisis: Congress is not just a part of the problem but also a part of the solution. Indeed, Congress can precipitate a crisis when it withdraws its support for the president, but it can, by the same token, also contribute to the solution. In such contexts, legislatures in presidential regimes play a similar role to that traditionally reserved for the legislative branch in parliamentary systems, namely, the formation of government. In any case, the analogy stops here, since it does not make reference to a routine function but rather to an exceptional one. Put differently, it is only under certain special circumstances that congresses in presidential systems intervene to remove presidents and form governments.16

30

Ana María Mustapic

Second, my analysis suggests that it is appropriate to include in the definition of presidentialism the impossibility of dissolving Congress. Until quite recently, conventional wisdom had it that the resignation of a prime minister played a central role in the architecture of parliamentary regimes, because it decisively shaped their political dynamics. By contrast, it was understood that the resignation of presidents was an exception and not part of the routine attributes of presidential regimes. Moreover, the view was that its absence was a structuring feature of executive-legislative relations. In the light of what happened in Latin America since the 1980s, this view needs to change. We must include the possibility of presidential resignations in our characterization of the political dynamics of presidentialism; it is an alternative that actors take into account when making their calculations and working out their strategies. In my view, the fact that it is possible to force a president to resign can contribute to promoting cooperative practises between executives and legislatures, in the same way that motions of censure do in parliamentary regimes. If the goal is to remain in office, the president should be the party most interested in promoting such practises. In 1949, Karl Loewenstein wrote that the president “cannot control or impose his will on the legislature except by persuasion and through the intermediation of the party machine he controls,” adding that “Most important, he does not have power to dissolve congress” (1949, 451). This goes to the crux of my argument. Thus, in contrast with the argument put forward by Arturo Valenzuela (2004), for whom presidential resignations are a sign of the instability of presidential regimes, I underline instead their potential contribution to boosting stability. Presidential resignations should not be seen, as Valenzuela does (2004), as an “unworthy” situation but, as under parliamentary regimes, as evidence of the loss of parliamentary support for the president. The presence of this innovation in the practise of presidentialism in Latin America may generate a favorable framework for a process of political learning. Clearly, as this learning process evolves, two things can happen: either government instability increases, or the risks of conflict diminish. In the first instance, the learning process will lead to the strengthening of congressional factionalism; in the second, mechanisms for closer cooperation between executives and legislatures may emerge, particularly through the broadening of coalition governments. The latter possibility bolsters the arguments of political scientists who, from another point of departure, argue that presidentialism does not necessarily lead to government stalemate or paralysis and, consequently, to democratic rupture (Cheibub 2002; Negretto 2006).

Notes * I would like to thank Constanza Figueroa Schibber for her Research assistance. A prior version of this chapter was presented at the workshop on “Parliamentary

Presidentialism and Early Exits

31

Practices in Presidential Systems: Perspectives on Parliamentary Power in Latin America,” European Consortium for Political Research Joint Sessions of Workshops, Nicosia, Cyprus, on April 25–30, 2006, and was published as Mustapic (2006). 1. In accordance with the constitution after the 1994 reform, in Bolivia Congress must choose from among the two candidates who have received the most votes (before the reform it was the three most voted candidates). In Argentina, the election of the president used to be indirect, by an Electoral College. If the latter failed to elect a president with an absolute majority of votes, Congress had to choose from among the candidates receiving the most votes. As it turned out, in contrast with Bolivia, it was never necessary to apply this rule while it was in force. 2. I borrow this distinction from Immergut (1992). 3. The authors define democratic and semi-democratic regimes thus: Regimes in which the government is elected in free and fair elections, in which civil liberties are protected, in which the electorate includes the majority of the adult population, and in which popularly elected authorities are not interfered with by the military or other nonelected power, are democratic. If one or more of these attributes is partially compromised, as when there is electoral fraud, a regime is considered semi-democratic (Mainwaring and Pérez-Liñán 2005, 15). 4. The government can fix the date of elections. In this case, it resorted to the longest legally permissible deadline to call elections before the end of mandates. Until that date, the minimum deadline (60 days) had been used. 5. Vice President Carlos Álvarez had renounced in October 2000, so that when President de la Rúa resigned, the so-called acephaly law (ley de acefalía) was applied. According to this law, congress must appoint a successor. Congress first nominated Adolfo Rodríguez Sáa, who only remained in power for a week, and then it appointed Eduardo Duhalde. 6. I take impeachment to be a type of parliamentary exit because my aim is to underline the role that congresses and presidents play in the resolution of crises. This is not incompatible with Marsteintredet and Berntzen (2008), who consider impeachment to be a type of presidential breakdown. However, the frame of reference for them is the contrast between presidential and parliamentary regimes. From that perspective, just as a political trial is a presidential form of breakdown, the motion of censure is the parliamentary mode of breakdown. By contrast, my analysis focuses only on presidential regimes in crisis situations. When Congress plays a dominant role in resolving the crisis, I call this a parliamentary exit; and when the executive plays a key role, the salida is presidential. 7. Fernando Collor de Melo and Rafael Cubas resigned before the conclusion of their impeachment; by contrast, Carlos Andrés Pérez was removed from office as a result of an impeachment. 8. The impeachment of U.S. president Nixon, for example, was only possible because the Republicans also supported it. 9. I have not included Rodríguez Sáa in this list, as he was only in office for a week. He had been nominated president by Congress with a majority vote by his party, the Justicialista Party, but he quickly lost that support. 10. On this point, see Marsteintredet and Berntzen (2008). 11. The Ecuadorian case is curious: when Bucaram was deposed, there was no constitutional provision that could be used to resolve the problem of acephaly. An ad hoc solution was found, with the nomination of the president of Congress

32

12. 13. 14.

15. 16.

Ana María Mustapic to head the presidency; this decision was then legitimated with a plebiscite (Mejía Acosta 2004). The legislation that is now in place was inspired by the U.S. model, and seeks to avoid acephaly by ensuring that there is always a vice president in place: if there is no vice president, congress must immediately elect one by a majority from a list presented by the president. Indeed, before the 2005 reform elections were only called if there were three years of the mandate left to complete. An example of this is the Venezuelan constitutional reform of 1999, on the basis of which general elections for all elected posts were called in July 2000. Shugart and Carey (1992) examine the issue of the dissolution of Congress, but they view it as a significant rupture of the principle of separation of powers rather than an issue connected with the fixed presidential term. This clause was eliminated with the return to democracy and after the reform of these constitutions. It should be noted that, depending on the case in question, the greater influence of Congress is also linked with the role played by the vice presidency. Here we must keep two circumstances in mind: on the one hand, in contexts of crisis, if there is no vice president, the president is even more exposed and has even fewer resources to deal with congressional fluctuations. Further, if the president resigns, a situation of acephaly arises that is resolved in some Latin American cases by granting Congress the capacity to elect a successor. On the other hand, the loyalty of a vice president is also relevant for government stability (see the introduction to this volume). When the vice president has his or her own ambitions, they almost inevitably use Congress to boost their own position. The analysis of resignations, then, sheds a new light on the vice presidency—a post that is usually treated as secondary in the literature.

2 Presidential and Democratic Breakdowns in Latin America: Similar Causes, Different Outcomes* Michael E. Álvarez and Leiv Marsteintredet

Introduction Are presidential breakdowns in Latin America simply contemporary expressions of democratic breakdown or are they a new phenomenon altogether? Despite the obvious differences between presidential and democratic breakdown, there are similarities, and much of the literature on presidential breakdown has found inspiration in studies of democratic breakdowns. Both phenomena end governments, but only one spells the end of democracy; both are shattering events for the regimes in question and are the result of deep political conflicts, but the military tends to solve one, and civilians the other. In this chapter, we examine whether the causes of democratic breakdown in Latin America resemble the causes of presidential breakdown in the period 1950–2005. The goal is not to rediscover the causes of either phenomenon but rather to compare the underlying economic, institutional, social, and regime factors that influence the likelihood of presidential and democratic breakdowns, and determine the degree to which these causes are similar in type and magnitude. We undertake a statistical analysis of socioeconomic, regime and political institutional data for the postwar period to answer our questions. Why compare these phenomena? Several authors have based their analyses of presidential breakdowns on models of democratic breakdown, placing particular emphasis upon political institutional variables as underlying causes (e.g., Negretto 2006; Valenzuela 2004). Further, coup attempts today tend to be nonstarters: with the new post-cold war zeitgeist and following

34

Michael E. Álvarez and Leiv Marsteintredet

the fall of communism, U.S. policy for Latin America shifted in the 1990s, and the Organization of American States (OAS) adopted a stronger commitment to democracy. Democratic breakdown, as one of the prevalent “standard” outcomes of political conflict in Latin America during the Cold War, has become less viable. Democratic conflicts are now usually resolved under the constraints imposed by democratic rules. However, political conflict has not ceased, so it has been hypothesized that the underlying causal factors that formerly resulted in democratic breakdown are today more likely to lead to presidential breakdown. Thus, the democratic regime is preserved, but the president is removed. Democracy has managed to survive various instances of presidential breakdowns, which may indicate either that democracy has grown stronger or that the crises leading to presidential breakdown are milder than those resulting in democratic breakdown. Our exploration proceeds as follows. First, we discuss the existing scholarly literature and identify the causes of presidential and democratic breakdowns. We then make some caveats regarding the causal comparison of these two distinct phenomena. Finally, we outline and discuss our results.

Presidential Breakdowns and Democratic Breakdowns A presidential breakdown constitutes an extraordinary and premature change of president without the breakdown of the democratic regime; by contrast, a democratic breakdown marks the end of the democratic constitutional order and the beginning of nondemocratic rule.1 A “premature” change is one in which the president leaves office before the expiration of a fixed term. We consider the change “extraordinary” to exclude regular changes established by an electoral calendar, such as permanent changes to presidential terms through constitutional amendment. Both presidential and democratic breakdowns occur under regimes that are at least minimally democratic.2 The occurrence of presidential breakdowns may be more or less constitutional and democratic, but democracy as a regime type survives, so the outcome of a presidential breakdown is qualitatively different from that of a democratic breakdown. There are also differences between the types of actors who are normally involved in the two phenomena. Until the “third wave” of democratizations, the military acted as the moderating power (poder moderador) and intervened in politics to interrupt or end democracy on the grounds that it had a duty to restore social order, wage war against a real or incipient communist threat, and protect the constitution. Today, by contrast, congresses, civil societies, and street protesters are the most active and important actors removing presidents from power, 3 and it has been suggested that civil society has replaced the military as a poder moderador (Hochstetler 2006, 403). Despite the differences in outcome, process, and

Breakdowns in Latin America

35

in terms of the actors involved, presidential breakdowns may nonetheless be caused by underlying factors similar to those that influence the likelihood of a democratic breakdown. Studies analyzing the breakdown of democracy tend to focus on economic variables when assessing the effectiveness of a democratic regime and how its performance affects the potential for breakdown (Mainwaring and PérezLiñán 2005; Przeworski et al. 2000). Przeworski et al. (2000, 117) have concluded that rich democracies do not die, and that economic crises in poor countries increase the likelihood of democratic breakdown.4 Mainwaring and Pérez Liñán’s (2005) tour de force analysis of democratic breakdowns, transitions, and erosions concludes that in addition to the level of economic development, several economic performance variables such as economic growth and inflation rates are significant predictors of breakdown. Furthermore, and relevant for our purposes, the predictions generated by their statistical model indicate that the likelihood of democratic breakdown after 1978 has not decreased or changed significantly relative to the past (Mainwaring and Pérez-Liñán 2005, 36). The sharp global decrease in democratic breakdowns after 1978 is explained not by the underlying forces specified in their statistical model but rather by the new regional context (changes in U.S. foreign policy, the new orientation of the OAS, 5 the position of the Catholic Church, and attitudes toward democracy). This suggests that presidential breakdown may be caused by the same factors that explain democratic breakdown, and that the new regional context explains why similar causal forces have produced different outcomes in the post-1978 period. It makes sense, then, to build on Mainwaring and Pérez Liñán’s work and compare the causes of presidential and democratic breakdowns. Indeed, the presidential breakdowns experienced by several Latin American countries are the closest that several of these regimes have come to a full democratic breakdown.6 Since Linz’s (1978) seminal work on democratic breakdown, democratic failure has been linked to concepts such as efficacy, effectiveness, and the legitimacy of the political system. Efficacy and effectiveness refer to a regime’s performance—its ability to identify and implement successful public policy solutions to solve basic social problems. Democratic legitimacy refers to the perception among the citizenry that democracy is the least evil form of government. This perception is affected by the efficacy and effectiveness of the system, and may have an indirect effect upon these variables (Linz 1978, 18–24). The idea is that political actors gradually lose legitimacy as they fail to find public policy solutions for problems such as inflation, unemployment, or violence, leading different sectors of the democratic opposition to adopt semi-loyal or disloyal behavior that eventually topples the regime and ends democracy. O’Donnell (1988, 22–24, ch. 9) shows how economic fluctuations and crisis preceded the 1976 coup in Argentina, arguing that capitalism had ceased to function normally in the Southern Cone before the advent of bureaucratic-authoritarianism in the region. Similarly, we could argue that regime performance matters, not simply for the survival of democracy but

36

Michael E. Álvarez and Leiv Marsteintredet

also for the survival of presidents, since various presidential breakdowns have occurred in a context of poor economic performance. The fall of Ecuadorian president Jamil Mahuad in January 2000, for instance, may be linked to the country’s abysmal economic performance, its international debt default in the fall of 1999, the inability to defend the national currency, and, just prior to his fall, the president’s very unpopular decision to dollarize the economy. Likewise, the Argentinean presidential breakdowns in 1989 and 2001–2002 occurred during harsh economic times. A second factor that is often linked to regime performance and also affects the survival of presidents is social and popular mobilization. Military interventions in civilian democratic rule are often preceded by social mobilization and street protests; similar actions have preceded presidential breakdowns as well. In Brazil, students had protested in support of President Goulart and his reforms for some time before the military coup of March 1964 (Wedge 1969). O’Donnell (1988, 295) observes that there was a clear increase in street demonstrations in Argentina during the early 1970s. In Chile, mobilization and left-right polarization increased tremendously prior to the September 1973 coup. Similarly, other presidential breakdowns have been linked very clearly with street mobilization and protest against the presidents (Hochstetler 2006; Pérez-Liñán 2008; Valenzuela 2004). With the exception of the fall of President Balaguer in 1994–1996 (chapter 10, this volume), popular mobilization (triggered by a variety of factors) has been a shared feature of presidential breakdowns throughout Latin America over the past twenty-five years. A third factor affecting presidential survival is institutional design. The dominant debate among students of Latin American democracy and political regimes over the past twenty years has been about the workings and performance of political institutions under presidential regimes and the question of what kind of institutional regime promotes greater democratic stability (Jones 1995; Linz 1994; Mainwaring 1990; 1993; Mainwaring and Scully 1995; Shugart and Carey 1992; Shugart and Mainwaring 1997; Cheibub 2002; 2007; Cheibub and Limongi 2002; Cheibub, Przeworski, and Saiegh 2004; Przeworski et al. 2000). It has long been argued that presidentialism, particularly when combined with proportional electoral systems that tend to generate minority governments and/or undisciplined parties, promotes intractable congressional-executive conflicts because of fixed terms of office and the competing sources of legitimacy of presidents and legislators. Political effectiveness is hampered, and in the absence of institutional mechanisms to resolve executive-legislative conflict and deadlock, democracy is endangered. Valenzuela (1994) blamed presidentialism and the party system for the fall of democracy in Chile, arguing that, together, they exacerbated the crisis of democracy by creating repeated interinstitutional conflicts and weak minority governments.7 In Peru, Fujimori’s autogolpe in 1992 was also preceded by intense conflict between the Congress and president (see Weyland 2006, 22; Cameron 1997).

Breakdowns in Latin America

37

It has also been argued that regime type and political institutions matter for understanding not only democratic breakdowns but also presidential breakdowns. Valenzuela (2004) holds that the difficult combination of presidentialism and minority governments can explain the fall of presidents just as well as it can explain democratic breakdown (Linz 1978; 1990; 1994; Valenzuela 1993). Negretto (2006) demonstrates that minority governments, particularly those that do not control the median voter in Congress, are vulnerable to collapse. Pérez-Liñán (2007) notes that, in Latin America, the presence or absence of a “legislative shield” can explain the success or failure of impeachment proceedings against a president. Clearly, institutional factors are an important factor in presidential breakdowns, if only because impeachment procedures require a qualified majority (the removal of Brazilian president Collor de Melo in 1992 is a case in point). In sum, institutions are central explanatory factors for presidential breakdowns, just as they are for democratic breakdowns. Our survey of the existing literature indicates that a variety of economic, social, and political factors cause presidential and democratic breakdowns. Students of presidential breakdowns have sought inspiration from theories of democratic breakdown and have adapted them to explain the more recent phenomenon of presidential breakdown. This preliminary ad-hoc comparison indicates that there are easily identifiable parallels between the two phenomena. In what follows, we undertake this comparison in a more systematic manner.

How Can We Compare Two Distinct Phenomena? Following Goertz’s (2006, 50–53, ch. 9) view of three-level concepts and two-level theories, we hypothesize that certain basic-level causal factors have similar effects upon both phenomena. Our purpose is to determine whether presidential and democratic breakdowns are actually triggered by the following four basic-level conditions: economic performance, regime characteristics, civil society mobilization, and political institutions. To concretely measure the impact of these basic-level forces, we specify in our statistical models various combinations of secondary-level empirical indicators (two indicators for each factor). Goertz (2006) maintains that the indicators at the secondary level are substitutable, whereas the causal factors at the basic level, from which our hypotheses emerge, are often not substitutable. Thus, for example, poor economic performance (a basic-level causal factor) may increase the risk of presidential or democratic breakdown via secondary-level indicators such as negative growth, increases in unemployment, deprivation of real salaries, and hyperinflation, among other factors. We use statistical tools to analyze the effects of the secondary-level indicator upon the two different phenomena—presidential breakdown and

38

Michael E. Álvarez and Leiv Marsteintredet

democratic breakdown—by comparing the estimated regression coefficients and t-statistics generated by the regression analyses. We can thereby infer whether the basic-level causal factors have similar consequences for the likelihood of presidential and democratic breakdowns. Theoretical reasoning, data availability, and a variety of specification tests led us to measure the four basic-level causal factors as follows. We measure economic performance according to rates of economic growth and inflation, which are two standard indicators of economic performance. These indicators are particularly appropriate, since the longest time series are available for these indicators and we are looking at events that extend as far back as 1950. Regime characteristics are measured by two indicators: first, by the classification of the regime in the year prior to breakdown as either a “full” or a “semi” democracy; second, by a count variable that shows the number of past breakdowns of democracy.8 We measure civil society mobilization by the number of antigovernment demonstrations and the level of strike activity, both measures drawn from the Arthur Banks Cross-National Time-Series Archive. There are more reliable data on civil society mobilization available for more recent years, but Banks’ data extend across a longer time series. Further, that dataset distinguishes between general strikes and antigovernment demonstrations. We hypothesize that the former had a more negative effect upon government survival before the “third wave” and the end of the Cold War, and the latter may prove to be relevant when explaining more recent presidential breakdowns.9 Finally, the consequences of institutional design are measured by variables such as the number of political parties (Laakso and Taagepera 1979) and the share of legislative seats held by the party of the chief executive, which is also a proxy for polarization (Mainwaring 1993). Again, there are more reliable indicators for executivelegislative relations, such as direct measures of deadlock (Jones 1995) and the size of the presidential coalition rather than the presidential party, but we were limited by the availability of data. Our methodology is to test statistically the effects of these indicators on the likelihood of presidential and democratic breakdowns and then to compare the estimated regression coefficients in terms of their signs, statistical significance, and marginal effects. Does the same causal model explain both phenomena? Do the estimated regression coefficients for matching independent variables have the same signs and similar magnitudes of effect upon the two quite different outcomes? If the answers to these questions are affirmative, then the hypothesis that presidential and democratic breakdowns are caused by the same underlying social, economic, and political factors is strengthened. Given that there have been fewer democratic breakdowns since the late 1970s, we may also therefore infer that other exogenous forces may explain why—given similar underlying causal forces—coups have become nonstarters and presidential breakdowns have become more common instead. The event that we are interested in is whether or not a president survives and whether—if a president is removed from power—democracy survives or

Breakdowns in Latin America

39

not. Our dependent variable is therefore a limited dependent variable with three possible outcomes coded as follows: 0 = the president survives (base category); 1 = a presidential breakdown occurs; 2 = a democratic breakdown occurs. There are several variants of multinomial probit and logit models available to model statistically the factors that influence the likelihood of each of these outcomes. A key consideration concerns the nature of the relationship between the three potential outcomes. Is there an increased “degree” of instability when one compares presidential breakdowns to democratic breakdowns? If so, ordered logit estimation is the preferred technique. We conducted a wide variety of estimations utilizing rare events logit and ordered logit.10 Ultimately, we opted for standard multinomial logit with corrections for heteroskedasticity (which is marginal in any case) and with robust standard errors.11 The reason for excluding ordered logit merits a short discussion. It is theoretically conceivable to order the outcomes of presidential and democratic breakdowns, under the assumption that the former is caused by a lower level of economic crisis, interinstitutional conflict, or civil society mobilization than the latter. For this reason, we tested whether our outcomes could actually be ordered. It was encouraging to find that the results generated by ordered logit and standard multinomial logit concerning the general patterns of effects of the independent variables were quite similar. Unfortunately, the ordered logit models tended to estimate very small magnitudes and levels of significance of the effects of the independent variables upon outcome two of the dependent variable regardless of whether outcome two was coded as presidential breakdown or democratic breakdown. Clearly, then, the data themselves do not support an assumption that instability or the specific form of presidential removal is of a “higher degree” for democratic breakdowns than for presidential breakdowns. This is an important finding in and of itself: presidential breakdowns are not “light” versions of democratic breakdowns. The panel data set that we use covers democratic rule in nineteen countries spanning the years 1946–2006, generating 692 country-year observations, and for which we count sixteen presidential breakdown and twenty-five democratic breakdowns.12 The number of observations actually utilized in the estimations, however, is reduced somewhat by limitations on data availability for several of the variables. In effect, this means that the large majority of results presented below are based upon a set of 616 observations that include sixteen presidential breakdowns and nineteen democratic breakdowns.13 The data on economic variables in our statistical analyses are drawn from the Penn World Tables (Heston, Summers, and Aten 2006) and the World Development Indicators (WDI) from the World Bank Group (WDI 2008); institutional and regime data come from Nohlen (2005), Przeworski et al. (2000), Mainwaring, Brinks, and Pérez-Liñán (2001), and Cheibub (2007);14 and protest and strike data, from Arthur Banks’ Cross-National Time Series Data Archives (Banks 2008).

40

Michael E. Álvarez and Leiv Marsteintredet

Results and Analysis: Factors That Influence the Likelihood of Presidential Survival Table 2.1 below presents our primary set of findings and serves as the point of reference for the detailed discussion of the respective effects of each independent variable that follows. Following a discussion of the control variables, we present the findings on societal, regime, political, institutional, and economic determinants of presidential and democratic breakdowns. Table 2.1 Determinants of presidential and democratic breakdowns15 Characteristics in numerator of Prob Y = 1 Presidential Breakdown Variable

Coefficient

t-statistics

–5.5517 –.0003 .9429 .2409 –.4075 .4095 .4463 .1381 .6366

–4.549 –1.953 1.673 1.063 –.516 2.714 4.041 .444 .769

Constant Level Prolonged Recession Past Breakdowns Semi-Democracy Number of Parties Demonstrations Strikes (lagged) Post-1978

Characteristics in numerator of Prob Y = 2 Democratic Breakdown Variable Constant Level Prolonged Recession Past Breakdowns Semi-Democracy Number of Parties Demonstrations Strikes (lagged) Post-1978

Coefficient

t-statistics

–4.1408 –.0004 –.4185 1.0297 1.7944 .1971 .2429 .5017 –3.5672

–4.116 –1.345 –.515 1.909 2.747 1.057 1.469 1.609 –2.773

Pseudo R-squared

.26

Marginal Effects Averaged Over Individuals Variable Constant Level ProlongedRecession Past Breakdowns Semi-Democracy

Y=0

Y=1

Y=2

.2244 .0000 –.0109 –.0304 –.0345

–.1248 .0000 .0220 .0048 –.0106

–.0996 .0000 –.0111 .0255 .0451 Continued

Breakdowns in Latin America Table 2.1

41

Continued

Marginal Effects Averaged Over Individuals Variable Number of Parties Demonstrations Strikes (lagged) Post-1978

Y=0 –.0139 –.0159 –.0153 .0724

Y=1 .0093 .0101 .0028 .0171

Y=2 .0046 .0058 .0124 –.0895

Definitions of Variables. Level: real GDP/capita; Prolonged Recession: dummy coded 1 for negative growth experienced in both years t-1 and t-2; Past Breakdowns: count variable for number of past breakdowns of democracy; SemiDemocracy: dummy coded 1 if regime was semi-democratic in the previous year; Number of Parties: Laakso-Taagepera index of effective number of parties; Demonstrations: number of antigovernment demonstrations, as defined by Banks (2008); Strikes (lagged one year): number of general strikes in the previous year, as defined by Banks (2008); Post-1978: dummy coded one for country-years after 1978. Sources: Level and Prolonged Recession from Penn World Tables 6.2 (Heston, Summers, and Aten 2006); Past Breakdowns (Cheibub 2007); Semi-Democracy (Mainwaring, Brinks, and Perez-Liñán 2001); Number of Parties: Cheibub (2007), Nohlen (2005), electoral reports Electoral Studies (various issues) and official electoral authorities Web sites for the most recent elections (the same sources were used for the share of seats of the president’s party in Congress/lower chamber).

Control Variables: Level of Economic Development and Period Effects The very strong and positive correlation between the level of per capita gross domestic product (GDP) and the percentage of regimes that are democratic compels us to control for this variable. In addition, the hypothesis popularized by modernization theory (e.g., Lipset 1959), that democracies are more likely to survive as they reach higher levels of economic development, also bolsters the need for specifying level of economic development as an explanatory variable in models that attempt to explain the breakdown of presidencies and regimes. As expected, there is a statistically significant positive relationship between the level of economic development and the likelihood that a president will survive in office—fewer presidents are removed in richer countries. While this positive relationship holds for both varieties of presidential removal, the significance of the estimated coefficient did not always reach the 90 percent level for democratic breakdowns, but it consistently reached the 99 percent level of significance for presidential breakdowns. However, the magnitudes of the estimated regression coefficients and elasticities of probabilities averaged across individuals were very similar for the two different outcomes.16 Thus, we conclude that both presidential and democratic breakdowns are less likely in richer countries, but that the relationship between income and presidential breakdowns is the more significant of the two.

42

Michael E. Álvarez and Leiv Marsteintredet

In all the models estimated for this study, we also specify a variable intended to mark the historical disjuncture established by the third wave of democratization. By controlling for this, we minimize the artificial inflation of the significance of other specified variables, which could result from the omission of time controls.17 The results presented in table 2.1 above and in table 2.2 below indicate the much lower likelihood of a democratic breakdown in the post-1978 period, and the absence of period effects on the likelihood of a presidential breakdown.18

The Mobilization of Civil Society: Antigovernment Demonstrations and General Strikes As discussed above, studies of presidential and democratic breakdowns highlight the role played by civil society mobilization. Higher levels of mobilization are linked to an increased likelihood of presidential or democratic breakdowns. We use two indicators to test this relationship, which also enables us to determine whether different types of mobilization produce different outcomes. The variable measuring the number of antigovernment demonstrations captures a phenomenon that is widely considered as relevant in initiating presidential breakdowns. The variable measuring the number of general strikes reflects labor union behavior and ideological polarization between the left and right. These are forms of mobilization that one would expect to have a stronger and more polarizing effect during the Cold War era, ushering in democratic breakdowns. Our findings are strongly consistent with Hochstetler’s (2006) observations, as the demonstrations variable proves to have a highly significant (with a t-statistic greater than 4.0) positive effect on the likelihood of a presidential breakdown. This is not surprising considering the protests observed before almost every presidential breakdown since 1985. By contrast, the effect of demonstrations upon the likelihood of democratic breakdown is less robust, failing to reach the 90 percent level of significance.19 Antigovernment demonstrations are not instrumental for democratic breakdown, a finding consistent with what we know about military interventions and the breakdown of democracy. The question remains whether the effects of antigovernment demonstrations operate with a lag. One could hypothesize that increased demonstrations could, over time, provoke military action to topple a government and democracy with it, or that a presidential breakdown might not occur immediately after antigovernment demonstrations. But there is no evidence to support this, as the t-statistics for demonstrations lagged one year were merely –.69 for presidential breakdown and .13 for democratic breakdown. Thus, our findings indicate that the effect of demonstrations operates in the short term and is meaningful for presidential breakdowns. The short-term effect of antigovernment demonstrations on presidential breakdowns stands in sharp contrast to the finding that general strike activity in the current year does not influence the likelihood of either presidential or democratic breakdowns (with t-statistics of .70 and 1.36, respectively).

Breakdowns in Latin America

43

However, there is evidence that strike activity has a lagged effect in influencing the likelihood of a democratic breakdown (t-statistics around of 1.6), but not a presidential breakdown. Military interventions to overthrow a sitting president in Latin America are often at least partly motivated by efforts to stop the spread of “communism.” Militaries have often viewed labor parties, union militancy, and left-oriented efforts to advance social justice as signs of advancing communism. So it is not surprising that general strike activity affects the possibility for democratic breakdown more strongly than broader antigovernment demonstrations. But this does not explain the finding that strikes operate with a lagged effect while demonstrations do not. Antigovernment demonstrations in the current year increase the likelihood of presidential breakdown; strike activity in the previous year increases the likelihood of democratic breakdown. The difference can be explained by the fact that demonstrations tend to be more spontaneous, without a conscious plan to overthrow the government and take control of executive offices and state power. Indeed, after presidents fall, there is often confusion about who should lead government. Only in one case of presidential breakdown (Mahuad in Ecuador in 2000) did a group claiming to represent the protesters immediately take power, but it only lasted a couple of days before the civil-military junta fell. Democratic breakdowns, on the other hand, require planning because the group toppling the president also plans to take over government for a more or less extended period.

Regime Characteristics: Level of Democracy and Past Regime History Democratic regimes vary with respect to the degree to which they are free and competitive, so we consider whether the nature of the existing democratic regime affects the chances of presidential removal. We capture this effect with a dummy variable, semi-democracy, coded 1 if the previous year was classified as “semi-democratic” by Mainwaring, Brinks, and Pérez-Liñán (2001). The semi-democracy variable has a very positive and highly robust effect on the likelihood of a democratic breakdown, with t-statistics consistently of a magnitude of about 3.0. Our methodology cannot identify the causal mechanisms, but one can hypothesize that a semi-democratic regime is more likely to degenerate into an authoritarian regime (as with the 1992 autogolpe or self-coup in Peru). Another causal route could be military intervention ostensibly to “save democracy” in a context of deteriorating rights and freedoms, particularly under left-wing governments. In contrast to its effect upon democratic breakdown, semi-democracy has no effect upon the chances of a presidential breakdown. Often, semi-democracies or delegative democracies are limited by an overly dominant chief executive and the restriction of civil liberties so that antigovernment demonstrations may be illegal or hard to organize effectively. A country’s history of democratic breakdown has a very significant impact on the likelihood of present or future breakdowns. The past breakdowns

44

Michael E. Álvarez and Leiv Marsteintredet

variable counts the number of times that democracy has broken down in a given country. While it has a significant positive effect on the chances of another democratic breakdown (with t-statistics ranging from around 1.6 to just over 2.0, depending upon model specification), it never matters for presidential breakdowns. It is noteworthy that several traditionally stable democracies or nondemocracies, such as those of Paraguay, the Dominican Republic, and Venezuela, have nonetheless experienced presidential breakdowns. Thus, the legacy of democratic breakdown may not matter for presidential breakdowns because the latter are not driven by the actors that traditionally interrupted democratic rule. Another explanation may be that scandals, which have toppled several presidents, are distributed randomly with respect to regime legacy.

Institutional Factors: Number of Parties and Presidential Party Share of Seats Presidents are more likely to survive when they are successful, and we assume that success depends on the effectiveness of public policies, which are in turn affected by the size and determination of the opposition. Given this assumption, we use the effective number of parties and the share of seats of the president’s party in the lower (or single) chamber as proxies for a president’s ability to govern successfully and thereby enhance the prospects for survival. 20 Both variables prove to be very significant for the likelihood of a presidential breakdown; 21 by contrast, they have no effect on the possibility for democratic breakdown, which is not consistent with previous findings (Mainwaring 1993). While the positive effect of the number of parties variable on the likelihood of presidential breakdown is significant at the 99 percent level, its t-statistic in the democratic breakdowns equation barely exceeds 1.0. Furthermore, the former’s marginal effect and elasticity are double the latter’s. The effects of the president’s share of seats variable are similar: it is significant at the 99 percent level for explaining presidential breakdowns, but irrelevant for democratic breakdowns (t-statistics of around –.19), while its marginal effect on presidential breakdown is ten times that of its effect on democratic breakdown. These findings are curious considering that the literature on presidential breakdown grew out of the literature on democratic breakdown, which emphasized the role of institutions. One reason may be the distribution of data on our variables across time, which indicates that since the late 1970s, presidents have existed under conditions that tend to make the opposition stronger. Our data show that the average effective number of parties has increased considerably since the mid-1980s, corresponding to the historical shift away from a prevalence of democratic breakdown to a prevalence of presidential breakdown. The consequences of this distribution are magnified by the fact that a larger number of country-years in our sample falls in the

Breakdowns in Latin America

45

latter historical period due to recent waves of democratization. Thus, the pre-1979 average number of parties is 3.14 and the average share of seats for a president’s party is .46, while the respective averages after 1979 are 3.43 and .41. Even more telling, the data indicate that circumstances that can magnify the strength of the congressional opposition have been more challenging in years when presidential interruptions have occurred than in years of democratic breakdown: the average effective number of parties and share of seats for presidents’ parties in years of presidential breakdown are 4.43 and .34, respectively, while the corresponding numbers for years of democratic breakdown are only 3.05 and .5. This pattern of distribution of observations across time is also illustrated by the fact that the data show that there was a low effective number of parties and rather high presidential support during democratic breakdowns before 1960. It is revealing that the statistical model presented in table 2.2 below, which includes an inflation variable in its model specification and excludes country-years before 1960 (due to data limitations), produces more significant estimates of the effects of institutional factors on democratic breakdown than those found in table 2.1. A final explanation for the weak effects of institutions on democratic breakdown is that societal mobilization driven by the left-right ideological cleavage (as measured in our models by general strikes) is probably more important than institutional factors for explaining democratic breakdown. By contrast, institutional factors should be more significant for presidential breakdowns than for democratic breakdowns: in the former case, Congress removes presidents more in accordance with constitutional rules, and it therefore plays a more important role in presidential breakdowns.

Economic Performance Factors: Growth and Inflation Does economic performance affect the likelihood of presidential removal? We approached this problem from a number of angles, and the results were mixed: rather than the rate of growth of per capita GDP, it is negative growth that has consequences for the likelihood of presidential or democratic breakdown. We specify the rate of growth of the per capita GDP variable as lagged one year, as theory suggests that it is not growth that generates effects but rather the consequences of growth, such as employment, income, and the distribution of income, that matters (as noted above with respect to the lagged effect of general strikes). Thus, the forces that may affect executive change come into play some time after growth.22 Yet, while controlling for the other explanatory variables identified above, growth lagged one year produces t-statistics of only –.16 and .41 on the likelihood of presidential and democratic breakdown, respectively.23 Economic crises as measured by negative rates of growth may prove to be more decisive for determining the survival of presidents. There is no

46

Michael E. Álvarez and Leiv Marsteintredet

clear definition of what constitutes a crisis, and the results from a battery of dummy variables flagging different levels of negative growth proved to be wildly inconsistent. On the other hand, civil society or the military may forgive a short period of poor economic performance but may be less forgiving during prolonged recessions. Following this logic, we found that if a country experiences negative growth in the two previous years, then the likelihood of a presidential breakdown increases. By contrast, this variable proved to be irrelevant for democratic breakdowns. Unlike the other tested measures of growth, the significance of the variable prolonged recession was very robust for presidential breakdowns across model specifications and estimation procedures, generating t-statistics of about 2.1. The marginal effect of a prolonged recession on presidential breakdowns is also about double the magnitude of the corresponding statistic for democratic breakdowns. Intuitively, it is not surprising that a measure of economic performance that highlights a prolonged crisis could prove to be influential. Given the lagged effects of growth, it will only be felt in real terms in the following year. If negative growth persists, generating more pessimistic expectations about future performance, opposition to the president will only be intensified. The World Development Indicators (WDI) database contains data on inflation from 1960 onward. Given this shorter time series, we ran separate analyses to explore the effects of inflation on presidential survival. Table 2.2 below specifies a statistical model that is identical to the one presented in table 2.1, to which we added the inflation variable. This analysis is based on a sample consisting of only 544 observations (compared to 616 in table 2.1) and excludes one presidential breakdown (Brazil in 1955) and two democratic breakdowns (Argentina in 1951 and Guatemala in 1954). Our approach to the study of the effects of inflation resembles our analysis of economic growth in that we proceed from the expectation that it is specifically higher rates of inflation that should have a pronounced effect on presidential removal. Given that inflation has an immediate impact upon social well-being (in contrast to the lagged effect of growth), we expected—and our analyses confirmed—no lagged effects of this variable. We found that both the natural log of the annual rate of inflation and a dummy variable flagging rates of inflation greater than 20 percent have similar consequences for presidential removal. Both have a strong positive effect on the likelihood of democratic breakdown (with t-statistics of 2.61 and 2.19, respectively), but not presidential breakdown (t-statistics of only –.24 and –.55). 24 An examination of the data reveals that for twelve of the fifteen presidential breakdowns, rates of inflation were 31 percent or lower. 25 By contrast, there were generally higher rates of inflation in years of democratic breakdown: eight of the eighteen cases experienced inflation rates greater than 25 percent. Of these, five experienced rates above 69 percent, and among these, two had rates of around 400 percent (Argentina in 1976 and Chile in 1973). This should not come as a surprise, as problems associated with extreme rates of inflation began to affect the region

Table 2.2

Determinants of presidential breakdowns and democratic breakdowns

Characteristics in numerator of Prob Y = 1 Presidential Breakdown Variable Constant Level Prolonged Recession Past Breakdowns Semi-Democracy Number of Parties Demonstrations Strikes (lagged) Inflation Post-1978

Coefficient

t-statistics

–5.2177 –.0003 1.3609 .2650 –.8766 .3906 .4987 –.0374 –.3855 .4604

–3.329 –1.946 2.072 1.105 –1.116 2.382 4.191 –.119 –.554 .415

Characteristics in numerator of Prob Y = 2 Democratic Breakdown Variable Constant Level Prolonged Recession Past Breakdowns Semi-Democracy Number of Parties Demonstrations Strikes (lagged) Inflation Post-1978 Pseudo R-squared

Coefficient

t-statistics

–3.2618 –.0008 .0180 1.3850 1.7846 .2702 .3891 .9271 1.6744 –5.5603

–3.265 –2.005 .022 1.728 2.918 1.546 1.925 2.389 2.188 –2.377

.34

Marginal Effects Averaged Over Individuals Variable

Y = 00

Y = 01

Y = 02

Constant Level Prolonged Recession Past Breakdowns Semi-Democracy Number of Parties Demonstrations Strikes (lagged) Inflation Post-1978

.1906 .0000 –.0318 –.0358 –.0180 –.0148 –.0199 –.0190 –.0270 .1087

–.1217 .0000 .0322 .0055 –.0218 .0091 .0116 –.0014 –.0101 .0141

–.0689 .0000 –.0004 .0304 .0398 .0057 .0083 .0204 .0371 –.1227

Note: Inflation is a dummy variable coded 1 if the current rate of inflation is greater than 20 percent. See the notes for table 2.1 for the other variables. Source: World Development Indicators (WDI 2008).

48

Michael E. Álvarez and Leiv Marsteintredet

in 1973 (with great intra-regional variation), and ended effectively in 1994 when Brazil experienced its final year of hyperinflation. Since 1994 there have been ten presidential breakdowns and only one democratic breakdown. The period before 1979 exhibited a much higher regional average rate of inflation than the period after 1994, with greater variation and higher maximum levels. The varying levels of inflation across different time periods are, of course, linked to different models of development and economic management (import substitution versus neoliberal models), which again loosely correspond to periods of democratic and presidential breakdown. In sum, as predicted in theory, both presidential and democratic breakdowns are affected by economic performance—as measured by the rate of economic growth and the rate of inflation. Different indicators of the same overarching basic-level determinant seem to affect both phenomena. Inflation reached higher levels and growth tended to be higher in earlier periods when democracies still broke down. In the neoliberal era (at least since 1994), hyperinflation has been less of a problem, but growth has stagnated. Thus, presidents who want to survive should work hard to avoid recessions and high rates of inflation.

Conclusions Our conclusions depend upon the level of analysis that we emphasize. Two of our four basic-level determinants of presidential survival (economic performance and civil society mobilization) have an important impact on the likelihood of both presidential and democratic breakdown. It is at this basic level that we create hypotheses of causal relationships. We conclude that, with respect to economic performance and civil society mobilization, our two forms of presidential removal are affected by the same factors in similar ways, lending support to the hypothesis that presidential breakdowns are contemporary outcomes of causal forces that caused democratic breakdowns in earlier periods. At the same time, different indicators of these basic-level determinants were significant for the types of outcome. With respect to economic performance, the differences are linked to the dominant economic model: import substitution for democratic breakdowns and neoliberal for the most recent period of presidential breakdowns. Civil society mobilization affects both outcomes, but only antigovernment demonstrations are relevant for presidential breakdown, while general strike activity in the previous year matters for democratic breakdown. This difference can also be explained by the clustering of the two respective outcomes in different time periods. In contrast to the period of democratic breakdown, unions in the post-1979 period have been weakened by neoliberal reforms although civil society has remained active (Arce and Bellinger Jr. 2007). Thus, different varieties of

Breakdowns in Latin America

49

civil society mobilization explain the removal of presidents in the two different periods. As regards regime characteristics and political institutions, the picture is a bit different. Our indicators of regime characteristics (past transitions to authoritarianism and the nature of the existing regime) are both important determinants of democratic breakdown but not of presidential breakdown. These results appear to indicate a qualitative difference between the two varieties of presidential removal considered here: authoritarian regimes are more likely to emerge from political regimes that manifest higher degrees of democratic failure and they may therefore be a reaction to more limited “freedoms” or to regime “radicalization” (as per the ideological worldview of the military). By contrast, presidential breakdowns may reflect a focus upon a more temporal and personal (rather than systemic) target—a failed or scandal-ridden president. Only political institutional design has a significant impact on the likelihood of presidential breakdown. As the number of political parties increases, and/or the share of legislative seats held by the party of the head of government falls, so the likelihood of a presidential breakdown increases. However, these factors have no influence on the chances of a democratic breakdown. Above, we discussed how the historical evolution in the number of parties and the share of seats might produce this surprising finding. Here, we can conclude that the “perils of presidentialism” as outlined by Linz may not be such a threat to democracy in Latin America after all. This matches Cheibub’s findings (2002), which were rather striking when they were initially published: executive-legislative deadlock does not increase the likelihood of democratic breakdown in presidential regimes; other factors are more important (Cheibub 2007). Paradoxically, while deadlock may not threaten the survival of a democratic regime, Linz’s analysis does contribute to our understanding of the causes of presidential breakdown. So, are presidential breakdowns in Latin America simply more limited contemporary expressions of democratic breakdown? Although there is little strong evidence to support this hypothesis, our results fail to support the claim that presidential breakdowns are an entirely new phenomenon. By comparing the causes of both types of outcomes, we have gained some insight into how and why presidents fall in different ways. While it is clear that the causes of democratic and presidential breakdowns differ at the indicator level, our findings also demonstrate that some basic-level determinants such as social mobilization and economic performance affect both phenomena similarly. By contrast, the quite different impact of political institutions on each outcome provides surprising insights into Linz’s perils of presidentialism: Linz was wrong about presidential systems causing democratic breakdown but he was right that they are unstable; and he was right that deadlocks cause presidential breakdowns but wrong to predict that such conflicts lead to the breakdown of democracy.

50

Michael E. Álvarez and Leiv Marsteintredet

Notes * The authors thank Mariana Llanos, Tor Einar Holvik-Skinlo, and Adam Przeworski for comments on previous versions of this chapter, and William H. Greene for advice on our statistical models. All errors remain our own. 1. Democratic breakdowns do not necessarily involve a change of the head of state. Autogolpes are a case in point. 2. Here, a minimally democratic regime is at least semi-democratic according to the index of democracy in Latin America in Mainwaring, Brinks, and Pérez-Liñán (2001). 3. The military has not been entirely absent in recent times (chapter 4, this volume), but it has been less important. 4. The richest democracy in the world ever to fall was Argentina in 1976. Argentina can also lay claim to being the richest country ever to experience a presidential breakdown (in 2001). Measured in 2000 dollars, Argentina’s GDP per capita in 1976 (10,200) and 2001 (10,700) differed by only 500 dollars. 5. The role of the OAS has changed considerably since 1965, when the organization backed a U.S. invasion of the Dominican Republic that clearly thwarted potential democratic progress (see Hartlyn 1998, 89). This shift is linked to the 1991 adoption of Resolution 1080. OAS action helped thwart the coup in Guatemala in 1993 (Villagrán de León 1993), to negotiate the 1994 salida of Balaguer in the Dominican Republic (chapter 10, this volume), and dissuaded General Oviedo from attempting a coup in Paraguay in 1997 (Valenzuela 1997). 6. Some presidential breakdowns clearly resemble transitions from grey zone to more democratic regimes. The impeachment of Raúl Cubas in Paraguay in 1999 involved a transition from a semi-democracy to a somewhat more democratic regime (see chapter 8, this volume, and Abente-Brun 1999), as did the negotiated salida of Joaquín Balaguer in the Dominican Republic, and the end of the Fujimori regime in 2000 (Levitsky 1999; Cameron 2006; Carrión 2006). 7. For a contrasting view, see Faúndez (1997). 8. See note 2 above. 9. The Banks general strikes variable counts any strike involving more than 1,000 industrial or service workers and more than one employer protesting against national authorities. The antigovernment demonstrations variable includes any peaceful gathering of more than 100 persons to voice opposition. 10. Rare events logistic regression analysis controls for the uneven distribution in the dependent variables: there are very few positive as compared to negative observations (King and Zeng 2001a; 2001b). This technique is preferable for studying these phenomena in separate probit equations (Mainwaring and Pérez-Liñán 2005; Negretto 2006; Pérez-Liñán 2007), but for our purposes the multinomial model is superior. Rare events logistics work with dichotomous dependent variables. Thus, in order to study presidential breakdowns we would either have to code democratic breakdowns as a nonevent (clearly wrong), or exclude the case (country-year) from the analysis (clearly undesirable). 11. An assumption built into the multinomial logit model is that the results generated for any given choice alternative (say, presidential breakdown) are independent of the addition of an irrelevant alternative (IIA). More specifically, the IIA assumption holds that “the relative odds of choosing [one alternative] i over [a different

Breakdowns in Latin America

12.

13.

14. 15.

16.

17.

18.

51

alternative] k are the same no matter what other alternatives are available or what the attributes of the other alternatives are” (Train 2003, 49–50). If this assumption were proven not to hold for our data, then the results generated would be significantly less reliable. However, our tests reveal that the results generated in binomial models using both standard and rare events logistic regression for both phenomena do not vary appreciably from the results generated by the multinomial logit model, which indicates that the outcomes are independent of irrelevant alternatives (Train 2003, 53). The following country-years are registered as presidential breakdowns: Brazil 1955 and 1992; Ecuador 1961, 1997, 2000, and 2005; Bolivia 1985, 2003, and 2005; Argentina 1989 and 2001; Guatemala 1993; Venezuela 1993; Dominican Republic 1994 (1996); Paraguay 1999; and Peru 2000. The following country years are registered as democratic breakdowns: Argentina 1951, 1962, 1966, and 1976; Bolivia 1964 and 1980; Brazil 1964; Chile 1973; Colombia 1949; Costa Rica 1948; Dominican Republic 1974; Ecuador 1963 and 1970; Guatemala 1954; Haiti 1999; Nicaragua 1948 and 1968; Panama 1963 and 1972; Peru 1948, 1962, 1968, and 1992; Uruguay 1973; and Venezuela 1948. The country-year format has certain advantages, such as increasing the sample size, but it also has drawbacks. If we have learnt anything from the volumes edited by Linz and Stepan on the breakdown of democracy it is that breakdowns take time and several causes interact, creating endogeneity problems. We try to avoid endogeneity problems as far as possible, but we admit that analyses based on country-year as the unit of analysis may model more effectively the timing of breakdowns rather than identifying the full list of causes and processes of a breakdown. Pérez-Líñán has also helped to fill the blanks on some pre-1978 institutional variables. Regression coefficients reported are log-odds, and are not easily interpretable in their untransformed state (although the t-statistics can be understood in the standard manner). Therefore, we also present the marginal effects of each respective independent variable upon the probability of observing a presidential or a democratic breakdown. We also explored whether specific levels of economic development (GDP/capita between 3,000 and 5,000, for instance) matter more than others for presidential and democratic breakdowns, but we did not find any evidence to support this hypothesis (Lipset, Seong, and Torres 1993). Adding post-1978 to the full model specification without any time controls, for example, produced only very minor changes in the statistics measuring the significance of the other independent variables for explaining presidential breakdowns. By contrast, when post-1978 was added to the model explaining regime breakdown, the impact proved to be more significant. For example, the t-statistic of prolonged recession fell from –1.1 to –.5, and its marginal effect fell by half. The significance and marginal effects of past transitions also fell somewhat, but the biggest change concerned the t-statistic and marginal effect of demonstrations, which increased from –.5 to 1.47 and from –.003 to .006, respectively, after the addition of post-1978. This finding suggests a trend but does not necessarily contradict Mainwaring and Pérez-Liñán’s apparently different finding. Their conclusion—an inference based on the underlying causal factors specified in their statistical model—is

52

19.

20.

21.

22.

23.

24.

25.

Michael E. Álvarez and Leiv Marsteintredet that the predicted likelihood of a democratic breakdown after 1978 has not decreased. Therefore, since the causal model that they specify does not predict a decrease in the likelihood of a democratic breakdown, the actual decreasing trend that we observe in the actual data must be due to some exogenous factors not specified in their statistical model. One of these factors could be the more recent pro-democratic slant in U.S. foreign policy. See table 2.2, in which a statistical model including the inflation variable reveals a slightly stronger effect of demonstrations upon the likelihood of a democratic breakdown. Due to the high correlation between the number of parties and the head share (.81), these two variables were never specified together in the same equation. Rather, the full model specification was estimated with each of these variables specified separately. None of the meaningful statistics that corresponded to the other independent variables were affected at all by the choice of specifying either the parties’ variable or the head share variable. (None of the other variables specified in our statistical models exhibited a degree of correlation with each other greater than .3). This is consistent with findings presented in studies that test the effects of other measures of institutional factors (Negretto 2006; Pérez-Liñán 2007; Valenzuela 2004). There is also a statistical argument against using growth in the current year: possible simultaneity bias. Given the difficulty of confirming the exogeneity of growth in the current year, any variable that measures growth should at least incorporate a lagged component. A variety of moving averages (2-year; 3-year; 5-year; 3-year weighted) for growth also fail to reveal any effects—none of these variables produced a t-statistic greater than 1.1 for either form of breakdown. Again, there is no rule to follow when exploring the effects of inflation on presidential survival. Several inflation thresholds were tested via a series of dummy variables that marked levels of inflation greater than 10, 20, 30, 50, and 100 percent. The results indicate that the effects of inflation on political stability are particularly strong at rates greater than 20 percent (with consequences that remain roughly the same at the higher rates of 30 and 50 percent, but drop significantly at rates above 100 percent). Specification tests also revealed that the dummy variable marking inflation rates greater than 20 percent was most robust across model specifications, displaying a consistently positive impact on the likelihood of a democratic breakdown at a 99 percent level of statistical significance. The three extreme cases above this level were Argentina in 1989, Brazil in 1992, and Bolivia in 1985, with annual rates of inflation of 3,000, 1,000, and 12,000, respectively.

II

Cases of Repeated Breakdown

This page intentionally left blank

3 Presidential Breakdowns in Argentina* Mariana Llanos

Introduction With the resignation of two of the five presidents elected during the third wave of democratization, Argentina became an obligatory reference in the comparative literature on presidential breakdown (Hochstetler 2006; Hochstetler and Edwards 2009; Kim and Bahry 2008; Mainwaring and Pérez-Liñán 2005; Negretto 2006; Pérez-Liñán 2003; 2007; 2008; Valenzuela 2004). The breakdown of the presidencies of Raúl Alfonsín (1983–1989) and Fernando de la Rúa (1999–1901) provided empirical support for the three explanatory factors most highlighted by this literature: divided government, economic adversity, and social protests. First, presidents Alfonsín and de la Rúa, both belonging to the Radical Civic Union (Unión Cívica Radical, UCR), had to coexist with a Peronistdominated Senate from the beginning of their administrations; they initially counted on a majority in the lower chamber, but lost it after midterm parliamentary elections. Thus, the breakdown of their presidencies can be associated with the shortage of institutional resources that typically occurs in systems of dual legitimacy, and with the so-called propensity to conflictive relations between presidents and congresses that is intrinsic to these systems (Linz 1990; 1994). Second, the two breakdowns occurred in critical contexts: amid hyperinflation in the case of Alfonsín, and in the midst of the rigors imposed by the convertibility straitjacket in the case of de la Rúa. They confirm how economic variables—typically, high inflation and lower rates of growth—as well as certain policy orientations, such as neoliberal policies or policies involving austerity measures, shape the fate of a president. Third, lootings took place in the provinces and in Greater Buenos Aires following the burst of hyperinflation in May 1989, and it made a comeback in December 2001, this time in a context of broader protests, both episodes

56

Mariana Llanos

ending with deaths. The comparative literature notes that social protests and mobilizations are the main force capable of overthrowing corrupt or unpopular governments (Pérez-Liñán 2007, 210), and that no parliamentary efforts to remove presidents from office before their term is over have succeeded without such protests (Hochstetler 2006, 410). But the two above-mentioned breakdowns were not the only ones that took place in democratic Argentina. The post-de la Rúa presidents appointed by Congress, Adolfo Rodríguez Saá and Eduardo Duhalde, failed to complete the terms assigned to them by respective parliamentary majorities. These cases do not fit the definition of presidential breakdown, as this is a phenomenon involving the early departure of elected presidents, but they certainly add evidence for the thesis of “a new pattern of instability” in the region that does not involve regime collapse (Mainwaring and Pérez-Liñán 2005, 49). Further, their parliamentary origin has led to claims that there is a process of regime parliamentarization underway (Mustapic 2005; Schamis 2002), strengthening the argument that there is a de facto flexibilization of Latin American presidentialism. In effect, for some authors, presidential breakdowns are actually a response to one of the most serious flaws of the presidential system—fixed terms, or their rigidity (Pérez-Liñán 2007, 57; chapter 1, this volume)—and confirm that presidential survival is no longer independent from the legislative and popular will (Marsteintredet and Berntzen 2008). This chapter revisits the two above-mentioned Argentine presidential breakdowns and, briefly, the two caretaker governments. It draws on press reports and on both the comparative and Argentine scholarly literature on the topic.1 My view is that, on close inspection, these cases confirm that every variable matters, but that there being a minority government (a well acknowledged variable) matters more, as do president-centered explanations, which remain underdeveloped in the comparative literature. Below, I explain how these variables connect with presidential instability in Argentina. I then describe the Argentine presidential crises and early departure from office of both presidents and caretaker presidents. Finally, I focus on the aftermath of these presidential breakdowns.

Midterm Elections and Presidential Survival The 1853 Constitution established a federal presidential republic with a bicameral legislature. According to the Constitution, presidents could coexist with various configurations within the legislature (a majority in one, both, or none of the chambers) during their mandates: they were elected by an Electoral College for a six-year term; half of the Chamber of Deputies was renewed every two years; and a third of the Senate was elected every three years by the provincial legislatures. After the 1994 constitutional reform, the now directly elected president had a shorter four-year term but could

Presidential Breakdowns in Argentina

57

be reelected, and presidents had to face one midterm election in their second year in office to renew half of the Chamber of Deputies, and a directly elected third of the Senate. Minority governments are certainly less pervasive than Linz had thought (Linz 1990), but the cases of Alfonsín and de la Rúa illustrate when they can endanger presidential stability (Valenzuela 2004, 14). In effect, presidents governing in the post-1983 democratic period did not face “recalcitrant” legislatures (Cox and Morgenstern 2002, 455) until they were confronted with midterm electoral defeat. Winning the midterm elections became an asset for presidents, and the consequences of losing them proved to be very damaging. Presidential defeats occurred on three occasions: under Alfonsín in 1987, under Menem in 1997, and under de la Rúa in 2001. When majority president Carlos Menem suffered defeat, his bills began to fail in Congress and he was openly challenged by other leaders of his party (Llanos 2002, 155–178). Menem’s “imperial” actions (Cox and Morgenstern 2002, 455) during those years—he governed mostly by decree and he engaged in an (unconstitutional) quest for a third presidential term—obscured the fact that this was a president weakened by a loss of popularity, and being challenged by potential successors after his electoral defeat. However, Menem did not fall, as Alfonsín and de la Rúa, who both had much weaker institutional resources, did. De la Rúa stepped down two months after the midterm elections, but, as we shall see below, Alfonsín also found it hard to recover from the loss of legitimacy crystallized in September 1987. In other words, midterm election defeats (an institutional factor that is not much present in the comparative literature) reinforced existing institutional constraints and paved the way for the early handover of power in both cases. However, institutional variables do not capture the different transition paths and institutional solutions that characterized these presidential successions: a longer, presidentially monitored and negotiated transition took place between the Alfonsín and Menem presidencies; by contrast, there was a short, Peronist-managed transition between de la Rúa and Rodríguez Saá from which the president was notably absent. A number of Argentine scholars point out that the president himself constitutes a key explanatory factor of a breakdown—an observation that also applies to caretaker presidency breakdowns. Certainly, it is acknowledged in the comparative literature that presidential corruption (and concomitant scandals), the ability or inability of presidents to form coalitions, and performance in terms of economic management are significant predictors of the fate of a presidency (Hochstetler and Edwards 2009; Pérez-Liñán 2007). However, the Argentine breakdowns and interpretations thereof persuasively show how the convictions and skills of presidents (or lack thereof), translated into governmental decisions (or omissions), interact with other structural or institutional features of the political system. Although president-centered features appear anecdotal or accidental at first glance, they actually seem to express a more general pattern found even beyond Argentina. Bringing

58

Mariana Llanos

them into a theoretical category at a higher level of abstraction (Pérez-Liñán 2008, 118, citing Sartori 1984) is a pending task, but probably one worth pursuing.

Raúl Alfonsín (1983–1989) Raúl Alfonsín won the 1983 presidential elections with 51.75 percent of the vote. His victory broke the “iron law” of Argentine politics by demonstrating for the first time since the 1940s that Peronism could be defeated in an open and fair electoral contest. Alfonsín gained control over the majority of seats in the Chamber of Deputies, but he won only 39.1 percent of senate seats, and only seven out of twenty-two governorships. Alfonsín won the presidency with various promises, ranging from the reestablishment of constitutional and rule of law government to a commitment to social justice. While the military regime had illegitimately concentrated both political power and economic wealth in the hands of a privileged few, democracy, so went the message, would reverse that process. However, it soon became clear that the new democracy would have to coexist with a difficult economic situation. Alfonsín encountered an oversized, inefficient, and exhausted state apparatus, the multiple activities of which had been mainly financed in the 1970s though inflation and large-scale loans from private banks. Mexico’s debt moratorium crisis in August 1982 ended the wave of easy international credit for developing countries, increased interest rates in industrial countries, and, given decreasing export revenues, put acute pressure on the balance of payments. International pressures emerged to service the debt and to stabilize economies. Thirteen months after its inauguration, the government found itself on the defensive, amidst growing inflation, union pressures, business discontent, and falling public approval ratings (Torre 1993, 78). A new economic team, led by Economy Minister Juan Sourrouille, launched a stabilization programme in June 1985, the Austral Plan. This was the first in a series of heterodox packages implemented in Latin America. The goal of the Plan was to stop inflation whilst maintaining levels of employment and economic activity; in contrast to orthodox plans, the Plan aimed to avoid regressive redistribution and social tensions that would have interfered with Alfonsín’s overarching aim of securing the transition to democracy. In other words, Alfonsín’s convictions about the painful social consequences that deep structural reforms could have (Palermo and Novaro 1996) defined the content of the Plan and prevented the government from taking some tough but much needed economic reform measures (Machinea 1993, 133). Although they were far from harsh, Alfonsín’s economic measures were regarded suspiciously by the parties in Congress. Congressional activity was intense in the 1980s (Mustapic and Goretti 1993) and the president enjoyed high levels of legislative approval, 2 but he also had to make decisions

Presidential Breakdowns in Argentina

59

unilaterally. On some occasions, the president resorted to necessity and urgency decrees, an instrument practically unknown until those years, the use of which was justified by the critical context and the nature of the measures adopted (notably, the Austral Plan). Although these decrees were used relatively cautiously (especially in comparison with their rampant abuse under President Menem), their use was seen as a high-risk insulating strategy at the time (Torre 1993, 75), which might initially help the president avoid opposition, but which would not protect him as challenges emerged at the later stages of the policy process when the ties between Alfonsín and his party, on the one hand, and business sectors and unions, on the other, proved to be a major problem. The Austral Plan was moderately successful for over a year, allowing the Radicals to reprise their victory over Peronism in the 1985 midterm elections. But from the middle of 1986 onward, inflation proved persistent and even worsened. With economic indicators performing badly, and under mounting pressure from the unions and business sectors, the government did not pass the second midterm electoral test in 1987. It lost control of the lower chamber and of most governorships—retaining only Córdoba and Río Negro, and the Federal Capital. Alfonsín was a lame duck president in 1987, and he knew it. As Morales Solá recounts, the election results were devastating for the president, who fell into a deep depression and even offered his resignation to the cabinet (1990, 14). This did not happen, but there was a cabinet reshuffle—enhancing the positions of the technocratic Economy Minister Sourrouille and new Minister of Public Works Rodolfo Terragno—which was intended to signal the start of a process of (albeit timid) state reforms and privatizations.

Midterm Elections and the End of Alfonsín The 1987 electoral defeat put paid to Alfonsín’s dreams of constitutional reform (and reelection), and hurried important political and institutional decisions that seemed to mark the path of a presidential transition. 3 By the end of 1987, two years before the constitutionally mandated end of the presidential term, the official presidential candidate was confirmed. Eduardo Angeloz—the governor of Córdoba and one of the few successful Radical figures in an otherwise increasingly shattered political landscape—was nominated as the candidate of the Radical Party. At the time, Carlos Menem, then governor of La Rioja, proclaimed himself the Peronist pre-candidate, challenging the official and progressive leadership of his party—headed by Antonio Cafiero—in the primary elections. 1988 presented a complicated scenario for the government: rising inflation, tense negotiations with international creditors, striking unions, military rebellions, and mounting public discontent. In July 1988, when Carlos Menem won the Peronist primaries, the Radical government was inaugurating a new stabilization package based on price agreements with some

60

Mariana Llanos

industrial sectors. President Alfonsín decided to take advantage of the initial success of the Spring Plan to favor the chances of Radical candidate Angeloz: in October 1988, the president enacted the decree that set the presidential Election Day for May 14, 1989—seven months before presidential handover due in December 1989. At the time, this was not an unreasonable gamble, since the Peronists held the lead in opinion polls but Angeloz was coming closer and closer in second place. The polls showed that only a meager 16 percent viewed Alfonsín positively, compared to the 69 percent who felt that Angeloz had done a good job as governor. Angeloz was becoming a haven for all those who feared Menem’s populist image, and his promises of a “debt moratorium, massive salary increases, a productive revolution, and Latin American unity.” For many business agents and regular voters, Menem’s old-style political discourse, and his partnership with the most recalcitrant Peronist factions, merely contributed to the fears of future fiscal indiscipline and, thus, higher inflation. At the beginning of 1989, the inflation rate began to rise again. Because of pressure from different sources, the government had not managed to get a grip on spending. Neither could it, in the context of an electoral campaign, implement further economic adjustment measures, because tax increases and privatizations would not get congressional approval. To make matters worse, there was a rush on the dollar toward the end of January, after information reached the press that the World Bank would not be disbursing a significant part of promised credits.4 On February 6, with reserves almost exhausted, the authorities suspended foreign currency sales. The value of the dollar on the free market began to rise, affecting prices across the board. Soon, the government was abandoned by its business partners and the unions returned to a “state of alert.” At the end of March, with black market value of the dollar escalating, Economy Minister Sourrouille and his team resigned. This came as a direct response to the demand by Radical candidate Angeloz that the team should resign. As part of his electoral campaign, Angeloz blamed the economy minister for the country’s economic woes, and advocated privatization, deregulation, and fiscal adjustment as alternatives. Angeloz expected the president to send a positive sign to financial markets by appointing a neoliberal economy minister, but Alfonsín was unable to betray his principles (Morales Solá 1990, 61) and chose Juan Carlos Pugliese, a veteran Radical politician who was then president of the Chamber of Deputies, to head the economy. Pugliese was a respected political figure who, the president hoped, could generate a new consensus. Three weeks before the May 14 presidential election, however, the value of the dollar rocketed from 15.50 to 83 australes in less than a month, interest rates rose above 66 percent a month, and the forecast for inflation was a minimum of 40 percent. Isolated and fearing electoral defeat, on April 21 President Alfonsín addressed the nation, calling on “all the presidential candidates to ensure an ordered transition” from his administration to a newly

Presidential Breakdowns in Argentina

61

elected government due to take office in December (LAWR 1989, 17). When elections confirmed the victory of Carlos Menem, Alfonsín immediately approached the elected president and the Peronist party in order to discuss the terms of the transition, which included the possibility of an anticipated handover of power (Lázara 1997, 167 and 358). However, conversations with the victorious Peronists did not prove to be easy. First, the Peronists were deeply divided over Alfonsín’s call for a smooth transition. Some wanted ongoing talks with the government, while others wanted to “hang tough” until the crisis “bottomed out.” Second, because of Menem’s distance from Peronist party structures, the government faced a dual, or even multiple, negotiation. Menem tended to disregard the results of interparty talks conducted by members of his own party. Third, as Lázara (1997, 412–413) documents, Menem’s economic plan, as proposed by the Bunge y Born holding, and his economy minister (who was a member of the holding), was already confirmed a couple of days after the May elections, but Menem needed “time and political conditions” (Lázara 1997, 283) to push ahead with his plans, so he refused to act immediately in conjunction with the Radical government. During the last week of May 1989, lootings erupted, mainly in Rosario and Greater Buenos Aires, initiated by elements of the most deprived and unprotected social sectors, which had suffered the most from the effects of hyperinflation. However, extreme left activists were immediately blamed as instigators, although it has also been argued that organized groups, involving Peronist right-wing factions, military carapintadas, and the provincial police, were responsible for organizing the first popular outbursts (Lázara 1997, 297–298). Whatever the case, it was this dramatic episode of social turmoil that left fifteen dead and more than eighty people injured, that led to the arrest of thousands, and prompted Alfonsín to withdraw from power. During the first week of June, Menem announced his future cabinet and there was progress (albeit not exempt from difficulties) in the negotiations between Alfonsín and Menem’s delegates. On June 12, President Alfonsín announced his resignation. A few days later, the agreements regarding the political transition were sealed. The terms of the agreement were twofold: there was an arrangement for the early transfer of the presidency, and rules were set to govern a period of party cohabitation in Congress. As both parties had agreed to shorten the transition period from December to July, the parliamentary assembly approved the resignation of the president and vice president and appointed Menem as the president elect. As regards the second aspect of the deal, December remained the changeover date for Congress, and the Radicals, which constituted the first minority in the Chamber with 114 deputies, agreed to cooperate to form the quorum. Thus, Menem obtained the Radicals’ commitment to facilitate the approval of all economic laws that he would send to Congress between July and December. As per the agreement, the Radicals attended the session in numbers sufficient to provide a quorum, but there was never to be more

62

Mariana Llanos

than seventy-five in attendance, so as to allow the Peronists to dominate the Chamber (Llanos 2002, 76–77). On July 8, Menem took over the presidency. Hyperinflation broke down resistance to the government undertaking major economic adjustments and state reforms, whilst the political agreements for the presidential transition provided the new administration with invaluable institutional resources to launch its programme. In 1989, one of the greatest privatization programmes of the neoliberal wave of the 1990s was initiated, although the battle against inflation was only won in 1991 with the introduction of the so-called Convertibility Law.

Fernando de la Rúa (1999–2001) Fernando de la Rúa won the presidential elections with 48.35 percent of the votes as a candidate of the Alliance, a loose coalition between the UCR and the newly established Country in Solidarity Front (Frente País Solidario, FREPASO), a conglomeration of small left-wing political groups and former Peronist factions. The Alliance was a successful electoral strategy that profited from the national organizational structures of the UCR and the popularity and communicational efficacy of Carlos “Chacho” Álvarez and Graciela Fernández Meijide, the main leaders of FREPASO (Novaro 2002, 43). However, the Alliance lacked a well-finished coalition project; it was not much more than an arrangement for a shared presidential formula for de la Rúa and Álvarez to run as president and vice president, respectively. In addition, the concomitant congressional and provincial elections gave the Alliance only a narrow majority in the Chamber of Deputies, and gave the Peronist opposition a resounding majority in the Senate (with 55.5 percent of the vote) and control over fifteen of the twenty-four provinces (including the three most important ones, Buenos Aires, Santa Fé, and Córdoba). De la Rúa’s mandate was to correct the legacies of Menem’s ten-year rule. The aim was to maintain the Convertibility Plan, but to remedy the social costs of market reform and to improve democratic governance mechanisms. Approved in 1991, the Convertibility Plan had controlled inflation and, together with measures (such as privatizations) to tackle fiscal problems, had generated the conditions for economic recovery and rapid growth.5 But a large part of the population paid a high price for the model that led Argentina to be held up as one of the most successful cases of structural reform in Latin America. When the Alliance assumed power, recession and fiscal deficit went hand in hand with a growing public debt and increasing difficulties in obtaining external financing. Further, the previous government had a poor record when it came to transparency and accountability, and the Alliance thus came to embody popular hopes for an alternative way to engage in politics (Charoski 2002, 198–211). However, the Alliance government found it hard to deliver on its electoral promises. Sustaining convertibility and promoting renewed growth proved

Presidential Breakdowns in Argentina

63

to be ambitious and contradictory economic goals; other promises also fell by the wayside. If the tasks facing the government were colossal in and of themselves, they were further complicated by de la Rúa’s personal interpretation of the role of the presidency (Llanos and Margueritis 2006). The president’s political strategy was isolation. He disregarded the coalition when managing the business of government and relied only on an inner circle of close advisors (friends and members of his family), who neither compensated for his weak base of support within his party nor promoted dialogue with coalition members.6 In particular, the way the president dealt with a corruption scandal in the Senate, in which members of the government had allegedly bribed senators to get the labor reform law approved, triggered a conflict with vice president Álvarez, which eventually led to the latter’s resignation in October 2000.7 Although it was Álvarez and not the whole of his party, the FREPASO, who departed from the government, his resignation weakened intra-coalition trust, and it also raised doubts among international investors and financial agencies about the president’s ability to control the Alliance. Further, the corruption scandal dealt a crushing blow to the credibility of the Alliance, which had been voted into office on the promise of eradicating the widespread corruption of the Menem years. A complicated cabinet reshuffle in March 2001 further damaged the government’s credibility on all fronts. As economic indicators and international economic risk assessments did not improve, Economy Minister Machinea resigned. But de la Rúa’s newly appointed economy minister, López Murphy, and his announcement of a new wave of adjustments provoked the resignation of three “political” ministers and other Alliance officials. Eventually, Domingo Cavallo, the father of convertibility during Menem’s years, assumed command of the economy. He immediately demanded the delegation of powers from Congress, which he got. However, neither the revenue-oriented measures nor a harsh new round of spending cuts to promote a zero public sector deficit, for which Cavallo also obtained congressional support, proved to work (Gervasoni 2003, 29–30). Although economic policy was certainly a source of tension within the coalition, the parties in Congress did not block the president’s policies until the midterm elections. De la Rúa often resorted to necessity and urgency decrees to pass economic measures, which was quite consistent with his unilateral vision of the exercise of power (Mustapic 2005, 270). But when the choice was to ask for congressional support, he got what he wanted on most occasions. De La Rúa’s bill approval rate was close to 40 per cent, similar to that of majority president Carlos Menem during his first six months in office (Calvo 2007, 269). He managed to pass some important laws, such as the restrictive 2000 budget, the law to delegate powers to Minister Cavallo, and the deficit zero law. In an analysis of roll-call votes, Jones, Hwang, and Micozzi (2009) show that, despite growing unrest among Radicals over the policy direction of the administration, the position of Radical deputies remained relatively homogenous and they were generally supportive until de

64

Mariana Llanos

la Rúa’s final two months in office. The support from FREPASO fell in the fourteen months following the resignation of the vice president, but it did so gradually.8 With the midterm elections approaching in October 2001, there were protests in the provinces, the economic situation had worsened, capital flight had reached new heights, and public discontent was widespread. The political environment was tense. Lacking the wherewithal, the national government failed to give the provinces the minimal monthly sum that was part of the revenue-sharing tax scheme, leading provincial governors to accuse the national government of “retaining what belongs to the provinces and thus causing social outbreaks” (Governor Ruckauf, La Nación November 10, 2001), and ultimately, even taking de la Rúa to court.9

Midterm Elections and the End of de la Rúa A distinctive feature of the October 2001 elections was that they expressed a great disaffection with party politics in general. For the first time in Argentine history, electoral participation declined to less than 75 percent of the electoral roll, while the number of blank and null votes increased significantly (Escolar et al. 2002, 25–28). Political disaffection did not hit all parties equally, though. Compared with the total obtained in the preceding presidential election in 1999, the Alliance in government lost 4,939,935 votes (61.8 percent), while Peronism lost only 27.2 percent (Torre 2005, 176). The 2001 elections also made it apparent that the political consensus around the convertibility rule, which had been breaking down for some time, had definitively come to an end. Immediately after the October electoral defeat, former president Raúl Alfonsín declared that “Minister Cavallo had finished his cycle” (La Nación October 19, 2001), and the Radical party leadership as a whole demanded a change of economic policy and declared that it was the unanimous decision of the party not to back any further adjustments (La Nación October 18 and 24, 2001). The government was left alone, as all the political forces supporting it (with the exception of Domingo Cavallo’s party) demanded a change of course. The Peronist blocs in Congress became progressively more combative. By the end of November, some of them began to talk publicly about the socalled acephaly law (ley de acefalía) (discussed below), and the Senate promoted the nomination of the just-elected Peronist senator, Ramón Puerta, to the presidency of the upper house. This decision broke with the old political tradition whereby only the party in charge of the presidency was entitled to conduct the work of the Senate independently of the chamber’s partisan composition. Although the government accused the Peronists of carrying out an “institutional coup,” at the time it was certainly difficult to say there was one (or, a single plan) plan to take over the government: who would then become the next president? Peronism had been in disarray since Menem had left power. There was no legitimate representative for the party as a whole and

Presidential Breakdowns in Argentina

65

“every Peronist governor (was) a potential presidential candidate,” as former president Menem said (La Nación November 24, 2001). By the middle of December, at the peak of the crisis, the Peronists used their majority in both congressional chambers to extend the ordinary parliamentary sessions (in order to prevent de la Rúa from governing alone during the summer) and to revoke the special faculties that Congress had extended to Minister Cavallo eight months before. Throughout these weeks, President de la Rúa leaned heavily on the dominant figure of Cavallo in all kinds of negotiations. The president’s actions were limited to a couple of marginal changes to the cabinet, the public defense of convertibility, a timid call for a multi-sector compromise, and, ultimately, and under the pressure of his party, for a government of national unity with the Peronists, which the latter rejected. De la Rúa’s timid attitude did not measure up to the gravity of the situation during the last three weeks of his government, which were marked by the collapse of economic activity, Cavallo’s desperate attempts to save the convertibility scheme, massive withdrawals of, and a consequent freeze on, bank deposits, the threat of debt default, mounting public discontent, street protests and looting, and the reluctance of the International Monetary Fund (IMF) to pay out a tranche due in mid-December. The social commotion that erupted in December included a wave of food riots, with thousands of people blockading roads and bridges throughout the country and banging pots and pans in the main plaza of Buenos Aires. Eighteen people were killed, either by the police or by store owners, hundreds more were seriously injured, and thousands were arrested (Auyero 2006, 249).10 De la Rúa decreed a state of siege to recover control, but the social commotion continued. The president finally resigned on December 20.

The Caretaker Governments: Two Further Breakdowns When de la Rúa resigned and with the vice presidency vacant, there was no doubt that his successor would be a Peronist. However, in contrast with 1989, there was no elected candidate to assume presidential duties indisputably. In this context, Article 88 of the Constitution and the ley de acefalía of 1975 (Law 20982) establishing the line of succession in the absence of a president and vice president were set in motion. According to this law, the presidency was to be held for forty-eight hours, successively, by the president of the Senate, the Chamber of Deputies, and the Supreme Court, until the congressional assembly agreed on a candidate (a senator, a deputy, or a governor) to act as caretaker until a new president could be elected. As under a parliamentary regime, the law gave Congress the prerogative to form a government (Mustapic 2005, 272). The congressional assembly decided to call elections to fill the posts of president and vice president who would take office in April 2002, and appointed

66

Mariana Llanos

the governor of San Luis, Adolfo Rodríguez Saá, as provisional president until an elected president took power. The terms of the decision resulted from an informal agreement among Peronist governors,11 and it violated the provisions of the electoral code as it allowed internal partisan factions to compete with each other on the basis of a ley de lemas scheme, whereby factions within the same party can compete with each other in a general election (Congressional Assembly records, December 23, 2001, www.senado. gov.ar). Non-Peronist forces were critical of the obviously unconstitutional nature of this decision as well as of the controversial appointment of a provincial caudillo to be the caretaker president, so Rodríguez Saá only received the votes of the Peronist blocs. During his eight days in power, unfortunate political decisions—to default on the debt, to appoint obscure (and corrupt) personalities, which caused renewed street protests, and Rodriguez Saá’s decision to state his desire to stay in power longer than agreed—alienated the caretaker president from his own party. In addition, the decision of the Supreme Court to reverse the ruling of a federal judge to lift the freeze on bank accounts upset the middle classes, setting off a new wave of demonstrations and lootings, this time directed against the Supreme Court, the government, Congress, and the banks, which forced the resignation of the whole cabinet (Clarín December 30, 2001). In view of the situation, the president called an emergency meeting of Peronist governors, but only six of the fourteen showed up, and the president eventually resigned. In his resignation note, Rodriguez Saá acknowledged that “some governors, who have not understood the seriousness of the situation, withdrew their support for me” (Congressional Assembly records, January 1, 2002, www.senado.gov.ar). On January 1, the congressional assembly met again to select a new president. This time it decided to extend the mandate of the appointed president until December 2003, to complete de la Rúa’s term. The appointment of Senator Eduardo Duhalde, former governor of the powerful province of Buenos Aires, was the result of an inter-partisan agreement including Radical and FREPASO legislators. A group of Peronist governors, initially reluctant to accept a solution that did not involve elections, ultimately opted to give Duhalde their support. They acknowledged and respected Duhalde’s broad parliamentary coalition but refused to get themselves involved in the business of government. At the time, Duhalde’s cabinet included a couple of non-Peronist positions reserved for the UCR and FREPASO, and a majority of ministers representing his faction (Levy Yeyati and Valenzuela 2007, 60). Acknowledging the fragility of his position, Duhalde evinced a sober and consensual attitude to the government, one that Rodríguez Saá had failed to adopt. He immediately put an end to a decade of convertibility. Duhalde’s devaluation and pesificación of the economy occurred in a context of upheaval, with a “state of alert” on the streets, unemployment and poverty indicators at their lowest historical rates, with the reluctant support of the governors, as well as a cabinet crisis in April 2002, which led the economy minister to resign.12 In June 2002, six months after the start of his presidency, just when the economy was showing timid positive signs

Presidential Breakdowns in Argentina

67

and an agreement with the IMF seemed likely, the horizon darkened when two protesters were killed on the street by police forces. Fearing that the episode would revive social violence and threaten his leadership of the political transition (La Nación July 3, 2002), Duhalde announced early presidential elections for March 2003 and his resignation for May 25, 2003 (Decree 1399/2002). The decision triggered the presidential succession within a context of a highly fragmented Peronist party. After several months of internal confrontations, the party’s primaries were cancelled, after which three Peronist candidates and other opposition candidates competed for the presidency. Néstor Kirchner, Duhalde’s preferred candidate, was proclaimed president with a narrow 22.24 percent of the vote after Menem dropped out of the race fearing defeat in the second electoral round.

Presidential Breakdowns: Causes and Aftermath Much of the scholarly debate on the presidential breakdowns in Argentina has revolved around two interpretations: one emphasizing the problems posed by the institutional design; and another maintaining, on the contrary, that weak institutional enforcement is at the heart of the problem of instability. The “institutional view” is that it was technically impossible for Alfonsín and de la Rúa to form congressional majorities because of the institutional features of the political system: the constitutional overrepresentation of the least populated provinces, the nature of electoral rules, and the geographic distribution of votes. Together, it is argued, these factors resulted in the Peronist Party more easily winning seats in the upper and lower congressional houses and provincial governorships. In other words, Peronists have an electoral advantage in congressional and provincial elections regardless of which party controls the national executive (Calvo and Murillo 2005, 213–214). For these authors, the peculiarity of the Argentine system is that minority governments and consequent executive instability only occur under non-Peronist rule, a puzzle they have labelled “the new iron law” of Argentine politics. From another institutional standpoint, Spiller and Tomassi observe that the configuration of the country’s political rules—mainly the unilateral resources of the presidency, the electoral system, and federalism—and the dynamics of provincial-federal relations, in particular, shorten the horizon of Argentine political actors and generate policy and executive instability (Spiller and Tomassi 2007, 119). The other interpretation is that Argentine institutions cannot be taken as explanatory factors because they suffer from serious enforcement limitations (rules are not complied with in practise) and from chronic instability (they do not survive fluctuations in the distribution of power and preferences) (Levitsky and Murillo 2005; O’Donnell 1994). In low-institutionalized democracies such as Argentina’s, the argument goes, presidents do not govern through

68

Mariana Llanos

congressional majorities, as Congress and courts are seen as nuisances and thus disregarded (O’Donnell 1994, 60). Instead of searching for political consensus, presidents facing crises often resort to unilateral strategies to implement economic decisions. At the same time, society expects the president to solve problems and provide socially valued results within the time period of a presidential term (Ollier 2008, 77). Failing to do so exposes presidents to social criticism and causes presidencies to breakdown, especially if low levels of presidential popularity are converted into social mobilizations (Ollier 2008, 86; Perez-Liñán 2008, 112–113). There is truth in both the institutional and noninstitutional explanations of the Argentine crises. As the institutional literature suggests, institutional resources can cushion the impact of critical situations, while the perils inherent in minority government are latent and can be activated at specific junctures. In normal times, working with Congress implies that the president should dictate policy directions and manage the legislative activities (of the chamber(s) in which he has a majority), while legislators should negotiate legislative support for presidential proposals—often following the instructions of provincial party bosses—in exchange for a say on what is placed on the legislative agenda, on the features of the policy in question, and on more concrete benefits such as transfers, subsidies, government posts, and pork (Eaton 2005; Jones and Hwang 2005; Llanos 2002; Mustapic 2002). However, as shown above, there are situations in which interbranch relations are threatened by gridlock: first, under critical economic conditions—when there is more to adjust than to distribute—political deals can become too costly for national fiscal accounts, and the government may begin to regret the lack of a sizeable number of loyal legislators; second, when a defeat in midterm elections questions presidential leadership and conflicts begin to outweigh cooperation in the relationship between the president and his party and the opposition. Notwithstanding institutional constraints, presidents have convictions and skills that can either mitigate or compound their objective weaknesses. As already noted, the “subjective resources” of the president (Ollier 2008, 77) mark the pace of politics; and whatever the constraints imposed by any given structure, a capable governmental team can nonetheless have some control over the nature and tempo of policies (Bonvecchi 2004, 90), and this applies to the demise of a presidency as well. The different performances of Alfonsín and de la Rúa may also be of some interest for the comparative literature. First, in the case of de la Rúa, the resignation of the vice president can be seen to broaden the scope of “scandal” factors, since the corruption scandal that culminated in Álvarez’s resignation affected figures close to the president and dealt a hard blow to the credibility of the government. In my view, this and other intra-government crises, which are overlooked in comparative studies, suggest that the errors and weaknesses of presidents may also contribute to presidential instability. Second, regarding the case of Alfonsín, existing analyses highlight the role of the minority status of

Presidential Breakdowns in Argentina

69

governments, adverse economic contexts, and lootings in presidential breakdowns, but they fail to show, for instance, how Alfonsín’s commitment to the reestablishment of democracy may also have played a role in his downfall. This commitment may have shaped his cabinet politics (see the section on Alfonsín above), played a role in his unwillingness to adopt tough economic decisions (Machinea 1993, 137), and shaped a range of “survival-oriented” decisions (such as the nomination of presidential candidate Angeloz two years early, the call for anticipated elections a year early, moves to work with recently elected president Menem, and the transition pact). Regarding this last point, as a counterpoint to the concepts of presidential breakdown (and equivalents) and presidential challenge (attempts to depose a president), I see no reason we cannot speak of counterbalancing presidential “survival strategies,” looking at the problem from the point of view of the president. Finally, the 2001 crisis brought about the breakdown of two more presidents, neither of whom had a problem with minority status in Congress or the typical troubles of dealing with “dual legitimacy” problems, since both were appointed by parliament. On the one hand, the Peronist breakdowns can be regarded as the tail-end effects of the huge crisis that put an end to the convertibility scheme. Similarly, the parliamentarization of the political regime in these circumstances can be seen as a flexible response to the challenges posed by a context of crisis. On the other hand, these breakdowns may have longer-term features with both negative and positive connotations. The manipulation of the electoral calendar and the threat of resignation have come to broaden the arsenal of weapons that presidents can use to organize political support and maintain control over governments in critical moments, provided that a history of resignations has rendered credible such a threat. The fact that we really do not know whether these parliamentary-like features are here to stay is probably because they can also be associated to the well-known phenomenon of the low level of internal institutionalization of the Peronist party, and its lack of respect for formal institutions (Levitsky 2005). From a positive angle, however, early resignation is now part of the political memories of citizens and presidents as a step that may be taken, regardless of the fixed-term constitutional rules, and they may be the most important way that presidents have found to get around the problem of poor or fading political support.

Notes * I am very grateful to Ana María Mustapic, Leiv Marsteintredet, and Vicente Palermo for their insightful comments on previous versions of this chapter. Any remaining errors are mine alone. 1. In contrast with other cases in the region, there are many varied and interesting accounts of the causes of the Argentine crises (some covering both crises). See, for instance, various chapters in Lewis and Torrents (1993) on the Alfonsín

70

2.

3.

4.

5.

6.

7.

8.

9.

Mariana Llanos crisis; the summary and critical assessment of the de la Rúa crisis in Bonvecchi (2006); and the collection edited by Levitsky and Murillo (2005). Other works are referred to throughout the chapter. Important bills were negotiated in Congress, such as the reform of the tax shared between the federal state and the provinces, and the two laws passed to appease military unrest. Calvo documents that Alfonsín initiated his presidency with an extremely high approval rate for government bills (almost 80 percent), but that this fell sharply toward the middle of his mandate (Calvo 2007, 268). More broadly, Calvo demonstrates that the Argentine Congress closely follows changes in public opinion and that the popular approval of presidents influences their success in congress, regardless of the size of the presidential contingent in congress. Alfonsín had wanted to incorporate parliamentary features into the Argentine constitution. Remarkably, he had been fond of the idea of including the figure of a prime minister, who could act to defuse crises. The decision of the Bank to withdraw support for the Argentine government was allegedly promoted by the mission to Washington in January 1989 of Domingo Cavallo (Palermo and Novaro 1996, 112). The Convertibility Plan was meant as a disciplinary measure that would eliminate monetary discretionality, control inflation, and enhance investors’ confidence. The plan basically pegged the peso to the U.S. dollar, establishing one-to-one parity and full convertibility between the two currencies. To make this work, the discretionary powers of the Central Bank were curtailed, and the Bank was required to keep foreign reserves in an amount equivalent to the domestic monetary base. The immediate result was positive: the monthly inflation rate quickly decreased. But the currency board also imposed a straitjacket on policy, mainly because it ruled out devaluation as an option. He was certainly distrustful because he did not have the leadership of his party, which still remained in Alfonsín’s hands, but, as his officials have noted, he also failed to provide the government with direction, and he “refused to commit himself until it became unbearable,” avoiding intervention or decisions, even when the situation was extremely serious, until he was aware of the explicit plans of the others involved (Novaro 2002, 25). He had a low profile and sober public image, and he was seen as a moderate, although one with conservative views. Disregarding Álvarez’s demands for an inquiry, de la Rúa’s cabinet reshuffle on October 5, 2000 left the suspects untouched. See, for instance, Morales Solá (2001) and Granovski (2001) for details about the conflict. The three deputies belonging to the Democratic Socialist Party (Partido Socialista Democrático, PSD) formed a separate bloc in October 2000 (La Nación October 12, 2000); the one member of the Christian Democratic Party left FREPASO (and so the Alliance) in November (La Nación November 22, 2000); when Cavallo took up his post as minister of economy in March 2001, another ten left the Alliance. Shortly after the 2001 midterm elections, the only FREPASO minister still in government, Juan Pablo Cafiero, resigned, and by the end of October, the FREPASO bloc in the chamber had split from Radicalism, thus putting an end to the Alliance. For some authors, the real institutional constraints to the Argentine presidential figure have not come from Congress, which they regard as an unimportant arena in the policymaking process (Spiller and Tomassi 2007, 7), but from

Presidential Breakdowns in Argentina

71

the prominent individual and collective role played in national politics by the governors. Accordingly, the fixed minimum floor of provincial revenues guaranteed by the national government through fiscal pacts was such a constraint for the latter that it can be attributed central responsibility in the collapse of the Argentine economy and de la Rúa’s government in 2001 (Eaton 2005, 108; Spiller and Tomassi 2007, 98). For others, this argument ignores the explanatory value of another institution—the convertibility regime—the logic of which was the ultimate endogenous cause of fiscal and external vulnerabilities (Bonvecchi 2006, citing Palermo 1995). 10. Auyero (2006) regards the lootings as a “grey zone,” in which the deeds and networks of looters, political entrepreneurs, and police officials meet and mesh. If Peronist brokers and police agents actively participated in the lootings and took advantage of the opportunities afforded by collective violence, the author maintains that evidence is too thin to sustain that they caused them, and that they conspired to depose the president. (see also Auyero’s interview in Página 12, November 25, 2007). 11. For Malamud (2006), the Argentine governors can be held directly responsible for the fall of de la Rúa. He refers to the presidential succession as a “political takeover.” 12. For the economically and politically troubled period under President Duhalde, see Levy Yeyati and Valenzuela (2007).

This page intentionally left blank

4 Parliamentary Solutions to Presidential Crises in Ecuador1 Andrés Mejía Acosta and John Polga-Hecimovich

Over the past decade, Ecuador has witnessed growing social and economic unrest that has led to the ousting of three constitutionally elected presidents and generated severe political instability. Although Ecuadorian presidents are historically familiar with political instability and frequent regime change, the fundamental feature of contemporary political crises is that instability did not lead to regime breakdown (Pérez-Liñán 2007). What is more interesting is the fact that Congress played a critical role in the unfolding of all these crises: first, by endorsing widespread demands for presidential resignation from pressure groups including indigenous organizations, the military, citizens in general, and political parties themselves through a de facto “vote of no confidence”; and second, by appointing a successor to the president’s office. We argue that this type of congressional intervention resembled a parliamentary “vote of no confidence” on the executive, in the sense that a partisan majority removed the executive in office and replaced the government to break a political deadlock. While this quasi-parliamentarian vote allowed greater institutional flexibility, it violated two fundamental principles of presidential regimes: the functional independence of the executive and legislative branches, and the election of executives for a fixed term in office. Understanding the links between the durability of a coalition and government survival in presidential systems has become even more relevant since the 2005–2006 cycle of general elections either produced minority or coalition governments in the majority of Latin American countries. In this chapter, we argue that presidential crises in Ecuador were linked to, and accelerated by, the rapid erosion of legislative coalitions. We claim that institutional reforms intended to strengthen presidential authority at the

74

Andrés M. Acosta and J. Polga-Hecimovich

expense of congressional influence in the mid-1990s affected the duration of legislative coalitions. Furthermore, the rapid erosion of government coalitions has been a critical factor precipitating the demise of three presidents in the past decade. Other factors usually associated with presidential crises, such as the role of the military, indigenous and popular mobilization, and economic crises, contributed to it but were not decisive factors. We examine coalition duration patterns in Ecuador and focus on the nature of coalitions (programmatic or clientelistic), the composition of their membership, and the nature of the electoral cycle. In the first section, we explore the links between coalition formation, durability, and presidential instability in Latin America. In the second, we put forward our argument that the adoption of institutional reforms undermined the duration of legislative coalitions. In the next section, we outline the links between the erosion of legislative coalitions and government removal in our analysis of the ousting of presidents Bucaram (1997), Mahuad (2000), and Gutiérrez (2005), and briefly comment on the survival of the three succeeding caretaker presidents. In the last section, we explore other factors contributing to presidential crises.

Coalition Formation, Erosion, and Government Instability in Latin America Nearly two decades ago, prominent scholars warned against the economic and political perils associated with highly fragmented presidential regimes (Linz 1990; Mainwaring 1993). The underlying premise was that presidents who lacked single party support in the legislature were more prone to situations of policy deadlock, which were, in turn, associated with poorer economic performance, citizen discontent, and greater regime instability. Successive electoral cycles have confirmed that minority governments have not only survived over time but have become a very common form of democratic government in Latin America (Zovatto 2008). In most such instances, presidents have been able to assemble explicit or implicit government coalitions with opposition parties. Existing explanations claim that opposition parties have three types of incentives to collaborate with a government. First, they can influence policymaking by participating in government cabinets. Multiparty cabinets contribute to the stability of government coalitions and allow opposition parties to benefit from government policy successes (Amorim Neto 2002). Second, opposition parties are likely to form government coalitions if there is a prospect of future reward for cooperation. Thus, legislators will be more likely to support government coalitions if their own political future is dependent on the government’s survival and policy success (Carey and Shugart 1995); conversely, cooperation incentives will disappear as new elections approach, especially if an executive faces term limits or its popularity erodes (Coppedge 1994; Altman

Presidential Crises in Ecuador

75

2000; Pérez-Liñán 2007). A third argument, found in a more recent body of literature, focuses on the governments’ ability to distribute and allocate selective incentives or investments (pork) to individual legislators as a common mechanism to form and sustain legislative coalitions. The distribution of clientelistic payoffs to government allies may help sustain policy coalitions provided that there are effective mechanisms to monitor the transparency of transactions and to ensure coalition compliance (Pereira and Mueller 2004; Alston, Melo et al. 2006). These explanations offer valid but static understandings of coalitionmaking patterns; they are insufficient to explain why some presidents are more likely than others to form partisan cabinets; why some coalitions are formed around particularistic rather than programmatic arrangements; or why some government coalitions remain stable over the presidential mandate whereas others are affected by the tyranny of the electoral cycle. More importantly, it is not only relevant to ask if the erosion of government coalitions may lead to policy deadlock, but also to ask whether eroding coalitions may contribute to regime instability as well. There is no definitive empirical evidence showing that minority presidential governments tend to induce deadlock situations more often than their parliamentary counterparts, or that parliamentary democracies are more fragile than presidential ones (Cheibub 2002). Political events in Latin America over the past two decades suggest that there is an increasing number of parliamentary responses to presidential crises, in the sense that opposition parties may be able to formally or informally “dismiss” the government to overcome policy deadlock without necessarily causing a regime breakdown (Pérez-Liñán 2007). On the basis of an empirical survey of third-wave Latin American democracies, Marsteintredet and Berntzen (2008) argue that presidential interruptions led by Congress show “a flexibilization” of two defining presidential features: the government’s independent origin, and the independent survival of the executive and legislative branches. The new phenomenon of Congress-led presidential interruptions raises two questions relevant to the scholarly literature: Why and when do government coalitions erode in the first place? And when do eroding coalitions contribute to presidential instability? The first question shifts attention away from coalition formation to explain the sources of coalition survival. We claim that there are two elements that help explain the durability of a coalition or its erosion in a fragmented setting: first, the finite or limited availability of coalition incentives; and second, the perceived value of coalition currencies by coalition partners over time. The first dimension refers to the nature of political bargaining between presidents and opposition parties, exploring, for example, whether government coalitions are formed around programmatic or ideological power-sharing arrangements, or whether coalitions are sustained through ad hoc particularistic payoffs made to individual legislators. Programmatic

76

Andrés M. Acosta and J. Polga-Hecimovich

coalitions tend to be more resilient over time, but when a coalition collapses, new deals are harder to achieve given the limited space to make new policy concessions to other parties, or because of the limited availability of coalition incentives (cabinet positions or patronage). The second dimension takes into account the expected value of cooperation payoffs according to legislators’ own time horizons. At the start of the presidential mandate, coalition incentives offered by the government are more valuable to parties and legislators, given their long-term cooperation frameworks (and lower discount rates). As time elapses and new elections approach, the value of coalition incentives depreciates and impatient legislators have higher discount rates. This is especially true when legislators and/or presidents face term limits. Thus, we argue that a legislative majority is likely to mobilize against the government when pivotal legislative actors perceive that available cooperation incentives (cabinets, patronage, or pork) are insufficient to justify their voting with the government. A second cause of coalition breakdown occurs when a party’s perceived value of current payoffs is lower than the expected utility of bargaining with a new government.

Institutional Reforms, Weak Coalitions, and Eroding Cooperation in Ecuador A consistent challenge for Ecuadorian presidents has been to form reliable legislative coalitions to promote policy change. Since the transition to democracy, multiple institutional reforms to the electoral system, political parties, and of the balance of powers have been adopted with the goal of strengthening presidents’ legislative influence. Given the depth and magnitude of reforms, there are two distinct coalition-making periods. The first period extends from the country’s transition to democracy until the end of the Durán Ballén administration (1979–1996), and this is characterized by a relative availability of coalition-making incentives, although legislators are constrained by the short-term horizons. The second period goes from the start of the Bucaram administration until the end of the Gutiérrez-Palacio government (1996–2006) and features the opposite combination of longerterm horizons for political agents with fewer coalition currencies available. The next section focuses on the impact of institutional reforms on coalition maintenance. 2

The Making of Ghost Coalitions (1979–1996) The proliferation of congressional parties after the democratic transition posed significant challenges to the formation of government coalitions. Despite the highly contentious nature of policymaking, presidents were able to form coalitions for reform with political parties and individual legislators.

Presidential Crises in Ecuador

77

To this end, presidents relied on a broad array of formal and informal incentives to secure, albeit temporarily, the support of coalition partners. Some of these mechanisms included the use of constitutional powers (such as decree or veto authority), the allocation of cabinet positions and government postings to allied parties, and the more informal distribution of particularistic payments, from discretionary spending accounts (gastos reservados) to finance—in principle—legislators’ investment projects (Mejía Acosta 2009). The “value” of these coalition incentives eroded as parties—and legislators—perceived an increasing liability of cooperating with the governments’ agenda, especially as new elections approached. 3 To sustain political cooperation, government and opposition parties crafted “ghost coalitions” or clandestine alliances, which enabled them to trade government benefits for legislative support4 while maintaining a public discourse of political independence (Mejía Acosta 2009). Although these informal practices did not necessarily produce long-term or programmatic exchanges, they allowed politicians to break policy deadlock and promote significant reforms. Coalition compliance was secured thanks to a series of informal “checks and balances” between the government and the opposition, such as legislative threats to impeach cabinet ministers, or presidential dismissal of government officials from opposition parties. This informal but functional coalition-making equilibrium came to an end during the Durán Ballén administration. Having won the 1992 general election, President Sixto Durán Ballén led a conservative government composed of the Republican Union Party (Partido Unión Republicana, PUR) and the Ecuadorian Conservative Party (Partido Conservador Ecuatoriano, PCE). The PUR-PCE alliance, which retained only eighteen of the seventyseven congressional seats, sought the support of the party they had defeated in the runner-up election, gaining a further twenty-one seats from the conservative Social Christian party (Partido Social Cristiano, PSC). For the next two years, the conservative coalition promoted an ambitious agenda of market-oriented reforms—including the Public Sector Budgets Law, tax reform, the State Modernization Law, deregulation of the intellectual property rights and foreign investment regimes, the reform of the financial system, a Capital Markets Law, a Hydrocarbons Law, and an Agrarian Reform Law. In exchange, the PSC demanded access to state resources, including budgetary allocations in excess of US $100 million for the PSCcontrolled provinces of Guayas, Manabí, Los Ríos, and Esmeraldas, and direct cash transfers from the executive’s discretionary spending fund to PSC legislators. The PSC also played a predominant role in the congressional nomination of Supreme Court judges and members of the Electoral Tribunal (Dahik 1995; Mejía Acosta 2006). By 1995, Vice President—and chief coalition manager—Dahik believed that PSC demands vastly exceeded the contents of the original, secret agreement and defended that government resources were not a “bottomless pit”

78

Andrés M. Acosta and J. Polga-Hecimovich

(Dahik, author interview (AI) 2006). After failed conversations to renew the terms of the alliance with the PSC, Dahik decided to “go public,” and suggested during an interview that Supreme Court justices linked to the PSC had sought bribes from the executive to rule in favor of government interests. The strategy backfired and Febres Cordero’s PSC party filed a legal action against Dahik to accuse him of the illegitimate use of millions of dollars in government funds to purchase the loyalties of—ironically—his own party members! (NotiSur October 13, 1995). The congressional impeachment process acquitted Dahik of any wrongdoings, but the political scandal harmed both sides. 5 Dahik resigned shortly thereafter and sought asylum in Costa Rica to end what he considered a political persecution.6 The corruption scandal also affected the PSC’s political reputation and undermined its performance in the 1996 general election. The corruption scandal prompted the government to adopt constitutional reforms (approved through referenda in 1995) to curb the legislators’ ability to demand and channel budgetary allocations for their provinces and to reduce the presidential use of discretionary off-budget spending accounts (gastos reservados). These reforms, intended to reduce the availability and discretionality of coalition incentives, further obstructed coalition formation and accelerated legislative erosion in the following years. For example, the approval of budget allocations by sectors instead of by provinces made it more difficult for presidents to target government support to specific parties and legislators in exchange for votes. Furthermore, the abolition of term limits for legislators but not for presidents created long-term cooperation incentives amongst legislators, but not necessarily with the government. In the following section, we argue that this set of contradictory reforms affected coalition dynamics.

Failed Institutional Reforms and Increased Coalition Erosion (1996–2006) The second period under analysis is defined by an interesting coalitionmaking paradox. The institutional reforms adopted in the aftermath of the Dahik corruption scandal not only failed to strengthen presidents’ policymaking power significantly but also eliminated key legislative currencies that had helped policymakers overcome policy deadlock. Some constitutional reforms sought to increase presidents’ agenda-setting powers by granting the president additional veto and urgency legislation (decree) powers, or by protecting cabinets from legislative impeachments. Other reforms sought to curb corruption incentives by eliminating legislators’ ability to bargain district specific budgetary allocations (pork), and allowed the dismissal of legislators who defected from their parties (usually to join a government coalition) or peddled political influences.

Presidential Crises in Ecuador

79

These reforms severely undermined the presidents’ power to bargain coalition exchanges and constrained the ability of political parties to deliver public goods to their constituents. In the absence of long-term coalition incentives, new coalitions demanded greater bargaining effort, including the revival of clandestine coalition-making practices, such as party-switching. Under the new dispensation, coalitions became significantly shorter and more fickle than those formed during the previous period (Mejía Acosta 2009). Overall, greater conflict in executive-legislative relations ensued, thus creating the conditions for the advent of what some authors have called an “anti-party moment” in Ecuador’s democratic history (Freidenberg 2008). Table 4.1 summarizes some key indicators of legislative performance between 1979 and 2006. The first two columns show the proportion of (ordinary) executive-initiated bills and the proportion of executive economic urgency laws that were approved by Congress.7 Although this measure cannot reflect the quality or amendments of adopted bills, it is nevertheless a reasonable measure of presidential efforts to influence the policy agenda (Morgenstern 2002; Saiegh 2008). The data confirms the expectation that presidential success rates are nearly doubled when presidents use urgency laws (decrees) compared to ordinary bills (an average of 66 to 36 percent), but success rates remain virtually the same before and after the adoption of reforms (35 to 38 percent for ordinary bills, and 68 to 64 percent or executive decrees).8 Although success rates remain similar in both periods, it is interesting to note that executives dramatically changed their strategies after 1996, from initiating ordinary bills to legislating by decree. The third column shows that presidents initiated fewer ordinary bills after 1996 (a drop from 9.3 to 6.3 bills per year) and compensated their legislative firepower with decrees (nearly a twofold increase, from 3.2 to 5.9 bills per year). The interpretation is that presidents used more decrees to compensate their legislative weakness, but the strategy probably backfired as legislators became more reluctant to cooperate with an imperial president. Table 4.1 Coalition duration and proxies by time period, 1979–2006 Period

Share of Executive-led bills (total)

Share of urgency decrees (total)

Bills/ decrees Initiated per year

Average coalition duration (months)

Cabinet Partyturnover switching ratea rate b

1979–1995 1996–2006

0.35 0.38

0.68 0.64

9.3/3.2 6.3/5.9

19.2 12

0.52 0.58

0.12 0.09

Total/ Average

0.36

0.66

12.6/12.3

15.6

0.54

0.11

Notes: (a) Cabinet data is available until 2003; (b) Party-switching data is available until 2002.

80

Andrés M. Acosta and J. Polga-Hecimovich

The next three columns confirm the notion that coalition-making in Ecuador became a costlier endeavor after 1996, as reflected by short-lived coalitions, higher rates of cabinet turnover, and constant levels of partyswitching. In the Ecuadorian context, we define a government coalition as the formal or informal agreements arrived at between the president and opposition parties, formed around a specific agenda of policy reforms as reported by the press.9 The fourth column shows that coalitions after 1996 were short-lived compared to the previous period: the durability of the average coalition fell from 19.2 months before 1996 to just 12 months in the second period. The cabinet turnover rate, measured in number of ministers per month, shows the share of ministers who abandoned office either because they were impeached by Congress, resigned, or were removed by the president.10 The turnover rate is used as an indicator of coalition volatility insofar as a higher turnover reflects more frequent bargaining with opposition parties, and vice versa. Following the institutional reforms in 1998 that introduced higher barriers for Congress to impeach and dismiss government ministers, we would expect greater ministerial stability in the second period. Despite institutional efforts to shield ministers from political volatility, the turnover rate shows a 11.5 percent increase, from 0.52 ministers per month pre-1996 to 0.58 ministers per month post-1996. All things being equal, higher cabinet volatility reflects a greater political struggle between the executive and the legislative opposition to form and renew policy coalitions. The last column in table 4.1 looks at the incidence of party-switching as a proxy of coalition-making. It is expected that individual defections (camisetazos) would become less frequent as policy coalitions tend to stabilize over time. The switching rates are calculated by dividing the number of members of Congress who switched parties over the total number of deputies for each year. The data shows that camisetazos experienced a very small decrease, from an annual rate of 12 percent between 1979 and 1996 to 9 percent in the following period. Again, this is a very minor change considering that specific legislation was adopted in 1998 to sack legislators who “sold” their votes in exchange for particularistic rewards, usually from the government. But more importantly, the persistence of this informal practice shows its critical role for cementing government coalitions when governments lack reliable support from political parties (Mejía Acosta 2009).

Did Coalition Erosion Contribute to Presidential Demise? We claim that the lack of coalition incentives and faster-eroding coalitions after 1996 increased the probability of policy deadlock between the government and the legislature, but it also increased the likelihood that an adversarial congressional majority would dismiss and replace the government. This

Presidential Crises in Ecuador

81

section discusses three cases of presidential crises in Ecuador between 1997 and 2005. Paradoxically, these instances of political instability came after the adoption of institutional reforms intended to strengthen the presidents’ policymaking influence over Congress. While each of the crises needs to be fully understood in their specific political context and in light of their interaction with other critical variables (Pérez-Liñán 2007), we claim that in all cases Congress played a contributing role to initiate the governments’ demise or eventually formalize their fall. To complement the argument, we discuss why the other three (interim) presidents were not affected by Congress.

Bucaram’s “Mental Incapacity” The first case of termination of a presidential mandate led by Congress took place in February 1997, when President Abdalá Bucaram was ousted from office by a congressional resolution only seven months after his inauguration. Bucaram, the traditional leader of the Roldosista Party (Partido Roldosista Ecuatoriano, PRE), won the 1996 presidential election as a political outsider with an eloquent antielite, antiestablishment discourse, gaining only twenty of seventy-eight seats in Congress (Freidenberg 2003). Minority status forced him to seek support from the largest and most cohesive legislative bloc in Congress, the conservative PSC. As with the 1992 clandestine agreement with Durán Ballén, the PSC vehemently denied any form of government collaboration but played a predominant role in electing Supreme Court judges and voted with the government to approve its Electric Sector Regime Law.11 PSC leader Heinz Moeller insisted that a PRE-PSC alignment on a hotly debated law was simply a “favourable coincidence” (1996a). The clandestine coalition fell apart on November 1996 when the PSC openly opposed the government’s proposal to raise taxes and end value-added tax (VAT) exemptions in the 1997 fiscal budget (1996b). Traditionally, the Budget Committee was in charge of amending the presidents’ budget before it was voted through by a simple majority in a thirty-five-member plenary chamber (plenario de las comisiones legislativas), a permanent legislative chamber made up of the five most important committees. Given that Bucaram could no longer use discretionary spending, or offer budgetary concessions to the “traditional” parties (such as the PSC) represented in the plenario, he decided to bypass them and push for budget proposal with the help of maverick legislators. Bucaram co-opted six out of seven Budget Committee members and secured the approval of his budget with eighteen of the plenario votes, eight from Bucaram’s party and another ten “independent” legislators. The traditional parties, which were uncomfortable with the idea that a self-portrayed “outsider” was unwilling, or unable, to negotiate agreements with them, formed a “Patriotic Front” to oppose Bucaram’s government. The Front rejected several fiscal austerity measures and price increases on basic goods (electricity, gas, and phone services) promoted by the president. The proposed economic adjustments, coupled with repeated accusations of

82

Andrés M. Acosta and J. Polga-Hecimovich

government corruption and nepotism from different sectors of society, triggered a series of popular protests calling for the president’s resignation. Led by the PSC, the parties in the legislature were keen to attack Bucaram but realized that they lacked the necessary two-thirds majority to impeach and dismiss him, so they resorted to a legislative resolution that declared the president “mentally incapacitated” to govern. Bucaram was ousted by a controversial vote of forty-four against thirty-four on February 6, 1997, and after an initial power struggle around presidential succession, the president of Congress, Fabián Alarcón, was declared interim president (Pérez-Liñán 2007).

Mahuad’s “Perfect Storm” Following an intense period of political instability and constitutional reform, Jamil Mahuad won the 1998 presidential election with 51 percent of the popular vote over banana tycoon Álvaro Noboa. No one expected that this Harvardtrained technocrat, strengthened with newly endowed presidential powers, would encounter a “perfect storm” scenario, that would end his government only eighteen months later. Mahuad’s party, Popular Democracy (Democracia Popular, DP) won 35 of 123 congressional seats, and, like his two failed predecessors Dahik and Bucaram, the new president sought the support of the PSC to assemble his governing coalition. Mahuad favored the alliance because the PSC was ideologically close, sharing a similar vision of the need to implement market-based economic reforms, but also because the party offered a reliable source of support in the legislature (Mahuad, AI 2002). The DP-PSC coalition, or the “steam roller” (aplanadora) as it became known in the media, was instrumental in producing a legislative vote ratifying the Ecuador-Peru peace treaty in October 1998; a PSC-led initiative to adopt a 1 percent tax on all financial transactions (the Impuesto a la Circulación de Capitales, ICC) to help finance public works projects in PSC districts (such as Malecón 2000 in Guayaquil); and the government’s Tribute and Finance Reform, which included new taxes, budget balancing, and reforms to the Agency of Deposit Guarantees (Agencia de Garantía de Depósito, AGD). The coalition jointly appointed the Attorney General, the People’s Attorney or Ombudsman, the Banking Superintendent, and the directorate of the Electoral Tribunal. The initial challenge to the PSC coalition came from government party ranks; some believed that this alliance imposed a very high price for the intended direction of reforms. Finance minister Fidel Jaramillo, for example, resigned when the government sided with the PSC to oppose fiscal reforms (a rise in VAT from 10 to 15 percent) that were needed to finance the mounting fiscal deficit. The handling of the economic crisis (the Asian financial crisis, falling oil prices, and the collapse of the Ecuadorian banking system) further polarized policy differences inside the government. When the government had to declare a banking holiday in March 1999 and imposed a

Presidential Crises in Ecuador

83

yearlong freeze on accounts to avoid the collapse of the banking system, the supporters of the steam roller coalition disbanded: while PSC allies criticized the governments’ decision to freeze bank accounts, government officials criticized the decision to bailout the banks. In the aftermath of the financial collapse, Mahuad sought to form an alternative coalition with the left to regulate the banking system and restore balanced public finances. However, cementing a reform coalition with the left, namely, the Pachakutik Plurinational Unity Movement (Movimiento Unidad Plurinacional Pachakutik, MUPP), and the Democratic Left (Izquierda Democrática, ID ), became a costlier endeavor as Mahuad had already allocated strategic posts to the PSC (in the Judicial Courts and the Electoral Tribunal), and no other party wanted the liability of collaborating with an unpopular government. As expected, the coalition with the left proved insufficient to ensure the approval of a broader package of economic reforms in June 1999 (Ley Marco). At the end of the year, the president resorted to a last-ditch alliance with Bucaram’s PRE to approve the 2000 budget in exchange for particularistic concessions, including a revision of the penal code to allow the return of the former president from his Panamanian exile. The legislative agreement with the PRE also meant that Mahuad alienated the support of the partisan left and right after only eighteen months in office. The absence of a “legislative shield” made Mahuad more vulnerable to mobilization and protest from different social groups, including doctors, indigenous people, labor unions, and the military, all of which complained about the consequences of the economic crises (Pérez-Liñán 2007). Social discontent materialized on January 21, 2000 when a military-indigenous alliance took the presidential palace and demanded the president’s resignation. The congressional majority assembled in Guayaquil approved a resolution to “accept” a never-proffered resignation from Mahuad and, in accordance with article 168 of the constitution, appointed Vice President Gustavo Noboa as his constitutional successor.

Gutiérrez’s “Forajidos” The rise and fall of the Lucio Gutiérrez administration (2003–2005) follows a strikingly similar pattern, and he met a similar fate to that of his three predecessors. The president invested heavily in a reform coalition with the PSC early on, the subsequent coalition fallout undermined the prospect of crafting a new coalition, and the erosion of last-resort government coalitions ultimately affected the government’s survival. President Gutiérrez was elected in November 2002 under the banner of the 21 January Patriotic Society Party (Partido Sociedad Patriótica 21 de Enero, PSP) in an electoral alliance with the leftist Popular Democratic Movement (Movimiento Popular Democrático, MPD) and the indigenous MUPP. The new government reflected the same indigenous-military alliance that had helped to overthrow Mahuad three years earlier.

84

Andrés M. Acosta and J. Polga-Hecimovich

After a few months in office, Gutiérrez betrayed his campaign allies and sought market-oriented reforms to privatize the petroleum sector, electricity, and telecommunications, as well as fiscal austerity measures to maximize profits from rising international oil prices. Between August 2003 and October 2004, Gutiérrez formed a ghost coalition with the PSC to advance this policy agenda, which included the negotiation of a bilateral free trade agreement with the United States, the approval of the 2004 fiscal budget, and a PSC-led reshuffle of Supreme Court judges. The nature of the PSP-PSC agreement only became evident when a series of alleged corruption scandals and embezzlement accusations between Gutiérrez’ cousin, Renán Borbúa, and the PSC’s leader León Febres Cordero were revealed.12 Replicating a decade-long strategy, the PSC led an opposition alliance with the ID and the MUPP to start impeachment proceedings against Gutiérrez for corruption (initiated by the PSC) and for jeopardizing state security (initiated by the MUPP and the ID). The congressional opposition lacked the two-thirds majority required to dismiss the president, but the government nevertheless tried to form a legislative shield with the support of Bucaram’s PRE and Álvaro Noboa’s Institutional Renewal Party of National Action (Partido Renovador Institucional de Acción Nacional, PRIAN).13 In the absence of “conventional payoffs” such as budgetary allocations or government offices, Gutiérrez agreed to reshuffle oversight authorities for the benefit of his new political allies. However, the very notion of “judicial independence” enshrined by the 1998 constitution ultimately constrained political bargaining between the government and the opposition, as judges had lifelong appointments. So on December 9, 2004, Gutiérrez illegally removed all of the Supreme Court justices (including some alleged PSC associates) and replaced them with political cronies sympathetic to the PSP, the PRE, and the PRIAN. He also reorganized the Constitutional and the Supreme Electoral tribunals. The new president of the Supreme Court dropped charges against former president Abdalá Bucaram in March 2005, allowing him to return from exile. These constitutional violations and Bucaram’s return angered a broad social segment including urban middle class protesters (forajidos), who rejected Gutiérrez’s authoritarian style. In the context of widespread protests, the Armed Forces Joint Command publicly declared their withdrawal of support for the president. A legislative session was convened by opposition parties on April 20, and it voted 60-0 (with two abstentions) to declare that there had been an “abandonment of office” and it appointed Vice President Alfredo Palacio as interim president. Gutiérrez initially sought political asylum in several neighboring countries, but eventually he returned to Ecuador to reclaim an active role in politics.

The Caretakers During the period under analysis, weak and eroding legislative support was a common contributing factor for the early termination of all three elected

Presidential Crises in Ecuador

85

presidents. Two of the corresponding interim presidents, Fabián Alarcón (1997–1998) and Alfredo Palacio (2005–2006), survived in office due to the specific and short-term nature of their mandate.14 Neither had sufficient time or incentives to push for ambitious reforms until the next election (less than eighteen months later), so they both opted to “muddle through” and did not send any substantive pieces of legislation to Congress. The short window of time gave the opposition enough time to organize and prepare for the next election without much concern for the interim government. In the case of Gustavo Noboa (2001–2003), he inherited a concrete and ambitious mandate from Mahuad: to implement and promote the economic dollarization package. To fulfil this mandate, Noboa used the limited coalition toolbox available: he formed a center-right coalition (including the PSC, the DP, the Alfarista Radical Front (Frente Radical Alfarista, FRA), the Concentration of Popular Forces (Concentración de Fuerzas Populares, CFP), and the PRE) to approve the first dollarization segment (known as Trole I); he approved Trole II with the support of an informal party organization made up of independent legislators;15 and he failed to form a working coalition to pass Trole III in 2001. Aware of his limited political capital, President Noboa decided to muddle through legislative politics and to avoid initiating significant new legislation during his last year in office.

Alternative Explanations of Presidential Instability In his comparative analysis of presidential crises in Latin America, PérezLiñán (2007) appropriately warns against assuming that the only relevant actors in presidential impeachment are the executive and the legislature. Indeed, contemporary accounts of presidential crises have explored other contributing factors, including the intervening role of the military, the mobilization of indigenous or broad-based protests, the emergence of investigative journalism, the mobilization of citizens against alleged corruption cases, the presence of economic crises, and (failed) government efforts to adopt economic reforms (Hochstetler 2006). We acknowledge that such factors affected presidential demise in different ways and times, and that none of these is sufficient to explain executive removal. However, we argue here that the erosion of legislative coalitions was a common factor that systematically provoked executive-legislative conflict and affected presidential survival in Ecuador after 1996. Students of civil-military relations in Latin America agree that contemporary military interventions have not been the cause of presidential crises, but have rather been an institutional mechanism to restore civilian order. Analyzing the role of the military during the 2000 coup against Mahuad, Samuel Fitch argues that military intervention occurred because of unique situational factors such as the fiscal and banking crises, corruption scandals allegedly involving the president, threats to some military interests, and the

86

Andrés M. Acosta and J. Polga-Hecimovich

erosion of the democratic principle within one faction of the military hierarchy. Fitch goes on to argue that military coups à la Mahuad are likely to be a last-resort strategy, while civilian coups and quasi-coups are more likely (Fitch 2005). The subsequent crisis that culminated in the removal of Gutiérrez in 2005 suggests that the armed forces are more likely to play a mediating and arbitrating role in times of political uncertainty, rather than actively promoting the removal of civilian leaders (García Gallegos 2005). These interpretations are consistent with the findings of Pérez-Liñán (2007), who notes that since the end of the Cold War, armed forces across Latin America have played a role in shaping the outcome of presidential crisis without provoking the crisis themselves. A second explanation for presidential instability has to do with the negative effect of economic crises. However, such interpretations fail to explain whether it is an actual economic recession or rather the (usually failed) government efforts to adopt unpopular market-oriented reforms that trigger political crises and ultimately the premature ousting of presidents. It is wellknown that economic crises undermine the survival of young democracies, as they lead people to become more risk-accepting and become more willing to overthrow a regime (Acemoglu and Robinson 2006; Haggard and Kaufman 2005; Przeworski, et al. 2000). Weyland (2002) used insights from cognitivepsychological theory to argue that risk-seeking or risk-accepting individuals (those most affected by economic downturns) are more likely to support governments carrying out draconian market-oriented reforms. The Ecuadorian experience does not support either view. The “crisis-led coup” explanation may be applicable to the Mahuad administration, as economic policymaking became unsustainable because of the combined presence of a triple financial, fiscal, and banking crisis between 1999 and 2000. The crisis was triggered partly by a sharp fall of the international prices of Ecuador’s main export, oil, and an imminent run on the banking system. Financial difficulties directly affected government spending, escalated inflation, prompted a freeze on bank deposits, and eroded people’s confidence in Mahuad’s ability to handle the policymaking challenges. However, it is difficult to isolate economic motivations as the source of political instability as there were many other factors that contributed to the “perfect storm” scenario, including military factionalism, deep-seated regional divisions, indigenous mobilization, and mounting congressional opposition. Moreover, the removal of presidents in 1997 and 2005 took place during a moment of relative bonanza when there was macroeconomic stability and moderate policy success. President Bucaram benefited from a period of accelerated economic growth and macroeconomic stability initiated during the previous administration of Durán Ballén, and Gutiérrez benefited from record-high oil prices, macroeconomic stability, and a favourable fiscal balance. A third source of instability is the mobilization of social and ethnic groups. Organized indigenous groups in Ecuador, most notably the Confederation of Indigenous Nationalities of Ecuador (Confederación de Nacionalidades

Presidential Crises in Ecuador

87

Indígenas del Ecuador, CONAIE), have also played important opposition roles against elected presidents, but in no case have these groups been the sole cause of the crisis (Pachano 2005; Pallares 2006). In Latin America, indigenous mobilization has challenged the state through demands to the right to effective citizenship (Yashar 1998). In the past two decades, these movements have succeeded in shaping political debates, demanding economic reforms, and mobilizing cross-class coalitions. More recently, these movements have coalesced into political parties to seek greater political representation and promote institutional change from within the formal political system (Van Cott 2003; 2005). Although indigenous movements led mass protests and organized opposition against all three deposed Ecuadorian presidents, it is difficult to ascertain a direct causal link between indigenous activism and political instability. Rather, activism is likely to be the social response to poor economic performance, increasing levels of poverty, inequality, and state weakness. In the long run, Madrid (2005) argues, the emergence of indigenous parties has the potential to stabilize democratic competition insofar as it broadens political participation, improves political representation, reduces party fragmentation, and deepens democracy.16 In Ecuador, CONAIE played a marginal role during the 2005 protests that brought down the Gutiérrez administration (Pachano 2005), and it is now unlikely that the indigenous movement or corresponding political party will play a destabilizing role in the near future. Some scholars have argued that the presence of other types of social protests and urban mobilization, usually sparked by alleged corruption scandals and/or poor economic performance, have been critical to trigger a presidential crisis (Hochstetler 2006; Pérez-Liñán 2007). The argument is that large street protests or broad social coalitions demanding the president’s removal are likely to prompt congressional action to dismiss the executive. It was emphasized above that economic crises were not a decisive factor, and that there have also been protests during times of economic bonanza. People do react angrily to corruption scandals, but, as shown in this chapter and elsewhere, scandals in Ecuador emerged as a response to (or in retaliation for) the absence of legislative agreements (Mejía Acosta 2006). Here, we argue that coalition breakdown was a common factor explaining presidential removals, either because it led to poor government performance or because it uncovered the corruption scandals that angered street protesters.

Conclusions The inevitable challenge of understanding presidential breakdowns is the need to analyze multiple causal factors with the limitations of having few comparative cases. Ecuador is one of two Latin American countries (Bolivia being the other) where extreme political volatility has provoked a frequent removal of governments: since 1996, three presidents have been ousted from office, and until 2009 no elected president had been able to complete his mandate.

88

Andrés M. Acosta and J. Polga-Hecimovich

Diverse explanations, including the role of social or ethnic groups, military intervention, economic crises, and corruption scandals, have been offered, but none are decisive or systematic across all cases. While it is clear that exogenous factors interact with inadequate institutional arrangements to produce severe forms of political instability (chapter 3, this volume; Marsteintredet 2008a), we focus on a single institutional feature, the rapid erosion of legislative support for the president, to explain how governments become more vulnerable to the impact of a military revolt, a perceived or observed economic crisis, or the demands and anger of organized social groups. We argue that Ecuador’s fragmented democracy resembled a parliamentary regime in that the formation of governing coalitions is critical not just to promote policy reforms and avoid deadlock, but also for presidents to survive in power. Despite important constraints, Ecuadorian presidents were able to form clandestine alliances, or ghost coalitions, with opposition parties to pass policy reforms in exchange or political benefits between 1979 and 1996. Paradoxically, a series of constitutional reforms adopted in the mid-1990s, originally conceived to strengthen presidents’ policymaking abilities and to curb incentives for corruption, in practice blocked the formation of policy coalitions, paving the way for greater political instability. These reforms abolished political incentives for coalition-making and increased politicians’ impatience with unpopular presidents. In the absence of credible coalition incentives, empirical findings suggest that, after 1996, Ecuadorian presidents had to make much greater legislative efforts to maintain or renegotiate the same levels of political support achieved before 1996. The qualitative analysis of presidential removals shows how presidents were unable to avoid the rapid erosion of political support, while the use of informal coalitionmaking practices, such as party-switching, ultimately triggered corruption scandals that affected the credibility of policymakers and made presidents more vulnerable to being removed from office. The resulting poor economic performance, the revelation of corruption scandals, and the arbitrary and unconstitutional ways in which presidential removals were carried out, helped to crystallize citizens’ disgust with the “traditional” political elites. This perception of the electorate was confirmed when voters elected a complete outsider, Rafael Correa and his Country Alliance (Alianza País, AP) movement, to lead the movement against the “traditional parties.” Ironically, Correa’s political reforms have sought to strengthen further presidential policymaking powers, to reduce or eliminate cooperation incentives for legislators, and virtually to eliminate or demobilize the legislative opposition.

Notes 1. Authors wish to thank Pablo Andrade, Mariana Llanos, Leiv Marsteintredet, and Aníbal Pérez-Liñán for their comments.

Presidential Crises in Ecuador

89

2. The distinction between these two coalition-making scenarios, and their implications for policymaking have been explored in Mejía Acosta, Araujo, et al. (2008). 3. Coalitions eroded further as presidents and legislators faced term limits and short-term mandates. Provincial legislators (85 percent of the congress) were elected for two-year periods, whereas presidents and national legislators were elected for four-year terms, in all cases with no possibility of reelection. 4. The term “alliance” is used in Ecuador to describe an informal, short term political agreement between the government and the opposition, or amongst opposition parties. Here, we use the terms “alliances” and “coalitions” interchangeably. 5. A pivotal actor for Dahik’s survival was Abdalá Bucaram’s PRE, which abstained from the final impeachment vote. When asked about his vote, PRE party leader Abdalá Bucaram responded: “Dahik should go to prison for corruption, but it is not the political trial that will put him behind bars, it is the criminal trial” (NotiSur 5, October 13, 1995). 6. Days later, Supreme Court president Carlos Solórzano Constantine (reportedly a PSC crony) ordered Dahik’s preventive detention after the trial, citing possible criminal charges for the illegal use of the discretionary accounts (Weekly News Update on the Americas 298, October 15, 1995). 7. The data is drawn from the Congressional Archives of Ecuador. It analyzes 336 executive-led bills introduced to Congress, of which 117 were labelled as economic urgency laws, the functional equivalent of a decree: if Congress does not approve, amend, or reject the executive bill in fifteen days, it automatically becomes law. Before 1998, Congress had up to thirty days to amend such bills. 8. Some of the more recent legislative data (from 2003 onward) may underestimate executive success if the legislative coding does not include bills that were formally “rejected” by Congress, but the validity of the congressional decision may have been appealed and/or subsequently approved via the Public Ministry. Such information would not be available on the congressional dataset, and would have to be verified against the publication of the bill in the Nation’s Official Registry, as was done for bills adopted between 1979 and 2002. 9. The information on coalitions has been drawn from firsthand interviews with legislators, secondary media sources, including Diario Hoy and El Comercio, as well as independent reports from the Latin American Weekly Report (LAWR) and the Andean Group Regional Report. To calculate the average coalition duration, we divided the total number of months included in the time period by the number of coalitions; by extension, to calculate coalitions per year, we divided that number by the number of years for the time period. 10. We tallied the total number of ministerial changes per year and divided it by twelve months, with the exception of 1979, which was a five-month year. Not included in this count are ministers who prematurely left office after the ousting of a president. 11. On August 30, the partial veto of the “electric law” by the PSC was surprisingly supported by the PRE bloc, which had previously opposed the move. The partial veto meant it was possible to avoid a three-year waiting period in favor of immediate enactment of the privatization law. 12. Borbúa publicly stated on April 13 that, “[PSC Deputy, Xavier] Neira is behind all petroleum contracts, Pacifictel and Andinatel” and accused the PSC hierarchy of belonging to a “dark circle” controlling contract disbursements. Febres

90

13.

14.

15.

16.

Andrés M. Acosta and J. Polga-Hecimovich Cordero’s PSC responded with charges of “trafficking in influences” against Borbúa in the Constitutional Tribunal (Diario Hoy April 14, 2004). The opposition formed by the PSC (25), the ID (15), Pachakutik (7), and the MPD (3) still lacked fifteen more votes necessary to censure the president (Diario Hoy November 4, 2004). The legislative resolution that declared Bucaram was “incapacitated” to govern also contained a specific mandate for the interim presidents to call for a constituent assembly and call elections in 1998. The group, the Movement of National Integration (Movimiento de Integración Nacional, MIN), was mostly formed by party dissidents from the DP and the PSC. Birnir and Van Cott (2007) demonstrate a statistical decrease in legislative fragmentation in Ecuador from before Pachakutik’s appearance in 1995 to the present.

5 Civil Society, Social Protest, and Presidential Breakdowns in Bolivia1 Miguel A. Buitrago

Introduction On the evening of October 17, 2003, while an aide delivered his resignation letter to Congress, President Gonzalo Sánchez de Lozada made his way to the Santa Cruz international airport to board a flight into exile. His second presidency had just ended after only fourteen months in office. In early October, Bolivians had mobilized demanding that the president resign. The government’s heavy-handed response to social protest had cost the lives of too many civilians. Vice President Mesa and the president’s political allies in the governing coalition had withdrawn their support. This was neither the first nor the last time a Bolivian president had been forced out of office in this manner. Hernán Siles Zuazo had experienced a similar fate in 1985, as had Carlos Mesa in 2005. This chapter deals with these three presidential breakdowns. There has been much debate about the roots and causes of presidential breakdowns in a political context in which government but not regime tends to break down (Pérez-Liñán 2007; América Hoy 49, 2008). Coalition breakdowns, minority governments, social protests, presidential scandals, and certain kinds of economic policy have been put forward as explanation of presidential breakdown. I seek to move beyond this debate and ask how the factors contributing to presidential breakdowns interact, and whether any single factor plays a more determining role than all others. I argue that institutional factors were insufficient to cause a presidential breakdown in Bolivia. They are necessary to destabilize a government, but a breakdown only occurs when civil society and social protests play a determining role in forcing presidents to leave office prematurely. In Bolivia, this has happened

92

Miguel A. Buitrago

for two reasons: first, social protest intensifies to high levels as a result of a government’s unwillingness to compromise with protesters; second, a broad coalition of organizations loses faith in the government and decides that it must change. The Bolivian case is exceptional because of the politically active role that civil society has played historically. Bolivian civil society has used strategically and systematically social protest as the main tool to interact with the state. In this chapter, I compare three cases of presidential breakdown since 1982. Although there are four cases in recent Bolivian history, I focus on the fall of Hernán Siles Zuazo, who was president from 1982 to 1985; the fall of Gonzalo Sánchez de Lozada’s during his second presidential term, from 2002 to 2003; and on the fall of Carlos Mesa, who was president from 2003 to 2005. Carlos Mesa is included because, as vice president for Sánchez de Lozada, he was popularly elected, and his mandate was to finish the term that Sánchez de Lozada was unable to complete. I do not include the fall of Hugo Banzer because his early departure from office was due to a terminal illness. The chapter is structured as follows. In the first section, I develop a working definition of social protest and of the intensity of social protest in Bolivia. I also highlight the role that civil society and social protest play in Bolivian politics and society. In the second section, I explore the theoretical links between social protest and presidential breakdown. In section three, I trace the political events of the presidencies under analysis, and in section four, I explore the multidimensional interaction between the different variables advanced in the literature and the instability of these Bolivian presidents. In section five, I analyze the role that civil society and social protest play on presidential breakdown.

Working Definitions and the Protagonism of Civil Society Although this is not the place to delve deeply into a discussion of social movements or collective action, I propose that social protest or direct action is the means for a group of people to exert political pressure on the government (Raschke 1988). Moreover, protest actions generally refer to a state of conflict with the government and have the aim of asserting, or in some situations gaining, political rights. In contrast with Carter’s (2005) definition, 2 I include civil disobedience, demonstrations, confrontations, and violent acts. I consider these as variations in intensity rather than in the nature of protests. I consider two levels of intensity: high-intensity protests include self-mutilation, suicide, physical confrontation between protesters and security forces, roadblocks, hunger strikes, and the occupation of public buildings; low-intensity protest includes largely symbolic and nonphysical protests such as marches, gatherings of people, town hall meetings, and boycotts,

Presidential Breakdowns in Bolivia

93

among other such demonstrations (Carter 2005). Distinguishing between varying intensities provides a set of criteria to mark the point at which social protest moves from a passive engagement to active and direct confrontation with the state. In Bolivia, civil society has played a predominant role in political life. Two developments explain this. First, Bolivian civil society has a rich tradition of involvement in political life, which reached its zenith during the national revolution when armed miners and peasants forced the military to retreat and hand over power to Paz Estensoro in April 1952. As a consequence, labor unions from the mining industry of the provinces of Potosi, Oruro, Sucre, and La Paz emerged as the most politically relevant organizations (Lora 1977). These unions have used street protests to make their demands heard and to apply pressure on the government. During the military dictatorships of the 1970s, the labor movement played the role of “defender of democracy.” Miners’ unions staged regular marches, strikes, and street protests to repudiate the rupture of the democratic process. This historical role remains in the collective memory of the labor movement today. In the 1990s, these organizations were demobilized through the government’s privatization programs. Tens of thousands of workers became unemployed and emigrated to other cities, where they reorganized, and the cities of La Paz, El Alto, and the coca-growing Chapare region in Cochabamba became the new centers of action. Second, civil society developed and institutionalized a dual strategy involving social protest as the preferred means to engage with the state. The strategy of confrontation and negotiation consisted of negotiating with the government over demands, on the one hand, and applying pressure by staging strikes, marches, and demonstrations, on the other. This strategy has often left little room for the government to maneuver, particularly when the intensity of protest rises as demands become more radical. The success of this strategy during the 1952 revolution, when resisting the military regimes and, most recently, when applying pressure on democratically elected governments, has validated its ongoing use. Civil society has instrumentalized and, in the process, institutionalized this strategy. Indeed, in its effort to gain political relevance, civil society has systematically used the confrontation-negotiation strategy as a tool over the past sixty years (Laserna and Villarroel 1999). Social protest has thus become an indispensable means to achieve results.

The Role of Civil Society: Social Protest and Presidential Breakdown This section briefly explores the theoretical linkages between the role of civil society, social protest, and presidential breakdowns.3 Arturo Valenzuela (2004) noted an interesting dynamic related to street protests and presidential breakdowns. First, he highlighted that Latin American presidencies tend

94

Miguel A. Buitrago

to be politically weak. It is often the case that even the president’s own party members are either reluctant to support a weak president or are uncomfortable with supporting an overly powerful president. Weak presidents are more easily removed from office. Second, Valenzuela notes that Latin American presidents have been affected deeply by protest movements, which sometimes present insurmountable obstacles to regime governability. Protests and protestors can seriously test a president’s power. If the president does not use force to control disturbances, these may become uncontrolled, but if force is applied, this can affect the president’s legitimacy. Either way, presidents get the blame, which is usually reason enough to remove them from office. Challenging the Linzian view, Hochstetler (2006) links social protests with presidential breakdowns, arguing that one must look beyond regime and institutional factors to understand presidential breakdowns in the region, and noting that street protest is more dangerous for the survival of a president than congressional challenges. Challenges to a president are also linked to neoliberal policies, scandals, and minority presidents. Thus, PérezLiñán, who focuses on scandals and impeachments, also points out that one must look beyond institutional variables and include extra-institutional causes such as social protests and economic policies. He further argues that popular protest has replaced military intervention as the extra-institutional instrument generating or marking the end of a presidential crisis, stating that “. . . the ability of the legislature to remove the president from office ultimately hinges on the degree of popular mobilization against the government. When a broad social coalition takes to the streets to demand the resignation of the president, the fall of the administration is usually in sight” (2007, 2–3). The same author also highlights the complexity of the breakdown process, probably induced by a combination of factors or what he calls a complex causality (Pérez-Liñán 2008). Following these authors, I suggest that we must analyze the interaction between institutional factors and civil society. When presidents become weakened either because of internal problems in the administration or because of deadlocks and interinstitutional conflicts, a civil society may increase the intensity of protests to extract more concessions from the government. If a weakened government does not accede to such demands, it is a short step for a coalition of civil society organizations to demand that a president leave power. Likewise, a president weakened by intense social protests can more easily be removed by Congress.

Three Cases of Presidential Breakdown in Bolivia Hernán Siles Zuazo: October 1982–August 1985 Siles Zuazo was elected president by Congress on October 10, 1982,4 an event that marked the end of the era of military dictatorship and the return of democracy. The Siles alliance, the Democratic Popular Union (Unión Democrática

Presidential Breakdowns in Bolivia

95

Popular, UDP), had received popular support in the general elections in 1979 and 1980 (Romero 2003), but Siles’ presidential term only began in 1982, since the democratic process had been interrupted by two military coups.5 The Siles Zuazo government was composed of a broad leftist alliance including the Bolivian Communist Party (Partido Comunista de Bolivia, PCB), the Left Nationalist Revolutionary Movement (Movimiento Nacionalista Revolucionario de Izquierda, MNRI), the Revolutionary Left Movement (Movimiento de Izquierda Revolucionaria, MIR), and the Christian Democratic Party (Partido Demócrata Cristiano, PDC). In addition, he enjoyed the support of the three most powerful labor organizations in the country, the Central Workers Union (Central Obrera Boliviana, COB), the associated miners’ union (Federación de Sindicatos de Trabajadores Mineros de Bolivia, FSTMB), and Confederation of Peasant Workers of Bolivia (Confederación Sindical Única de Trabajadores Campesinos de Bolivia, CSUTCB).6 In Congress, the governing coalition held a minority with 36 percent of the seats, while the opposition, made up by the Historical Revolutionary Nationalist Movement (Movimiento Nacionalista Revolucionario Histórico, MNRH) and the Nationalist Democratic Action (Acción Democrática Nacionalista, ADN), controlled 64 percent of the seats (CNE 2007). When the government took office, it faced a dire economic situation. Growth rates were negative, there was a substantial budget deficit, and it was becoming increasingly difficult to pay the foreign debt. In addition, the collapse of the mining industry, the most important source of income, was having a devastating effect on employment, foreign reserves, and government earnings (Sachs 1987). From the outset, the government strove to control hyperinflation, whilst simultaneously meeting the demands of the labor movement (which led to more inflation). In effect, the government was trapped by the overly high social expectations, especially of the labor unions. A major union demand was the introduction of cogestión—the participation of workers on the boards of public companies and the government’s cabinet.7 While the unions sought majority representation in the decision-making process through cogestión, the government proposed equal representation. On February 20, 1983, a radical sector of the president’s own party, the MNRI, forced the president to exclude MIR militants from his cabinet. The radical faction had been unhappy with the distribution of ministries and with the policies that MIR militants had been advocating. The immediate effect was twofold: first, Vice President Jaime Paz Zamora and leader of the MIR remained in office, but distanced himself from Siles Zuazo, thus weakening the governing alliance; second, the constellation of political forces changed in Congress, as the MIR joined the opposition. On April 19, the FSTMB occupied several mining companies and imposed cogestión. Consequently, the COB began to demand political participation in government with the support of the CSUTCB. In November 1983, the government issued another economic package, setting in motion several developments. The COB started a new wave of national strikes, accusing

96

Miguel A. Buitrago

the government of working for the International Monetary Fund (IMF). Meanwhile, the Senate censured the entire cabinet, forcing Siles Zuazo to undertake a major reshuffle. In turn, this redoubled calls within Congress for an impeachment process, given the president’s “incapacity to govern.” In December 1983, the leader of the ADN, Hugo Banzer, stated in an interview that Siles Zuazo had two alternatives: he could either call early elections in 1984, or he could resign and allow the vice president to take over.8 On April 10, 1984, the MIR reentered the government coalition, and the PCB abandoned it. MIR demands that Siles Zuazo leave office subsided. At the same time, there was a significant change in the leadership of the COB. The directory—mainly made up of PCB militants—changed hands in favor of the more radical Unified Revolutionary Directorate (Dirección Revolucionaria Unificada, DRU). With this change in leadership, the labor movement opened a second opposition front against the government, which soon became apparent on the streets. Lazarte and Pacheco report that protests increased by 192 percent from 1983 to 1984, and that the number of protests demanding higher wages increased by 277 percent in the same period (1992, 117). The situation became dramatic when President Siles Zuazo was kidnapped as part of a failed military coup on June 30, 1984. The kidnapped president was freed and the government arrested General Cayoja, the alleged leader of the coup. At the same time, the opposition in Congress intensified its attempts to remove the president from office “constitutionally.” In September 1984, ADN senator Heberto Castedo read a letter to Congress in which he outlined three paths to remove the president: voluntary resignation, early elections, and impeachment. This led the opposition to attempt impeachment on the grounds that President Siles Zuazo had authorized a meeting between a government official, Rafael Otego, and a well-known drug lord, but on September 20, 1984, the Supreme Court ruled that the motion was without merit. On November 14, 1984, the Catholic Church organized a meeting between eleven political parties, the government, and members of Congress, excluding the COB and the Socialist Party One (PS-1). The opposition (the ADN and the MNRH) and government coalition partner, the MIR, pushed the government to call early presidential elections. Siles was outflanked by Congress, the opposition, and a wave of street protests. Given that the COB had decided to shift from “cooperative” opposition to a “confrontational” stance (Lazarte and Pacheco 1992, 120–121) and thus added to the wave of street protest, Siles had no choice but to shorten his government by one year. Congress then decided to call early elections by July 14, 1985.

The Second Presidency of Gonzalo Sánchez de Lozada: August 2002–October 2003 Gonzalo Sánchez de Lozada took office in August 6, 2002. Although elected under the so-called pacted democracy arrangement, his presidency marked

Presidential Breakdowns in Bolivia

97

a change in the country’s political environment. In the wake of the 2002 elections, the long-standing dominance of the traditional political parties, the MNR, the MIR, and the ADN, came to an end. The unexpected success of anti-systemic parties such as the Movement toward Socialism (Movimiento al Socialismo, MAS),9 led by Evo Morales, and of the New Republican Force (Nueva Fuerza Republicana, NFR), led by Manfred Reyes Villa, took many politicians by surprise. Sánchez de Lozada’s coalition included the traditional parties, the MIR, the ADN, and the Civic Solidarity Union (Unión Cívica Solidaria, UCS). Initially, the government enjoyed a relatively comfortable majority, and after the NFR joined the government on August 2003 it was able to increase its majority to two-thirds control of Congress. The government’s agenda was dominated by the adverse economic setting and sociopolitical problems, and it proposed a three-pronged policy: the reactivation of the economy, promoting citizen security, and combating corruption. One of the key economic problems was a large budget deficit of 8.5 percent of GDP (LAWR February 4, 2003). The government sought to reduce it by raising revenue through taxation (income tax and property tax, primarily) and by freezing public sector wages, in accordance to IMF guidelines. The main sociopolitical problem consisted of the long-standing conflict with the anti-neoliberal coca growers union from the Chapare region over the government’s attempts to implement a neoliberal agenda, including an ambitious privatization plan. The government engaged Morales and the MAS in talks from September to December 2002, hoping to arrive at an understanding over the coca eradication conflict. This was important because Morales, as leader of the opposition in Congress and head of the MAS, had a significant say “on the streets.” Just as the government came close to reaching an agreement with the coca unions, it became known that the Sánchez de Lozada government and the U.S. administration had agreed to let Bolivian textiles into the U.S. market in exchange for the continued eradication of coca fields. This provoked a fourteen-day confrontation between coca growers and government forces, in which twelve people were killed and many more wounded (Lazarte 2005, 424–425). In February 2003, Evo Morales sought to force a congressional motion of censure against six ministers who, he alleged, had been responsible for the deaths of the protesters two months earlier. On February 12 and 13, police security forces staged a mutiny in front of the government building. Although this was a reaction to a proposed reform of the National Police Force that, among other things, transferred many important police functions to the military, in public the police demanded a pay raise and rejected the government’s income tax plan. Many officers of the Special Security Forces unit dressed in civilian clothes protested against the government. Soon, a student march joined the police protest, and there was an armed confrontation between police and military forces. The civilians were caught in the crossfire, leaving seventeen dead and hundreds wounded (Rocabado 2006). Most damaging for the government were the

98

Miguel A. Buitrago

images broadcast by the media of military and police snipers aiming and shooting at the crowds (OAS 2003). In June 2003, NFR leader Manfred Reyes Villa issued the first call for the president’s resignation. In July, the MAS made a successful attempt to gain control of the CSUTCB. Román Loayza, a close collaborator of Evo Morales, was elected president of the confederation. With this, the MAS increased its mobilization power well beyond Cochabamba. In the beginning of August, the NFR agreed to join the government coalition. Sánchez de Lozada reshuffled his cabinet to include three ministers from NFR, which consolidated the government’s control of Congress.10 In view of the new coalition, the MAS and Evo Morales in particular realized that the scope for opposition in Congress was limited, and from then on the streets became the main stage for their actions. However, the president did not profit from the oversized coalition. His decisions tended to antagonize Congress, and even his coalition partners (the MIR and the NFR) remained critical. When hydrocarbons vice minister Mario Requena stated his preference to export Bolivian gas through a Chilean port on September 16,11 the MIR made its support for the initiative conditional upon the negotiation of sea access for Bolivia, while the NFR proposed that this should be decided in a national referendum.12 The government’s plans to export gas provoked an escalation of street protests by the COB, the MAS, and the Pachakuti Indigenous Movement (Movimiento Indigena Pachakuti, MIP), making the country ungovernable. Soon, groups as varied as the Landless Movement, Felipe Quispe’s MIP, the CSUTCB, coca growers, pensioners, neighborhood associations, and even the government of Potosi were staging roadblocks, marches, and strikes. The demands were as varied as the participants, but there was common ground: protesters wanted a Constituent Assembly convened, the new income tax plan repealed, no export of natural gas through a Chilean port, and the resignation of President Sánchez de Lozada. On September 20, the government rescued a group of seventy-five tourists who had been held hostage by protesters in the town of Sorata, to the north of La Paz. The operation backfired and resulted in the death of seven protesters (Rocabado 2006, 211–214). On October 2, the Santa Cruz region demanded regional autonomy and the concomitant reorganization of the state. On October 12, the government repressed a large demonstration in El Alto, to the north of La Paz, with a reported twenty-six deaths resulting from police repression.13 A day later, Vice President Mesa publicly broke with the government and called on Congress to intervene. The coalition partners, the MIR and the NFR, initially supported the government, alleging that the opposition in Congress wanted to take power “through the streets.” On October 17, several unit commanders around the nation expressed their disagreement with the president. Soon thereafter, a NFR faction in Cochabamba, the UCS, together with the MAS, the COB, and a human rights organization, called for Sánchez de Lozada to resign. In the evening of that day, the NFR and the MIR finally announced that they were withdrawing their support for

Presidential Breakdowns in Bolivia

99

the government. Gonzalo Sánchez de Lozada found himself besieged by the protests and without the support of his political allies in Congress and in government. Amidst the social and political chaos, he resigned and left the country.

The Presidency of Carlos Mesa: October 2003–June 2005 In 2002, Carlos Mesa had been elected vice president on the Sánchez de Lozada-MNR ticket.14 After the president had been forced to leave office on October 17, 2003, Mesa was appointed interim president. In contrast to Sánchez de Lozada’s repressive response to social protest, Mesa promised that he would not respond to protests with police violence. Once in power, Mesa declared himself an independent in an effort to distance himself from the traditional parties.15 In the first months of his presidency, although Mesa lacked political support in Congress, his popularity rating was above 80 percent (EIU January 2004). However, while Congress remained a significant obstacle for Mesa, there were also some marked divisions among opposition forces. Despite internal disputes, the largest opposition party, the MNR, was united against Mesa’s attempts to dismantle the policies of the Sánchez de Lozada presidency, while the MIR and the NFR gave the new government half-hearted support. The government of Carlos Mesa faced a loaded agenda, including consideration of a Constitutional Assembly to reform the constitution, the revision of the hydrocarbons sector law, and the organization of a referendum on the export of natural gas.16 The MAS and the MIP, the main parties capable of organizing mass mobilization, gave the new government a ninety-day truce. In late November 2003, the government began talks with the different political forces in Congress. The opposition was willing to support the president on noncontroversial issues such as the Constitutional Assembly, but not on issues such as the 2004 budget. The general view was that Mesa was unwilling to commit to any specific policy stance, and so the major parties were reluctant to support him openly. On February 20, 2004, Congress approved a set of fifteen popular constitutional amendments to increase popular participation and promote forms of direct democracy. In contrast, the government was unable to pass economic initiatives such as tax reforms and the elimination of fuel subsidies. In late March, a wave of protests hit the country. Miners, pensioners, teachers, health workers, transport workers, and university students staged mass strikes, debilitating the government even further. On March 30, 2004, a miner demanding pension rights blew himself up in the lobby of Congress (EIU May 2004). With this, the government’s lame duck image and the inefficiency of Congress became all too apparent. In April 2004, alleging that the newly passed constitutional reforms allowed him to bypass Congress, Mesa set the date for the referendum on energy policy by presidential decree for July 18. The president blamed

100

Miguel A. Buitrago

Congress for taking too long to pass the law. The opposition in Congress countered by alleging that the referendum did not go far enough: the NFR, together with the MAS, called for the full nationalization of the hydrocarbons sector. Clearly, Mesa was going to have difficulties passing any legislation that might result from a referendum. In May and June 2004, social protests picked up pace again. As before, a plethora of groups, including teachers, peasant farmers, pensioners, former state miners, and the landless peasant organization mobilized nationally. The most disruptive action was a five-week teachers’ strike, which began in May when the vice minister of education was taken hostage (EIU August 2004). Civil society was encouraged by Mesa’s decision not to use police force to repress protesters. A U.S.-Bolivian agreement signed in May, exempting U.S. soldiers from prosecution by international tribunals split the MAS, and eight MAS senators abstained from voting on the agreement in Congress. The parties and party system were now in disarray. On July 6, 2004, Congress passed a law ending the monopoly of political parties on representation, in order to increase political participation. The MIR officially split into three groups, one led by Jaime Paz, another by cement baron, Doria Medina, who created the National Unity Front (Frente de Unidad Nacional, FUN) and another led by the El Alto Mayor, José Luís Paredes, who established the Progress Plan (Plan Progreso, PP). Divisions within the MNR deepened, with a group loyal to Sánchez de Lozada opposed to another seeking a change of leadership. At the end of 2004, Mesa attempted to form a voting bloc in Congress. Its most prominent members were the six dissident MAS senators. It was reported that this bloc controlled around a quarter of the votes of the 157 seats in both chambers (EIU February 2005). After the December 5, 2004 municipal elections, the MAS emerged strengthened by the rural vote, but it failed to gain control of any of the ten most important municipalities (EIU February 2005). In December 30, 2004, the president eliminated fuel subsidies, provoking an immediate social reaction. Amidst protests against the imminent rise in fuel prices, the citizens of El Alto initiated a protest to force the nationalization of the city’s water services company. On January 13, 2005 Mesa rescinded the contract with the French-Swiss company, and six days later he reintroduced fuel subsidies. Beset by protests, and with the legislature unwilling to work with him, Mesa submitted his resignation to Congress on March 6. Congress promptly rejected it because of the popular support Mesa still enjoyed. On June 6, Mesa submitted his resignation again. This time, Morales and the president of the Confederation of Private Entrepreneurs in Bolivia (Confederación de Empresarios Privados en Bolivia, CEPB), two strange bedfellows, agreed that it was time to let Mesa go. Jaime Solares, head of COB, had also demanded Mesa’s resignation at this point.17 Consequently, this time Congress accepted Mesa’s resignation. In accordance with succession rules, Congress had to appoint either the president of the senate, Hormando Vaca Diez, or the president of the lower

Presidential Breakdowns in Bolivia

101

chamber, Mario Cossio, as president. Both were highly unpopular with the MAS and Evo Morales, however, and protesters blocked the entrance to Congress, forcing legislators to move the session to Sucre. Different civil society organizations and MAS supporters then surrounded the building in which Congress was meeting in Sucre and forced the legislators to skip the first two names and appoint the president of the Supreme Court, Eduardo Rodríguez, as interim president until new elections could be held. Rodríguez was sworn into office on June 9, 2005.

On Presidential Instability This account of three presidential breakdowns in Bolivia supports the validity of most of the variables that the specialized literature claims have had an effect on presidential instability, although at different levels, with differing degrees of intensity, and within different time frames.18 Table 5.1 below collates these variables for all the so-called third-wave presidencies, after which I discuss the interaction among them and highlight the constant presence of high-intensity social protest throughout all Bolivian cases of presidential breakdown.

Economic Policy The neoliberal era, which began in 1985 after Victor Paz Estensoro issued Decree 21060 in response to the country’s macroeconomic crisis, had succeeded in bringing about stability but had failed to deliver on the promise of a better life for the population. In particular, as a result of the partial privatization of state companies initiated during the first Sánchez de Lozada government, the bulk of the labor movement had lost their jobs and ended up deeply dissatisfied with the government. That dissatisfaction manifested itself in the form of social protest, of which the most symbolic events were those that brought down Sánchez de Lozada, namely, “Black February” and the “Gas Wars.” By contrast, neither the Siles Zuazo nor Carlos Mesa presidencies were affected by neoliberal policies. Siles Zuazo sought, without success, to bring hyperinflation under control by applying timid price controls, which only exacerbated the situation. Mesa began to reverse the privatization of public companies by drafting and signing into law a nationalization law for the hydrocarbons sector.

Minority Share of Seats and Executive-Legislative Conflict The Siles Zuazo government was born out of a compromise arising from the deadlock in the 1980 Congress when political forces had failed to elect a president. His government was a minority government, controlling only 36 percent of the seats in Congress (CNE 2007). This hindered governability and forced the new administration to govern by decree. The opposition in

Table 5.1

Causes of presidential instability Hernán Víctor Paz Jaime Paz Siles Estensoro Zamora Zuazo (1985–1989) (1989–1993) (1982–1985)

Gonzalo Sánchez de Lozada (I) (1993–1997)

Hugo Banzer Suárez (1997–2001)

Jorge Quiroga Ramírez (2001–2002)

Gonzalo Sánchez de Lozada (II) (2002–2003)

Carlos D. Mesa Gisbert (2003–2005)

Neoliberal policies Minority governments Coalition breakdown Social protest High-intensity social protest

N

Y

Y

Y

Y

Y

Y

N

Y

N

N

N

N

N

N

Y

Y

N

N

N

N

N

Y

N/A

Y

Y

Y

Y

Y

Y

Y

Y

Y

N

N

N

Y

N

Y

Y

Outcome

Breakdown

Term completed

Term completed

Term completed

Term not completed*

Term completed

Breakdown

Breakdown

Y = yes; N = no; N/A: non-applicable. Source: Own elaboration. * Hugo Banzer did not complete his term due to terminal illness.

10.1057/9780230105812 - Presidential Breakdowns in Latin America, Edited by Mariana Llanos and Leiv Marsteintredet

Presidential Breakdowns in Bolivia

103

Congress not only blocked any action the government wanted to take, but it sought to undermine the government by censuring the entire cabinet, repeatedly calling for the president’s resignation, and attempting to impeach him. For his part, Siles Zuazo exacerbated the situation by excluding the vice president’s party from the cabinet in response to demands from his own party. The presence in government of the PCB was a major concern as well. In contrast to Siles Zuazo, Sánchez de Lozada had a comfortable majority in Congress. Indeed, after the NFR joined the government, he enjoyed a two-thirds majority. However, the government did not profit from this overwhelming majority because intra-coalition relations remained tense. Further, the government had emerged from the democracia pactada model, which for more than a decade had generated majority governments with solid congressional support. Paradoxically, however, in the case of Sánchez de Lozada, this model actually contributed to the president’s demise since it was by then perceived as a mechanism for the traditional parties to retain power by sharing it between them. Especially damaging were the public negotiations between the government and the NFR on the distribution of ministries. As for the opposition parties, after the NFR joined the government during the latter half of 2003, the opposition was reduced to the MAS and Evo Morales. Both realized the limitations of acting within Congress and decided to take the protests to the streets. Mesa was exceptional in the sense that he did not have support in Congress when he took over government. Not only did he seek to distance himself from traditional politicians by declaring himself independent, but, as a journalist, he upset legislators by using the media to pressure Congress to support his initiatives. Congress cooperated on issues with significant popular support such as the nationalization of natural resources, but was reluctant to support other important initiatives, such as passing the budget, for example. At another level, Congress lost effectiveness due to the crisis in the party system, which at the time was in a state of constant realignment in and outside the parliamentary arena. Many parties, such as the MNR, the ADN, and the MIR, began to split or even dissolve, which made it easier for Mesa to form ad hoc legislative coalitions, but the latter did not meet with great success.

Coalition Breakdowns Closely related to the above, coalition breakdowns are another factor, which, according to the relevant literature, contributes to presidential instability. In the period under consideration, most administrations have been postelection coalition governments (i.e., including more than one political party). According to the literature, these kinds of governments tend to be unstable (Deheza 1998). In this respect, Bolivia is no outlier. The governing coalition under Siles Zuazo entered into crisis in early 1983 when the president excluded the MIR from his government. The president was forced to reshuffle his cabinet, and although Vice President Jaime Paz remained in his post,

104

Miguel A. Buitrago

he became a fierce critic of the president. In early 1984, Paz returned to government, but this time the PCB had left the coalition. Sánchez de Lozada’s coalition was also prone to divisions. Even after the NFR joined the coalition, the president continued to be criticized by legislators, especially about the issues of natural gas exports through Chile and the government appointment of public officials. The more Sánchez de Lozada adopted a hard-line position regarding social protest, the more fragile the coalition became. In October 2003, several factions from the NFR in the Cochabamba region openly criticized the government, ultimately forcing the NFR and the MIR to abandon the administration.

On Civil Society and Social Protest Institutional factors do not account fully for the removal of Bolivian presidents. The factors constant across presidential breakdowns in the country have been high-intensity social protest and street challenges mounted against presidents. In effect, the three presidential breakdowns examined here occurred during periods of either high interinstitutional conflicts (Siles Zuazo and Mesa), or when a majority coalition disintegrated. Thus, different institutional factors help to generate a crisis of governability, which makes civil society lose faith in government. If a government is considered a lame duck or gridlocked, civil society has little to gain in the negotiationconfrontation strategy that has been a mainstay for unions and social movements for decades. As noted above, Bolivian civil society has played a determining role in political life. Because it has been such a constant presence, it has the potential to make or break a president. Laserna and Chacón (2006) state that hunger strikes, road blocks, and marches have been efficient weapons to physically and politically isolate central governments, and to force a president either to accede to popular demands or to leave office. In particular, they note that high-intensity protests increased in the period between 1994 and 2005, from 42 percent to 78 percent as a proportion of total protests. The main targets of high-intensity protest were the Siles Zuazo, Sánchez de Lozada (second presidency), and Mesa presidencies (Laserna and Chacón 2006). In addition, the rate of success of the social protests has been high. The same authors argue that in the three above-mentioned cases, social protests reached 50, 25, and 24 percent of success rates, respectively. When we consider that the aim of the protests, especially by the end of each presidency, was to remove the president, the meaning of success becomes powerfully clear. Social protests in Bolivia seem to acquire a dynamic of their own and reach a point of no return. This dynamic is apparently set in motion by frustration and lack of confidence in the political system. Government delaying strategies and repression probably add to this explosive mixture. The turning point seems to come when, first, most social actors decide they want a

Presidential Breakdowns in Bolivia

105

change in leadership because they think the president is no longer a trustworthy negotiator; and, second, when other groups seize the opportunity to press for their demands. In other words, the protests become widespread and it is no longer just workers, the underrepresented, the poor, or the most radical who take to the streets, but also large parts of the middle class and even the private sector join the ranks of protesters. Under these circumstances, the intensity of protests is generally high. During the Siles Zuazo presidency, Bolivia was falling rapidly into ever deeper recession and faced hyperinflation. At the same time, it was clear that the government was stalling the implementation of the cogestión model. As a result, the labor unions shifted from passive support to active resistance. The COB had begun to call for cogestión in December 1984, but had also decided to intensify the pressure by carrying out more marches, strikes, road blockades, and protests. The turning point came in November and December 1984 during the negotiations sponsored by the Catholic Church. The government was overwhelmed by the opposition parties within government buildings and by the rising wave of union protests outside them. From early 1985 onward, the streets of the largest cities (La Paz, Cochabamba, and Santa Cruz) were filled with road blocks and marches. Each of the participating groups had its set of demands, but they all agreed on one thing: that it was time for the president to go. There was very intense and dramatic social protest against the second presidency of Sánchez de Lozada as well. The president’s problems started when the protests and ensuing repression culminated in civilian deaths, as in the case of the fourteen-day confrontation between coca growers and police in 2002 (Lazarte, 2005), the armed conflict between military and police forces in February 2003, and the October 2003 crisis in El Alto, which resulted in more than twenty deaths.19 The turning point came sometime between September and October 2003, when other groups such as the autonomy movement in Santa Cruz joined the MAS and the opposition demanding the removal of the president. Under the Mesa government, the turning point came in December 30, 2004, when the government announced the end of fuel subsidies. From then on, the intensity of protests escalated very rapidly. In June, after Mesa threatened to resign for the second time, Congress reconvened, accepted the president’s resignation, and sought to nominate a successor in accordance with constitutional rules. Ultimately, neither the first nor the second in line had popular support, and protesters forced legislators to ignore the established line of succession and nominate the president of the Supreme Court.

Conclusion Bolivian civil society has been a significant player in national political life, adopting a strategy of negotiation and confrontation in its relations with

106

Miguel A. Buitrago

the state, and using social protest as its main weapon. Taking that modus operandi into consideration, I have tried to demonstrate here how crucial civil society and social protest have been in the dynamics of presidential breakdown in Bolivia. As in other presidential breakdowns in the region, difficult relations between minority presidents and Congress, intra- government conflicts related to coalition politics, and even social protests against particular policy measures contribute to weakening the president in office and can be a source of presidential instability. But it is only when there are generalized and high-intensity protests against weak presidents that they are forced to step down. In such instances, the generalization of protests and loss of confidence in the government become powerful forces contributing to presidential breakdown. Indeed, contrary to Hochstetler and Edwards (2009) and in accord with Valenzuela (2004), it appears that once civil society mobilizes against a president in Bolivia, that president is damned if he responds to protests with violence (like Sánchez de Lozada) and damned if he does not (like Mesa).

Notes 1. I wish to thank Mariana Llanos and Leiv Marsteintredet for their insightful comments. 2. For April Carter, direct action consists of nonviolent means of noncooperation, obstruction or defiance to exert pressure on governments and other powerful institutions (2005, 1–5). 3. For more on this debate, see Valenzuela (2004), Pérez-Liñán (2007, 2008), Hoschtetler (2006), and chapters 1 and 2, this volume. 4. According to Articles 86 and 90 of the 1967 constitution, the president and vice president are elected by direct vote for a period of five years. If no candidates attain an absolute majority (50+1 percent), Congress is entitled to choose between the three most voted candidates. Because this meant that different congressional political forces had to seek alliances with one or another candidate, this mechanism was called democracia pactada (pacted democracy), and the Bolivian presidential system has been called “parliamentarized presidentialism” for that reason (Mayorga 2001). The 1994 constitutional reform established that Congress had to choose between the two most voted candidates (CPE 2004, Article 70). 5. Hernán Siles Zuazo won the general elections of 1979 and 1980 with a relative majority. However, since Congress was unable to elect the new president from among the three most voted candidates, Alberto Natusch Busch (1979) and Luís García Meza (1980) interrupted the democratic process, ostensibly to restore order. In 1982, General Guido Vildoso, having taken the presidency through a coup, recalled the Congress that was elected in 1980 in order for it to elect a new president. After much negotiation, Congress nominated Siles Zuazo, the winner of the 1980 popular vote. 6. Siles Zuazo had been a major player in the April 1952 national revolution, together with Victor Paz Estensoro. Siles Zuazo had also served as president from 1956 to 1960.

Presidential Breakdowns in Bolivia

107

7. This idea was a legacy of the 1952 revolution, when the miners’ unions and government had sought to increase union participation in the management of state companies. 8. A New Kind of Coup, And Several of the Old Kind, Latin America Weekly Report (LAWR) December 2, 1983. 9. In the general elections of 2002, the MAS and the NFR each received 20 percent of the vote, while the MNR received 22 percent (CNE, 2007). 10. Goni se fortalece con un pacto a la medida de las cuotas de poder, La Razón August 6, 2003. 11. The export of Bolivia’s natural gas became a controversial topic because of the historical disputes with Chile over Bolivian access to the sea. 12. Gobierno aboga para exportar gas por Chile, Bolpress, at http://bolpress.com/ art.php?Cod=2002072096&PHPSESSID=3a16bc2c299aba03134d3681f729f9 46. September 16, 2003. 13. El Presidente no renunciará a su cargo, pese a la violencia y a Mesa, La Razón October 14, 2003. 14. The Bolivian constitution stipulates that if a president cannot finish a term in office, the first in line to succeed is the vice president, followed by the president of the senate, followed by the president of the chamber of deputies, followed by the president of the Supreme Court. If the latter is selected and the president has been in office less than three years, early elections must be called (CPE 2004, Articles 90–93). 15. As expected, as a result of the events leading to the breakdown of Sánchez de Lozada’s presidency, public image of the traditional parties (the MNR, the MIR, and the ADN) as well as the NFR (which was associated with the government) was badly tarnished. 16. The social movements called this the “October agenda.” 17. LAWR 22, June 7, 2005; Notisur May 27, 2005. 18. Presidential scandals have not been relevant in presidential breakdowns in Bolivia. 19. For more details, see Lazarte and Pacheco (1992), Mayorga (1987), and Lazarte (2005).

This page intentionally left blank

III

The Cases of Impeachment

This page intentionally left blank

6 The Collor Impeachment and Presidential Government in Brazil Argelina Cheibub Figueiredo*

Introduction In the fall of 1992, Fernando Collor de Melo, the first popularly elected president in Brazil after twenty-nine years of military rule, was impeached by the Chamber of Deputies and he was deprived of his political rights (cassado) by the Senate. At the time, the prospects for Brazilian democracy seemed dim. Later developments showed otherwise, however. Subsequent presidencies have managed to achieve a far from trivial record of social and economic improvements carried out in a political context of broad freedom of the press and association, and expanding electoral and social participation, in particular. What kind of political phenomenon was the breakdown of the presidency of Collor? Does it represent a new trend in Latin American presidentialism? What have been its consequences for Brazilian democracy, and how do they relate to the causes of the demise of President Collor? I seek to answer these questions in this chapter by examining the impeachment and its aftermath in the light of two opposing interpretations about presidential breakdown in Latin America: the theory of “de facto flexibilization” of presidentialism, on the one hand, and the “intermittent checks and balances” model, on the other. According to the first interpretation, the frequent and diverse interruptions of presidential mandates in Latin America signal a transformation in the nature of presidentialism. Thus, the emergence of different mechanisms to remove presidents, associated with different systems of government, provides evidence of a “flexibilization” of a basic defining aspect of presidential systems: the independent origin and survival of the executive and

112

Argelina Cheibub Figueiredo

legislative branches (Marsteintredet and Berntzen 2008, 85; chapter 1 in this volume). According to the second interpretation, “the multiplication of impeachments” signals the emergence of a peculiar Latin American “model of checks and balances,” whereby horizontal accountability is “politicized and spasmodic” and occurs “in a context marked by a traditional Latin American dynamic—executive dominance punctuated by presidential failure.” Thus, impeachments prove “that legislators [are] strong enough to hold the president accountable when media scandals and popular protests gave them enough leverage, but [are] not strong enough to discourage the occurrence of presidential abuse on a regular basis” (Pérez-Liñán 2007, 209). My goal in this chapter is twofold. First, I assess whether these interpretations explain Collor’s impeachment. Following Marsteintredet and Berntzen (2008), I argue that the Collor dêbacle did not signal a crisis of presidential democracy. The impeachment followed established procedures and was independent of the state of democratization. Indeed, such a proceeding could happen in an old and new democracy alike for similar reasons.1 I then trace the course of events leading to the removal of Collor, showing that it was a typical case of successfully using a constitutional mechanism designed specifically to deal with an exceptional, albeit anticipated, situation that can predictably arise in presidential systems. 2 Although a successful impeachment certainly depends on political factors, the process can be launched and completed only if established formal conditions and procedures are fulfilled. In the second section, this argument is borne out in my examination of the relationship between Collor and Congress. The president’s inability to obtain and sustain legislative support played a crucial role in the success of the impeachment. Indeed, the Collor government had what has been seen as a typical conflictive relationship with the legislature (Cox and Morgenstern 2002; Amorim Neto 2006; Amorim Neto, Cox, and McCubbins 2003; Carey 2003). Collor governed with minority support and was further isolated by his extreme policy preferences and unilateral strategy. However, the “intermittent checks and balances” model does not account for the behavior of Congress and its attitude toward the president. The Brazilian Congress acted with independence, not as an obstacle, or as a “recalcitrant” legislature (Cox and Morgenstern 2002, 455). The opposition majority in Congress pursued an institutional strategy using all the available procedural mechanisms to achieve its goals and preferred policy options. Opposition parties acted in opposition, but not irresponsibly. This can be observed with regard to the stabilization plan (the Collor Plan), but also in other areas, particularly social policy. Congress was able to set limits on Collor’s attempt to go beyond his constitutionally defined decree prerogatives and bring his unilateral strategy to a halt. Collor was also compelled to broaden his coalition, bringing political parties into the fold. These are hardly signs of a weak legislature as implied in the “intermittent checks and balances” model.

Presidential Government in Brazil

113

In the final section, I show that Collor’s removal from the presidency had some, mostly positive, consequences for subsequent governments and for Brazilian democracy. The aftermath of Collor’s impeachment reinforces the argument that such an event does not signify a crisis of presidential democracy. Rather, it was a sign that society was prepared to pursue a democratic choice, and that the country’s political institutions, although far from ideal, were solid enough to sustain it. Subsequent presidents were elected with broad support for two terms and commanded majority government coalitions. They have made extensive and continuous use of decree powers, and enjoyed significant party support nonetheless. In contrast with Collor, they never faced a legislative opposition majority. There are no presidential breakdowns lurking on the horizon, and it does not seem that Brazilian presidential democracy is embarking on a process to “flexibilize” the mechanisms to remove presidents. However, Brazil’s government structure is perhaps an extreme example of a presidential system that exhibits characteristics often associated with parliamentarism, notably coalition governments (Linz 1990), extensive executive agenda-setting powers (Tsebelis 1995, 2002), and integration between the executive and the legislative branches of government (Cox and Morgenstern 2002). Coalition governments in presidential systems can be viewed as a type of “flexibilization,” in which presidents seeking to increase their support in the legislature change the party composition of their cabinets and thereby initiate a new government. Presidential agenda-setting powers, on the other hand, are useful to solve conflicts among supporting parties, 3 a characteristic that is reinforced by greater integration between the executive and the legislature. The combination of all these features gives rise to a model of government that departs significantly from the pure presidentialism found in the United States. The study of the Brazilian case can therefore shed new light on how governments in presidential systems work, and on the impact of key institutional mechanisms under different systems of government.

The Rise and Fall of Collor An account of the events leading to the fall of Collor suffices to show that the accusations against the former president by his own brother, and the strong evidence of his involvement in corruption schemes, were necessary, though not sufficient, conditions for his impeachment. The impeachment process followed constitutional and legislative procedures strictly, as did the succession with the appointment of the elected vice president. As the governor of the small state of Alagoas, Fernando Collor de Melo made national headlines as the self-entitled maharaja-hunter (caçador de marajás), marajá being the colloquial Brazilian term for corrupt government employees. In his 1989 presidential campaign, he ran for the tiny recently created National Reconstruction Party (Partido da Reconstrução

114

Argelina Cheibub Figueiredo

Nacional, PRN) and portrayed himself to the electorate as the defender of the “shirtless” (descamisados) and the enemy of the “elites.” Of the twentyone candidates running that year, he took 30.5 percent of the vote in the first round and 53 percent in the second round against the candidate of the Workers’ Party (Partido dos Trabalhadores, PT), Luís Inácio Lula da Silva. Having clinched the presidency in a nonconcurrent election, the Collor administration was inaugurated on March 15, 1990. Congress was to be renewed in elections the following November and its elected members would have to wait until the following year to take their seats. Collor’s party had less than 5 percent of the seats in the Chamber of Deputies and none in the Senate. He forged government coalitions but only obtained minority support from Congress during his tenure. The main parties in the government coalitions were the Social Democratic Party (Partido Democrático Social, PDS) and the Liberal Front Party (Partido da Frente Liberal, PFL), both offshoots of the National Renovation Alliance (Aliança Renovadora Nacional, ARENA), the party created under military dictatorship to support the regime. The Brazilian Democratic Movement Party (Partido do Movimento Democrático Brasileiro, PMDB), the main opposition party during the dictatorship and the largest centrist party in Congress, led the opposition against the Collor government, allowing the formation of center-left coalitions. Collor targeted inflation as his number one enemy and promised to bring it down “in one fell swoop.” On the day of his inauguration, he presented Congress with a stabilization plan, the Collor Plan. The Plan changed the currency, froze prices and salaries, slashed public spending with massive public sector job cuts, and mandated the closure of various public government agencies. The government also initiated the privatization of public enterprises and tariff reductions. Radical as they were, these measures did not deliver the expected outcomes. On the contrary, they caused recession, unemployment, and, consequently, popular dissatisfaction. During the first year of his presidency, Collor had little to celebrate. The Brazilian economy shrank by 4.4 percent, the largest slowdown since measurement of this statistic had begun. Inflation was rampant, increasing by 450 percent in the first year, and by 1,100 percent in the second. In addition, from 1990 to 1992, average home incomes decreased significantly (from 304 to 264 cruzeiros), as did the real minimum wage (from 232 to 175 cruzeiros). Poverty also increased (the percentage of poor people rose slightly from 40 to 40.4 percent), as did the unemployment rate (from 4.4 to 6.8 percent).4 At the beginning of 1991, the government launched the second Collor Plan (Plano Collor II), but this also failed, provoking even deeper popular dismay. Polls on the popularity of the president reflect the widespread dissatisfaction with the government’s economic policy. When the government was inaugurated, expectations were high: 71 percent of the population expected an excellent or a good government; only three months later, that assessment had plummeted to 36 percent. By March 1991, the proportion of those who

Presidential Government in Brazil

115

considered the government to be either excellent or good had fallen to 23 percent, and by March 1992, the number had fallen to 15 percent. The number of those deeming the government either a bad or a terrible one grew at an inversely proportional rate, from 19 to 24 percent and then to 48 percent, respectively. 5 Collor’s growing unpopularity affected all social cohorts. It was in this context of economic crisis that rumors of corruption spread. It was denounced that there was a scheme masterminded by Collor’s campaign treasurer, Paulo César Faria, to benefit several top-tier government officials and Collor himself. The coup de grâce came in May 1992 when the president’s brother, Pedro Collor, who was not a politician but shared the family business, confirmed many of the accusations in one of Brazil’s largest weeklies. In response to this, Congress installed a Joint Parliamentary Commission of Inquiry (Comissão Mista Parlamentar de Inquérito, CPI) on June 1, 1992, and less than three months later, on August 24 1992, the CPI presented conclusive evidence against Collor in its final report. The first public protests against the government were staged in the middle of August, and grew as evidence of Collor’s corruption mounted. In an attempt to save his mandate, Collor appealed directly to the population in a nationally televised speech, asking them to show their support for him the next day by wearing green and yellow, the colors of the national flag. His attempt backfired, however: students clad in black organized a massive protest demanding his impeachment. The final report of the CPI sufficed to initiate impeachment proceedings. The lower chamber set the impeachment in motion, in accordance with the 1988 Constitution, which provides the guidelines to try presidents, and defines which presidential actions constitute “responsibility crimes” against the public administration. On September 29, 1992, in a 441 to 38 vote, the deputies authorized the initiation of the trial by the Senate. Vice President Itamar Franco temporarily assumed government. On December 29, on the eve of the Senate’s last trial session, certain of indictment, Collor resigned. The Senate went ahead anyway, condemning him for “responsibility crimes” and suspending his political rights for eight years.

Collor and Congress In a comprehensive study of Latin American impeachments processes, PérezLiñán singles out three factors that, given an international context in which military interventions are strongly discouraged, determine the fate of an executive: the role of the press, the president’s ability to bring together a legislative coalition against impeachment, and, finally, the extent of popular protest (2008, 36–37). The sequence of events that led to the fall of Collor shows that all three ingredients for a successful impeachment were present. In this section, I focus on just one, namely, the president’s ability to obtain and sustain congressional support.

116

Argelina Cheibub Figueiredo

An analysis of executive-legislative relations during Collor’s presidency shows that they shifted between three strategic models: confrontation, negotiation, and isolation (Pérez-Liñán 2008, 168). During most of his government, Collor went back and forth, sometimes confronting and sometimes isolating himself from Congress. Paradoxically, it was soon after he tried to adopt a strategy of negotiation, yielding to the demands of parties and coming close to attaining a majority coalition in Congress, that he was dealt a lethal blow by his brother’s accusations, followed by the work of the CPI. Three factors account for the different turning points of the Collor presidency: first, his government formation strategy and the coalition changes made during his tenure; second, a consequence of the first, his formal legislative support as represented by the share of seats held by the parties in the government coalition; third, the presidential strategy of passing laws by decree (known as “provisional measures” or medidas provisórias, MP), the most effective executive prerogative provided for by the 1988 Constitution. Table 6.1 compares these dimensions for the nineteen government coalitions formed by the five presidents who ruled Brazil from October 1988 to December 2007. This framework permits a more nuanced and complex analysis of Collor’s relationship with Congress, distinguishing not only presidential strategies but also how Congress responded to them. Collor initially opted for a confrontational strategy. Having trumpeted his feats and virtues, when he took power he made it clear that he intended to govern without the parties. The only party included in his first cabinet was the PFL, in addition to one ministerial post assigned to a member of his own minute party. The new president also drastically reduced the number of ministries and appointed inexperienced politicians or experts of little renown to the ministries that remained. Collor’s legislative strategy was unilateral. He passed the stabilization plan and all measures to reduce public administration by using MPs, which are decrees that come into effect immediately. This was not a novelty in Brazil: previous plans had been passed similarly. President José Sarney (1985–1990) relied heavily on decree law (decreto lei) power established under military rule to pass the Cruzado Plan, the temporarily successful economic plan launched after the reestablishment of democracy. With the passage of the 1988 Constitution, decree laws were replaced by provisional measures, and Sarney used them to achieve stabilization with the Summer Plan (Plano Verão) launched in January 1989. However, in contrast with the past, to reject some MPs in the Collor Plan, such as the ones mandating a change of currency and the confiscation of financial assets, had the potential to wreak economic havoc. The parties in Congress acted as cautiously as the situation required, and it is hard to pin responsibility on Congress for the failure of the Plan.

Table 6.1 Government coalitions, parliamentary support, and executive law-making strategy in Brazil, 1988–2007 Government Coalition

Sarney Collor 1 Collor 2 Collor 3 Collor 4 Franco 1 Franco 2 Franco 3 Cardoso I1 Cardoso I2 Cardoso II1 Cardoso II2 Lula I 1 Lula I 2 Lula I 3 Lula I 4 Lula I 5 Lula II 1 Lula II 2

Coalition

Seats Lower House

Coalition Attributes

Duration (Months)

Start Date

% Pres. Party

% Govt. Coalition

Partisan Ministers

17 7 4 14 6 11 5 11 16 32 38 10 13 12 4 2 19 2 9

10/1988 03/1990 10/1990 02/1991 04/1992 10/1992 08/1993 01/1994 01/1995 04/1996 01/1999 03/2002 01/2003 01/2004 02/2005 05/2005 07/2005 02/2007 04/2007

41.4 5.1 6.1 8 6.2 –* – – 12.5 16.6 18.3 18.2 18.2 17.7 17.7 17.7 17.7 16.0 16.0

63.03 29.70 30.51 33.20 42.15 60.04 59.64 55.27 56.14 77.19 73.88 45.22 42.88 62.38 57.7 58.28 69.59 60.04 68.03

77.3 18.8 25.0 18.8 42.1 57.1 52.4 23.8 66.7 65.2 69.0 46.2 87.1 87.1 83.9 83.9 80.0

Law Making Strategy

Ministers % Ministries / % Bills Decrees with Leg. Seats (Monthly (Monthly Mandate Average)** Average) 36.4 15.4 21.4 14.3 35.3 52.4 42.9 19.0 30.0 38.1 33.3 20.0 41.9 45.2 45.2 46.7 35.7

0.77 0.19 0.25 0.19 0.42 0.48 0.44 0.20 0.56 0.59 0.64 0.46 0.59 0.53 0.54 0.53 0.53

5.8 5.1 4.5 9.0 5.2 6.5 6.2 3.6 4.8 3.7 3.8 5.3 2.4 4.5 2.5 4.5 1.6

Source: Banco de Dados Legislativos, CEBRAP; Figueiredo (2007). * Franco left his party, the PMDB, to run for the vice presidency, and remained unaffiliated during his presidency. ** The first column excludes the budget laws altering the annual budget law, which the constitution allows only the executive to introduce.

10.1057/9780230105812 - Presidential Breakdowns in Latin America, Edited by Mariana Llanos and Leiv Marsteintredet

7.3 8.0 4.5 0.6 0.5 2.0 5.6 8.2 2.3 3.8 3.6 6.9 4.5 6.3 3.0 5.0 3.7

118

Argelina Cheibub Figueiredo

Congress dealt with the Plan with a better-defined institutional framework than that which existed under Sarney. The two main gray areas in executive-legislative relations found in the succinct constitutional text on “provisional measures” had been solved when the Summer Plan had been implemented. One of the solutions favored the legislature, allowing it to amend executive-sponsored MPs; the other one favored the executive, giving it the prerogative to reissue an MP if Congress did not vote on it within a constitutionally defined thirty-day period (Figueiredo and Limongi 1999, 160–172). As a result, to reissue an MP, the executive only had to prevent an opposition majority from forming in Congress. Further, because of the nature of the presidential measures, Congress had limited opportunities to propose modifications. At any rate, Congress responded promptly to the barrage of MPs issued in the initial days of the plan. Only two were reissued, which means that Congress deliberated on all the others within the thirtyday period. Indeed, none of the Collor Plan-related MPs spent more than sixty days in Congress. This belies the notion that Congress was unable “to keep pace with the decrees emanating from the Planalto Palace” (Power 1998; 2009). The PMDB led the negotiations with the government and was able to extract some concessions. The PMDB adopted an institutional strategy: as the largest party in Congress and because of its ideologically centrist position, it controlled the crucial votes necessary to approve government initiatives. The party took advantage of all the institutional instruments at its disposal, particularly agenda-setting powers and the ability of party leaders to control legislative procedures (as outlined in the new internal regulations governing the Chamber of Deputies). The PMDB occupied all positions with review powers (relatorias) and affirmed itself as a privileged negotiator between the government and other parties. Support for agreements was secured by party leaderships’ control over the agenda and over the main voting procedures. Although congressmen presented a considerable number of amendments (3,279), only thirty-eight were considered. With the disciplined votes of its ranks, the PMDB was able to consolidate its position as the broker of political agreements with the government. The government also resorted to veto power to nullify the modifications of PMDB reviewers, so that the measures implementing the core guidelines of the Plan survived unscathed (Figueiredo and Limongi 1999, 173–175). The confrontational strategy was taken a step further after the defeat of the PMDB when Congress voted on an alternative congressional project to replace MP 168 on changing the currency and freezing financial assets, including personal savings and investments. The PMDB tried to modify this MP by changing the rules on freeing frozen assets, but it was defeated by the government by a slim margin. The government made it clear that it would make no more concessions to the PMDB in the future. However, in May the PMDB and other opposition parties rejected provisional measures 184 and 186, also by slim majorities (the latter measures suspended

Presidential Government in Brazil

119

the execution of certain court rulings to “prevent damages” to the public economy). Collor defiantly reissued measure 185, renaming it MP190, to show that he meant to bulldoze over congressional resistance. The political conflict between the PMDB and the government escalated into a clash between the two branches, and the judiciary had to intervene. The Federal Supreme Court (Supremo Tribunal Federal, STF) declared the MPs unconstitutional on the grounds that they infringed on the individual right to resort to the courts and undermined the independence of court rulings, but by then Congress had already rejected them (Power 1998, 209). Congress also tried to limit recourse to extraordinary legislation through a complementary bill (projeto de lei complementar) to regulate the issuing of MPs (Figueiredo and Limongi 1999, 141–143). In this first phase of his government, Collor issued an average of eight provisional measures per month, well above his five per month average for bills (projetos de lei). This ratio became more balanced toward the end of the year, but after the beginning of the second legislature in February 1991 Collor changed his strategy toward Congress. In terms of legislative initiatives, this shift is apparent in the drastic fall in the number of MPs issued and the increase in the number of ordinary bills (leis ordinárias)—for each MP, the president submitted an average of nine bills. Even when Collor II was launched, just a day before the new members of Congress were sworn in, the use of provisional measures was contained. The second Plan proposed to freeze prices and salaries again, control the financial market, unpeg the currency, and reduce import tariffs to foster industrial sector competition. A few months later, it was clear that the Plan had failed. The lack of political support for the economic cabinet headed by Zélia Cardoso de Melo, and the criticisms of exaggerated interventionism and centralization, led to her dismissal in May. Zélia’s departure marked a crucial shift in the style of government inaugurated by Collor. The new minister of economy, Marcílio Marques Moreira, was known for being more inclined to negotiate, and he declared his opposition to “shock treatments” (Baer 2003, 205–206). In January 1992, as a reaction to accusations of corruption, Collor dismissed the ministers of health, social action, and labor. More experienced politicians were appointed to head these ministries, consolidating the path initiated with the change in the economy ministry. But the main indication of Collor’s willingness to redefine the terms of his relationship with the parties and Congress was the creation of the Government Secretariat (Secretaria de Governo) to head up the political coordination of the government. In a clear gesture of goodwill toward the parties, Collor appointed experienced PFL politician Jorge Bornhausen, who enjoyed good interparty relations, to head the new ministry. The new presidential strategy culminated with more profound changes in the composition of government. In March, Collor increased the number of ministries and collectively dismissed all ministers to promote a full-scale ministerial reshuffle. When the new cabinet had been appointed, none of the original ministers remained. The formation of this cabinet

120

Argelina Cheibub Figueiredo

followed two basic principles. First, the president brought in well-known and respected technical authorities, such as the renowned surgeon, Adib Jatene (Minister of Health), academics José Goldenberg (Education), Hélio Jaguaribe (Science and Technology), and Celso Lafer (Foreign Affairs)—the latter two from the ranks of the PSDB. Second, the president brought in as many parties as possible, including the opposition parties. Bornhausen was assigned this task. The replacements in January had partly done the job, with the appointment of skilled PFL politicians such as Ricardo Fiúza and Reihnold Stephanis. But the government also managed to make inroads with the opposition PSDB and PDT. Indeed, the PSDB voted on the issue and the party was split over it.6 Ultimately, only two new parties, the PTB and the PL, joined the government coalition, but Collor managed to broaden his party support, gaining 42 percent support in the Chamber of Deputies. At this time, the president’s popularity was low among all social classes. Various sectors of the economic elite were dissatisfied, as was much of the media, including those who had given him solid support during the electoral campaign. Even so, the parties did not rule out cooperation. The PSDB, for instance, allowed two of its members to participate in the government, and the PDT voted systematically with the government. This new coalition showed a greater degree of integration between the two branches than previous ones. As shown in table 6.1, this coalition had the highest proportion of ministers participating with the endorsement of their parties (42 percent); the highest degree of proportionality (0.42) between the number of ministry positions occupied and the number of seats in the Chamber of Deputies; and, finally, the one with the highest proportion of ministers who simultaneously held seat in the Senate or the Chamber of Deputies (35 percent). This new strategy was inaugurated a month before the Pedro Collor interview. Clearly, then, in April 1992, parties had not yet abandoned Collor to his fate. It was not until the accusations made by the president’s brother, and particularly until the investigative parliamentary committee had begun its work and started to produce evidence that Collor’s hopes of building a stronger political base in Congress were definitely dashed. At the end of August, when the final report of the CPI was voted on, it was expected that there would be defections during the impeachment vote from the coalition supporting the government. The government tried to change the voting procedure to make the impeachment authorization vote secret rather than open. As the PMDB president of the lower chamber Ibsen Pinheiro observed, “This crisis did not originate in party politics nor will it have a partisan ending. No party is responsible for the grave acts gathered by the investigative committee. By acknowledging this I do justice to the parties supporting the government. [. . .] The representative votes will have an ethical and moral motivation . . .”7 Indeed, Collor’s attempt to get the parties to collaborate occurred when his prestige was already faltering. His failed attempts to stabilize the tattered economy had already corroded most public support. The polls show that his

Presidential Government in Brazil

121

popularity plummeted. Nonetheless, there had been no public demonstrations until Congress had initiated its investigation. The CPI was proposed by the opposition parties that had the required number of signatures to establish it (a third of the members of each house). Popular pressure was not the driving force behind the congressional initiative to investigate Collor, as is argued in most studies (Weyland 1993; Hochstetler 2006; Pérez-Liñán 2007). In the “intermittent checks and balances” model, the legislature’s ability to make the executive accountable depends on popular pressure: “popular uprisings, rather than legislative action, are the main force behind the collapse of elected administrations. Congressional action provides the framework [. . .] but it often plays a merely reactive role” (Pérez-Liñán 2008, 206). But in the case of Brazil, it is difficult to argue that Congress was merely reactive. It was the swift and efficient work of the CPI that proved Collor’s misdeeds, which triggered popular demonstrations. The upcoming municipal elections of October 1992 may have compelled Congress to resolve the situation more speedily; moreover, public demonstrations did help to undermine support for Collor among conservative politicians seeking to improve their electoral prospects, as Weyland points out (1993, 19). But, always and everywhere, congressional action has an eye on the public and on elections (Arnold 1990). In this sense, what is at stake is the balance of power between the executive and the legislative branches, and its consequences for democratic government. This question requires more than an empirical answer. But it is worth remembering that in 1992 the Brazilian Congress could hardly be considered a “new” institution, as it had been working continuously since 1946 (except for a period of less than a year in 1969).8 On the other hand, most of the executive prerogatives granted by the 1988 Constitution had not existed when the 1946 democratic constitution was in force. Theoretically, however, it is very difficult to assign an optimal equilibrium between the two branches of government. For the time being, “further reflection on the role of the public” is necessary, although not just under presidential systems, as Hochstetler (2006) suggests, but under democratic government in general. As the experience of “older” presidential and parliamentary democracies shows, the mix of executive-legislative balance is diverse enough to warrant more cross-system comparisons.

The Aftermath of the Presidential Breakdown What impact did the impeachment have on subsequent governments? To what extent was the behavior of Congress a reaction to public protest? What are the institutional limitations on Congress as an agency of horizontal accountability? These questions require distinct answers. The president who replaced Collor, Itamar Franco (1992–1994), formed a majority government coalition with wide-ranging support from various

122

Argelina Cheibub Figueiredo

parties. Even the PT, which usually expelled its undisciplined members, did no more than warn deputy and former mayor of São Paulo, Luiza Erundina, for participating in Franco’s cabinet. After various failed attempts to stabilize the economy and a succession of economy ministers, inflation was finally curbed after the appointment of Fernando Henrique Cardoso to the economy ministry, an achievement that gave him the ticket to the presidency in a single round election. President Cardoso governed for two consecutive terms (1995– 1998 and 1999–2002) and pushed through an extensive agenda of economic and social security reform, which required significant constitutional changes that he achieved with solid support from the parties in Congress. The current president, Luís Inácio Lula da Silva, whose second term in office ends in 2010, also won two highly competitive elections. His first government coalition fell short of a majority, but his policy agenda was supported by the opposition parties that had been in government before. This gave him a legislative majority but led some left-leaning parties to abandon his coalition. The inclusion of the PMDB in the coalition at the beginning of Lula’s second year in office ensured that he had majority support in the legislature. He was then able to execute his policy agenda, and his success guaranteed his reelection in 2006. In the middle of his second term, 80 percent of the population felt that Lula’s government was good or very good. Both Fernando Henrique and Lula experienced crises and threats of impeachment by the opposition, but neither came close to being a real challenge. Commanding center-right coalitions, Cardoso was able to neutralize the left, particularly the PT. He prevented the approval and formation of a CPI proposed by his opposition. By contrast, Lula faced two CPIs that functioned for almost ten months investigating the so-called mensalão scandal.9 In May, 2005, as the parliamentary commission created to investigate corruption in the Brazilian Postal Service was poised to start its work, Deputy Roberto Jefferson revealed the alleged existence and inner workings of a mensalão scheme, giving rise to another CPI. The accusations involved the PT first and foremost, but also affected high-ranking government members, including José Dirceu, minister of the Casa Civil and former president of the PT, who was mastermind of the alliance that had led Lula to power and Lula’s “strongman.” Unlike those established to investigate Collor, the investigations of these CPIs lasted approximately ten months, their activities were televised, and both provoked a significant decline in the president’s popularity. At one point during the investigations, the opposition parties called for the president’s impeachment, but the idea, however, fell on deaf ears. If one hypothesizes that impeachment processes may signal a “flexibilization” of the presidential system, this would have been the moment to watch in Brazil. However, the evidence produced by the commissions was not as conclusive as that against Collor, and, in February 2006, the president’s popularity began to rise again.10 The government coalitions of the last two presidents display a high degree of integration between the legislative and executive branches (see table 6.1).

Presidential Government in Brazil

123

The coalitions they forged show no significant difference in terms of the rate of proportionality between ministry positions occupied and seats controlled by parties in the Chamber of Deputies, although this ratio was slightly more proportional under Cardoso. Given how the proportionality index is calculated, differences can be attributed to several factors. During both Cardoso terms, these rates were the result of a greater proportion of ministries not being occupied by party-endorsed politicians. During Lula’s mandates, they are the result of the greater proportion of ministries occupied by the PT, as well as the fact that some of the government parties, such as the Brazilian Communist Party (Partido Comunista do Brasil, PCdoB), are quite small. The proportion of ministries occupied either by senators or deputies are much higher in the Lula government, with more than 80 percent for all coalitions. This data indicates greater proximity and engagement between government and parties, making negotiation and collaboration between the two branches easier. In Congress this translates into votes for legislative initiatives by the executive.11 However, the two last columns in the table show that all governments regularly and systematically resorted to decrees, except the Collor administration for most of its time in office. This shows that decrees are instruments of majority governments in Brazil. Thus, in this case, it is difficult to claim that the president circumvents legislative majorities. If a majority opposes a decree, it can reject or modify it. After the implementation of the Real Plan and when Cardoso took charge of the economy ministry under the third coalition of the Itamar Franco government, the number of MPs surpassed the number of ordinary bills submitted by the executive. In fact, the Real Plan surpassed the Collor Plan both in terms of MPs issued and in terms of the proportionality between this instrument and ordinary bills (see the last two columns of table 6.1). Furthermore, the practice of repeatedly reissuing MPs was consolidated after the Real Plan. Significant modifications could be made each time a MP was reissued, without requiring the creation of a committee to evaluate them. This type of decision-making, away from the plenary and the spotlight, maximizes the power of the party leaders to “protect” those supporting government from scrutiny by their electoral bases. It is a low-visibility decision-making process that inhibits accountability. For this reason, the ability to reissue MPs indefinitely has been very favorable for the executive and its parliamentary base. It has facilitated deals on policies or any other “side payment,” thus contributing to the cohesion of the government coalition. Notwithstanding, at the end of Cardoso’s second term, Congress approved constitutional amendment 32 in September 2001, as it was concerned with public opinion, which increasingly viewed Congress as being inactive. With this amendment, an MP must be voted in the lower house forty-five days after it is issued, otherwise it blocks the voting of all bills scheduled on the plenary agenda, which also happens in the senate, after the MP has been voted in the

124

Argelina Cheibub Figueiredo

lower house. Further, the voting deadline was extended from thirty to sixty days. Whether or not this decision can be viewed as consistent with a strong or independent Congress depends on how one interprets the motivation of legislators to delegate legislative powers to the president. Shugart and Carey (1992, 165), for instance, suggest that Congress members are motivated by their desire freely to pursue particularistic and short-term policies, leaving broader longer-term public policies to the executive. The evidence from the Brazilian case, however, reinforces the interpretation that takes into account the interests of the parties in the coalition supporting the president. The continuous reissuing of MPs during the administrations analyzed here depended on the collaboration of party leaders in the government coalition who could, in turn, weigh in on the modifications to the original MPs. The fact that the executive relied heavily on decree power does not mean that presidents were governing against or without Congress. Even presidents with strong legislative powers cannot govern against the majority, as Collor discovered in the face of PMDB opposition. On the other hand, agenda powers and a centralized decision-making process make the action of the majority that supports the president—when this majority exists—much easier. Presidential action is constrained, as it should be under a democratic regime, by the distribution of seats in the legislature. In some circumstances, such as in situations when highly unpopular decisions are taken, the majority may prefer a MP to an ordinary bill. Constitutional decrees, like any other agenda powers, may be used to neutralize the incentives for opportunistic behavior by government coalition members.

Conclusions The analysis of Collor’s impeachment and its consequences for democracy in Brazil clearly shows that it should not be included in the group of countries that, as Marsteintredet and Berntzen have demonstrated convincingly, have been adopting the functional equivalents of parliamentary institutions in order to change government by removing the chief executive. In reality, in their list of the various removal mechanism employed, these authors exclude impeachment, and as regards the cases they study, they conclude that the three cases of successful impeachment considered (Brazil, 1992; Venezuela, 1993; and Paraguay, 1999) “cannot be interpreted as indications of a crisis of presidential democracy as such, although it took a political crisis and a deadlock to trigger the impeachment processes. On the contrary, these cases may be seen as proof that the constitutional instrument available for resolving conflicts between the executive and the legislature in presidential system is in fact working” (Marsteintredet and Berntzen 2008, 91, emphasis added). Brazil may not be an example of the “flexibilization” of presidentialism with regard to the early removal of presidents, but it does offer some lessons regarding the functioning of government. The practice of coalition

Presidential Government in Brazil

125

governments, which dates back to 1946 when the democratic period was inaugurated, was reestablished under Sarney, and subsequently reinforced, including under Collor. The latter’s first attempt to ensure the survival of his government amidst mounting accusations of corruption consisted of a ministerial reform to broaden his party support in 1992. The formation of a new coalition will not lead to a change of presidency, but it certainly implies changes in government. The independent election of the chief executive justifies his continued leadership of the government formation process, but it does not prevent him from promoting the necessary ministerial changes that would allow for more effective government. Although the “fusion” between the executive and legislative branches is not complete, they have become increasingly integrated. As shown elsewhere, this integration is not limited to the distribution and command of ministerial positions: it also extends to other areas of government, particularly budget policy (Figueiredo and Limongi 2008). Finally, the president’s extensive legislative powers, especially the power to issue decrees with the force of law, grants the executive extraordinary control over the legislative agenda. Further, in presidential systems, this institutional mechanism can be used to do more than solve “vertical conflict” between the executive and legislature: it can become a powerful instrument for “horizontal bargaining” between presidents and parties in their governing coalitions (Huber 1996). In this case, contrary to the prevailing view, the employment of these instruments cannot be seen as an example of the executive trampling on the legislature, and neither would it be true to see the acquiescence of Congress as a sign of weakness. On the other hand, the Brazilian Congress does not seem to fit with the “intermittent checks and balances” model. Congress functioned as an effective oversight institution, not only during the impeachment of Collor, but on many other occasions thereafter.12 In any case, it is difficult to assess whether the ability of Congress to oversee the executive is continuous or intermittent, since oversight does not necessarily imply “policing” controls but rather effective action when necessary (McCubbins and Schwartz 1987).13 In addition, James Madison’s fear of “stable divisions of political conflict” has proved to be correct. When political parties become the main basis of government, incentives to check the executive diminish. In sum, the differences in the basic structures of government systems—their defining characteristics—are not the main determinants of how governments function. Contrary to what Moe and Caldwell have argued (1994, 172 and 192), government systems do not come with a genetic code predetermining all their “traits.” Institutions and procedures, although generally associated to a certain system of government, exist independently and can have a significant impact on the way governments are run in democratic societies. In reality, the transformation of the presidential system underway in some Latin American countries is not incompatible with the historical origins of parliamentarism. Parliamentary institutions emerged from the

126

Argelina Cheibub Figueiredo

secular transformation of the relationship between monarchs, the executive, and the parliament, giving rise to the cabinet, a parliamentary executive. Presidentialism, in contrast, is an institutional arrangement designed to introduce the separation between these powers, but it is not immune to the practical difficulties that governments face, and certainly not to institutional diffusion. A final note of caution regarding generalizations about “Latin American presidentialism” is appropriate, however. There is still insufficient systematic evidence of a regional trend, and the theoretical framework of some analyses remains unsatisfactory. The data collected by Marsteintredet and Berntzen (2008, 90) on Latin American presidential breakdowns, using functional equivalents of parliamentary institutions (early elections and no-confidence votes) or recalls, show that eleven out of thirteen cases occurred in only four countries (Argentina, Bolivia, Ecuador, and Peru). In addition, there was only one impeachment in the other cases observed, with the exception of Paraguay, where one attempt succeeded and another failed. All of this suggests that the inclination commonly found in the literature to identify these traits with Latin American presidentialism may downplay the role of other, perhaps noninstitutional, factors that make these countries more susceptible to these kinds of outcomes than others. Or it may obscure the fact that these changes can be associated with more general historical trends leading to a convergence in the way that governments work.

Notes * I would like to thank Fernando Limongi, José Antônio Cheibub, and Marcus Figueiredo for their comments on an earlier version of this chapter, and Mariana Llanos and Leiv Marsteintredet, who helped to improve this chapter above and beyond their duties as editors. Responsibility for any shortcomings is mine, of course. 1. I cannot think of any factor related to the stage of democratization of the United States and Brazil that might account for differences in the removal of Nixon and Collor. 2. Brazil is part of a group of Latin American countries that followed the model established by the U.S. Constitution of 1787 (Pérez-Liñán 2005, 59). 3. See Huber (1996). For Brazil, see Figueiredo and Limongi (2008). 4. See the Web site of the Institute of Research on Applied Economics (Instituto de Pesquisa Econômica Aplicada, IPEA), at www.ipeadata.gov.br. 5. See Pesquisa DataFolha, Folha de São Paulo, http://datafolha.folha.uol.com.br/ 6. Veja April 8, 1992, 28. 7. Veja August 26, 1992, 8. 8. It is true that congress was purged during the military dictatorship, but the opposition party, the MDB (now the PMDB), had its first electoral victory in 1976 and grew in strength from that time on (Kinzo 1988).

Presidential Government in Brazil

127

9. The scandal became known as the mensalão because of the allegation that deputies from the government coalition were being paid on a monthly (“mensal” in Portuguese) stipends by the president of the PT to secure favorable votes from them. 10. Avaliação de governo, Datafolha, Opinião Pública, http://datafolha.folha.uol. com.br/ 11. On the support of the coalition in the Chamber of Deputies, see Figueiredo and Limongi (2008); and in the Senate, see Neiva (2008). 12. There are many good examples of this. In 1993, for instance, a parliamentary investigation of the budget process led six members of congress, including the speaker of the lower house, to lose their mandate. It also led to significant changes in the rules regulating congressional consideration of the budget (Figueiredo and Limongi 2008). Another example occurred in 1999, when congressional pressure forced President Cardoso to fire one of his strongest ministers, Communication Minister Luís C. M. de Barros, immediately following the minister’s testimony in congress about allegations that he had favored one group competing in the auction for the privatization of a telecommunication company. On other occasions, opposition calls for the impeachment of presidents Cardoso or Lula, as during the mensalão scandal, failed not because Congress was unable to pursue that option but because a congressional majority could not be persuaded to support such action. 13. These authors contrast “police patrol” to “fire alarm” types of controls, the first implying systematic and routine work, the later being intermittent, only activated when necessary.

This page intentionally left blank

7 The Impeachment of Carlos Andrés Pérez and the Collapse of Venezuelan Partyarchy Rickard Lalander

Introduction On December 4, 1988 Carlos Andrés Pérez, popularly known as CAP, was elected president of Venezuela for the second time, representing the country’s hitherto strongest political party, Democratic Action (Acción Democrática, AD). At sixty-six years of age, Pérez took over the presidency with 53 percent of the vote. He inherited a state on the verge of bankruptcy, after a decade of deepening economic crisis with social and political repercussions. Pérez had promised to bring back the golden era of the oil bonanza, which had coincided with his first presidency in the mid 1970s (CAP 1). At that time, oil revenues had quadrupled and Pérez had used export income to expand the state and to provide employment and state subsidies, thus contributing to improved socioeconomic conditions. There was, then, a remarkable contrast between the contexts surrounding the two CAP presidencies. The second government (CAP 2) became the most dramatic presidential term since the establishment of democracy in 1958. Shortly after his inauguration in February 1989, the president introduced a neoliberal reform package (officially known as the paquete) to confront the economic crisis. It was backed by International Monetary Fund (IMF) technocrats but was not supported by the majority of the national AD leadership. There was popular protest in all major Venezuelan cities in almost immediate response to the CAP package, culminating in a weeklong binge of looting, burning of vehicles, and riots that is known by Venezuelans as the caracazo. Pérez ordered the military and police forces to control looters, and in the resulting tumult hundreds of citizens were killed.1 In 1992, there were two military coup attempts against Pérez, also reflecting the growing discontent

130

Rickard Lalander

with the government and, more broadly, the political system. Pérez did not complete his presidential term. In 1993 he was impeached for mismanaging state funds. It was the first and only time in Venezuelan democratic history to date that a president failed to complete his term in office. The crisis of the Venezuelan political system that culminated with the impeachment of Pérez is the subject of this chapter. In comparative research, the Pérez case is noted for its peculiarities: the president was elected with an absolute majority, and with at least initially strong party support in parliament. Further, the scandal that led to his impeachment was only the culminating point of a process marked by popular protest (Pérez-Liñán 2008; 2007; Marsteintredet 2008b; Valenzuela 2008). 2 The factors identified as causes for presidential breakdowns are, among others, related to regime type, levels of parliamentary support, coalition status, power-building (through inclusion or isolation), relations between the president and Congress before and during breakdown, public opinion, popular protest, and scandals (Pérez-Liñán 2007). These factors are considered in this chapter, but I argue that one must analyze the political actors and the political party system within which they operate in greater depth to understand the presidential breakdown. In what follows, I provide a historical account of Venezuelan democracy and its party system, with a focus on AD. I then explore Venezuelan “exceptionalism” and the economic crisis of the 1980s, after which I analyze the Pérez presidency, particularly his neoliberal and decentralizing reforms, and popular unrest, and the issues relevant to the impeachment process. The final section focuses on the aftermath of the impeachment.

Partyarchy, Factionalism, and Crisis To better understand the scope of the political crisis, we must examine the historical context and dominant actors in the political system. Modern Venezuela has long been considered a political and economic exception in Latin America: politically, because of its early democratization; economically, because of the massive income generated by the oil industry. While many of its neighbors were governed by military regimes, Venezuela’s democratic system seemed to function quite smoothly. Free and competitive elections were held regularly after 1958 following the overthrow of the dictator, Marcos Pérez Jiménez. From that time on, AD and the Christian Democratic Committee of Independent Electoral Politics (Comité de Organización Política Electoral Independiente, COPEI), two strongly centralized political parties, dominated the national political arena, often through strategic pacts and alliances. In the period 1973–1988, between them these parties captured 81–93 percent of the vote in every presidential election. The political parties have been the channels mediating political participation to such an extent that the system has been called a partyarchy

Collapse of Venezuelan Partyarchy

131

(partidocracia). Indeed, Michael Coppedge’s influential work on the behavior, functioning, and structure of the AD under democracy labels Venezuela the most extreme case of partyarchy in the democratic world; that is, “a democracy where political parties monopolize the formal political process and politicize society along party lines” (Coppedge 1994, 18). The political rules of the game were established with the 1958 Punto Fijo Pact. With it, AD and COPEI, together with the most influential representatives of Venezuelan society, agreed on the principles that would govern the political system and on methods to avoid challenges from the left and right of the political spectrum. The Venezuelan Communist Party (Partido Comunista Venezolano, PCV), an erstwhile AD and COPEI ally in the struggle against dictatorship, was excluded from the pact. 3 Since the pact, Venezuela can be described as a multiparty democracy even though only AD and the COPEI governed until 1992. AD propounded a social democratic doctrine focusing on the lower middle class in particular, and abandoned its formerly more leftist traits, while the COPEI emphasized social justice and focused on the entrepreneurial and urban middle classes (Lalander 2004). Political stability was achieved with a party-mediated patronage system through which the parties with access to state economic resources provided support to particular social groups, such as peasant and workers’ organizations. AD held a near hegemonic position within the national unions, dominating the Confederation of Workers of Venezuela (Confederación de Trabajadores de Venezuela, CTV), the main union confederation. The business and industry sectors have been principally represented by the Federation of Chambers and Associations of Commerce and Production (Federación de Cámaras y Asociaciones de Comercio y Producción, FEDECAMARAS). Both the CTV and FEDECAMARAS functioned as consultative partners and as the exclusive representatives of their corresponding social sectors. Until the 1980s, the Venezuelan economic system was based on state capitalism and import substitution interventionism. As regards the protagonist of this chapter, Carlos Andrés Pérez was born in the Andean state of Táchira in 1922, but his family moved to Caracas in the 1930s where the young Pérez finished his studies in philosophy and law. In 1938 he was among the cofounders of the Venezuelan Juvenile Association as a member of the National Democratic Party (Partido Democrático Nacional, PDN), which became AD in 1941 (Pérez et al. 2006, 1–22). From his adolescent years, Pérez was among the core leaders of the AD, and in 1945, when AD first came to power in a civil-military alliance, he was nominated secretary to President Rómulo Betancourt (the key founding leader of AD) and appointed to the Council of Ministers. After the decade of dictatorship under Marcos Pérez Jiménez (1948–1958), Pérez retained top positions within AD. After 1961, he was appointed minister of domestic affairs during the second Betancourt government. AD factionalism was apparent already at this time. Initially, and until 1958–1959 in particular, AD was an anti-imperialist socialist party. During

132

Rickard Lalander

the first decade of democracy, however, the party split twice, in 1963 and 1967, leading to the formation of the Revolutionary Left Movement (Movimiento de Izquierda Revolucionario, MIR) and the People’s Electoral Movement (Movimiento Electoral del Pueblo, MEP), parties that were both to the left of AD. With the formation of MEP, AD control of the CTV was threatened seriously for the first time. The MEP included former AD militants (adecos) among its supporters, including a majority of AD trade union leaders. This division and the ensuing internal crisis helped Rafael Caldera and the COPEI to win the presidential elections in the period 1969–1973. This was the period when the left was integrated smoothly into national politics, following Caldera’s legalization of the PCV and the signing of a peace agreement with guerrilla forces in 1969.4 With the election of Pérez to the presidency in 1973, there had been a consensus within AD regarding the nomination process. In the 1978 electoral campaign, however, the conservative faction, or old guard, backed Jaime Lusinchi, while the reformist faction (the pérecistas) supported Luís Piñerúa. A split was ultimately avoided, but Piñerúa lost to COPEI candidate Luís Herrera Campíns, who governed Venezuela between 1979 and 1983. In 1983, AD successfully launched Lusinchi as a presidential candidate, this time following a consensual nomination process. 5 Before the 1988 elections, AD reformists and conservatives were divided again, although the party ultimately rallied under Pérez for a second term.6 Despite the official consensus backing Pérez, the factionalist divisions and personal rivalries remained and proved to be crucial during the CAP 2 presidency. Although the 1980s is remembered as the “lost decade” in Latin America, an overview of political and institutional developments tells another story. Democracy in the Andean countries spread to other countries in the region, and Venezuela and Colombia lost the “exceptional” democratic status they had enjoyed since 1958. This put Venezuela, as well as Colombia and Mexico, under pressure. Venezuela’s relatively stable democratic system came to be seen in a new light after democracy took root elsewhere in Latin America. Venezuelan society ceased to be exceptional and became “Latin Americanized” in the 1980s: from being considered the continent’s political and economic exception, Venezuela was suddenly seen to suffer from the same problems as other Latin American countries, including poverty, class grievances, unequal income distribution, and an inefficient overly bureaucratized state. During the bonanza years, it had been relatively easy for the centralized party leadership in the state administration to satisfy social and sectoral demands with oil income. For the Venezuelan population, the first blow came during the COPEI government of Luis Herrera Campíns. On “Black Friday,” on February 18, 1983, the currency was dramatically devalued to deal with the foreign debt crisis. The devaluation also had a negative impact on AD and COPEI clientelistic networks. This led to growing social discontent with the traditional political leadership, which found expression in increased electoral abstention, particularly from 1988 onward. Taking a

Collapse of Venezuelan Partyarchy

133

longer-term view, both AD and the COPEI had failed to improve the lot of the urban poor. The Lusinchi government had drastically cut subsidies when the economic situation had begun to deteriorate. Increased unwillingness of citizens to tolerate political and business corruption worsened the situation, as opposition politicians, academics, and ordinary citizens alike linked socioeconomic decline with corruption.

Reforms and Unrest: The Return of Carlos Andrés Pérez The year 1989 was a turning point in Venezuelan democratic history. Pérez had been elected with a strong parliamentary support, and AD was the dominant party with a plurality of 48 percent of the seats in both chambers. When CAP returned to the presidency, the rivalry between him and centralist hard-liners within the party was already rampant. Further, Pérez felt that imposing strong party discipline might backfire, and he consulted mainly with reliable friends. He formed a cabinet dominated by radical pro-market and technocratic academics from the Superior Institute of Administration Studies (Instituto de Estudios Superiores de Administración, IESA), including Ricardo Hausmann, Moisés Naím, and Miguel Rodríguez.7 The neoliberal reform package to address the acute economic crisis was announced as “the great turnaround” (el gran viraje) and included the privatization of public enterprises, currency devaluation, tax reform, deregulation, and tariff reduction. The price of goods produced by state industries were also increased, including that of oil and, consequently, gasoline, which led to increased public transportation tariffs.8 Initially, the passage of the economic measures, which mirrored those proposed in the COPEI programme (Lalander 2004, 130), was facilitated by the popular disillusionment caused by the crisis of credibility of AD and the COPEI. But the reform process was fragile, as CAP was isolated within his own party, and he lacked party and legislative backing. The shock caused by the package was aggravated further because the reforms did not reflect Pérez’s recognized political profile. In 1988, he had publicly stated that he would reject neoliberalist adjustment policies if he was elected. From 1987 onward he had attacked the International Monetary Fund in the mass media, calling it “a bomb that only kills people” (in Romero 1999, 42–43). The “great turnaround” was not the only dramatic reform Pérez introduced in 1989. In December 1989, following a presidential decree on political decentralization, the first direct elections were held for regional state governors and municipal mayors.9 Before the reform, the centralized leadership of the governing party had systematically blurred the distinctions between the party and public spheres, and regional state governors were nominated by the president or the national executive board of the governing party. With the decentralized 1989 elections, new left-wing parties triumphed in key states

134

Rickard Lalander

such as Bolívar (Causa R under Andrés Velásquez) and the industrial state of Aragua (the Movement to Socialism, Movimiento al Socialismo, MAS) under Carlos Tablante, shattering the myth of a classless consensual Venezuela that had dominated during the AD-COPEI heyday. Causa R also triumphed in the Caracas Libertador municipality with Aristóbulo Istúriz, who became a key minister in various Hugo Chávez cabinets after 1999. Thus, largely as a result of decentralization from 1989 onward, leftist (including radical) politicians began to enter government and state institutions at distinct politicalterritorial levels. One can interpret decentralization as an attempt to save democracy by sacrificing the status quo (the partyarchy). It is not surprising that such a dramatic political reform occurred in 1989: after the caracazo, pressure for political reform had grown, and social mobilization promoted by leftist parties such as the MAS and Causa R played a crucial role in forcing Pérez to decree decentralization. There was also international pressure from the multilateral banks, and Pérez also used decentralization for resolving internal AD conflicts (he was supported by adeco leaders at the regional and local levels, hoping to benefit with the reform). Both AD and the COPEI were divided about decentralization. Subnational level party members had pressed for an opening of the political system since the early 1980s, but many establishment leaders feared that they would lose social, political, and economic status (as it turned out, events after 1989 proved them right). We should not see decentralization as a measure adopted by a president panicking in a difficult political climate: the reform was a strategic maneuver to handle factional disputes within AD and a compromise of sorts with leftist actors. The economic and political credibility also contributed to the reform (Lalander 2004). Decentralization probably saved Venezuelan democracy from its worst sociopolitical crisis ever, probably dampened the scope for social uprisings during the 1992 coup attempts, and allowed alternative actors to enter the state and political arenas; but the reform sacrificed the old system and failed to save CAP. Venezuelan society experienced rapid change and its political institutions fell behind, failing to adapt to the new context. The pattern of selective inclusion that characterized the consultative politics of partyarchy was seriously questioned. In May 1989, the CTV promoted the first general against an AD president in thirty-one years. Although the CTV had been the key labor representative under the Punto Fijo regime, Causa R union activism had challenged the hitherto hegemonic position of the CTV (Crisp 2000, 174–178; Lalander 2004).10 Despite the difficulties, it is important to note that the great turnaround did have a positive macroeconomic effect. By early 1991, economy was growing, inflation decreasing, investment returning, and fiscal deficit controlled. For a brief period, CAP enjoyed significant public approval, but by the fall of 1991 the president had come under fire again. Street protests by students and other social groups grew, another CTV-promoted general strike took

Collapse of Venezuelan Partyarchy

135

place, and the public and entrepreneurial sector became increasingly critical of the government despite the economic improvements. I agree with Corrales that the most important explanatory factor for the regime crisis was the hostile relationship between the executive and the ruling party. The president lacked party support in Congress; indeed, AD behaved like an aggressive opposition party (Corrales 2002, 146–148; Figueredo, author interview (AI), September 26, 2007).11 President Pérez refused to abandon his reform programme throughout 1992, despite the growing social and political pressures in the form of strikes, demonstrations, military coup attempts, and criticism both from citizens and politicians of all stripes. Pérez acted against the mandate he gained in the 1988 election by implementing the neoliberal reform package. In contrast with Argentina, where President Carlos Saúl Menem was able to adopt a messianic pose as the leader who would save the country from hyperinflation through market-oriented reform, Venezuelan political culture did not allow Pérez to strike a similar pose. Also, in contrast with Pérez, Menem eventually secured party support for his reforms. Corrales (2002) concludes that the structure of AD was too rigid, whereas the Peronist Party in Argentina was sufficiently flexible to accept the reforms. The attempted coups in 1992, the first led by Lieutenant Colonel Hugo Chávez, Commander Francisco Arias Cárdenas, and other officers of the Revolutionary Bolivarian Movement 200 (Movimiento Bolivariano Revolucionario, MBR-200)12 reflected the very widespread discontent with the AD-COPEI political establishment. Chávez himself later admitted that the coup was triggered by the caracazo and the way in which police and military force was used against the population. Chávez was arrested after the coup attempt of February 4, but he managed to address the nation on television in what became a highly symbolic act, delivering the message of Bolivarian revolution even though the coup had failed “for now.”13 Many analysts suggest that Chávez indirectly won the 1998 presidential election as a result of that short televised moment in 1992. It is important to underline the hugely important role that Causa R played in the Chávez political project. Initially, his movement was intimately connected with Causa R, particularly the group led by Pablo Medina, and many of its members were appointed to important posts in Chávez’s governments (Buxton 2001, 5; Lalander 2004, 198–201). Chávez has admitted that Causa R and MEP sectors were committed to the coup. Also highly important are the chavista elements in the discourse of Rafael Caldera. At the time of the attempted coup, when AD and COPEI leaders expressed lynch-mob feelings about the rebels, Caldera defended the rebel officers in a Senate meeting, which, along with his antitraditionalist approach, played in his favor in the 1993 presidential elections. Pérez never recovered the initiative after the first coup attempt of February 1992. His popularity declined, and the MAS and Causa R took advantage of negative public feelings toward the political establishment, insisting on the urgent need for constitutional reform,

136

Rickard Lalander

including the establishment of mechanisms to remove a sitting president and other elected authorities. The impeachment of Fernando Collor de Melo in Brazil in 1992 also raised hopes among those opposed to CAP who believed that a similar process was possible in Venezuela. In this political climate, many Venezuelans interpreted the economic crisis as the direct result of corruption and power abuse by the leaders of the two dominant parties. Citizens and politicians blurred the distinction between the president’s mandate, the hierarchic party system, corruption, and economic policies, and CAP came to embody Venezuelans’ political disillusionment and regime crisis.

The Impeachment of Carlos Andrés Pérez As noted above, a particularity of the Venezuelan process is that a specific public scandal was not the first event that eventually led to the breakdown of the presidency. There had been social unrest and protests long before 1993. Already in the late 1970s, at the end of his first presidential mandate, Pérez had been accused of corruption. At that time, the biggest scandal involved his purchase of an overpriced refrigerated container ship, the Sierra Nevada, although the case did not lead to judicial action against the president. Repeated public scandals generated unrest among the population, but since corruption cases remained unsolved, public anxiety was counteracted by a much deeper perception that this was a familiar and immovable reality. Powerful people rarely ended up in jail; they remained powerful and rich, and were even admired for their ability to preserve their position and assets. Historian Fernando Coronil cites a Venezuelan journalist, who noted during the trial of one scandal in the late 1970s that: “In Venezuela only people who are fools (pendejos), or broke, go to jail. Clever people (vivos) never do” (Coronil 1997, 358). During the Lusinchi presidency, corruption and mismanagement problems intensified, which also affected the efficiency of state institutions and industry. The most criticized source of corruption in the 1980s involved the office of Differential Exchange Rates Regime (Régimen de Cambios Diferenciales, RECADI) created in 1983 at the end of the Herrera Campíns government. The RECADI established a differential currency exchange system, and AD and COPEI politicians and business leaders systematically used the favorable exchange rates for private gain. During its six years of existence, under Lusinchi, the RECADI became Venezuela’s largest source of illegal enrichment ever (Ellner 2008, 82–83; see also Capriles 1992). In November 1989, corruption charges were presented against former president Lusinchi and four of his ministers for involvement in a seven billion dollar RECADI fraud. The charges against the ministers were dropped in 1990, but in late 1991 the senate confirmed the accusation against Lusinchi for corruption and abuse of power during his presidency, although he was never sentenced by a court and only morally and politically condemned.

Collapse of Venezuelan Partyarchy

137

Pérez adopted various strategies in addition to decentralization to recover popular credibility. In 1990 he called on a group of distinguished Venezuelans—los notables—including novelist Arturo Uslar Pietri, to establish a consultative board to find a way out of the crisis. However, the notables soon became critical and mobilized against the government. In October 1991 they wrote a public letter—the carta de los notables—to the Supreme Court of Justice, calling for the removal of the Court’s judges because they had been unable to punish corruption such as that affecting RECADI. The letter had an immediate public impact, and only months after the coup attempt in February 1992, the court was reconstituted and AD lost its previous control over it (Pérez Perdomo 1993, 105–106; Ellner 1993, 14). Pérez had abolished the RECADI when he had introduced his package proposing unified exchange rates in February 1989. The day before the agency closed, however, the Ministry of Domestic Affairs and the Presidential Secretariat converted 250 million bolivars (approximately US$ 17 million), held in a presidential fund for secret operations, into dollars with the preferential exchange rate, even though they were not authorized to manage these funds.14 On November 8, 1992, leftist journalist and politician José Vicente Rangel publicly denounced the operation, only nineteen days before the second coup attempt against Pérez. Andrés Galdo, an investigative journalist for the El Nacional daily (which had previously backed Pérez), also contributed decisively to fomenting popular outrage and to the decisions later adopted by the Supreme Court and Attorney General. Each journalist had a different interpretation of what had happened, but both shaped the political climate and the conditions to prosecute the president. Galdo claimed that even if the US$ 17.2 million were a mere “drop in the ocean” for Pérez, one of the richest men in Latin America, the money had been laundered through intermediaries who bought up or invested in foreign companies so that Pérez would not appear to be involved. The president denied the accusations and stated that he did not have any foreign bank accounts (Ellner 1993, 14–15). Rangel, backed by leaders of Causa R (Lalander 2004; interviews, Benítez Tello, AI, June 29, 1997; July 15, 2007), claimed that Pérez had used the secret funds for political reasons, to support allies on the continent, such as providing security for President Violeta Chamorro in Nicaragua. He also alleged that some of the money was used against domestic enemies, and that Salvadorian mercenaries had been contracted to eliminate persons on a hit list that included Attorney General Ramón Escobar Salom, Galdo, and Rangel himself (Ellner 1993, 15). Rangel’s accusation became crucial in the judicial process against Pérez and was included was evidences, and Rangel acted as a key witness for the prosecution. After the second coup attempt and the Rangel and Galdo accusations against CAP in November 1992, Congress debated a constitutional amendment to shorten the presidential term, but AD and the COPEI, which faced declining popular support, feared that this would work against them. Since Pérez’s “notables strategy” had failed to generate credibility, he nominated

138

Rickard Lalander

two COPEI members to his cabinet after the November coup attempt to promote a government of unity (Crisp 2000, 179), and in March 1993, CAP also proposed a constitutional reform, desperately hoping to salvage the situation, but it gained no traction. Finally, on March 20, 1993, Attorney General Escovar Salom charged Pérez with the mismanagement of state funds.15 Two months later, on May 20, the reconstituted Supreme Court approved the accusation, declaring that there was sufficient evidence to prosecute the president. A day later, the Senate decided unanimously to strip the president of immunity and automatically suspend him from office. The Attorney General played a crucial role in the prosecution of CAP. Escovar Salom had been appointed by Congress in 1989 and, unlike previous attorneys general, he rapidly showed his nonpartisan credentials. While his predecessors had tended to be discreet figures dependent on the executive, Escovar Salom took a public stand on the caracazo and other politically delicate issues, particularly those related to corruption. After the decision to impeach the president, there was a new crisis over the power vacuum in the presidency. The president of the Congress, Octavio Lepage, who had served as minister during CAP 1, replaced Pérez on May 21 as provisional president of the Republic while Congress debated how to deal with the remainder of the presidential term (at that time the post of vice president did not exist).16 According to Article 180 of the 1961 Constitution, an interim president could remain in power for ninety days. Although Article 187 stipulated that Congress had to replace the interim president within thirty days, the Lepage solution lasted only two weeks, and on June 4 Congress elected independent Senator Ramón J. Velásquez as interim president to conclude the term. It is important to emphasize that Velásquez was seen as a compromise approved by AD and the COPEI. The AD Congress deputies were generally more supportive and granted Velásquez decree authority to legislate. In July 1993 Senator Cristóbal Fernández Daló of the MAS presented the Attorney General with photocopies of bank transactions by Pérez and his private secretary, mistress and future wife Cecilia Matos. Hundreds of thousands of dollars had been transferred between accounts in the United States, France, the Philippines, Hong Kong, and Switzerland. Since Matos had also denied having secret accounts, these documents outraged the public because both she and Pérez were caught lying (Ellner 1993, 14–15). On August 30, 1993 the Congress removed Pérez from office permanently as the court evidence against him was clear. The impeachment confirmed the high level of political discontent and seemed to justify the most aggressive criticisms of the CAP government, although some believe that Pérez was AD’s political scapegoat (Lalander 2004; Figueredo, AI September 26, 2007). Pérez reacted to the condemnation by denying the charges and declaring in a public document that he was the victim of political justice and had acted solely in the interests of the Republic. He reminded people of his early

Collapse of Venezuelan Partyarchy

139

efforts to promote freedom and democracy as a member of AD, and his more recent role in promoting decentralization: I put all my efforts into the political reforms. In this way we began to convert the Presidency of the Republic from an absolute power into a moderate one. Four parties share, or have shared, power throughout this presidential period. Two elections of governors and mayors have taken place in these four years. I reclaim my special leading role in this reform process, which aims toward the democratization of power and unmistakable national participation. (Pérez 1993, 2) Pérez thus took much of the credit for the process of democratization aspects and reform of the two-party system through decentralization. His ties between decentralizing reform leaders within AD, such as Antonio Ledezma and Claudio Fermín, confirm this, as do the former president’s familiarity with international trends and his personal annoyance with the Alfaro Ucero-wing of AD. Regarding the supposed mismanagement of the 250 million bolivars, CAP denied any personal enrichment, although he confessed that he had used state funds to support political activities in Nicaragua, El Salvador, and Bolivia: “All that I could do to help democracy in Nicaragua and to end the war in El Salvador, well, I did it.”17 From a legal standpoint, the arguments of the Supreme Court suffered from various weaknesses. Pérez Perdomo notes the Court’s decision was understandable because the president had become such an unpopular figure and his removal was greeted by the public with relief. However, in the 250page decision of the Court, hardly any references were made to why and how the use of 250 million bolivars from the secret presidential security funds constituted a crime, and the right to defense was violated, as Pérez was not heard in court (Pérez Perdomo 1993, 111). In effect, he was impeached and imprisoned for a crime that had not even been clearly defined. CAP himself calls the judicial process part of a civic conspiracy to overthrow him from power: It is important to emphasize that the campaign of the celebrated notables did not concentrate exclusively on the President, or AD, but was also directed against the system, the political parties, the electoral system and all the mechanisms of democracy. Before the coup attempt of February 4, there were declarations by Arturo Uslar Pietri and other notables about the possibility of a coup d’état, which opened the floodgates, the vestibule of the military adventure . . . There is no doubt concerning the participation of Caldera . . . Caldera’s discourse gave much greater consistency, deeper cohesion, to the destabilizing action . . . Caldera’s discourse unified the groups that were acting against the government, although seemingly disconnected. They dared

140

Rickard Lalander

to heedlessly call for the renunciation of the president of the Republic and the Supreme Court of Justice, for a repeal referendum, and furthermore to incite public opinion against Congress. It was a coordinated action. (Pérez et al. 2006, 385–386)18 Pérez’s impeachment can be seen as resulting from declining levels of tolerance of corruption, and as a civilian coup against the president. Equally, it is important to emphasize the link between future presidents Caldera and Chávez, regardless of whether there was a conspiracy or not, as CAP and his defenders and others, worried about regime crisis and the condition of democracy in Venezuela, have done. Pérez Perdomo concludes that, from a short-term perspective, the decisions by the Supreme Court of Justice regarding Lusinchi and Pérez were popular among citizens and politicians, and that the Venezuelan institutional system was put under pressure to show that judicial decisions could in fact also affect the most powerful (Pérez Perdomo 1993, 111). While awaiting trial, CAP had been placed under house arrest, as Venezuelan law prohibits incarcerating persons over seventy years of age. However, in May 1994 the Supreme Court of Justice declared the arrest and detention of Pérez, who was dramatically moved from his home to the El Junquito prison. CAP insisted that he should remain in a public jail and therefore spent the first seventy days there before returning to house arrest. On May 30, 1996, the Supreme Court issued its final sentence, after two years and four months of house arrest, accusing the former president of relatively minor mismanagement of state funds, and of using the secret funds to finance military protection for Nicaraguan presidential candidate Violeta Chamorro, which he had already confessed to. After the sentence, CAP was also expelled from AD. In recent comparative research on presidential impeachments in Latin America, Pérez-Liñán concludes that impeachment is most likely to occur when a president does not control Congress (either because the president’s party is in a minority or because a rival faction of the presidential party works against him), and when the mass media exposes a political scandal systematically (Pérez-Liñán 2007, 3). These conditions existed in Venezuela, but other factors, such as the behavior of opposition forces, transformation of civil and political society, and changes in political culture at the time, were also important.

The Aftermath: From Caldera to Chávez During the Velásquez interim period in 1993, the Supreme Court indicted former president Lusinchi for using military funds in 1988 to purchase jeeps that were distributed in exchange for political favors, although Congress eventually decided to decriminalize Lusinchi’s misuse of public funds.19 With the victory of Rafael Caldera in the presidential election of 1993, AD and the COPEI lost control of the presidency for the first time since

Collapse of Venezuelan Partyarchy

141

1958. Caldera had left COPEI and formed a new party—Convergencia— which joined with the MAS and a group of smaller, mainly leftist, parties. The party system was very fragmented and volatile compared with what it was in 1988, when AD and the COPEI together captured 93 percent of the vote. The difference between the votes obtained by AD (23.6 percent), COPEI (22.7 percent), and Causa R (21.9 percent) was very small indeed. On its own, Caldera’s party would have come fourth, but with the 10 percent of votes for the MAS and the leftist alliance, he gained enough votes to win (30.5 percent).20 Causa R was perceived as a dangerous threat to AD and the COPEI, given its electoral promise to radically reform political structures. Several scholars, politicians, and citizens claim that a potential Causa R victory was averted by electoral fraud committed by AD and the COPEI (Buxton 2001, 5; Lalander 2004, 178–179). Causa R was the party closest to Chávez, and a large number of radical militants, unionists, and parliamentary deputies later joined the ranks of chavismo. The growing mystique around the imprisoned Chávez was also a sign of widespread disillusionment with the political establishment. Caldera amnestied Chávez, Arias Cárdenas, and other rebels and coup leaders in 1994 to dampen leftist antiestablishment sentiments.21 In the economic field, the tax reform initiated by Pérez had continued under the interim presidency of Velásquez, but Congress was reluctant to legislate on the issue, including an unpopular value added tax measure and had granted the president decree authority to legislate on economic and financial matters. Caldera had campaigned on an anti-VAT strategy and after his election he eliminated the tax but made up for it by increasing taxes on companies and individuals. The Caldera government was much more interventionist than either CAP or Velásquez, establishing price and exchange rate controls (Crisp 2000, 186). In 1994, a financial crisis provoked a banking collapse, the government took over the administration of eighteen private banks, and 322 bankers fled the country with huge sums of capital to escape arrest for economic fraud (Ellner 2008, 100). From 1995 onward, the ConvergenciaMAS government took a neoliberal turn, as did the orthodox (anti-Pérez) faction of AD. In fact, Caldera introduced a package of IMF-backed economic reforms in 1996, the so-called Agenda Venezuela. Caldera nominated Teodoro Petkoff, the legendary former communist guerrilla leader and cofounder of MAS, to the post of minister of planning from 1996 onward, partly as a strategy to maintain the support of the left. Although the Caldera alliance formed the government, there existed a tacit and sometimes open alliance between AD and Caldera. Thus, although officially independent and although he had won the election with his antitraditionalist discourse and by defending the coup leaders and Chávez in 1992, Caldera was not actually hostile to the traditional political parties. In this context, the parliamentary alliance between Caldera and AD was rather contradictory. In response to this, the COPEI, the MAS, and Causa R allied in Congress, establishing the “triple alliance.” This was particularly confusing for the MAS, which was

142

Rickard Lalander

both in opposition and in government. Overall, the situation of the left was unclear by the end of Caldera’s government, and in 1997 and 1998 both the MAS and Causa R splintered, with the more leftist factions supporting the presidential candidature of Hugo Chávez in 1998. New legal proceedings were initiated against CAP in 1998, who was placed under house arrest again. However, with the 1998 elections, CAP was elected senator on the slate of his new party Apertura (Opening). This gave him senatorial immunity only briefly, since the Chávez government suspended immunity and reopened the case against him and his mistress Cecilia Matos. The CAP case is still open. Since 1998, Pérez has lived in exile in Miami, New York, or the Dominican Republic and has continued to contribute regularly to the political debate, frequently attacking the Chávez government and his own former party, AD. In February 2005, the Caracas Attorney, Indira Josefina Mora Padilla, presented a new arrest warrant against Pérez for allegedly ordering the Armed Forces to quell riots during the 1989 caracazo. 22 In April 2008, already seriously ill, Pérez expressed a desire to return to Venezuela for “humanitarian reasons,” but he has been unable to return as judicial proceedings are ongoing.

Conclusions This chapter outlines the probable causes of the impeachment of Venezuelan president Carlos Andrés Pérez in 1993 by examining the role of the political actors involved in the process. Particular emphasis has been placed on the collapse of the AD-COPEI partyarchic model. Partyarchy once guaranteed the Venezuelan political system relative stability, but it also provoked its own demise. It hindered the institutionalization of political accountability between the electorate and political leaders. There was no really effective opposition and the two dominant parties efficiently obstructed access to political society for alternative actors, which resulted in a loss of legitimacy for the system and of credibility for traditional political leaders. It is important to look for the deeper causes of presidential breakdowns, impeachments, and even political scandals. The presidential breakdown in Venezuela was the result of complex causes. After the oil bonanza years in the 1970s, the economic situation worsened throughout the 1980s, with negative social and political repercussions. From the late 1980s onward, the hegemonic bipartisan system was undermined and challenged by new political actors, some emerging from the decentralization after 1989. The caracazo, the two military coup attempts of 1992, and the popular sympathy elicited by the rebels or coup leaders provide evidence of widespread social and political discontent. It has also been noted here that the case of Venezuela is not that typical, since the impeachment was not the direct result of a scandal. Corruption

Collapse of Venezuelan Partyarchy

143

charges against CAP in 1993 (the scandal) functioned rather as an emergency exit from the acute regime crisis. The scandal helped the actors who wished to remove CAP and was an excuse to proceed with impeachment, probably much more so than in the case of Collor de Melo in Brazil or Cubas in Paraguay. Further, officially Pérez had strong parliamentary representation, but AD chose not to support the president’s reform policies. President Pérez thus lacked party and grassroots support and suffered from the hostile relationship between the executive power and the ruling party that facilitated the judicial process. The impeachment procedures were also reviewed here, as were some of the political consequences of presidential breakdown. The judicial procedure was problematic: on the one hand, it signalled that the powerful were not immune; on the other, it is unclear whether there was a clearly defined crime or whether Pérez was simply sacrificed to punish the corruption of the political system in a context of regime crisis. Clearly, CAP was impeached as an indirect result of the exhaustion of partyarchy. Economic crisis and deteriorating socioeconomic conditions caused political discontent and protest. The Pérez economic package was perceived as a betrayal by citizens and members of AD. The conservative wing of AD argued that Pérez had stabbed them in the back by his isolated policymaking, although others argue the opposite: that AD betrayed Pérez (Steve Ellner, AI December 29, 2008). CAP was certainly a scapegoat of the collapse of partyarchy and of the revenge against the political culture of corruption. There are also several direct links between the political events under CAP 2 and the left-wing Chávez government after 1999, not only dating from the first coup attempt, but also connected with decentralization, the second Caldera government, and shifts in political organizations and protest strategies. The 1989–1993 period—the most dramatic years of Venezuelan democracy until then—can be interpreted as a first phase of a protest cycle that contributed to the rise to national power of chavismo and as the beginning of the end of the AD-COPEI partyarchic era.

Notes 1. According to some sources, deaths were in the thousands (Ellner 2008, 91). 2. It is relevant to consider these issues in other countries, as in the case of Raúl Cubas in Paraguay. 3. Initially three political parties participated in the pact, but the Republican Democratic Union (Unión Republicana Democrática, URD) withdrew from the alliance early on and its political importance declined. In 1958–1968, the PCV operated clandestinely as a guerrilla force. On different aspects of the Punto Fijo regime, see, among others: Coppedge (1994), Ellner (2008), Buxton (2001), Crisp (2000), Lalander (2004), Coronil (1997), and Corrales (2002). 4. Almost immediately, the PCV divided, and two new parties were formed in 1971, the MAS and Causa R, which became important new oppositional actors.

144

Rickard Lalander

5. For a more detailed analysis of the presidential nomination strategies within AD, see Coppedge (1994). 6. According to the 1961 Constitution, presidents must wait two terms before they can run for the presidency again. 7. Hausmann and Naím were later recruited to multilateral banks. 8. The measure that most angered the population was increased public transportation costs, a situation that was aggravated when bus drivers raised prices even more on their own initiative to cover their own rising costs. According to Steve Ellner, some observers believe AD militants actually promoted the initial uprising of February 27, 1989 (Ellner, AI December 29, 2008). 9. In 1984, a commission for political and administrative reform, the Presidential Commission for the Reform of the State (Comisión Presidencial para la Reforma del Estado, COPRE) was set up under the Lusinchi presidency to examine government decentralization, but reform was halted until 1989 by politicians who feared that decentralization would threaten their economic and/or social status. 10. FEDECAMARAS also faced opposition from competing business associations. 11. At this time, AD was divided between the weakened reformist wing of Pérez and two orthodox factions; one that was more conservative and the other more moderate and previously closer to the Pérez faction at times. In the October 1991 AD central executive council elections, the two orthodox factions united to remove the pérecistas from the party leadership. 12. The MBR-200 was formed in 1983 (the name alludes to the 200th anniversary of the birth of the Liberator, Simón Bolívar). The MBR-200 can be described as nationalist, populist, and antiestablishment. Officially, the movement supported Bolivarianism (the ideas of Bolívar), combating corruption, the establishment of a “moral” political power (in addition to the traditional executive, legislative, and judicial branches), and the promotion of popular participation. The MBR gained strength in 1989 as it took advantage of the instability caused by the caracazo. The second coup attempt on November 27 was led by a group of generals, among them Air Brigadier General Francisco Visconti (who was neither Bolivarian nor chavista). 13. An important aspect of Chávez’ televised speech was that he took responsibility of the “failed” coup. Venezuelans were not used to high ranking leaders taking the blame for failed actions, so this increased Chávez’s symbolic power. 14. In the 1960s, the secret operations fund had been used to combat the leftist guerrilla. 15. The two ministers involved in the case, Alejandro Izaguirre (Domestic Affairs) and Reinaldo Figueredo (Presidency), were also charged. 16. The vice presidency was introduced with the 1999 Constitution. 17. Interview with Pérez in Cambio 16, quoted in Urbina (2003, 163). 18. Pérez also emphasized that pressure from Causa R contributed to events, although he does not consider the radicals part of the conspiracy. 19. Lusinchi fled to Costa Rica in 1993, after his immunity as senator was suspended, and met with his mistress and future wife Blanca Ibáñez (who was also prosecuted for corruption). By 1997 the “Jeeps case” against Lusinchi was dropped. 20. AD, COPEI, and Causa R candidates were launched mainly for their merits as mayor (Claudio Fermín for AD) or governors (Oswaldo Álvarez Paz for the

Collapse of Venezuelan Partyarchy

145

COPEI and Andrés Velásquez for Causa R) after 1989, which is evidence of the importance of decentralization and local/regional political leadership. 21. Not only was Arias pardoned by Caldera but he was also appointed president of the State Programme for Mother and Infant Nutrition (PAMI). 22. Venezuela Analítica 2005, http://www.analitica.com /global/xvas/xvade/ imprimir.asp, accessed November 26, 2008.

This page intentionally left blank

8 Paraguay: The President in his General’s Labyrinth Detlef Nolte

Introduction On March 26, 1999, faced with imminent impeachment in the Senate and violent protests on the streets of Asunción, Paraguayan president Raúl Cubas resigned. He took refuge at the Brazilian embassy, and later left the country. However, Cubas had not been the main protagonist of this crisis: in a complicated power game resulting from Paraguay’s peculiar transition to democracy, the main player had been General Lino Oviedo. At a first glance, the story of Paraguayan presidential breakdown resembles the plot of a thriller, featuring intrigue, betrayal, heroic young protesters marching in the streets, and even murder. The roles of the different political actors change sometimes, and political circumstances define whether they are labelled heroes or villains. In academic terms, the Paraguayan case contains most of the elements that have proved to have explanatory value in comparative studies of presidential breakdowns: a minority president (Negretto 2006; Hochstetler 2006; Kim and Bahry 2008); a fragile legislative shield against impeachment (Pérez-Liñán 2007; 2008); strategic mistakes committed by the president and weak presidential leadership (Ollier 2008; Pérez-Liñán 2007); a strong parliamentary opposition acting in unison (Morgenstern et al. 2008); the introduction of greater flexibility to presidential systems normally seen as rigid (Marsteintredet and Berntzen 2008); the supremacy of parliament in a situation of conflict between the president and Congress (Pérez-Liñán 2005; 2007); scandals (Pérez-Liñán 2007); and, finally, violent mass demonstrations against the government (Hochstetler 2006; Kim and Bahry 2008; Morgenstern et al. 2008, Pérez-Liñán 2008).

148

Detlef Nolte

The Paraguayan case, therefore, does not defy conventional wisdom or constitute a deviant case, although the relative weight of each factor in the presidential breakdown may differ from that of other cases. In this chapter, I argue that political parties or factions rather than street demonstrators were the principal drivers of the fall of President Cubas. Street protest came later, and the pressure they generated played a significant role in the political power struggle. I further argue that the role played by succession rules and possible presidential successors, the way in which conflicts were framed by the actors, and the symbolic elements of mass demonstrations were also important factors explaining the Paraguayan case.

Democracy by Default: The Paraguayan Transition Process In order to understand the complex events that led President Cubas to surrender power, we must turn to the specific mode of transition to democracy in Paraguay. Paraguay was a latecomer in the third wave of democratization in Latin America, coming at “the tail end of the democratic revolution” (Falcoff 1998). The transition was initiated in 1989 after thirty-five years of dictatorship under General Stroessner, and was the result of intra-elite power struggles. It was therefore characterized by substantial continuity in terms of the political elite and state bureaucracy (public administration, the military, and the judiciary). The whole country was covered by a tight-knit network of control and patronage administered by the Colorado Party (Partido Colorado, PC) (Abente 1995), and even after democracy was installed, the party remained in power, thus becoming one of the longest-governing parties in Latin America (it was in power between 1947 and 2008). In spite of the appearance of regime institutionalization—with the armed forces and the Colorado Party as its main pillars—the Stroessner regime had been essentially a personalist authoritarian regime. During the second half of the 1980s, under adverse economic conditions that strained the prebendary basis of the regime, clashes within the Colorado Party increased. At this critical juncture, choosing the man who would succeed the ailing dictator constituted a serious challenge for the survival of personalist rule, and, in the end, the regime elite split over the succession question, which was closely linked to the control of the prebendary system. Indeed, General Andrés Rodríguez, who led the coup that deposed Stroessner on February 2–3, 1989, had amassed a fortune during the Stroessner regime.1 One sector of the regime elite—the so-called traditionalist sector supported by General Rodríguez—tried to institutionalize the regime and to increase the influence of the Colorado Party in a non-personalist transition. Another sector of hardcore Stroessner supporters, the so-called militantes combatientes stronistas, planned to guarantee the continuity of the regime by promoting as successor Stroessner’s son, Gustavo. In 1987, the militant sector conquered the

President in his General’s Labyrinth

149

leadership of the Colorado Party in a fraudulent and violent party convention (Abente 1995, 310–311). Next, they tried to purge the military of traditionalists and, at the same time, pave the way for the ascension of Gustavo Stroessner. At the end of 1988 and the beginning of 1989, Stroessner père had presided over a major reshuffle of the armed forces, as a result of which many long-serving colonels had been forced to retire; his son, together with another fifty-five officers, had been promoted to colonel; and military hard-liners had been appointed to important commands. At the same time, some confidants of General Rodríguez, who was at that time the commander of the powerful First Army Corps, had been relocated or retired, and there were rumors that Rodríguez himself would soon be affected by the army restructuring (Sondrol 1992, 112). General Rodríguez reacted before that could happen. On February 2, 1989, a relatively small number of troops loyal to General Rodríguez and the navy commander started the coup to bring down Stroessner and his cronies. The dictator surrendered after a final battle with the presidential guard in which 150 to 250 soldiers died (Nickson 1989, 185; Roett 1989, 138). Stroessner was sent to Brazil where he died in exile in 2006. Interestingly, in his first proclamation defending the coup, General Rodríguez listed the following as the main reasons for deposing the dictator: the reunification of the Colorado Party; the restoration of the dignity of the armed forces; and only in third place, the beginning of a process of democratization (Martini and Mar 1997, 67; Sondrol 1992, 112). In an international context favorable to democracy, and in order to legitimize the coup at home and stabilize his rule, General Rodríguez opted for a democratic opening. Only seventy-two hours after the coup, General Rodríguez signed a decree dissolving Congress and calling for new elections on May 1, 1989. It was clear that the opposition would not have enough time to organize effectively. 2 Furthermore, the election was to take place with the same deficient electoral registry of the Stroessner era (Abente 1995, 315–317), the electoral authorities were totally controlled by the government, and the electoral law automatically gave a two-thirds majority to the winning party. However, General Rodríguez had become popular with the coup and would have won in any case. With electoral participation of around 50 percent, Rodríguez won the election with 74 percent of the votes against the candidate of the Authentic Radical Liberal Party (Partido Liberal Radical Auténtico, PLRA), Domingo Laíno. Although the clientelistic system remained largely in place and although the Colorado Party maintained its grip on power, the presidency of Andrés Rodríguez (1989–1993) did give rise to a process of democratization. Two important institutional reforms took place in those years: the electoral law (March 1990) was changed, and a new constitution (1992) was enacted. The new electoral code explicitly prohibited military personnel on active duty from party affiliation, and those already affiliated to a political party had to abstain from party activities.

150

Detlef Nolte

The Ouster of President Cubas: Setting the Stage Although the Rodríguez presidency (1989–1993) initiated the democratization process, the highly politicized military continued to play a troublesome political role throughout the 1990s. The institution viewed itself as part of the Colorado Party and openly took partisan positions. As Martini and Mar noted in the mid-1990s, “Until the contradiction of the existence of a civilian-military pact within this democratic state is resolved, there will be no shortage of other military figures prepared to continue the long Paraguayan tradition of generals holding political power” (1997, 71). Below, I introduce the central players of the 1990s. Lino Oviedo, a close collaborator of Rodríguez during Stroessner’s rule, tried to use his military position as a political springboard. The subsequent power struggle within the Colorado Party set the stage for the ouster of President Cubas and, paradoxically, brought to an end the era of military strongmen in Paraguayan politics. Lino Oviedo was one of the military officers who would have been retired if Gustavo Stroessner were further promoted. During the coup, he commanded an armored unit and played an active role in the events that led to the dictator’s ouster. Once the coup was over, and after Stroessner loyalists had been purged from the armed forces, Oviedo quickly ascended the ranks, becoming army general in only four years. He supported Juan Carlos Wasmosy as Rodríguez’s presidential successor for the Colorado Party in a fraudulent internal election that penalized Luís María Argaña. 3 After he became president in 1993, Wasmosy—a businessman and member of the landowner’s association who had become rich during the construction of the Itaipú hydroelectric power plant—rewarded Oviedo by appointing him commander-in-chief of the army. Luis Maria Argaña had been a member of the parliament and president of the Supreme Court under Stroessner, and was one the authors of the 1967 constitution. Nevertheless, in the final stages of the regime, as a representative of the traditionalist wing of the Colorado Party, he had distanced himself from Stroessner and had played an active part in his removal. Although he was one of the favorites to replace Stroessner, he ultimately had to settle for the post of foreign minister under President Rodríguez.4 He was dismissed on July 23, 1990 after declaring that he was even ready to conspire to ensure that the Colorado Party would not be displaced from power. 5 Argaña’s dismissal did not put an end to the conflicts within the Colorado Party between the now dominant traditionalist faction and other more reformist currents.6 During the Constitutional Assembly of 1992, the Argaña faction supported the inclusion of a non-reelection clause for the president and a ban on presidential relatives from running for the presidency (Costa and Ayala 1996, 39). For some, this was one of the causes of the estrangement between Argaña and Rodríguez (Bareiro 1999, 229). Furthermore, Argaña favored reducing the influence of the military within the party (Riquelme and Riquelme 1997, 58), which also pitted him against Oviedo.

President in his General’s Labyrinth

151

Although he had backed the coup against Stroessner, Argaña still retained good contacts with the former dictator’s hard-core supporters.7 Abroad, the majority candidate of the Colorado Party was regarded with suspicion and perceived as a “Stroessner nostalgic,” so there was no foreign protest over the electoral fraud that consecrated Wasmosy as the Colorado candidate. Although some of his political adversaries questioned the origin of his fortune, Wasmosy gave the party a more liberal image. Eventually, however, the general who had helped him become a candidate came to be seen as the power behind the throne. Oviedo made no secret of his ambition to become the next president of Paraguay, and, openly violating the rule prohibiting party political activities for active members of the armed forces and the police, he engaged actively in political proselytism. In August 1995, Congress passed a motion that called for Oviedo’s dismissal as commander of the army because of this violation of the electoral law, but judicial proceedings against him fizzled out because the judiciary was politicized and corrupt. Again and again, Oviedo tried to influence Wasmosy’s political maneuvering and nominations, and it was also rumored that the two men clashed over competing business interests.

Political and Judicial Engineering President Wasmosy, a civil engineer by profession, confronted his army commander in a risky maneuver. He believed that he could win if Oviedo dared to disobey his orders, and he managed to outmaneuver the general, proving that he was a talent when it came to political engineering.8 In the morning of April 22, 1996, President Wasmosy met with the commanders of the armed forces, including General Oviedo. After intense discussions, Wasmosy ordered Oviedo to retire. The general refused to obey and started to put pressure on the president to revoke the order (including threat of a coup). The president soon got the backing of the ambassadors of the United States and the Common Southern Market (MERCOSUR) countries, who supported Wasmosy’s decision and condemned all attempts to subvert democratic rule in Paraguay on the very day that Oviedo’s retirement had been ordered. In a press conference, the president confirmed that he had ordered the army commander to retire, but that General Oviedo had refused to comply and was in rebellion in the headquarter of the First Cavalry Division. The government could count on the support of the Presidential Guard, the police forces, the Navy, and the Air Force, but it was unclear whether some army units would support a revolt. In this uncertain context, the president took refuge at the U.S. Ambassador’s residence, and the vice president hid out in at the Marine Corps headquarters. Oviedo tried to negotiate the terms of his retirement, demanding that the president and the vice president should step down as well. During the night, crucial military commanders were contacted by foreign diplomats and politicians.

152

Detlef Nolte

In the early morning hours of April 23, Congress was able to muster a quorum to declare its support for the president. People kept up with events on the radio, on which calls were issued for street protests; and the labor unions called for a general strike. People came out onto the streets—mostly high school and university students—in defense of democracy. On the same day, the Permanent Council of the OAS approved a declaration condemning the threat to the constitutional order by General Oviedo, and giving its full and resolute support to President Wasmosy. Meanwhile, in Asunción, negotiations to settle the crisis continued. As a solution to the crisis, Wasmosy offered the Ministry of Defense to Oviedo; he accepted, hoping to pursue his political ambitions from his new post. While the president’s decision was initially supported by the foreign ambassadors, the secretary general of the OAS, César Gaviria, and leading politicians of the Colorado and Liberal parties (Costa and Ayala 1996, 108–121), it provoked protests in the streets and pressure from all parties,9 and President Wasmosy was forced to reverse it. On April 24, in a ceremony attended by the president, Oviedo resigned as commander of the army. With Oviedo out of the army and without a position in the government, the next round of the power struggle had both political and judicial components. A judge initiated an investigation of the events that had taken place between April 22 and 26, charged Oviedo with military rebellion and insubordination, and ordered his arrest. Oviedo’s lawyers used all judicial means at their disposal until an appeals court acquitted Oviedo in December 1996.10 Four month later, the Supreme Court overturned this decision, clearing the way for further legal action against Oviedo. Meanwhile, President Wasmosy tried to create an “institutionalist safeguard” through promotions and early retirement (Morínigo 1999, 188) to prevent Oviedo from making a comeback. When Oviedo was still in government, a power struggle had begun within the governing Colorado Party. President Wasmosy had tried to organize his own faction, but his adversary in the 1993 primaries, Luís Maria Argaña, still controlled a powerful faction.11 Argaña and his faction won control of the Colorado Party on April 28, 1996 against the Wasmosy candidate, the then vice president Seifart. Since Argaña had kept a low profile during the Wasmosy-Oviedo conflict, he probably also got votes from Oviedo’s supporters (Costa and Ayala 1996, 121). Soon after leaving the government, however, Oviedo organized his own faction, which he rather strangely named National Unity of Ethical Colorados (Unidad Nacional de Colorados Éticos, UNACE), to campaign for the Colorado presidential nomination for the 1998 elections. Competing against Luís María Argaña and the Wasmosy faction candidate, Oviedo won the internal elections in September 1997 and became the official candidate of the Colorado Party. His vice president was Raúl Cubas, a business man and a friend since their studies at military college. Cubas had resigned from his post as Wasmosy’s finance minister in April 1996, in solidarity with Oviedo.

President in his General’s Labyrinth

153

However, President Wasmosy had convened a special military tribunal to charge Oviedo for insubordination in April 1996, and on March 8, 1998, the tribunal sentenced the general to ten years in prison for rebellion and ordered that he be dishonorably discharged from the army. Oviedo appealed to the Supreme Court, which upheld the sentence in a divided vote (5-4) on April 17, 1998.12 As a result, Oviedo was barred from running in the impending presidential election. According to the law, Raúl Cubas would replace him as official candidate for the Colorado Party, and Luís Argaña would become his running mate. Running with campaign slogans such as “Vote Cubas and Free Oviedo” and “Cubas to the Government, Oviedo to Power,” Cubas promised to pardon Oviedo if elected. Oviedo’s adversaries quickly reacted: before the president-elect could take office, the outgoing Congress passed a law signed by Wasmosy on June 23, 1998, which restricted the president’s authority to pardon. According to the new law, convicts would have to complete at least half of their prison terms before a president could pardon them.

End Game: Blood on the Streets13 On May 10, 1998, the strange Cubas-Argaña Colorado Party ticket won the presidential election with a clear majority of 54 percent. In Congress, the Colorado Party had a comfortable majority in both chambers, even though it was highly fractionalized. Cubas had twenty-nine out of the eighty deputies in the Chamber, and the remaining fifty-one included sixteen deputies from opposing Colorado factions.14 On August 18, Cuba’s first working day as president and his third day in office, the president issued a decree freeing General Oviedo. Because he was legally barred from pardoning the general, he commuted the prison term to a shorter sentence (three month instead of 10 years), allowing for the immediate released of the general. At the same time, he began to purge the army of officers who had been loyal to former president Wasmosy during the 1996 coup attempt, and initiated measures to reinstate officers who had been removed from service at that time. Last but not least, the president changed the members of the extraordinary military court to retry Oviedo on the 1996 charges. The new court overturned the prior sentence and dismissed all charges against the general. Oviedo was back on the political stage. But the general’s adversaries did not stand idly by. On August 20, 1998, the majority of the opposition parties in Congress, together with the antiOviedo factions of the Colorado Party, approved a resolution that condemned the decision of president Cubas and called on the Supreme Court to assess the constitutionality of the presidential decree. On December 2, the Supreme Court cast a divided vote (5-4) overturning the presidential decree and ordering Oviedo’s return to prison.15 President Cubas send the Supreme Court verdict to the special military tribunal that had acquitted Oviedo, in

154

Detlef Nolte

an attempt to portray the conflict as a matter of conflicting judicial interpretations rather than as a conflict between different state powers. Meanwhile, on the political front, the leadership of the Colorado Party, dominated by the supporters of Argaña, decided to expel Oviedo from the party on December 5, while Oviedo and his supporters put pressure on the Supreme Court. Perhaps encouraged by positive public opinion surveys and the mobilization of his supporters, Oviedo issued some declarations threatening the Supreme Court. In a mass meeting outside Congress on December 10, he called for the resignation of the five Supreme Court justices who had voted against the constitutionality of the presidential decree that set him free.16 Molotov cocktails exploded outside the homes of two of the justices, and there were violent protests outside the Supreme Court building. Radio stations that sympathized with Oviedo agitated his supporters with violent diatribes against Congress and the Supreme Court. President Cubas called for the resignation of two presidential appointees on the Judicial Council, the institution that was responsible, inter alia, for nominating the Supreme Court judges. The Supreme Court struck back, however. On February 5, its president issued a direct order to President Cubas to arrest Oviedo within seventy-two hours. Cubas reacted by accusing the Supreme Court of attempting to overthrow the government. Moreover, he argued once again that Oviedo had been freed by a military tribunal and not by presidential decree. Thus, although Cubas insisted that the conflict was a matter of different judicial interpretations, he also depicted the conflict as a plot led by politicized Supreme Court judges, factions of the Colorado Party, and the parliamentary opposition. On February 11, when Cubas did not comply with the order, a joint session of both houses of Congress passed a resolution that supported the Supreme Court order with a substantial majority of seventy-four votes against thirtyeight.17 However, a second resolution to impeach the president did not get the necessary two-third majority in the Chamber of Deputies,18 so the political standoff continued. As foreseen by constitutional succession rules, a successful impeachment would give the presidency to Vice President Argaña. The latter therefore began to hold meetings with opposition parties and to discuss a government of national unity. Argaña also met the U.S. ambassador (Morínigo 1999, 203). At the same time, Oviedo’s opposition tried to generate the momentum to attain a majority for impeachment. As a first step, the Chamber of Deputies changed its own internal rules to improve the odds (Abente 1999). The Constitution stipulated that an impeachment had to be initiated by a two-thirds majority of the attending deputies, and that the Senate then had to decide by a two-third majority. The Chamber of Deputy house rules provided that when all eighty deputies were present, a two-third majority required fifty-four votes (mathematically, fifty-three), and that the speaker of the chamber had no right to vote. Both these rules were changed: the quorum was set at fifty-three deputies and the speaker was allowed to cast a vote.

President in his General’s Labyrinth

155

Social tensions mounted as the Chamber organized the impeachment vote for April 7. There had already been a general strike against the government’s economic policy in November 1998, and in January 1999 the economic situation had worsened after the Brazilian currency devaluation. Peasants had protested and demanded relief to compensate for falling cotton prices. There had been a deadlock in Congress between the president and the opposition majority on economic issues; moreover, given the background of conflict, Congress had blocked presidential military promotions. Conflicts had also escalated within the Colorado Party. On March 14 1999, a majority of the National Convention voted to extend the mandate of the party leadership, in which the Argaña faction had a clear majority. The Convention ended in chaos when supporters of Oviedo and Argaña attacked each other with chairs. Oviedo supporters left the place to organize a parallel Convention, and tried to occupy the party headquarters. They received the support of President Cubas. The escalating crisis came to a sudden end within a week, however. On Tuesday March 23, 1999, Vice President Argaña and his bodyguard were assassinated on the streets of Asunción by a commando dressed in military uniform. The political opposition and the media believed that General Oviedo and President Cubas were involved in the crime, and the voices demanding the resignation of President Cubas began to multiply. The labor federations called a general strike, and during the next days public transport in Asunción was partially paralyzed. Cubas tried to deflate the situation by placing General Oviedo under arrest, but the efforts to impeach the president gathered pace nonetheless. Moreover, Oviedo declared that he was not under arrest, but that he had taken himself into custody to clarify his judicial situation. A day after the assassination of Argaña, in a surprising early morning session initiated at 8:00, the Chamber of Deputies secured the two-third majority required for a presidential impeachment by a single vote: forty-nine deputies voted in favor, twenty-four against, while seven deputies (several of whom had previously opposed the impeachment) were absent or prevented from attending at the time of the vote. It was speculated that two Colorado deputies switched their vote in favor of the impeachment (Bareiro 1999, 231). Meanwhile, supporters of the impeachment—mainly young people from the Youths for Democracy movement (Jóvenes por la Democracia, JD)—had gathered outside Congress to prevent Oviedo’s followers from assaulting Congress. Coincidentally, the same day that Argaña had been assassinated, 20,000 peasants had arrived in Asunción to demand that their debts to state banks be written off.19 The congressional majority reacted in record time to this demand, approving the law and thus increasing peasant interest in supporting Congress in the institutional conflict with the president. As a result, during the most critical moments of the crisis there were between 3,000 and 5,000 peasants in the center of Asunción. Some questions regarding the mobilization remained unanswered. On the one hand, most Paraguayan narratives stress its spontaneous nature, and the

156

Detlef Nolte

fact that it was undertaken in defense of democracy and in favor of impeachment (Bareiro 1999). Interestingly, Argentine newspapers were more circumspect, and also mentioned acts of violence by pro-government demonstrators and vandalism by demonstrators from both sides. Further, the perception was that the Argaña-faction of the Colorado Party and the opposition parties played a greater role in organizing the demonstrations. Paraguayan publications also hint that the JD movement originated with youth members of the opposition parties and opposition Colorado factions, and that there were initially only very few protesters—around hundred (Bareiro 1999, 227). One Paraguayan observer wrote that “in the square, in strictly quantitative terms, there were no more than 800 citizens who participated as protagonists during the critical days. On the days of major mobilization, there were 10,000 people at most, including the peasants” (Morínigo 1999, 212). 20 The numbers issue is ultimately not that important, as it was not “street pressure” that brought President Cubas down. What mattered more was the symbolic power of the dominant presence of pro-opposition protesters in the square facing Congress. 21 From Tuesday, March 23, to Friday, March 26, there were skirmishes between antigovernment demonstrators and General Oviedo’s supporters. Although the latter were significantly outnumbered by the antigovernment demonstrators, the anti-government demonstrators were continuously harassed and attacked by police forces. On Friday, March 26, the situation escalated. Late in the evening, snipers opened fire on the crowd from a nearby building. In the early morning of Saturday, March 27, several hundred people were wounded, and seven of the demonstrators died. On Saturday, March 27, soldiers of the Marine Corps moved on to separate the pro- and antigovernment demonstrators. The Senate convened to continue impeachment deliberations and determined that it would convene again on Monday, March 29. The pressure on President Cubas to resign grew as the tide of demonstrations spread to other Paraguayan cities. The Paraguayan Church and foreign diplomats (such as the Brazilian and U.S. ambassadors and the Papal Nuncio) tried to mediate and find a peaceful solution to the crisis. Brazilian President Fernando Henrique Cardoso telephoned Cubas and offered to guarantee his and his family’s security if he were to step down. It was clear that the armed forces—purged of Oviedo supporters during the Wasmosy presidency—would respect the result of the impeachment procedure. Only the cavalry supported Cubas and Oviedo. Finally, President Cubas addressed the nation in a televised speech on the evening of Sunday, March 28. He anticipated a negative vote in the Senate, and he announced that he did not want to be responsible for more bloodshed in the streets and that he had decided to resign, adding that his resignation had been forced by a conspiracy. Two hours earlier, General Oviedo had escaped to Argentina, where President Menem had granted him asylum. 22 Cubas sought refuge in the Brazilian embassy first, and he was later allowed

President in his General’s Labyrinth

157

to go to Brazil where he was immediately granted asylum. 23 A couple of hours after his resignation, a joint session of both houses of Congress swore in Luís Ángel González Macchi, head of the Senate and first in the line of succession, and a member of the Argaña faction of the Colorado Party, for the remainder of the presidential term (until August 2003).

Why did Cubas become a Failed President? There are different interpretations of the causes and implications of the fall of Raúl Cubas, most of which view the course and the outcome of the political crisis in a positive light. For Abente, the events of March 1999 “marked the emergence of a strong citizenry” (1999, 5). Zagorski argues that popular mobilization forced Cubas and Oviedo to step aside and prevented the return of dictatorship (2003, 107–108). Pérez-Liñán attributes the downfall of Cubas to lack of support in Congress (2007). In his view, the legislative shield had been weakened by the president’s confrontational style, political scandals (the assassination of Argaña and the killing of demonstrators), and, most importantly, political mobilization in the streets. These interpretations are not mutually exclusive, but according to my reading of the Paraguayan case, President Cubas was akin to a chess piece eliminated in a complex power game. First, Cubas was a minority president and, anyway, Paraguayan presidents are constitutionally weak when faced with opposition in Congress. 24 Second, many interests converged to get rid of the president. Cubas’s successor(s)—Vice President Argaña and, after his assassination, Senate President González Macchi, both from the same opposition Colorado faction—enjoyed majority support in Congress. For the Colorado Party, therefore, the Cubas ouster did not mean losing power. This broadly indicates the importance of taking into account succession rules and possible successors when examining the causes of presidential breakdown (Ollier 2008). For the opposition, an Argaña or González Macchi government promised a better bargaining position and, in all likelihood, participation in government. Further, the still influential former president Wasmosy also had an interest in getting rid of Cubas: the judiciary had started proceedings against Wasmosy, which were supported by the government. In short, Cubas’s allegation that his resignation was the result of a political conspiracy may be exaggerated but it is not entirely unreasonable. Third, an initially internal Colorado Party political struggle turned into an institutional conflict between the president, the Congress (or the congressional majority), and the Supreme Court. In such conflicts, the president only prevails when the military supports a coup (Pérez-Liñán 2005), but as the armed forces were divided, an unconstitutional solution was not available. The Paraguayan case transmits a clear message for presidents: not to let a conflict with your political adversaries become framed as a conflict between

158

Detlef Nolte

government bodies (Congress or the Supreme Court), or as a constitutional transgression. Fourth, there were political mobilizations and social protests against the government. As comparative studies demonstrate, this factor is quite important in explaining presidential failure. The mobilization alone was not decisive in bringing down the president, but it may have helped to produce more votes in favor of impeachment and to discourage those undecided within the armed forces to support a coup. It was the specific interaction between the parties in Congress and the demonstrators in the streets that tipped the balance. The demonstrators did not force the political parties to act but rather supported the parties in their power struggle, a fact that differentiates the Paraguayan case from other cases. 25 This case also suggests that future research should focus more on how protests are framed, and on how one of the conflicting parties asserts or achieves discursive hegemony. The media played an active part in mobilizing people and in reporting continuously about the unfolding events (Otter 1999). It created a shield against the repressive activities of the police forces. Fifth, the president suffered a further loss of legitimacy because of the violence and bloodshed in the streets. We may never know who assassinated Vice President Argaña and who gave the orders to open fire on demonstrators, but it sufficed that most people believed that the Cubas-Oviedo duo was responsible for the acts of violence. In this sense, the Paraguayan case combines the element of scandal mentioned by various authors (Pérez-Liñán 2007; Hochstetler 2006; Kim and Bahry 2008), with the element of violence stressed by Hochstetler (2008). Some recent cases of failed presidencies in Latin America—such as that of Fernando de la Rúa in Argentina and Sánchez de Lozada in Bolivia—demonstrate that in a crisis situation the killing of protesters can be the last straw that breaks the back of a president. Finally, the international community, and the MERCORSUR neighbors in particular, did not support an unconstitutional solution to the crisis. Instead, they pressured President Cubas to resign and offered him a safe haven. The Paraguayan crisis of 1996–1999 demonstrates the growing importance of the Latin American community of democracies, and especially of the MERCOSUR, in promoting democratic stability. To some extent, the influence of external actors in the solution of the Paraguayan crisis resembles what happened in the Dominican Republic in 1994 (see chapter 10, this volume).

Conclusion: The Presidential Breakdown and its Aftermath The imprisonment of Oviedo and the failure of President Cubas provide only prima facie evidence that democratic institutions and the rule of law were strengthened and consolidated in Paraguay. General Oviedo was indicted

President in his General’s Labyrinth

159

and exonerated of the same crimes several times, only by different tribunals and at different times. Events since 1989 clearly show that what is regarded as a criminal offense in Paraguay depends a great deal on who controls the judiciary. Recent Paraguayan history recalls a phrase attributed to the former Brazilian president Getúlio Vargas, “for my friends anything, for my enemies the law.” After being sentenced to ten years of prison for military rebellion in 1998, the Supreme Court granted Oviedo habeas corpus in July 2007, and stet lifted the preventive arrest order pending against Oviedo for the assassination of Vice President Luís Argaña and for the later killing of seven student protesters (at the time of writing, in March 2009, these cases remain closed). Two months later, in September 2007, a military tribunal freed Oviedo on parole on the grounds that he had served more than half of his ten-year sentence without fomenting political unrest. Shortly after the release, in October 2007, the Supreme Court (in an 8-1 vote) absolved him of conspiring to depose President Wasmosy in1996. 26 Thus, Oviedo was able to run for the presidency again in 2008. It has been speculated that President Nicanor Duarte (2003–2008) had used his influence over the judiciary to free Oviedo and thus split the opposition and improve the electoral chances of the Colorado Party. The Colorado Party was weakened by the outcome of the conflict. In retrospect, the crisis was a step toward the change of government that occurred nearly ten years later, in 2008. The supporters of Lino Oviedo definitely left the Colorado Party and formed their own party, the National Union Party of Ethical Citizens (Partido Unión Nacional de Ciudadanos Éticos, PUNACE). The candidate of the opposition Liberal Party won the vice presidency in 2000, but the Colorado Party again won the presidential election in 2003. Five years later, in the presidential election of 2008, the Colorado Party candidate came second with 31 percent of the vote. Back in the political game, Lino Oviedo ended third with 22 percent of the vote, perhaps frustrating another Colorado Party victory. The fall of president Cubas also undermined the political influence of the armed forces and shifted the balance of civil-military relations. After the political meddling of General Oviedo, politicians of all stripes—including those of the Colorado Party—had an interest in reducing the independent political influence of the armed forces in politics. The military, too, wanted to move in the direction of greater professionalization, because the politicization of the institution had resulted in internal divisions and purges. This does not mean, however, that military promotions since then have not followed any political criteria. As regards the emergence of a strong citizenry during the 1999 crisis, clearly, civil society became a more important actor during the crisis. In 1999, when only ten years had passed since the end of the Stroessner dictatorship, an important motivation for the demonstrators was to defend democratic institutions. However, civil action did not seem to take place on a permanent basis. After the March 1999 events, there have been recurrent episodes of

160

Detlef Nolte

political mobilization against the government, especially in times of political scandal, or when leaders attempt to perpetuate their power. However, if we take Latinobarometer surveys into account, it becomes clear that support for democracy has been low among Paraguayans (between 30 and 50 percent), and only rose after 2005. Compared to the rest of Latin America, a high percentage of the population is open to authoritarian solutions (between 30 and 44 percent) (www.latinobarometro.org). Finally, I wish to stress that the Paraguayan case helps us to differentiate more clearly between the statistical probabilities that a certain event will occur and the historical contingent factors that influence final outcomes. Irrespective of a very adverse constellation, the failure of President Cubas was not predetermined: Cubas and Oviedo committed significant strategic mistakes; some institutional decisions against Cubas (by the Supreme Court and Congress) were very close; and, last but not least, the adversaries of President Cubas were able to capitalize on the coincidental presence of peasant demonstrators in Asunción during the critical moments of the crisis.

Notes 1. Many generals dedicated their time to economic activities unconnected with their professional duties. The statute of the armed forces explicitly allowed this. Ultimately, the low level of army professionalization may have facilitated the coup against Stroessner (Riquelme 1993, 192–193; Sondrol 1992). 2. The main opposition party was the PLRA, a multi-class, center-left party outlawed under Stroessner. The more leftist parties did not play an important role after the fall of Stroessner (Caballero 1993; Medina 1997; Riquelme and Riquelme 1997). 3. Argaña had won the primary elections of December 27, 1992 by a narrow margin, gaining 50 percent of the votes against 43 percent (Nolte 1994). Only after various extraordinary party congresses and the election of a new party electoral tribunal was Wasmosy declared the winner of the elections on March 4, 1993. 4. Argaña publicly supported Rodríguez after the coup, and it was believed that he would become the candidate of the Colorado Party in 1993 in exchange for the favour (Roett 1989, 139; New York Times February 8, 1989). 5. Latin America Weekly Report (LAWR), 1990–30, 12. 6. LAWR 1990–48, 8–9. 7. After the coup, Argaña tried to whitewash Stroessner, who, he felt, had made some mistakes due to his failing health and because he was poorly advised. In an interview Argaña said: “There were 32 or 33 years of great government. Stroessner was a great President for many years. Unfortunately, during his last two years, he was surrounded by a group of irresponsible and greedy politicians with no sense of nationalism” (Paraguay’s Chief Shakes up Army, New York Times February 8, 1989). 8. The following description of the events is based on Costa and Ayala (1996); Yuste and Valiente (1996); Valenzuela (1997); Zagorski (2003); the Paraguayan journal Análisis del Mes; and the LAWR. 9. Análisis del Mes 126, April 1996.

President in his General’s Labyrinth

161

10. Análisis del Mes 134, 1996, 3. 11. The Argaña faction (comprising 10 out of 20 senators, and 22 out of 38 deputies of the Colorado Party) often opposed President Wasmosy in congress (Nickson 1997, 186). 12. On April 14, 1998, Public Prosecutor Aníbal Cabrera had recommended that the Supreme Court should annul the military proceedings against Oviedo because they had been improperly conducted (LAWR Southern Cone Report, March 1998, 1). 13. This section is based on the electronic archives of the Argentine newspapers La Nación, Clarín and Página 12, which covered the events in Asunción with their own reporters; the LAWR; Abente (1999); Bareiro (1999); Morínigo (1999); Zagorski (2003); and Pérez-Liñán (2007). 14. See LAWR Southern Cone, October 1998, 6; La Nación February 12, 1999. 15. Clarín, December 3, 1998. 16. LAWR Southern Cone Report, October 1998, 6. 17. Clarín, February 12, 1999. 18. LAWR July 1999, 74. 19. The estimates of how many peasants were protesting vary from 3,000 to 32,000 depending on the source (Otter 1999; Bareiro 1999; Clarín, March 27, 1999; Morínigo 1999). Perhaps the difference can be explained by the fact that only some of the peasants stayed in Asunción after the protest ended. 20. Hochstetler writes that no president has failed without street demonstrations numbering at least 10,000 people and lasting at least a few days (2008, 65). This is narrowly confirmed by the Paraguayan case. 21. On this, I agree with Morínigo, who writes: “. . . the square will become a symbol of citizen support. The oviedista goal was to try to remove the youths from the square, and the aim of the youths and citizens in general was to remain there. The square became a fundamental point of reference of citizen support, in other words, whoever controlled the square did so because their actions generated consensus support among the population” (1999, 206) 22. Oviedo had to hide when Fernando de la Rúa was elected president, but was found and arrested in Foz do Iguazú in Brazil in June 2000. He fought against his extradition and the courts finally granted him asylum. In June 2004, Oviedo returned to Paraguay, where he was immediately placed under military custody. 23. Cubas returned to Paraguay in 2002 and was judged for conspiring to murder Argaña, but he was later acquitted. 24. The study by Payne et al. states that only the Nicaraguan president has less legislative power than the president of Paraguay (2007, 96). 25. The Paraguayan case does not fit with Pérez-Liñán’s argument that “popular uprisings, rather than legislative action, are the main force behind the collapse of elected administrations. Congressional action provides the framework—the best available framework—for a government collapse, but it often plays a merely reactive role” (2007, 206). 26. The Supreme Court verdict was based on the testimony of seven retired army officers, who denied that Oviedo had disobeyed the presidential order to surrender the army command (Latin American Brazil & Southern Cone Report, November 2007, 7).

This page intentionally left blank

IV

Presidential Breakdowns Restoring Democratic Order

This page intentionally left blank

9 Guarding Privileges and Saving the Day: Guatemalan Elites and the Settlement of the Serranazo Maren Christensen Bjune and Stina Petersen

In retrospect, it seems apparent that the government of Jorge Elías Serrano (1991–1993) was bound to break down. Political analysts, journalists, and other observers recognized the signs from Serrano’s first day in power: only 18 of the 116 members of Congress belonged to the President’s recently constituted party, the Solidarity Action Movement (Movimiento de Acción Solidaria, MAS). His presidency reads like a tale of a breakdown foretold. Serrano ultimately opted for a self-coup (autogolpe). And, indeed, there were many factors in Guatemala in the spring of 1993 that made a coup a definite possibility. A prolonged deadlock in Congress had left the president with little room for maneuver and few political allies, and his popularity was fading dramatically. Corruption scandals and protests and strikes against the government’s neoliberal economic policies aggravated the crisis. It was generally believed that in one way or another, President Serrano was on his way out. Interestingly, Serrano’s choice to break democratic rules failed. Within a fortnight of the self-coup, he had been forced to flee the country, constitutional order had been restored, and on June 5, 1993, a new interim president had been inaugurated. So if the authoritarian strategy failed, does this mean that democracy won? This chapter analyzes Serrano’s presidential breakdown and its impact on political developments in Guatemala. We argue that the rapid restoration of constitutional order was a result of the elite settlement of the crisis, and we particularly highlight the participation of de facto power structures in resolving the crisis: the armed forces and the business elite,

166

Maren C. Bjune and Stina Petersen

represented by the Coordinating Committee of Agricultural, Commercial, Industrial, and Financial Associations (Comité Coordinador de Asociaciones Agrícolas, Comerciales, Industriales y Financieras de Guatemala, CACIF). Analyses carried out immediately after the autogolpe largely see the failure of Serrano’s authoritarian plans as a significant step toward democracy (Jonas 2000a; McCleary 1999a; Cameron 1998; Villagrán 1993). We have the benefit of hindsight, so our conclusions are based on a longer-term perspective. We thereby challenge the optimistic assessments made in the immediate aftermath of the autogolpe. Looking at the same actors and choices now, the overall impact does not seem to be that uniform after all, considering, as we do, that, fifteen years later, Guatemalan democracy is still weak and fraudulent (Sánchez 2008). In order to explain Serrano’s failed coup and the impact of the presidential breakdown, we examine the role of the political, economic, and military actors of the regime. We underline the alliances they formed and the occurrence of an unprecedented civil society uprising. It is important to consider the role of President Serrano, his allies (or lack thereof), and his political enemies when assessing the outcome of the crisis. Today, Serrano is viewed as the embodiment of impunity in Guatemala, and, not surprisingly, most people hold him responsible for the political turbulence of May 1993.1 Problems with coalition-building and a corresponding instability and lack of cooperation in Congress (McCleary 1999a, 101–102) were striking features of the political situation leading up to Serrano’s attempted autogolpe. These are elements that are central in the Linzian perspective on the presidential system. In that view, the president’s opponents in Congress systematically choose conflict over cooperation and coalition-building, and presidents who, like Serrano, lack a majority both in Congress and in government, run a great risk of facing severe deadlock, instability, interruption, and, consequently, regime crisis. The difficulties in Congress may explain why Serrano—already notorious for adopting a confrontational line toward the legislature—chose authoritarian means in the attempt to stabilize his presidency. However, considering the flaws in the structures of the presidential system in Guatemala—a country where these structures are remarkably weak— will only lead to imperfect conclusions. We consider that at the time of the coup the Guatemalan regime was a tutelary democracy. 2 Before, during, and after the presidential breakdown, the power of the executive and legislative branches was significantly constrained by the de facto power of two nonelected and recognizably authoritarian institutions: the armed forces and the CACIF. Paradoxically, their efforts to protect their power structures and privileges during the serranazo were rewarded with widespread approval, as they successfully posed as guardians of the constitutional order and even the saviors of Guatemalan democracy.

Guarding Privileges and Saving the Day

167

From Military Rule to Tutelary Democracy: The Background of the Autogolpe Guatemalan institutions and political actors are variously described as “young,” “weak,” “unpredictable,” “corrupted,” “unstable,” “turbulent,” and “authoritarian.” The country has had a short democratic history and a long legacy of military rule and civil war. Following the brief experience of democratic rule between 1944 and 1954, 3 authoritarian regimes held sway in 1954–1985 based on an alliance between the armed forces and the oligarchy.4 After 1963, the armed forces established a military authoritarian regime and became a remarkably powerful institution. In 1982, however, divided by deep factionalism, the army initiated a controlled liberalization to enhance regime legitimacy and combat the guerrilla more effectively. In 1984, a constituent assembly was convened, and a year later, Vinicio Cerezo of the Guatemalan Christian Democratic Party (Democracia Cristiana Guatemalteca, DCG) was elected president of the Republic. His inauguration in 1986 marks the transition to civilian rule. The new constitution introduced democratic institutions such as a human rights ombudsman (Procuraduría de Derechos Humanos, PDH), a Constitutional Court (Corte Constitucional, CC), and a Supreme Electoral Tribunal (Tribunal Supremo Electoral, TSE). However, the center-left parties were still banned from participating in the general elections, and the army’s tutelary powers heavily constrained Cerezo’s government. Internal security was still in the hands of the army, and the president remained under the powerful control of the Presidential Military Guard (Estado Mayor Presidencial, EMP). 5 Indeed, there were two military coup attempts against him (Jonas 2000a; Schirmer 1998; Arévalo de León 1997; Aguilera 1993).6 The economic elite represented by the CACIF were also very influential. This corporate organization has acted as a pressure group on behalf of a small number of historically wealthy families (Valdez 2003; Casaús 1996; Dosal 1995) and has significantly circumscribed the powers of civilian governments. The CACIF favored political liberalization to increase the regime’s legitimacy and improve economic prospects, but it continued deploying strategies of confrontation and negotiation outside the democratic institutions (Fuentes and Cabrera 2005; McCleary 1999a).7 The CACIF has tended to stay out of party politics, but when President Serrano came to power in 1991, several ministers were recruited from the organization (Palencia 2002; Dosal 1995). The 1991 election of President Serrano is seen as an important step toward democracy, as it constituted the first handover of power from one civilian president to another following technically free and fair elections. Serrano, who was hitherto an unknown figure, set up his own party in 1987 to run for the presidency. His electoral victory was aided by the unpopularity of the

168

Maren C. Bjune and Stina Petersen

traditional parties and the disqualification of former general Ríos Montt as a candidate. Serrano’s party won only eighteen congressional seats, however, and the president relied on a clique of close friends in the cabinet, so his government remained fragile. Further, because of deep structural processes of political and economic liberalization, Guatemala was highly polarized and difficult to govern. According to opinion polls at the time, the population was deeply mistrustful of political parties, the legislature, and the judiciary. The country was experiencing a historical social and political apertura (opening) after three decades of civil war; social movements were growing in number and visibility; and a strategy of broad and inclusive peace talks was being promoted by the international community. The executive was also continuously challenged by the business elite and the military, both eager to control the liberalization and the direction and inclusiveness of the emerging democratic regime (Fernández Camacho 2004, 242; Jonas 2000b, 19; McCleary 1999a; Cameron 1998, 134; Martínez 1993; Villagrán 1993, 118).

The Autogolpe and Immediate Reactions It was in this context that President Serrano decided to stage the self-coup two-and-a-half years into his presidency, on May 25, 1993. Some key events influenced the timing of this radical decision. First, there was a stalemate between the resident and Congress that had left Serrano with little room for political maneuver. Throughout 1992, he had conveniently relied on a coalition with the DCG and the National Centrist Union (Unión del Centro Nacional, UCN). This had allowed him to approve legislation, including two bills to increase electricity rates.8 In February 1993, however, the DCG blocked a third executive proposal to raise electricity rates. Growing demands by legislators for bribes in exchange for votes contributed to the stalemate.9 Serrano was also under pressure from the head of the Supreme Court, who threatened to press corruption charges against him. According to McCleary, this threat was part of an attempt to force the president to ensure that the head of the court and his allies in Congress would obtain a valuable concession in the electricity sector. The coalition broke down in April, and in May the Christian Democrats ran for the municipal elections independently (1999a, 100–102). Second, Serrano’s efforts between January and May to relaunch the peace process, which was blocked over the issue of a truth commission, were unsuccessful. Serrano was under serious pressure from the CACIF and the military, which insisted on limiting the negotiation to the terms of a cease-fire. The army was especially concerned about the publication of the results of the independent truth commission in neighboring El Salvador in March 1993, and did not want the same to happen in Guatemala (Martínez 1993).

Guarding Privileges and Saving the Day

169

Third, the population was angry about the social impact of the neoliberal austerity measures, including high electricity rates and the allegations of corruption. From March onward, students, trade unions, and other social movements had carried out street protests. Just a week before the coup, demonstrations escalated when public sector workers decided to go on nationwide strike to show solidarity with the students (Aguilera 1993; Martínez 1993). Facing stiff opposition on all fronts and a failing peace process, the politically isolated and desperate president opted for a self-coup. On May 25, Serrano temporarily suspended the Constitution and ordered the dissolution of Congress, the Supreme Court, the office of the Attorney General, and of the human rights Ombudsman. All constitutional guarantees were temporarily suspended and heavy censorship was imposed on the media. The president further called on the TSE to convene a Constituent Assembly within sixty days. In the meantime, Serrano ruled by presidential decree. He argued that it was necessary to broaden executive powers to purge the state of corruption, particularly in Congress, which according to him hampered his ability to rule. Even before the coup, Serrano had governed with a confrontational style. He was known for his strong will and bad temper, and his style of leadership had already caused him trouble in his dealings with the media, the general population, the Catholic Church, the trade unions, some political parties, and even some of his ministers (Human Rights Watch 2004, 3–4; McCleary 1999a, 90; Cameron 1998, 133; Martínez 1993). Some therefore saw the decision to solve the crisis by carrying out a self-coup as an extreme outcome of Serrano’s confrontational style (Aguilera 1993; Martínez 1993). Guatemala’s civilian institutions were the first to be affected by this authoritarian move. The presidents of the Congress and Supreme Court were placed under house arrest and the Supreme Court was seized by the military. The police also took over Congress, which offered little resistance. Thereafter and throughout the crisis, the legislature was unable to hold sessions. This did not prevent the political parties from opposing Serrano’s actions from the outset; they issued public statements to the media and distributed leaflets, and a majority of the parties in Congress soon joined a broad civilian alliance against the coup, as explained further below.10 A number of congressmen also continued to meet in small groups in bars and hotels. However, because of disagreements about whether or not to purge the legislature, the Congress as an institution was unable to speak with one voice during the crisis. Other institutions that were forcibly closed down openly opposed Serrano’s actions: both the Constitutional Court and the TSE continued to hold sessions after the autogolpe (Aguilera 1993). Human rights Ombudsman de León Carpio escaped arrest and participated in demonstrations and public meetings (Human Rights Watch 1994; Villagrán 1993, 119–120). According to Serrano, he acted out of necessity to save Guatemalan democracy, and he promised to call for elections as soon as order was restored. Serrano’s coup is often compared to and put in the same category

170

Maren C. Bjune and Stina Petersen

as Fujimori’s successful coup in Peru a year earlier. However, Fujimori showed much greater political skills and had a much higher level of popularity than Serrano (Cameron 1998, 128–129). Further, when Fujimori staged his autogolpe in 1992, he successfully took advantage of a perceived national threat that was much clearer than the “chaos” that Serrano vaguely alluded to. Fujimori presented himself as the solution: he was a leader who needed room to carry out his duty as president, and he was backed by the army and intelligence services. By contrast, on the eve of the coup Serrano was politically isolated because of his confrontations on various fronts. Given its history of military coups and prolonged military authoritarian rule, Guatemalan society was initially unsure whether the coup had been staged or backed by the army. However, the military high command, including the minister of defense, the chief of staff, and the head of EMP were only informed by the president about the coup plans the day before it took place. Indeed, the president only relied on a small group of civilian advisors and security bodies to plan and execute the coup. During the first critical week following the self-coup, President Serrano became utterly isolated in his authoritarian efforts to establish order in a society in turmoil. From the very first day of the crisis, the CACIF condemned the self-coup in no uncertain terms and began work on a plan to reestablish constitutional rule. The quick mobilization of this powerful and experienced pressure group turned out to be vital for a prompt solution to the crisis. As a key power broker in national politics, the CACIF organized a broad civil society alliance against Serrano and mediated during secret meetings with the president and different army factions. The democratic institutions such as the political parties, the legislature, and the Constitutional Court were all young and were unable to take action against Serrano’s authoritarian measures without the assistance of the de facto power brokers, the CACIF and the army. On the whole, the CACIF had been satisfied with Serrano’s fiscal and economic policies. The president had established strong collaborative links with the organized private sector and demonstrated a commitment to International Monetary Fund (IMF) guidelines on privatization and credit. The autogolpe threatened the economic and political interests of the business elite, however. The elite feared that an authoritarian setback would hurt the national economy. Indeed, the United States warned there might be economic sanctions and suspended aid to Guatemala two days after the coup, and the country’s status under the Generalized System of Preferences (GSP) was at risk. Politically, the CACIF worried that the guerrilla might take advantage of the crisis and the high level of social mobilization. Thus, the CACIF sought to purge the regime of unwanted elements, including the president and his allies, and to reestablish constitutional order and political stability immediately. On the eve of the coup, the CACIF convened trade union and party leaders, academics, and lawyers to form a civilian alliance, the Multisectorial Forum

Guarding Privileges and Saving the Day

171

(Foro Multisectorial), later renamed the Instancia (McCleary 1999a, 109– 111). In line with its political strategy since the 1982 apertura (opening), the CACIF’s response to the coup aimed to enhance the legitimacy of a regime that was being contested nationally and internationally after decades of state oppression. The self-coup threatened the image of respect for the rule of law and civilian rule that the CACIF was promoting abroad to gain access to global markets. The self-coup met with broad international condemnation: the Organization of the American States (OAS) directly intervened in the crisis, sending a delegation to Guatemala to hold talks with all parties, including the military high command, in order to press for a return to constitutional rule (McCleary 1999b, 216; Jonas 2000a, 42; Villagrán 1993, 122). The OAS intervention, as well as U.S. economic sanctions, influenced the strategy adopted by the emerging opposition alliance, and the CACIF in particular. In the immediate aftermath of the self-coup, the army was divided about how to position itself: the hard-liners in the military leadership supported Serrano, arguing that the army would preserve its institutional privileges by supporting the coup. A few extreme hard-liners also argued that the army should seize state power. They were challenged by the soft-liners or “institutionalists,” who were headed by the vice chief of staff and the powerful director of military intelligence. Like the CACIF, these men were concerned about possible repercussions for the regime, such as international isolation. This faction had supported civilian rule since the mid-1980s to enhance the army’s legitimacy and preserve the army-as-institution, including its power and privileges. They did not want to take responsibility for President Serrano’s actions because they felt this would jeopardize the army’s legitimacy as an institution. They argued that the army would be better off not ruling the country and by supporting a limited political democracy (Schirmer 1998; Aguilera 1993). Because the military was divided, the president only received partial backing from the powerful army, which severely weakened Serrano’s power base from the outset (Jonas 2000a, 41–42; McCleary 1999a, 124). Serrano had not immediately shared his coup plans with the military high command, and this also worked against the president (Cameron 1998, 133). If one considers earlier coup attempts by the army, it is clear that the deep division within the military over the self-coup provides evidence of growing support for civilian rule within the army. Two days after the coup, the institutionalists began talks with the CACIF to resolve the political crisis and restore constitutional rule, so Serrano faced opposition from a powerful alliance between the two main power brokers in Guatemalan politics.

The Instancia: A Historical Alliance The president also faced a bottom-up mobilization by various social movements. On the day of the coup, indigenous movements, peasant organizations,

172

Maren C. Bjune and Stina Petersen

trade unions, human rights organizations, and the public university of San Carlos formed an ad-hoc center-left alliance, the so-called Multisectoral Social Forum (Foro Multisectorial Social), fearing that the autogolpe was the first step toward an authoritarian, army-based settlement of the civil war, particularly since the peace talks were stalled. These organizations adopted a confrontational strategy throughout the crisis because of fears fuelled by years of oppression. Serrano was soon also challenged by a broader civil society opposition alliance, the National Forum for Consensus (Instancia Nacional de Consenso). The Instancia emerged from the above-mentioned CACIF initiative and gathered together the strategic committee of the CACIF, a majority of the political parties, several trade unions, and, as the crisis evolved, various center-left social movements from the Foro Multisectorial Social.11 The Instancia was formed on May 30, held several meetings during the crisis, and drafted a proposal with six recommendations for a return to constitutional rule. This was later the basis for the elite settlement between the leaders of the Instancia and military hard-liners. The proposal included the election of a new president and vice president, the reform of electoral laws and political parties, a purge and reform of the three branches of the state, and the legal recognition of a forum along the lines of Instancia. The outgoing president had brought up the need to purge the legislature to combat the culture of corruption. Party leaders were convened by the Instancia on several occasions during the crisis to discuss the resignation of all legislators. Although most of the political parties participated in the Instancia, some were reluctant to cede their power in Congress. The DCG and the UCN, Serrano’s former coalition partners and the largest political parties in the Congress, were particularly opposed to resignations, arguing that they had received their mandate through free and fair elections. The issue remained unsolved for many months after the autogolpe, as discussed below. The allegations of corruption and the divergent opinions within Congress regarding its closure also explain why congressmen and political parties lacked the legitimacy and strength to lead the opposition against Serrano. The president, Vice President Espina, and the military hard-liners immediately opposed the Instancia’s call for Serrano’s resignation. However, after intense negotiations, on June 1, the Instancia and the minister of defense announced the president’s resignation from his office in the National Palace. The vice president instantly challenged the decision, declaring himself the new president of the Republic. But within days, the Instancia was able to marginalize Espina, whose only backing was a small group of military hardliners. By this stage, the Congress had also joined ranks with civil society to remove Espina. On June 2, the vice president tried to convene Congress to swear him in as president, but the deliberate absence of the DCG, the Guatemalan Republican Front (Frente Republicano Guatemalteco, FRG), the National Advancement Party (Partido de Avanzada Nacional, PAN), and UCN legislators hampered his effort. Finally, the Instancia presented

Guarding Privileges and Saving the Day

173

a list of three candidates for interim president to be elected by Congress, and on June 4, the hard-liners consented. Based on recommendations by the Instancia, the Constitutional Court declared the presidency legally vacant on June 4, and the following day the restored Congress elected former human rights Ombudsman, de León Carpio, as interim president of the Republic. The Instancia is considered a historical alliance, as it included social movements in a national decision-making process for the first time since the onset of civil war in the 1960s (Jonas 2000a; McCleary 1999a). Given that the line of confrontation had long been drawn between social movements and the regime, it is also notable that several of them joined with the CACIF and negotiated with military power brokers to resolve the crisis. Some organizations in the Foro Multisectorial Social, such as the indigenous movements and the Catholic Church, never joined the Instancia, however. In line with the position adopted by the Guatemalan National Revolutionary Union (Unidad Revolucionaria Nacional Guatemalteca, URNG) in the peace process, their condition for joining the coalition was that the Instancia had to address the structural problems of Guatemalan society, namely, socioeconomic inequalities. Others left the Instancia when the candidates for the interim presidency were discussed, as they felt they had not been allowed to influence that debate (McCleary 1999a; Villagrán 1993). Nonetheless, the participation of center-left organizations in the Instancia shows that the arena of civil society was being democratized, although the reluctance of some groups to participate also reveals how deeply divided Guatemalans were about the legitimacy of the regime and other structural issues.

The Role of the Army and the CACIF in the Elite Settlement However broad and diverse the composition of the Instancia, the CACIF and the institutionalists within the army sustained the civil society–based opposition with their de facto powers and therefore had a decisive hand in outcome of the crisis. The Instancia was based in the headquarters of the CACIF. All the parties involved held separate, secret meetings with CACIF representatives throughout the emergency period. These bilateral arrangements enabled the CACIF to force the president, the vice president, and military hard-liners to give up on their coup plans.12 Representatives of the CACIF also held several separate private talks with the military high command after the removal of Serrano to convince the hard-line faction to support the elite settlement. These were held before the official meeting between the Instancia and the hard-liners. On the other hand, the de facto power of the institutionalists was reflected in the physical removal of the president, which was planned and staged by soft-liners within the intelligence services. On June 1, army officials escorted the president from his office and put him on board a plane to El Salvador.13

174

Maren C. Bjune and Stina Petersen

The powerful alliance between the CACIF and the institutionalists was not a new phenomenon in Guatemalan politics. Representatives of both knew each other well from their two-year long engagement in the government’s peace commission. Trust between the parties paved the way for close collaboration in the aftermath of the self-coup and strengthened their role as power brokers (McCleary 1999a, 128–131). In sum, the strong engagement of a powerful, organized private sector and resistance to the coup by a powerful army faction played a critical role in ensuring that Serrano’s coup plans failed so quickly. To summarize, the serranazo is a case of a presidential breakdown in which the political regime survived but remained at the mercy of nondemocratic forces, which operated outside democratic institutions. Although the legislature played a key role in sidelining Espina and in electing the interim president, most stakeholders considered that the corrupt Congress had contributed significantly to the presidential crisis in the first place. Thus, in our view, Serrano’s failed coup attempt and the regime’s survival do not exemplify the flexibilization of the presidential system, as other cases of presidential breakdowns seem to do (Marsteintredet and Berntzen 2008). Guatemalan tutelary democracy appears to have survived in spite, rather than because of, the qualities of the constitutional institutions.

The Settlement and Democratic Development Self-coups severely threaten a constitutional and democratic order by violating core democratic principles (Cameron 1998). Thus, a failed autogolpe could strengthen and even accelerate the democratization of the regime. According to several scholars, Serrano’s failed coup confirms this assumption. Cameron (1998) argues that the events set in motion by the serranazo strengthened democracy and the rule of law and helped to reorder Guatemalan politics. Further, the inclusion of the protesting masses has been seen as a reflection of the regime’s increased respect for civil liberties. McCleary argues that the broad elite settlement constitutes the consolidation of Guatemalan democracy. Among others, she states that the response to the serranazo demonstrated that all segments of political and civil society were committed to the rules of the democratic game (1999b, 292). However, considering the legacies of Guatemala’s authoritarian history, and with the benefit of hindsight, some of these conclusions drawn on the basis of theoretical expectations and an overall optimism can now be challenged. We will examine the immediate postcoup period leading up to the general elections in 1995, which coincides with the presidency of de León Carpio. This period was affected by two interrelated transitions: from war to peace, and from military authoritarian rule to limited democracy. Throughout his presidency, the former human rights ombudsman de León Carpio acted as though he was indebted to the above-mentioned guardians

Guarding Privileges and Saving the Day

175

of the constitutional rule. The particular dynamics of the elite settlement severely constrained the interim president’s room for maneuver. During the serranazo, the institutionalists had replaced the hard-liners in the military high command, and they remained dominant in the military leadership until 1997 (Arévalo de León 1997, 27). Further, the new president lacked a political constituency. He had no political party to turn to, and Congress was beset by internal conflicts and mistrusted by the public, which helps explain why the president came under the influence of the military leadership. The president’s weak position vis-à-vis the army was demonstrated by his new human rights discourse. De León Carpio had initiated his presidency with a number of reforms in compliance with the UN and other international bodies. These included the investigation and prosecution of human rights cases hitherto protected by broad impunity laws, and a demilitarization of the national police. Given rising crime levels, however, only one year into his presidency de León Carpio rather allowed a renewed infiltration of military personnel into the police force (Jonas 1994).14 Further, during his first years in power he gave no sign of support for the prosecution of human rights abuses and took no clear stance on the issue of impunity,15 even though international bodies reported increased police violence, kidnappings, and frequent assaults on human rights advocates. From the start of the peace talks, the guerrilla had called for the disbanding of the paramilitary groups, the Voluntary Civilian Self-Defense Committees (Patrullas de Autodefensa Civil, PAC), responsible for a large proportion of human rights violations against the civilian population in the early 1980s (CEH 1999, 128). The army, which controlled the PAC, adamantly opposed their disbanding. Before assuming the presidency, León Carpio had proclaimed his support for the dissolution of these paramilitary groups, but as president he defended their existence, probably under pressure from the military high command (Human Rights Watch 1994; Jonas 1994, 21). That the government abandoned its initial commitments is further demonstrated by the fact that it ceased to cooperate with de León Carpio’s former office, the PDH (Human Rights Watch 1995). Again, pressure from the military high command proved stronger than the young political institutions and rule of law. In sum, the first years after the serranazo were characterized by continuity in power relations in Guatemalan politics, and constituted a prolongation of tutelary democracy rather than a democratic consolidation, as previously argued (McCleary 1999a).16 The turbulence generated by the serranazo also served to severely undermine the credibility of young civilian institutions and the political class in general. Police investigations into the coup-makers had revealed Serrano’s corrupt and rent-seeking activities, along with those of many members of Congress and other civil servants. De León Carpio’s first six months in office were affected by a conflict regarding the legality and the legitimacy of the regime in general and of Congress and the Supreme Court in particular.

176

Maren C. Bjune and Stina Petersen

Although the elite settlement had resolved the immediate symptoms of the crisis by physically removing the sitting president and electing an interim president, the conflict over the purge of Congress remained unsolved. As propounded by the Instancia, de León Carpio’s government did seek to purge both institutions. All 116 members of Congress and the justices of the Supreme Court were asked to leave their positions by the president. Even after the Constitutional Court declared that Congress itself would have to approve the dismissal of Congress to legalize such action, the president continued to promote a purge through a referendum. This generated another deadlock between the executive and the legislative branch, and shows that the settlement of the crisis had not contributed to any immediate improvements in the relationship between the two branches. Initially, the legislature was divided between the congressmen approving the dissolution of the Congress and a faction headed by the DCG that did not want to give up power. The pro-purge faction went as far as establishing its own bloc, the Group of 70, which worked closely with the Instancia and operated as an “alternative Congress.” For a brief period, then, Guatemala effectively had two functioning but equally inefficient legislatures. However, in the long run, the Instancia did not provide a legitimate solution. Although the president had wanted to use the coalition as his advisory council, he never managed to turn it into a legal entity, and key players such as the CACIF left the alliance. The Instancia was soon excluded from the negotiations between Congress and the president, and the coalition was unable to implement the remaining recommendations of the settlement. By November, Congress and the executive finally began to negotiate a solution to the crisis; half a year later, the parties came to an agreement, and constitutional reforms were approved by referendum in January 1994. However, the 15 percent voter turnout indicates that the population did not believe they had much influence on the process and on the regime in general. The reforms paved the way for an election of an interim Congress, but they did not include any measures to curb corruption in the legislature. A new Congress was not elected until August 1994. The fact that political turmoil persisted further undermined voters’ short- and long-term trust in democratic institutions and their representatives. Finally, the political turbulence generated by the serranazo proved to be a direct setback for the peace negotiations. In April 1991, President Serrano had become the first civilian president to initiate direct bilateral negotiations between the URNG and the government presided by a national chairperson. These negotiations had been severely constrained by the veto power of the military and the CACIF, and had been interrupted some weeks before the autogolpe. Serrano’s failed coup attempt partly explains why the progress of the Guatemalan peace negotiations was much slower than those of El Salvador and Nicaragua, and can therefore be seen as a moment of crisis in the process of democratization in Guatemala (Spence 2004, 43). Jonas finds that the

Guarding Privileges and Saving the Day

177

“May–June 1993 political crisis unleashed by the serranazo [. . .] interrupted the peace talks altogether. Paradoxically, even the resolution of that crisis, through the ascendance of former Human Rights Ombudsman Ramiro de León Carpio to the presidency in June 1993, further postponed the resumption of serious peace talks” (2000a, 41). In his first effort to relaunch the peace process, de León Carpio’s position mirrored the army’s and the CACIF’s demand that talks should focus only on the terms of a cease-fire (Jonas 1994, 21), which shows that the position of the CACIF and the army had been strengthened compared to the period of pre-serranazo negotiations. In early 1994, the president was nevertheless able to reinitiate the negotiations, this time with UN mediation and the support of the Grupo de Países Amigos.17 The Civil Society Assembly (Asamblea de la Sociedad Civil, ACS), which brought together representatives of civil society and the social movements, also participated (Rosada-Granados 1997). This time there was significant progress,18 but the parties delayed any talks about substantive issues such as land and security until after the 1995 general elections, issues considered substantial for the peace agenda. According to Asturias, a key URNG negotiator, the land and security issues were postponed because the de León Carpio government was too weak to implement any agreement that might be reached, and it was deemed that it was best to negotiate them with a government strong enough to carry out the terms of the accords while in power.19 For this reason, the URNG rather initiated peace talks with de León Carpio’s successor, Alvaro Arzú of the PAN, even before he won the elections. All of Guatemala’s elites, the leaders of the social movements and of the armed left, participated in the Guatemalan peace negotiations, since the goal was to accommodate all political and social sectors to ensure regime democratization. The process was halted by the autogolpe, but also by Serrano’s successors. Under de León Carpio, there was a new (un)balance at the negotiation table: the military and business elites had gained substantial strength by acting as “constitutional guardians” and because of the weakness of the interim president. Paradoxically, then, the historical settlement designed to preserve constitutional order threatened the peace negotiations and limited the inclusiveness of Guatemala’s emerging democracy.

Conclusions The serranazo is a peculiar variant of presidential breakdown, one in which it was the president’s own authoritarian behavior that initiated the crisis. The breakdown resulted from his failed autogolpe. We have emphasized the role of other actors involved in the serranazo and its aftermath, but the breakdown was a direct result of Serrano’s own decisions. The autogolpe was above all a symptom of an “unhealthy” regime, and its failure has therefore been applauded as a sign of democratic progress.

178

Maren C. Bjune and Stina Petersen

However, the postcoup settlement per se did not strengthen accountable and representative democratic institutions. As the “guardians of the constitutional order” operated outside formal democratic institutions, we hesitate to see the solution to the political crisis as an example of democratic forces gaining ground in Guatemalan society. Rather, the process of democratization has proved to be slow and flawed. Guatemala has undergone a peace process and the final peace accords were signed in 1996, but thirteen years later, the pace of implementation of the accords has been remarkably slow, and no governments have made this a priority (Torres-Rivas 2006). Although the business elite and factions of the army did officially collaborate with democratic civil society groups and other political institutions, they had the decisive hand in resolving the crisis. Anxious not to lose out to emerging political forces in Guatemalan society, the much anticipated golpe of Serrano gave them the opportunity to present themselves as guardians of the constitutional rule in Guatemala, while simultaneously consolidating their de facto power position. Their reserved domains of power were not challenged but rather prolonged with the settlement, leaving the regime under the same constraints that it had worked under before the autogolpe. Thus, we argue that these “saviours of the constitutional order” have stood in the way of strengthening young civilian institutions, and to limit the inclusiveness and further development of Guatemalan democracy. Today, many view the serranazo as an important moment in the political history of Guatemala. The post-autogolpe settlement showed that there was an overall commitment to the constitutional order within civil society (a lasting alliance of civil actors seeking political influence was formed) and among the de facto elites (the CACIF and the military). The involvement of the protesting masses shows that this was a historically unique political opening, and the coalitions and alliances that were forged facilitated the broader civil society participation that characterized the Guatemalan peace process. However, while the presidential breakdown did not culminate in an authoritarian setback, its settlement also shows that there was a continued commitment to old practices. Post-serranazo changes in popular participation and commitment to democratic practices were not accompanied by changes in de facto power relations, which have been characterized by continuity both in terms of the actors involved and their behavior.

Notes 1. See Quince años como símbolo de impunidad, Prensa Libre May 25, 2008, at http://www.prensalibre.com/pl/2008/mayo/25/239362.html, accessed November 23, 2008. 2. When discussing democratization in Guatemala, we have found it particularly useful to focus on the government’s ability to govern. As with other LatinAmerican post-authoritarian regimes, nonelected institutions like the military tend to constrain elected governments through reserved domains of power,

Guarding Privileges and Saving the Day

3. 4.

5. 6. 7. 8.

9.

10. 11.

12.

13.

14.

15.

16.

179

producing what are called tutelary democracies (Collier and Levitsky 1997, 443, 438; Przeworski 1992, 131). This ten-year period is referred to as the “ten years of spring” in Guatemalan history. The term oligarchy refers to the small but powerful group of families that have historically owned most of Guatemala’s economic resources (Valdez 2003; Casaús 1996; Dosal 1995). In the 1980s, the EMP included 730 civilian and military officers and became the most powerful unit within the military (Schirmer 1998). The military coups took place on May 11, 1988 and on May 9, 1989. In 1987, the CACIF organized a national labor strike to block the Cerezo administration’s fiscal policies (McCleary 1999a, 34; Fuentes and Cabrera 2005, 8). In addition to the 18 seats held by the MAS, the UCN had 41 seats, and DCG had 27. The coalition of parties held approximately 74 percent of the seats, or more than a two-third majority in Congress. Ironically, in a final effort to save his presidency, Serrano approached party leaders and legislators he had initially criticized for accepting bribes in exchange of support (McCleary 1999a, 140–141). The parties that joined the Instancia were the DCG, the FRG, the PAN, the Social Democratic Party (Partido Social Demócrata, PSD), and the UCN. Examples of center-left organizations that joined the Instancia are the Mutual Support Group (Grupo de Apoyo Mutuo, GAM), a human rights group, and the National Federation of Workers Unions of Guatemala (Federación Nacional de Sindicatos de Trabajadores de Guatemala, FENASTEG). Upon the president’s request, CACIF representatives first met with the president on the day of the autogolpe. The second meeting was held on May 27, there was a meeting with the vice president on May 29, and another meeting with Serrano on the evening of the same day. Another meeting was held on May 31 (McCleary 1999a, 195, 226, 235–236). Indeed, a renowned Guatemalan political scientist described the institutionalists’ prompt removal of President Serrano as a coup, pointing out that the military leadership gave civilians a few hours to resolve the crisis before intervening (Aguilera 1993). This was initiated by the newly nominated interior minister César Augusto Medina, who had replaced the more reform-minded Arnoldo Ortiz Moscoso. There was also a change of personnel within the national police directorate. Congress only ratified the joint Guatemalan-UN International Commission against Impunity in Guatemala on August 1, 2007. The Commission has a mandate to investigate the existence of illegal security groups and clandestine security structures in Guatemala in order to strengthen state institutions and the rule of law. The national police, the prison, and customs systems were set to come under investigation first (WOLA 2007; CICIG 2008). The commission has had noteworthy difficulties in cooperation with governmental institutions. It is worth mentioning some individuals who establish a line of continuity between the serranazo period and the present. Current president Álvaro Colom of the National Unity of Hope (Unidad Nacional de la Esperanza, UNE) party and his competitor in the last round of the 2007 presidential elections, former General Otto Pérez Molina of the right-wing Patriotic Party (Partido Patriota, PP), were both involved in the 1993 autogolpe, each on a different site. General Pérez

180

Maren C. Bjune and Stina Petersen

Molina, the then Director of Intelligence, headed the army institutionalists, while Colom, then vice minister of the economy and a member of one of the CACIF’s chambers, quickly slipped under the radar after the coup. See Serrano quiere regresar con el nuevo gobierno, Prensa Libre, November 8, 2007, at http://www. prensalibre.com.gt/pl/2007/noviembre/08/187209.html, accessed November 23, 2008. During the much criticized 2007 presidential campaign, the serranazo was brought up, particularly when Pérez Molina’s camp sought to connect Colom to the exiled former president, claiming that a vote for Colom was the same as voting for the return of Serrano and for authoritarianism. These two men embody the continuity in political agency and power relations in Guatemalan politics. 17. This group consisted of Colombia, Spain, the United States, Mexico, Norway, and Venezuela. 18. The following partial peace agreements were signed under de León Carpio’s presidency: the January 1994 Framework Agreement; the June 1994 Agreement on Resettlement of the Population Groups Uprooted by the Armed Conflict; the June 1994 Agreement on the establishment of the Commission to clarify past human rights violations and acts of violence that have caused the Guatemalan population to suffer, also known as the truth commission agreement (Jonas 2000a). 19. Author’s (S. Petersen) interviews with Rodrigo Asturias alias Gaspar Ilom, a former head of the URNG, and of the Revolutionary Organization of the Armed People (Organización Revolucionaria del Pueblo en Armas, ORPA), and a URNG peace negotiator (Guatemala City, January 19, 2005 and February 10, 2005). The interviews were done as part of the author’s master’s thesis fieldwork.

10 The Dominican Republic and the Fall of Balaguer 1994–1996: Presidential Breakdown or Democratic Transition?* Leiv Marsteintredet

On May 16, 1994, Balaguer of the Social Christian Reformist Party (Partido Reformista Social Cristiano, PRSC) won the presidential election through massive electoral fraud, securing a victory over his main opponent, Peña Gómez, who led the Dominican Revolutionary Party (Partido Revolucionario Dominicano, PRD), by a margin of 0.8 of a percentage point. The fraud was discovered between Election Day and the inauguration of the Balaguer administration on August 16. As a result, the country was thrown into its biggest postelectoral crisis since the transitional elections of 1978. As the country was filled with international electoral observers, the Dominican electoral crisis attracted international attention. The Central Electoral Board (Junta Central Electoral, JCE) did not publish the final results until August 2, when it declared Balaguer and Jacinto Peynado the winners and de facto president and vice president for the next four years. August 2 did not mark the end of this series of events. Despite emerging “victorious,” there was a process of negotiation between Balaguer and the opposition between that date and his inauguration on August 16, which culminated in a constitutional reform that shortened Balaguer’s term in office to two years and banned presidential reelection. The Pact for Democracy (Pacto por la Democracia) and the consequent shortening of Balaguer’s presidency have led observers to consider the fall of Balaguer from the presidency as a case of presidential breakdown. This chapter examines the fall of Balaguer and asks whether it is a case of presidential breakdown, defined here as the termination of a presidency through a forced and extraordinary shortening of the president’s term without terminating a democratic regime.1 I compare the fall of Balaguer with two

182

Leiv Marsteintredet

other presidential crises and examine the extent to which current theories of presidential breakdown can explain Balaguer’s exit (salida) in 1994–1996. In my view, the comparative literature on presidential breakdowns (Hochstetler 2006; Kim and Bahry 2008; Negretto 2006; Pérez-Liñán 2007; 2008; Valenzuela 2004) does not help to explain the fall of Balaguer. However, theories of transition and on hybrid regimes do help to clarify the case. Elite pacts, former regime types (e.g., Linz and Stepan 1996; Whitehead, O’Donnell, and Schmitter 1986), and the international dimension, particularly western leverage and linkages to the regimes in question (e.g., Whitehead 1996; Levitsky and Way 2005; 2006), are central elements in any explanation of the presidential breakdown in the Dominican Republic. By tracing key decisions during the summer of 1994, I seek to establish that the international dimension is particularly vital; and by focusing on Balaguer’s trajectory as an authoritarian leader, I argue that it is unlikely that he would have left power without external pressure. In my analysis, I emphasize various peculiarities specific to this case that explain why theories of presidential breakdown fail to account for the fall of Balaguer. Indeed, the case may be more accurately understood as one of transition from a hybrid regime to democracy. 2 I therefore consider whether the fall of Balaguer should be understood as a presidential breakdown at all. In the conclusion, I assess the outcome of the presidential interruption in terms of its consequences for Dominican democracy, focusing on the effects of the constitutional deal that shortened Balaguer’s term.

The Fall of President Balaguer President Balaguer signed and sealed his fate with the Pacto por la Democracia on August 10, 1994, and the constitutional reform on August 16 that shortened his term to two years. The immediate reason for this reform was the electoral fraud on May 16, 1994, which altered the outcome of the election in his favor. 3 Although the fraud received attention from national and international media and political actors, Balaguer was not removed from office. In contrast to other cases of interrupted presidencies in Latin America, Balaguer was given two years to pack his bags and leave the presidential palace. The 1994 fraud thus secured Balaguer an undeserved victory and two more years in power. The Dominican Republic has a long tradition of fraudulent elections under both democratic and authoritarian regime periods. During the dictatorship of Rafael Leonidas Trujillo Molina in 1930–1961, elections were held every four years. The civilian (but authoritarian) regime of President Balaguer (the so-called doce años, or twelve years, between 1966 and 1978) also organized controlled elections. Following a brief period of unproblematic elections, the 1990 election was fought in an uneven playing field and plagued by accusations of fraud, as a result of which Juan Bosch, the leader of the

Dominican Republic and Fall of Balaguer

183

Dominican Liberation Party (Partido de la Liberación Dominicana, PLD) was probably deprived of victory (Maríñez 1994, 243–255). After that election, at the behest of international observers and national political actors, the Balaguer government accepted reforms to the electoral system. Despite this apparent concession, Díaz (1996, 74–157) demonstrates how the PRSC government, together with the JCE, sabotaged the promised electoral reforms and planned the 1994 electoral fraud. On Election Day, the biggest problem was that the JCE altered the electoral register that had been sent to the political parties for verification. This led mainly to the disenfranchisement of PRD voters. Estimates of the number of disenfranchised voters range from the JCE’s official number of 45 thousand, to the PRD’s claim that 200 thousand were affected (National Democratic Institute (NDI) 1994; Sagás 1997). Whichever the case, the number of disenfranchised voters was clearly much higher than the margin of victory of around 20 thousand votes, and PRD voters were disenfranchised disproportionally compared to those voting for the other parties. In response to this, international observers, who had learnt from the experience of the 1990 elections four years earlier, coordinated their efforts, prolonged their stay in the country after the election (Graham 2008),4 and published their findings internationally. The PRD had also learned from the victimization of the PLD in 1990, and did a tremendous job of documenting their allegations of fraud. 5 Early on, the victims of the fraud, the PRD and its candidate Peña Gómez, wanted to take to the streets in protest. Pressed by close advisors and party followers, Peña Gómez threatened to promote national strikes and to paralyze the Balaguer administration on various occasions. But he was persuaded by international observers to avoid violence at all costs (Graham, author interview (AI), February 9, 2009; Merloe, AI, November 25, 2008), and was warned that street protests would lessen international goodwill toward him (Griner, AI, February 4, 2009). So although the months between May and August were tense,6 there were no strikes and only a few social protests7— fewer, in fact, than those that followed the electoral crisis of 1990. Because of the problems on Election Day, and because the JCE-nominated verification commission had to examine the electoral results, these were not officially published until August 2. The JCE ignored its own verification commission and declared Balaguer president for a new four-year term, while the army and police declared their full confidence in the electoral results.8 The PRD and PLD demanded new elections, Peña threatened national strikes again, the Catholic Church lamented the JCE decision, the U.S. administration expressed its disappointment with the JCE decision and gave its support for new elections (Hartlyn 1994, 122), various U.S. newspaper editorials criticized the electoral outcome,9 and international electoral reports confirmed the fraud (NDI 1994; OAS 1997).10 In a context of social tension and U.S. and international pressure, Balaguer was persuaded to meet the PRD and the PLD at the negotiating table. Having

184

Leiv Marsteintredet

been granted formal victory by the proper authorities and as president-elect, Balaguer had the upper hand and could benefit from the status quo in the negotiations. The result of the talks was the Pacto por la Democracia, which shortened Balaguer’s electoral term and prohibited his immediate reelection. The pact was signed on August 10, and its terms were incorporated into the constitution days later. Balaguer was sworn in as president on August 16. On May 16, 1996, new presidential elections were held. For the first time in thirty years, Balaguer was not one of the candidates.11

Explaining the Fall of Balaguer Theories of presidential breakdown call attention to several factors that are important when explaining the premature downfall of presidents, including the following: a president’s initial support in elections and in Congress; the president’s legislative shield and control of the median voter in Congress (Kim and Bahry 2008; Negretto 2006; Pérez-Liñán 2007; Valenzuela 2004); the implementation of neoliberal policies; whether a country is subject to an International Monetary Fund (IMF) agreement; presidential scandals; street protests; and economic crises (Hochstetler 2006; Pérez-Liñán 2007). Borrowing from the literature on democratic transition, I add a new variable in table 10.1 below: international pressure for governments to adhere to democratic rules. In what follows, I analyze the degree to which these variables have explanatory power in the case of the fall of Balaguer in 1994. To increase analytical leverage, I compare the 1994 case to a selection of negative cases in the Dominican Republic between 1978 and 2008 (that did not result in presidential removal,12 but which, according to theoretical expectations, could have led to presidential breakdown). Table 10.1 compares the three presidential challenges that have occurred since democratization in 1978. When identifying a presidential challenge, I include strikes and social protests that turn into riots, and scandal-related situations that threaten the survival of government. In addition to the positive case (the successful removal of Balaguer), the table presents two negative (but likely) challenges that did not end in presidential breakdown (the Easter uprising in 1984, in which social discontent caused by rising prices and neoliberal policies produced widespread public protest; and the 1990 election and subsequent protest against the alleged fraud). The table also summarizes the values for the central explanatory variables for each case. A cursory review of the table reveals that, according to theories of presidential breakdown, both the Easter uprising in 1984 and the 1990 electoral scandal were more likely candidates for presidential breakdown than the 1994 electoral scandal. In 1990, both Jorge Blanco and Balaguer had lower initial support than four years before; both led a country experiencing negative growth and high inflation rates; and both challenges led to strikes and

Dominican Republic and Fall of Balaguer Table 10.1

185

Challenges to presidents in the Dominican Republic

Government

Jorge Blanco 1982–1986

Balaguer 1986–1990

Balaguer 1990–1994

Case

1984 Easter Riot

1990 Election

1994 Election

Initial support Minority Neoliberal policies Negative growth Street challenge/riots Congressional challenge Scandal International pressure

46,7% (–5) no yes yes yes no no no (not relevant)

35% (–6.5) yes no yes yes no (partial) yes, electoral no, partial

42.3% (7.3) yes no no no no yes, electoral yes

Survival

Survival

Breakdown

Outcome

Notes: Coding based on Marsteintredet (2004) and the Latin American Weekly Report (1980– 2005). Initial support indicates the share of votes for the president, followed in parentheses, by a comparison of the share of votes of the president’s party from the previous to the current election. The “legislative shield” variable for impeachment is not relevant and thus excluded; the variable “control of median voter” is also excluded due to a lack of variation across cases. The variable “neoliberal policies” is coded according to Hartlyn (1998, chs. 6–7). The variable “negative growth” is based on the Penn World Tables 6.2 (Heston, Summers, and Aten 2006).

riots. Furthermore, Balaguer led a minority government, and although Jorge Blanco had a majority in Congress, his party, the PRD, suffered from bitter factionalism. In 1994, by contrast, Balaguer had strengthened his popular support compared to 1990, there were no street protests against the fraud, the economy was in better shape, and he was not challenged by Congress. As the table also shows, however, international pressure was present in 1994 and only partially so in 1990 (it was irrelevant in 1984). It would seem, then, that most of the explanatory variables drawn from current theories of presidential breakdown fail to explain the fall of Balaguer in 1994. In the next section, I analyze and explain the two negative cases. I then call attention to the “international pressure” variable, inspired by the transition to democracy literature, and focus on how this may be central to explain the different outcomes.

The Negative Cases: Social Uprising and Election Protest Jorge Blanco (PRD) was elected president in 1982, and his party won a majority in both chambers of Congress. In contrast with his predecessor, who had promised to install democracy, Jorge Blanco had promised that

186

Leiv Marsteintredet

there would be “social democracy.” The 1982 debt crisis hit the economy extremely hard, and the social democratic party (with strong union and popular backing) was soon forced to take the difficult decision of seeking an agreement with the IMF, a decision that would hurt its own supporters. In 1984, in the midst of economic stagnation, a credit crisis, and the strong devaluation of the peso, the Jorge Blanco administration took advantage of the Easter holidays to increase the prices of a series of essential products. The social climate was already near boiling point because of the economic situation, and the government-mandated price adjustments triggered social protests. On April 23, the Monday after Easter, a popular uprising and riot began in Capotillo, a poor barrio of Santo Domingo. Despite strong police repression, the protests spread to several parts of the capital and to thirty other urban centers across the country. On the second day of the uprising, the government censored all media and intensified its repression of protests (Maríñez 1994, 202–203). The uprising was quelled after three days, but more than 100 people had been killed, 200 injured, and 5,000 arrested by the police and army (Faxas 2007, 178). Although the 1984 protests were spontaneous and never intended as an organized demand for the resignation of the president, the harsh treatment of protesters by the army could have caused the latter to escalate. However, at that time, Jorge Blanco had strong army backing because he had promoted reforms to professionalize the institution, and he also enjoyed a majority in Congress. On the third day of the protests, the president appeared on national television surrounded by army officials. After the army publicly declared its loyalty to the president, Jorge Blanco lauded the response of the police and army to what he called a political conspiracy, and he condemned the violent acts of protesters and the looting (Faxas 2007, 173). The army closed ranks behind the president, and the government’s survival was ensured because civil society and the unions were weak and because the protests were quelled rapidly and violently. In 1986, Balaguer and the PRSC returned to power, defeating a factionalized PRD in fair elections. However, the PRSC did not win a majority in the lower chamber of Congress. Further, the socioeconomic situation was fraught and by the 1990 elections, four years later, growth was negative and inflation was at levels higher than during the Blanco presidency. In addition, polls taken in the months leading to the 1990 elections showed that Balaguer was losing to his political archenemy, Juan Bosch (PLD). In the aftermath of the 1990 election, the PLD alleged that it had been victimized by colossal fraud.13 International observers could not verify the fraud, but the Carter Center (1990) did cast doubt on the electoral outcome.14 The PLD organized several national strikes protesting the results. On August 15 and 16, a strike paralyzed the country, as a result of which twelve people were killed and 5,000 arrested; it organized another three-day strike in September, as a result of which a further twenty-three people were killed and 400 arrested. Because the opposition parties boycotted Congress, the

Dominican Republic and Fall of Balaguer

187

Supreme Court had to swear Balaguer in as president on August 16. For months after the election, unions, business groups, political parties, and professional organizations demanded that Balaguer should resign (Moya Pons 2000, 441–442). The situation seemed ripe for a presidential interruption: growth rates were even more negative than in 1984; inflation was at twice the 1984 level; social protests were intense; the electoral process had turned into a scandal; the president had low initial support (he had received only 35 percent of the vote), and he did not have a majority in Congress. As shown in table 10.1, almost all the ingredients for a presidential interruption were present. And yet Balaguer survived. The PLD failed to present hard evidence of fraud, leaving the international community without a solid basis to react, and the PLD lost its primary reason for removing the president. Furthermore, the PRD did not support the claims of fraud, creating a split in the opposition, which further weakened internal pressure on Balaguer.15

The Positive Case: Electoral Fraud and the Fall of Balaguer Conditions for a presidential breakdown seem less favorable when one compares the 1994 situation with the 1984 challenge against Jorge Blanco and the 1990 electoral crisis. Whereas both Jorge Blanco in 1982 and Balaguer in 1990 had lost electoral support and the economy was in a bad shape, in 1994 Balaguer had considerably greater electoral support, the economy was growing, inflation was under control, and the situation in Congress was about the same as four years earlier. In contrast to the challenges in 1984 and 1990, there were no street protests against the president in 1994 (there were only credible threats, as can be seen in table 10.1), and civil society did not mobilize (Díaz Santana 1996, 235). Theories of presidential breakdowns, then, fail to explain either the 1994 breakdown or the absence thereof in 1990 and 1984. However, a comparison between the two most similar presidential crises—the 1990 and 1994 electoral scandals—brings to light one major difference between the two events: in 1994, fraud was proven by the opposition beyond any reasonable doubt, and consequently international actors such as the OAS and the United States could pressure Balaguer to uphold democratic game rules. The following sections discuss how this pressure proved to be the key difference between the 1990 and the 1994 elections.

United States and International Pressure: Process-Tracing Balaguer’s Decisions Hartlyn (1998, 256), who was an electoral observer of the 1990 and 1994 elections for the NDI, argues that the U.S. government was a major but

188

Leiv Marsteintredet

not a decisive actor. I challenge this conclusion by process-tracing important decisions made by Balaguer during the summer of 1994 and linking these decisions to important meetings between Dominican and U.S. authorities and declarations of the U.S. administration at the time. I argue that U.S. pressure was not only of major importance but was indeed necessary and decisive for the fall of Balaguer. In the previous section, I demonstrated that national pressure proved to be insufficient to topple Balaguer in 1990. Although it is likely that, even in the absence of international pressure, street mobilizations after the 1994 election pressure would have reached the levels of 1990, table 10.1 shows that other factors were more favorable for Balaguer in 1994. Thus, it is unlikely that the protests alone would have resulted in his resignation. The potentially larger scale of fraud in 1994 and the fact that the PRD could prove that it had occurred were important differences between 1990 and 1994, but they were important insofar as they attracted international attention, rather than because of their local importance for generating street demonstrations. As the PLD and Bosch demonstrated, if people are to mobilize they must believe that fraud has been committed, and this belief was as strong in 1990 as in 1994. What distinguishes 1994 from 1990 is that in the former instance there was evidence of fraud, and the United States used its strong leverage to pressure Balaguer. The United States exerted this pressure during meetings with the Balaguer administration and with public announcements condemning the electoral results. The United States initially favored new elections and let Balaguer know that the relationship between the two countries depended on the resolution of the electoral crisis. Due to the proximity between the United States and the Dominican Republic there is a very high linkage between the countries, and the difference in size, economy, and development between the two nations also secure U.S. leverage over the small Caribbean nation (Levitsky and Way 2005; 2006). Almost all types of civil society linkages are extremely high between the two countries, which could not be more unequal in size and military importance. The only factor that could limit U.S. leverage over the Dominican Republic in 1994 was its ongoing conflict with Haiti’s military government.16 The situation with Haiti probably limited U.S. leverage over the Dominican Republic and made the United States willing to give in to the deal that unjustly secured Balaguer two years in power rather than firmly adhering to its previous stance on new elections. On the other hand, the United States had a strong commitment to democracy in the region at the time, and this is also what they demanded for Haiti. Thus, accepting a Balaguer victory would undermine its policy for democratic change on the western part of the island. Michael Skol (AI, February 6, 2009), principal deputy assistant secretary of state in the Bureau of Inter-American Affairs at that time, and point man for the State Department in the Dominican Republic after the elections, argues that the Haitian conflict actually increased the determination of the United States to prevent electoral fraud and stand firm against Balaguer’s attempts to steal the election. In this case, therefore, U.S. pressure proved to be effective.

Dominican Republic and Fall of Balaguer

189

As in 1990, Balaguer gambled that the United States would hesitate, particularly because of the situation in Haiti, and that it would be best to wait until things cooled down internationally before taking power again. It was Balaguer that first mentioned a possible shortening of the electoral term and of a pact on July 14. Balaguer had remained silent about the issue, saying only that it was the JCE that was responsible for counting the votes. The timing of his July 14 statement was not coincidental, however, occurring as it did only hours after the Dominican Ambassador to the United State, José del Carmen Ariza, was called to the State Department to meet with Skol, and only a day after U.S. ambassador to the OAS Harriet Babbit had mentioned in a congressional subcommittee meeting that if no acceptable solution were found, the application of Resolution 1080 and OAS sanctions would be considered (Graham 2008). The United States used blunt language in the meeting with the Dominican ambassador, making it clear that it had no doubts that the elections had been stolen, and that it wanted Balaguer to know that the United States was serious. Balaguer revealed that the Ambassador had made “a suggestion of certain importance” (Díaz Santana 1996, 218–219). Thus were shattered Balaguer’s hopes to stay in power as a quid pro quo in exchange for closing Dominican borders and upholding the UN embargo against Haiti.17 Between July 14 and August 2, Balaguer was again silent. He even made himself inaccessible to close collaborators within his own party and to the OAS delegation that stayed on to oversee the post-electoral process (Graham 2008). The declaration that Balaguer was the winner, as well as a meeting between the Dominican ambassador to the United States on August 1 and a statement from the U.S. State Department were crucial in prompting Balaguer to negotiate with the PLD and the PRD after August 2. Reacting to the JCE proclamation that Balaguer had won the elections, the U.S. State Department stated that the JCE’s decision would be a “detrimental factor” in the relationship between the two countries.18 It was after that meeting and these statements that Balaguer finally agreed to negotiate the shortening of his term in office. The timing of Balaguer’s statement may have been coincidental, but it does back my hypothesis that international pressure was a necessary factor in the outcome of the 1994 postelectoral crisis.

The Historic Falls of Balaguer Balaguer was not a populist president like Collor de Melo, Bucaram, or Fujimori, who all quickly lost all support when their power began to crumble (Pachano 1997; Weyland 1993, 2006). In 1994, Balaguer was sworn in as president for the seventh time since 1960. He was a historic leader in a new democracy who had (along with his party) maintained his authoritarian mindset despite the new post-1986 democratic climate (Taylor-Robinson 2001). It was therefore no surprise that the elections in 1990 and 1994 were (allegedly) fraudulent. Nor, given Balaguer’s history, was it a surprise that

190

Leiv Marsteintredet

international pressure was required to remove him from power. Balaguer had lost the presidency twice after 1960, first in 1962 and later in 1978. On both occasions, U.S. pressure had been necessary to remove him. This section focuses on the authoritarian nature of Balaguer’s leadership and on the role played by the United States. After Trujillo’s assassination on May 30, 1961, Balaguer ruled the country until mid-January 1962, when he relinquished power after a failed coup. During this period, he attempted to lift the international embargo, but was only successful after being pressed by the United States to accept a powersharing agreement with a Council of State on January 1, 1962. However, already in December 1961, the United States had abandoned its plan to maintain Balaguer in power until the next elections, and had started pushing for his ouster (Vega 1991, 386–392). The change in U.S. policy emboldened the opposition, which organized a protest that was violently suppressed on Balaguer’s orders. Five people were killed and many injured, leading the other members of the Council of State to call for Balaguer’s resignation. Balaguer sought support from military strongman, Air Force General Rodríguez, who organized a coup to establish a civil-military junta led by Balaguer. However, domestic and international pressure led Balaguer to decline leadership of the new junta, and the coup failed (Guerrero 1991, 151–168, 197–217). In 1966, with the help of the United States, Balaguer returned to power, presiding over an authoritarian electoral regime until 1978. In 1978, he lost the presidential election to PRD candidate Antonio Guzmán, but he did not go without a fight. The army and Balaguer supporters took control of the JCE on election night to stop the vote count (Guzmán was leading the count), with the intent of staging a coup in favor of Balaguer. Pressure from the Carter administration in the United States played a crucial role in aborting this coup in the making, in securing the handover of power to Guzmán, and in promoting a transition to a minimally democratic regime (Atkins 1981, 108–110; Hartlyn 1998, 122–126).19 Again U.S. involvement aborted Balaguer’s attempt to prolong his time in power. During his political career, which lasted almost seven decades, Balaguer left the presidency three times, and U.S. pressure was required to convince him to leave power each time.20 In 1994, therefore, Balaguer knew from personal experience that he could not rule against the will of the superpower in the north. He had been president during the fall of the dictator Trujillo, and he knew firsthand how dire the effects of U.S. pressure and of an international embargo could be.

Was the Fall of President Balaguer a Presidential Breakdown? A presidential breakdown is extraordinary because it interrupts a fixed term of office; and it is distinct from a democratic breakdown because democracy

Dominican Republic and Fall of Balaguer

191

survives. 21 Hochstetler (2006, 402) defines presidential interruptions as all instances in which “elected presidents [leave] office before their terms [are] completed, whether they [resign] or [are] impeached or otherwise forced out of office.”22 Balaguer was clearly forced out, but he did complete his term—albeit a constitutionally altered one. Thus, the fall of Balaguer does not qualify as a presidential breakdown as defined in the literature. His fall is unique because his is the only Latin American presidential breakdown that was formalized through a constitutional reform. Furthermore, despite his fraudulent electoral victory, Balaguer stayed in power for two years. The 1994 pact actually extended his tenure rather than shortening it, and thus fails to constitute a presidential breakdown. On the other hand, Balaguer did not accept the pact voluntarily, and was forced to accept the shortening of his term from four to two years. 23 Hence the difficulty of classifying this case: it fits with the definition of presidential breakdown given in the introduction, which emphasizes the extraordinary and forced shortening of a presidential term of office; but it does not fit with a definition that focuses on an uncompleted term of office. Also problematic is the electoral fraud, which raises questions as to whether the regime was democratic at all. Most studies characterize the Balaguer regime between 1986 and 1996 as a subtype of democracy rather than as authoritarian (e.g., Hartlyn 1998; Jiménez Polanco 1999). However, given the fraudulent election of 1994, one could argue that the regime crosses the threshold from democracy to authoritarianism. Carrión distinguishes “delegative” democracies (O’Donnell 1994) and electoral authoritarian regimes according to “whether the regime is willing to accept meaningful political competition” and its “overall record of respect for democracy” (Carrión 2006, 303). The Dominican Republic crossed that line in 1994, when the democratic principles of contestation and participation were violated during the presidential election (unlike the congressional elections, which were contested). The case is included as a presidential breakdown in several studies along with other dubiously “democratic” cases such as Guatemala in 1993 and Peru in 2000, even though the Dominican Republic received higher scores from the Polity Democracy Index than these countries for the years corresponding to the breakdowns and for the three years previous to the breakdowns. Thus, taking the regime-type criterion into account, one could argue for both the inclusion and the exclusion of the Dominican Republic as an example of a presidential breakdown. As discussed above and summarized in table 10.1, what is perhaps most troublesome is that the fall of Balaguer cannot be explained by existing theories of presidential breakdown. By including cases that fall in the “lower” end of the democracy scale, theories of presidential breakdown lose explanatory value because they were originally inspired by the democratic breakdown literature. If one defines electoral fraud as a presidential scandal, the theories presented by Pérez-Liñán on impeachment may help to explain the

192

Leiv Marsteintredet

fall of Balaguer. However, Pérez-Liñán’s (2007, 65) concept of scandals focuses on personal presidential scandals and corruption, and electoral fraud is a very different kind of political “scandal.” Electoral fraud is a phenomenon that exposes not only the personal character of a specific president but, more importantly, reflects the nature of the political regime itself; it not only threatens the survival of the president but imperils the longevity of a democratic regime. These considerations lead me to look elsewhere for theories to explain the Dominican case. The lesson from the Dominican case is that scandals that expose undemocratic behavior by key political actors—more likely to occur under hybrid regimes—can result in presidential breakdowns with consequences that more closely resemble those affecting democratic transitions rather than those pertinent to democratic breakdowns. 24 So the transition literature is more helpful when it comes to explaining the fall of Balaguer and the way in which he left power (through a pact). The fall of President Balaguer is a “fuzzy” example of a presidential breakdown, and of a transition from authoritarianism to democracy. I cannot clearly categorize this case, as there are valid arguments on both sides. However, by conceptualizing the fall of Balaguer as a presidential breakdown rather than as a democratic transition, we use a different frame of comparison, and this itself yields new insights not only about this case but also about the theories that are more or less helpful when it comes to explaining the phenomenon as a whole.

Conclusion: The Implications of the Fall of Balaguer The fall of Balaguer teaches us that analyses of presidential breakdowns can draw lessons from the literature on transitions from authoritarian to democratic rule. When presidential breakdowns are triggered by a president attacking key democratic institutions, international involvement becomes more important. Democracy has become an important global policy goal, and U.S. or OAS involvement is more likely, and often crucial, if the survival of democracy is at stake. The Dominican case also shows that it is hard to separate in conceptual and analytical terms presidential breakdowns under hybrid regimes, on the one hand, and transitions from hybrid regimes to democracy, on the other. The two phenomena may exhibit more similarities than differences, and comparisons between the two should be explored further. Given that the fall of Balaguer was pacted and constitutional, one can expect either modest or long-lasting consequences to arise therefrom. Although it marks the “beginning of the end” of the era of the three caudillos who have dominated Dominican political life since the death of Trujillo

Dominican Republic and Fall of Balaguer

193

(Juan Bosch of the PLD, Peña Gómez of the PRD, and Joaquín Balaguer of the PRSC), the existing regime retains a neo-patrimonial logic secured partly by the pacted nature of the transition. So in some ways it has been business as usual. But the salida of Balaguer also had clearly positive consequences for Dominican democracy. The breakdown was caused by an attack on a central democratic institution: free, open, and competitive elections. Such limits upon contestation are quite typical under electoral authoritarian regimes, and so the removal of the person or institution responsible for attacking democracy has positive consequences on democratic procedures. 25 Indeed, in the Dominican case, electoral contests have been free and fair, and there has been no serious electoral fraud since 1994. What are the implications of the breakdown for the political institutions of the Dominican Republic? The constitutional reforms included a ban on reelection, provisions for midterm elections, and a ballotage electoral system for presidential elections. 26 Midterm elections have created the problem of constant electioneering and have also diminished the presidents’ ability to plan for the longer term. The level of support the president gets in Congress changes after each midterm election, so that coalitions or political alliances must be renegotiated every two years. This has caused detrimental gridlocks in the political system on various occasions (Marsteintredet 2008a). The most curious implication, however, has been the fact that the reelection issue has remained at the center of political debate since 1994. In 2002, following two years of debate, the constitution was amended under President Mejía (PRD) to allow a president to be reelected and serve a maximum of two consecutive terms. The reform split the governing party, as it effectively deprived other PRD hopefuls from being nominated as presidential candidates, and paved the way for the reelection of former president Leonel Fernández of the PLD (1996 –2000). Leonel Fernández and the PLD also benefitted from the ballotage system in the 1996 election: after winning second place in the first round, he allied himself with Balaguer’s PRSC to beat Peña Gómez in the second round. In this way, the Pacto por la Democracia and the salida of Balaguer paved the way for a new caudillo. Leonel Fernández is now in his third presidential term, and, at the time of writing, Congress is considering a constitutional reform that would allow him to serve a fourth and fifth presidential term, after four years out of power. The Pacto may have saved democracy and ended the era of the three historic caudillos, but it did not change the political culture, and it has given birth to a new Dominican strongman.

Notes * The author thanks Michael E. Álvarez, Mariana Llanos, Kathryn Hochstetler, Andrew Schrank, and Einar Berntzen for their comments. The chapter also benefited from a conversation with Wilfredo Lozano, and e-mail exchanges with Bernardo Vega and John W. Graham. Funding for the research was provided by the

194

Leiv Marsteintredet

Fulbright-Hayes Foundation, and the Leiv Eriksson mobility fund at the Norwegian Research Council. 1. The definition emphasizes the extraordinary nature of this event so as to exclude a permanent change in the electoral calendar. 2. Examples are the so-called colored revolutions in Serbia, the Ukraine and Georgia, and there are similar failed attempts in Belarus and Azerbaijan. 3. The most thorough account of the 1994 election can be found in Díaz (1996). See also Hartlyn (1994; 1998), Lozano (2002, ch. 3), Spanakos (2000), Wiarda (1998), and Sagás (2000) on racism during the campaign. International electoral reports are also available from the Organization of American States (OAS) and NDI (NDI 1994; OAS 1997). 4. Santiago Cantón (2009), who observed both elections for the NDI, argues that the 1990 elections were crucial to prepare international observers for the1994 election, and that both the PLD and the PRD were much better prepared in the event of fraud. 5. That summer, delegates of the PRD went to an OAS meeting in Belém do Para, Brazil, met with EU representatives in Brussels, and with politicians and bureaucrats in the United States to present evidence of fraud and gather international support. Representatives from the OAS electoral observer team stayed on in the Dominican Republic between May and August of that year. 6. The international observers also felt the tensions: an observer for the International Foundation for Election Systems (IFES) had to leave the country due to death threats, the NDI sent home their representatives soon after the elections were held for security reasons, and observers’ telephone conversations were being tapped (Graham 2008, 2009; Merloe, AI, November 25, 2008). 7. The national newspapers, Listín Diario and Hoy, reported only minor protests for the May 16–August 16 period. Some occurred at the campus of the Santo Domingo Autonomous University (UASD) in June, and some in Eastern Santo Domingo in July. These protests happened just as much because of the disappearance of UASD professor Narciso “Narcizaso” González as the electoral fraud. “Narcizaso” was kidnapped by public security forces on May 26 after denouncing the fraud, and disappeared. 8. Despite the army’s declaration of support for Balaguer, there was uncertainty about whether it would support a PRD uprising. Although PRD leader Esquea Guerrero, AI, January 13, 2009, later said that the army would have supported such an uprising, it is more likely that the army was divided over the issue. 9. Washington Post July 16, 1994; New York Times May 20–24, 1994; August 22, 1994; Dallas Morning News August 5, 1994. 10. The IFES also observed the elections. 11. He did compete in the 2000 presidential election for the PRSC, however, winning 24.6 percent of the vote in the first round. 12. Kim and Bahry (2008) include the suicide of President Guzmán (PRD) in 1982 as a case of presidential breakdown. Guzmán took his own life in the midst of a corruption scandal involving his daughter, who worked for his administration. The reasons for Guzmán’s suicide are not known, and speculations abound (e.g., Balaguer 1988, 327–328). The scandal involving Sonia Guzmán never generated protests or went beyond rumors, and Congress never considered removing

Dominican Republic and Fall of Balaguer

13. 14.

15.

16.

17.

18. 19.

20.

21.

22.

195

the president. Because we do not know why President Guzmán committed suicide, it cannot be regarded as a case of presidential breakdown. For details see Mariñez (1994, 248–255). In my interviews with observers participating in that electoral mission, some stated that they were convinced that the 1990 elections had also been stolen (Cantón 2009; Merloe, AI, November 25, 2008). Furthermore, at the time Bosch was showing signs of dementia. Congressman and PLD moderate Norge Botello (AI, February 12, 2003) argued that many in the party were relieved when Balaguer was declared the winner because they did not believe that Bosch was fit to govern. The United States had pressed for democracy in Haiti since the coup led by General Raoul Cédras had ousted democratically elected President Aristide in September of 1991. The OAS had instituted a trade embargo on Haiti to force the military leaders to withdraw, but as it became clear that a trade embargo was insufficient to remove General Cédras from power, in 1994 the Clinton administration gained UN Security Council backing to consider tougher measures. In September 1994 the Cédras regime finally backed down, and Aristide was able to return to Haiti. Late in May, the Balaguer government agreed to close the borders and help enforce the UN embargo. A lighter version of the embargo had been in place for some time, but Balaguer had refused to uphold it. The sudden change in policy was an attempt for a quid pro quo with the United States, which had Haiti high on its Caribbean agenda. The goal was to give in on the Haitian embargo and persuade the United States to let Balaguer off the hook for the fraudulent elections (New York Times July 1, 1994; May 30, 1994). According to Pat Merloe, Peynado openly said that the Dominican government had no intention of upholding the embargo. Peynado and Balaguer became furious when it was made clear to them that the United States would not accept the election results in exchange for the Dominicans upholding the embargo. There were some symbolic border measures adopted, but Dominican-Haitian borders were never actually sealed. New York Times August 4, 1994. The army was divided and decided not to pursue a coup on May 20. Balaguer is thought to have stayed on the sidelines to keep his options open. But he did not want to leave power without maintaining some control, and the pact allowed his party to maintain control of the Senate (Guerrero 1999). Although he was forced out by the United States three times, the United States did not manage to curb economic policies that benefited Balaguer as president of a patrimonial regime (Schrank 2003). The following section benefited from e-mail communication with Kathryn Hochstetler and Andrew Schrank. I wish to thank Andrew in particular for challenging the idea that the fall of Balaguer is clearly a case of presidential breakdown. See Negretto (2006, 80–81), Valenzuela (2004, 6), and Kim and Bahry (2008, 811) for similar definitions. Negretto refers to presidencies that are terminated before the end of the constitutionally mandated term, while Valenzuela, Kim, and Bahry refer to “early removals” (which I assume refers to an anticipation of what is determined by the electoral calendar). Marsteintredet and Berntzen (2008) include it as a case of “early election,” which seems correct. However, the definition suffers from the same problems as those already mentioned.

196

Leiv Marsteintredet

23. Balaguer supposedly claimed that “these elections were the cleanest in Dominican history,” and he and his party never admitted to fraud (Joaquín Ricardo, AI, January 12, 2009). 24. Similar lessons could be drawn from the Peruvian case, and to some degree the Guatemalan and Paraguayan breakdowns (see chapters 8, 9, and 11, this volume). 25. An even more positive outcome would have been either a new election or the acceptance of Peña Gómez’s victory. 26. For a full list see Darío Espinal (2001).

11 Peru and the Fujimori Presidential Breakdown in 2000: Continuismo Gone Bad Einar Berntzen and Tor-Einar Holvik Skinlo

Introduction The regime that Fujimori established in Peru after his 1992 autogolpe has generally been characterized as an electoral authoritarian regime. Elections serve a dual purpose and have an inbuilt contradiction under electoral authoritarian regimes: on the one hand, they are a source of both internal and external legitimacy; on the other, they threaten incumbents, since they are by nature uncertain and provide an opportunity for political opponents to gain power. Elections therefore create a serious dilemma for authoritarian incumbents: they must choose between allowing fair play at the cost of possible defeat, and violating democratic rules at the cost of generating domestic protest and international repudiation, which may lead to regime crisis. In order to secure regime continuismo, Fujimori resorted to extensive electoral fraud in the 2000 elections. The leak of a videotape exposing how the Fujimori regime had blatantly changed the outcome of the congressional election by bribing opposition legislators-elect to join the president’s coalition in the legislature ultimately triggered not only the downfall of Fujimori’s presidency but also the breakdown of the regime and a transition to democracy. The fact that the 2000 presidential breakdown affected what was an electoral authoritarian regime also meant that Fujimori’s fall from power triggered a transition from electoral authoritarianism to democracy. We argue that the presidential breakdown of Alberto Fujimori following the 2000 election must be analyzed in light of both theories of presidential breakdown and theories of gray-zone regimes and transitions from (electoral)

198

Einar Berntzen and Tor-Einar H. Skinlo

authoritarian rule. Fujimori’s decision to shorten his third presidential term from five years to one year, and his abrupt resignation before finishing that abbreviated one-year term, qualify this as a case of presidential breakdown. But the literature on presidential breakdowns cannot explain the outcome of Fujimori’s fall: it led not just to a change of government but also to a change of regime, from electoral authoritarianism to democracy. Reference to the literature on gray-zone regimes and transitions from (electoral) authoritarian regimes allows us to account more fully for the outcome of Fujimori’s abrupt resignation, and complements the literature on presidential breakdowns. In what follows, therefore, we reassess the fall of President Fujimori in 2000 as a case of a presidential breakdown within the context of an electoral authoritarian regime. Taking regime type (electoral authoritarian) into account helps us to understand the mode of presidential breakdown (i.e., why Fujimori fell in connection with an electoral fraud). The chapter is divided into three sections. The first describes the fall of Fujimori. The second provides a brief review of the literature on presidential breakdowns and suggests how theories of regime types and transitions help to account more fully for the outcome of the presidential breakdown in Peru. The third section reassesses the causes and the modality of the presidential breakdown of Fujimori in light of the regime/transition and presidential breakdown literature. In the last section we focus on the outcome and consequences of the presidential breakdown: the restoration of the minimal procedures of democracy.

The Fujimori Presidential Breakdown The beginning of the end of fujimorismo occurred on September 14, 2000, when one of the only remaining independent television stations, cable outlet Channel N, broadcast a video showing Fujimori’s spymaster, Vladimiro Montesinos, paying opposition congressman-elect Luis Alberto Kouri US$ 15,000 in cash to switch sides and support the president (McMillan and Zoido 2004, 89). The so-called vladivideo exposed the illegitimacy of the congressional “majority” that Montesinos had established through bribery and effectively put paid to Fujimori’s already questionable electoral legitimacy. Although the scandal involved Montesinos, it affected Fujimori’s survival in office since the relationship between the two men was a mainstay of the regime. The release of the vladivideo destroyed Montesinos, but it also destroyed the informal network of power established by the president with Montesinos. President Fujimori immediately sensed the gravity of the situation. After failing to save himself by forcing Montesinos to resign on September 16, Fujimori announced that he would end his term in office a year later, on July 28, 2001, and hold new elections in which he would not be a candidate.

Peru and the Fujimori Presidential Breakdown

199

This pre-announced presidential breakdown by way of early elections was intended as a preemptive strike to stall for time and secure a controlled exit. With this flexible exit strategy, Fujimori hoped to save his government by getting rid of Montesinos, and to retain control of the transition process. Congress quickly approved a constitutional amendment (Law 27.365 of November 4, 2000), which abbreviated the terms of the incumbent president, vice presidents, and Congress, and prohibited immediate presidential reelection (Schmidt 2002, 355).1 The date of the elections was set for April 8, 2001 (Balbi and Palmer 2001, 69). Fujimori thus gave himself a year to wrap things up before he would have to leave the presidential palace. The decision to shorten his term from five years to one year parallels the decision by Dominican President Balaguer to shorten his term to two years in 1994 (see chapter 10 in this volume). However, whereas Balaguer finished his shortened term, Fujimori resigned before finishing his. The Organization of American States (OAS) and the United States helped Fujimori to get Montesinos out of Peru and facilitated asylum in Panama, as it was feared that there might be a military coup since Montesinos controlled the army. The OAS mission provided a framework for negotiations between the government and the opposition (Cooper and Legler 2001, 125). 2 But when the government demanded a blanket amnesty as a condition for implementing the electoral transition, the opposition broke off talks, and made its first attempt to remove the fujimorista leadership of Congress by a vote of censure. The fujimoristas won the vote by 60 to 56.3 Fujimori’s plan to control his exit from power was weakened by the “secret” return of Montesinos to Peru on October 23, which generated fears of an imminent coup.4 Montesinos’s return created divisions within the military, and triggered the quixotic rebellion led by army captain Ollanta Humala in Arequipa on October 29 to denounce corruption in the armed forces. 5 Fujimori’s authority was weakened further when First Vice President Francisco Tudela resigned in opposition to the proposed amnesty law and to protest Montesinos’s return. Sensing the weakness of the Fujimori government, politicians began to realign. The opposition camp grew as defections from Fujimori’s legislative coalition picked up pace, and Fujimori ultimately lost his legislative shield in Congress (Pérez-Liñán 2007). On November 16, the opposition, now in the majority, voted 64 against 51 to unseat the pro-government congressional leadership led by Martha Hildebrandt.6 There was a consensus that Hildebrandt should be replaced as president of Congress by the widely respected career politician Valentín Paniagua of the Popular Action Party (Partido de Acción Popular, PAP), one of the few remaining members of the traditional parties. President Fujimori thus faced an opposition-controlled legislature for the first time since 1992. In 1992, rather than enduring the difficult negotiations with the opposition in Congress, Fujimori had decided to stage the autogolpe that led to the dissolution of Congress, the dismissal of the judiciary, and the suspension of

200

Einar Berntzen and Tor-Einar H. Skinlo

the 1979 constitution.7 In 2000, having lost the confidence of the majority in Congress, he chose to resign rather than to complete the transition. Thus, when he could no longer play the political game according to his own rules, he left the playing field and went home to Japan, from where he notified Congress about his decision to resign by fax, on November 20. Because he resigned, Fujimori left office before completing his shortened one-year term. Thus, Fujimori’s presidency broke down in two steps: first, on September 16, he shortened his five-year term, announcing that he would resign after one year following early elections; second, he abruptly resigned on November 20, before the end of his new constitutional one-year term. Congress refused to accept Fujimori’s resignation by fax, and decided instead, in a 62 to 9 vote on November 21, that the presidency had been vacated due to the incumbent’s “moral incapacity” (Balbi and Palmer 2001, 70). However, since President Fujimori was formally removed from office by a congressional vote (Carrión 2006, 1) only after he had de facto resigned, the effect of the declaration of “incapacity” was merely symbolic (Pérez-Liñán 2007, 193; Marsteintredet and Berntzen 2008). According to constitutional procedure, First Vice President Francisco Tudela should occupy the presidency, but he had already resigned in October. Next in the line of succession was Second Vice President Ricardo Márquez, but he too had resigned shortly after Fujimori. On November 22, therefore, after the resignation of both the first and second vice presidents had been accepted, the constitutional succession put the president of the Congress in the presidency. Valentín Paniagua was sworn in as interim president until July 28, 2001 (Schmidt 2002, 355). Fujimori’s second presidential breakdown in November also led to the final breakdown of the electoral authoritarian regime and to regime change. With Fujimori out of the way, power and control of the transition passed entirely into the hands of the opposition. The democratic transition under Paniagua was not “pacted” as there was no agreement between the outgoing Fujimori authoritarian regime and the incoming democratic government. On April 8, 2001, new presidential and congressional elections were held. With the inauguration of the new president Alejandro Toledo on July 28, 2001, the Peruvian transition to democracy was completed.

Electoral Authoritarian Regimes and Presidential Breakdowns Some authors argue that institutional factors linked to the perils of presidentialism and divided government are central explanatory factors of presidential breakdown (Linz 1990; Valenzuela 2004). Others argue that a mobilized and hostile citizenry is more important to explain presidential breakdown than challenges by the legislature (Hochstetler 2006). Focusing more narrowly on impeachments, Pérez-Liñán (2007) observes that both mobilization

Peru and the Fujimori Presidential Breakdown

201

and legislative challenges are triggered by scandals involving the president. Finally, dissatisfaction with neoliberal economic policies, economic stagnation, and declining growth also play a role, since pressure to adopt economic austerity measures can produce overt public discontent (Hochstetler 2006; Kim and Bahry 2008; chapter 2 in this volume). Although helpful, these authors fail to explain fully the specific causes, mode, and outcome of Fujimori’s presidential breakdown. We need to draw on theories on transitions from (electoral) authoritarian rule and gray-zone regimes to gain a better understanding of this case, since the Fujimori breakdown had the characteristics of a transition from an electoral authoritarian regime to democracy. In the literature on transitions from authoritarian rule, one finds the argument that “cracks” within the ruling coalition that divide hard-liners from soft-liners are often a necessary condition for a transition to occur (O’Donnell and Schmitter 1986). This observation is also relevant for the dynamics of electoral authoritarian regime breakdown. In Peru, a “scandal” generated intra-regime tension (between Montesinos and Fujimori), which caused the breakdown and led to a democratic transition (Cameron 2006). Thus, the literature on transitions from authoritarian rule can help to explain the causes of the presidential breakdown in Peru. There is a more recent literature on regimes stuck in a “gray-zone” between authoritarianism and democracy (Carothers 2002). Terms such as “competitive authoritarian” (Levitsky and Way 2002) or “electoral authoritarian” (Schedler 2002) have been used to describe such regimes. Electoral authoritarian regimes hold elections and tolerate some pluralism and interparty competition, but incumbents violate minimal democratic norms so often and to such an extent that the regimes fail to meet conventional minimum democratic standards (Levitsky and Way 2002).8 These regimes are usually seen not so much as examples of “limited” democracy but more as examples of “mitigated” authoritarianism. After the fall of Peruvian democracy with the 1992 autogolpe, the United States and the OAS pressed for a return to democracy, and formal constitutional rules and elections were reestablished. There were elections in November 1992 for a Democratic Constituent Congress, which was charged with drafting a new constitution, approved in a referendum in October 1993 (McClintock 2006). Fujimori won the 1995 presidential election and a safe majority in the new single chamber Congress. His regime cannot be considered democratic, however, even by “Latin American standards,” because democratic principles were violated so systematically (Carrión 2006, 2). Most analysts of Peruvian politics (Carrión 2006a; McClintock 2006; Levitsky and Way 2002; Calderón 2001; Cameron 1998) therefore describe the Fujimori regime as electoral authoritarian. In such regimes, elections provide the regime with its main “democratic alibi.” In Peru, this “cover was blown” with the release of the vladivideo. The gray-zone literature helps us to understand the mode of the Fujimori presidential breakdown (the shortening of the presidential term and the

202

Einar Berntzen and Tor-Einar H. Skinlo

holding of early elections), the uneven playing field for government and opposition, and the blatant electoral fraud that characterizes electoral/competitive authoritarian regimes. Considering the regime type dimension also helps to explain why the outcome of the presidential breakdown was a transition to democracy. Certain kinds of authoritarian regimes are more vulnerable to breakdown than others (McFaul 2005). We posit that semi-authoritarian or electoral authoritarian regimes are more open to democratization than full-blown authoritarian regimes. Hadenius and Teorell (2007) find that the majority of transitions from electoral authoritarian regimes result in democracy, and that this subtype of authoritarianism is “the typical stepping-stone to democratization” (2007, 152). In Peru, the breakdown of Fujimori’s presidency triggered a transition from electoral authoritarianism to a minimally functioning democracy.

The Fujimori Breakdown as a Breakdown of an Electoral Authoritarian Regime Presidential reelection was the core issue informing the decision to carry out the autogolpe. The increasingly authoritarian character of the Fujimori regime was linked to Fujimori’s continuista project, namely, his desire to run for an unconstitutional third term. Article 112 of the 1993 Constitution prohibited the president being reelected a second time. In order to clear the way for this in 2000, Congress passed the so-called Law of Authentic Interpretation of the Constitution in August 1996. This stated that Article 112 could not be applied retroactively, so that Fujimori had only been elected once under the new constitution. This law established the legal foundations for Fujimori’s 2000 reelection bid. Despite the obvious unconstitutionality of Fujimori’s effort to legalize his second reelection bid, the opposition was unable to challenge it through institutional channels. Fujimori’s legislative majority limited the ability of Congress to check executive power. Other mechanisms of horizontal accountability, such as the courts and the electoral authorities, were corrupted by the vast network of espionage, bribery, and extortion managed by Fujimori’s eminence grise, Vladimiro Montesinos, who headed the National Intelligence Service (Servicio de Inteligencia Nacional, SIN). The greatest obstacle to Fujimori’s reelection project was the Constitutional Tribunal, which was charged with the interpretation of constitutional law. In January 1997, the tribunal ruled, in a 3 to 2 vote, that the Law of Authentic Interpretation did not apply to Fujimori. The fujimorista Congress retaliated on May 28, 1997 by voting to remove the three members of the Constitutional Tribunal who had voted in January against presidential reelection (Levitsky and Cameron 2003, 15). The justices were never replaced and the Constitutional Tribunal ceased to function (McClintock 2006, 251). For the first time in a decade, public discontent with the Fujimori administration

Peru and the Fujimori Presidential Breakdown

203

became visible as the opposition took to the streets to protest against the dismissal of the three justices. Without recourse to the legislature, the courts, or the electoral authorities, a new civil society movement was organized by the opposition in an attempt to call a referendum on the reelection issue.9 However, this effort was thwarted by another act of Congress, a law stating that forty-eight congressional votes were necessary for a referendum to be held. Since the opposition was unable to muster forty-eight votes in the legislature, this effectively meant that only the government could make use of referenda. Nevertheless, the opposition went ahead with the campaign to collect the 1.2 million signatures needed to call a referendum. It gathered more than 1.4 million signatures, but in August 1998, Congress voted sixty-seven against forty-five to block the referendum (Levitsky and Cameron 2003, 15). When the opposition’s referendum initiative was blocked, thousands took to the streets in nationwide protests (Carrión 2006). Most Peruvians seemed to oppose the government’s abuses, but notwithstanding the nationwide protests, opposition to the regime did not become a sustained prodemocracy movement. The protests fizzled out, and the Constitutional Tribunal and referendum issues soon faded from the national agenda. Thus, in December 1999, the fujimorista dominated National Election Board (Junta Nacional Electoral, JNE) officially approved Fujimori’s candidacy for the 2000 presidential election. Although the electoral playing field was heavily tilted in favor of the incumbent, Fujimori’s reelection campaign in 2000 became one of the dirtiest campaigns in Peruvian history. The divided opposition stood little chance of defeating the president, but opposition candidates were openly harassed and their political campaigns obstructed throughout the election campaign. With the campaign in full swing, evidence emerged that the incumbent regime was actively engaged in fraud (McClintock 2006) with the so-called signatures scandal. This involved the mass falsification of signatures in order to register Fujimori’s 2000 electoral vehicle, Peru 2000. Unable to collect the nearly 500,000 signatures required to register a new party, as many as 1.2 million signatures had been falsified (McClintock 2006, 258). The signatures affair seemed to confirm that Fujimori was prepared to go to any lengths to remain in power. However, the National Office of Electoral Procedures (Oficina Nacional de Procedimientos Electorales, ONPE), the body in charge of administering the count, was eventually forced to abandon its plan to declare Fujimori the “winner” already after the first round. The official results of the first round on April 9 gave Fujimori 49.8 percent of the vote and Toledo 40.3 percent. The ONPE would probably have declared Fujimori’s first-round “victory” had it not been for intense scrutiny by national and international observers and massive street protests on electionnight, and for three days thereafter, in support of Toledo and a second round (Taylor 2001, 12–13). The government adamantly refused to correct its abuses and establish a level playing field for the runoff election on May 28 (OAS 2000). The

204

Einar Berntzen and Tor-Einar H. Skinlo

opposition demanded that independent observers be granted access to the ONPE computing system. Instead, in mid-May, the electoral authorities announced the introduction of uncontrolled new computer software for the runoff, indicating that the government would rig the count if necessary. This led Toledo to withdraw from the race, leaving Fujimori to “win” the second round unopposed. Dislodging the president through electoral competition had proven impossible. The government’s manipulation and fraudulent behavior triggered public contestation of the validity of the 2000 elections. It was not just the opposition that denounced the tilted electoral playing field; even the OAS, which had been previously reluctant to criticize the Fujimori government, concluded in its final electoral observation report that: “. . . with respect to international standards, the Peruvian electoral process is far from being considered free and fair” (McClintock 2001, 137). In the wake of the 2000 reelection debacle, there were intense protests, culminating in the huge popular demonstration known as la Marcha de los Cuatro Suyos, which coincided with Fujimori’s July 28 swearing-in ceremony (Carrión 2006, 146). Contrary to Hochstetler’s (2006) findings, however, street protest did not dislodge the Fujimori regime or precipitate its fall. Both the U.S. government and the OAS observer mission declared the 2000 election unfair, but the OAS General Assembly decided not to question its legitimacy, and resolved to send a High-Level Mission to Peru instead, with a mandate to explore future reforms to “strengthen democracy” (OAS 2000; McClintock 2001, 138). Ultimately, both the United States and the OAS accepted Fujimori’s third term as a fait accompli. In short, the Fujimori government was unable to pass the test of victory in a free and fair election, and while this led to a loss of legitimacy and debilitated the regime, it did not undermine its stability. Indeed, Fujimori enjoyed a popularity rating of above 45 percent even after the 2000 elections (Cameron 2006, 283). So how are we to explain the downfall of the president and his whole regime? The event that set the process in motion was the vladivideo, although a prior arms-dealing scandal involving Montesinos had already diminished U.S. support for the regime. On August 21, at a stage-managed press conference, Fujimori and Montesinos announced the uncovering of a network of international arms traffickers who had sold weapons to the Revolutionary Armed Forces of Colombia (Fuerzas Armadas Revolucionarias de Colombia, FARC). Although Montesinos claimed the credit for this discovery, it soon became known that Montesinos and his army associates had orchestrated the gunrunning operation themselves (Tanaka 2005, 285–286). The revelation of his involvement lost him the support of the United States, for which defeating the FARC was a strategic priority. Thus, Montesinos ceased to be an asset and became a liability. The United States tried but failed to persuade Fujimori to fire Montesinos.10 Fujimori may have feared that there would be a coup against him if he moved against his advisor, who had solid backing

Peru and the Fujimori Presidential Breakdown

205

from the highest ranks of the armed forces.11 But Montesinos’ role in the arms deal may have strengthened the resolve of his enemies within the military. It appears that a group of disgruntled naval officers was involved in the release of the vladivideo, which suggests there were internal divisions within the military.12 In 2000, the government won only 52 of 120 seats in Congress. The failure to win a congressional majority led the Fujimori government to engage in further fraudulent behavior, with Montesinos “buying” individual opposition members of Congress in an attempt to reconstruct a majority (Levitsky and Cameron 2003, 21). The event that ultimately triggered the downfall of the regime was the revelation of this outrageous buying of legislators-elect (Kenney 2003, 1227). Had it not been for the television broadcast of the vladivideo, which exposed the Fujimori administration’s shameless attempts to pervert the outcome of the congressional election retroactively by buying a compliant majority, Fujimori may well have survived as head of the executive for another five years (Carey 2003, 128).13 To understand how the airing of the vladivideo on September 14 became so crucial in the demise of the regime, we must consider the regimerelated nature of the scandal. By showing Vladimiro Montesinos bribing a congressman-elect, the videotape exposed the total illegitimacy of the congressional majority that Montesinos had established. The scandal was regime-related because it revealed the corrupt nature of the regime, providing irrefutable “smoking-gun” evidence that Montesinos was running a secret “government within the government.” Thus, the event triggered a regime crisis rather than just a government crisis, as in other cases of presidential breakdown.14 The scandal involved Montesinos but affected Fujimori’s survival in office. It created a crack in the edifice of the regime because it produced a conflict between Fujimori and Montesinos within the regime. The literature on the breakdown of authoritarian regimes finds that splits and divisions within the ruling bloc are often a necessary condition for a transition to occur (O’Donnell and Schmitter 1986). Consistent with this literature, the vladivideo was the decisive event in the breakdown of the Fujimori regime because it generated a division within its ranks: the rupture between Fujimori and Montesinos caused the breakdown from within. Not surprisingly, the impact of the video on the public’s view of the regime was profound. In October 2000, Fujimori’s approval rating dropped to 36 percent, and in November, just before his faxed resignation, to 24 percent (Kenney 2003, 1235). The effect of falling public approval would have been limited, however, had it not been for the internal splits within the regime. Disgruntled elements within the military facilitated the revelation of the vladivideo, and the ensuing scandal unleashed a power struggle between Fujimori and Montesinos, and created internal tensions that triggered the collapse of the regime. Montesinos had kept a video archive to ensure his own survival, as it meant that Fujimori could not remove him without jeopardizing the entire regime.

206

Einar Berntzen and Tor-Einar H. Skinlo

The videos were the equivalent of weapons of mutually assured destruction. With the release of the vladivideo, the weapon had been deployed. It brought down Montesinos, but it also destroyed Fujimori and the entire informal network of power they had woven together. Montesinos became a liability, but Fujimori was unable to govern without him. President Fujimori immediately realized that he could not survive if he did not fire Montesinos. Caught between a rock and a hard place, Fujimori initially tried to save himself by firing his spymaster, but the latter refused to resign. Seeing no other way to force Montesinos out, on September 16, Fujimori announced that he would end his term in office a year later, on July 28, 2001, and hold new elections in which he would not be a candidate. With this political “hara-kiri,” a tale of preannounced presidential interruption by way of early elections, Fujimori hoped to get rid of Montesinos, save his government, and preside over a transition process under his ultimate control. The electoral authoritarian nature of the Fujimori regime thus sheds light on the mode of Fujimori’s presidential breakdown: the curtailing of his third term by four years and the holding of new elections. Fujimori fell as a result of an attack on the only truly democratic source of legitimacy of his regime. The vladivideo was the smoking gun that revealed an outrageous attack on the democratic institution of free and fair elections. After Fujimori’s announcement of his decision to shorten his term and hold new elections, which created an opening for a possible negotiated solution to the regime crisis, international factors began to play a role in the Peruvian transition process. The High-Level OAS Mission with a mandate to “strengthen democracy” had established its presence in the country in midAugust and provided a framework for negotiations between the government and the opposition (Cooper and Legler 2001, 125). In this sense, international involvement in the political crisis was coincidental, since the OAS was in the country before the crisis. In addition to providing a venue for negotiations, the presence of the OAS conveyed the message to key actors in the Fujimori government that the international community was keeping a watchful eye on developments in Peru, and that neither repression nor a military coup would be tolerated (McClintock 2001, 139). Given Montesinos’ control of and support from the army, the possibility of a coup against Fujimori could not be excluded.15 The OAS and the United States were instrumental in arranging Montesinos’ exit by facilitating his asylum in Panama. With Montesinos out of the country, Fujimori regained control, but the transition negotiations with the opposition soon broke down as the president tried to impose too many conditions. The issue of amnesty was used as a deal-maker by the Fujimori government, but it was equally a deal-breaker for the opposition: when the government demanded a blanket amnesty to protect itself from accountability for corruption as a condition for implementing the electoral transition, the opposition broke off discussions. There ensued mounting opposition demands for Fujimori’s resignation and for an immediate change in congressional leadership, but the opposition failed to remove

Peru and the Fujimori Presidential Breakdown

207

the fujimorista leadership in Congress through a censure motion (Hamilton 2006, 202–203). In mid-October, therefore, the Fujimori government still seemed to have a chance of surviving until a new government took office in July 2001. Montesinos’ return to Peru on October 23, however, put paid to Fujimori’s plan to oversee and control the political transition. This event unravelled the government and finally ended the electoral authoritarian regime. The stage was set for the final showdown between Fujimori and Montesinos. Fujimori had personally led a massive manhunt for Montesinos and had taken several of the general’s military associates into custody, but the president had failed to alleviate suspicions that his rupture with Montesinos was inauthentic.16 Montesinos’s return sparked divisions within the military. Army captain Ollanta Humala’s rebellion on October 29 in the military outpost of Arequipa was insignificant in military terms, but its psychological impact was important as it conveyed the impression of imminent regime collapse (Hamilton 2006, 206). Fujimori’s government was weakened further by the desertion of First Vice President Francisco Tudela, who resigned in opposition to the proposed amnesty law and in response to Montesinos’s return. Sensing the weakness of the Fujimori government, many began to defect from Fujimori’s legislative coalition and to swell the ranks of the opposition, thus losing Fujimori his legislative shield (Pérez-Liñán 2007). When Congress held a second vote to censure the fujimorista congressional leadership on November 16, the turncoats (transfugas) who were beholden to Montesinos contributed to Fujimori’s political demise by voting with the opposition (Hamilton 2006, 207). Fujimori, it turned out, could no more survive with Montesinos than survive without him. Fujimorismo without Montesinos was not viable. With Congress controlled by the opposition, Fujimori was no longer immune to Article 114 of the 1993 Constitution, which permitted Congress to remove the president by a simple majority vote on grounds of physical or moral incapacity (Hamilton 2006, 193). As soon as Congress was in the hands of the democratic opposition, Fujimori was exposed to imminent removal. But rather than wait for Congress to determine his fate, Fujimori preemptively faxed his resignation announcement from Japan to avoid impeachment by Congress and being declared morally incapacitated. By stepping down when the opposition gained control over Congress, Fujimori technically followed the logic of parliamentarism, stepping down when his party no longer commanded a congressional majority (Pérez-Linán 2005, 55; Marsteintredet and Berntzen 2008). But in the context of an electoral authoritarian regime, Fujimori’s resignation from the safety of exile followed the logic of authoritarianism: when he could no longer play the political game according to his own rules he chose escape instead of accountability. Fujimori’s presidential breakdown, resigning before he could complete his shortened one-year third term, marked the culmination of the disintegration of the electoral authoritarian regime and the start of a transition to

208

Einar Berntzen and Tor-Einar H. Skinlo

democracy. With Fujimori out of the way, power and control of the transition passed into the hands of the opposition. The transition under the leadership of Paniagua was not “pacted,” as there was no agreement between the outgoing electoral authoritarian regime and the incoming democratic government.

The Outcome of the Fujimori Breakdown: Transition to Democracy The interim regime was led by Valentín Paniagua, an experienced politician from the center-right PAP. The new cabinet of national unity consisted of respected figures covering the political spectrum. Former United Nations secretary general Javier Pérez de Cuellar became the new prime minister and foreign minister. Interim president Valentín Paniagua proved to be an able consensus-builder and capable of seeking points of agreement among the various players, so that the interim government was backed by all antiFujimori sectors. The first priority of the transitional government was to reestablish public confidence in state institutions by dismantling the mechanisms of corruption set up by the outgoing regime. The Fujimori-Montesinos power structure was quickly dismantled and the transitional administration restored stability to the political system (Taylor 2005). One of President Paniagua’s first acts was to reestablish the Constitutional Tribunal by restoring to their posts the three justices who had been removed in 1997 in response to the ruling that a third Fujimori presidency would be unconstitutional. The interim government also abolished the system of temporary judges and presidential judicial oversight commissions that Fujimori had set up in 1992 and 1993 to influence legal decisions. Fujimori had abandoned institutional procedures when promoting military officers in order to concentrate power in the hands of a small clique. This had marginalized the institucionalista camp, which was frustrated with the absence of promotion prospects for anyone who was not in Montesinos’ confidence. The interim government therefore moved swiftly to retire key Montesinos supporters in the armed forces and replace them with “institutionalists” in order to prevent open armed forces dissidence. Sixteen army generals, fourteen air force generals, and thirty-two police generals, as well as 138 other officers were retired (Balbi and Palmer 2001, 70). By swiftly implementing such significant changes, the interim government signalled its determination to tackle the abuses and corruption of the Fujimori regime. It was assisted by an outpouring of popular support, as Peruvians regained their confidence in the state. The transitional government was emboldened by the legitimizing effects of broad-based popular support, and by the almost daily revelations of corruption under the previous regime. The approval rating of the interim government stood at 73 percent in late

Peru and the Fujimori Presidential Breakdown

209

December 2000 (Balbi and Palmer 2001, 72). Faultlessly democratic, almost everybody viewed the Paniagua government as a success. Politically independent officers replaced Montesinos appointees in the JNE and the ONPE. The April and June 2001 elections marked a fundamental break with the authoritarianism of the Fujimori regime. In sharp contrast to the 2000 elections, they were held in an exemplary democratic fashion and praised by foreign and domestic observers alike (OAS 2002). Especially important was the changed attitude of the police and military, which went out of their way to support the election. The electoral results were devastating for the former ruling party: the vote for fujimorismo fell from 49.9 to 1.7 percent of the vote. The fujimoristas were also virtually wiped out in Congress, gaining only 4 of the 120 congressional seats (Schmidt 2003). In June 2001, Alejandro Toledo, who promised to continue to strengthen Peruvian democratic institutions, won the second round of the presidential election with 53 percent of the vote. However, the results in Congress showed that party loyalty was weak and political fragmentation strong. Clearly, the new executive would not enjoy a parliamentary majority and would have to engage in coalition-building (Taylor 2005, 579). The inauguration of Toledo as president on July 28, 2001, marked the successful completion of the Peruvian transition to democracy. By laying the basis for a successful democratic transition in 2000–2001, the presidential breakdown of Fujimori had a clearly positive effect on the Peruvian regime. The most important implication of the presidential breakdown for the presidential regime in Peru was the constitutional prohibition of immediate presidential reelection (Schmidt 2002, 355). In the Peruvian case, the presidential crisis occurred within the context of an electoral authoritarian regime. In such a context the outcome of the presidential breakdown is positive since the breakdown can help promote a transition to democracy. This stands in sharp contrast with the findings of the literature that links presidential breakdowns with the perils of presidentialism and democratic breakdown. Valenzuela is right about the authoritarian temptations inherent in hyper-presidentialism: “The cautionary tales of Fujimori, Aristide, and most recently Chávez in Venezuela, show how presidentialism can be perverted into quasi-authoritarianism or even dictatorship” (Valenzuela 2004, 14). However, Valenzuela (2004) warns against such dangers in the case of Peru’s failed presidency of 2000. His overly pessimistic view of presidentialism fails to take into account the implications of regime type. The perils of presidentialism may have been at work in the 1992 autogolpe, but because the presidential breakdown of 2000 affected an electoral authoritarian regime, Fujimori’s fall must be evaluated positively, at least in terms of its consequences for democracy. In many respects, the interruption of the presidency in 2000 can be seen as a democratic success within the framework of a presidential regime, as a corrupt president who had perverted democratic institutions fell, thus paving the way for a transition from electoral authoritarianism to democracy.

210

Einar Berntzen and Tor-Einar H. Skinlo

Levitsky (1999) was the first to note the counter-cyclical nature of Peruvian politics. The descent into electoral authoritarianism in Peru in the 1990s represented a counter trend in a regional context dominated by transitions to democracy. Similarly, the fall of Fujimori in 2000 and the breakdown of the electoral authoritarian regime that triggered the transition to democracy contrasts with an ongoing cycle of democratic erosion in the region (Mainwaring and Pérez-Liñán 2005) and of presidential breakdowns in electoral democracies (as in Ecuador), which have generated worries about the negative consequences for democracy. The Peruvian presidential crisis in 2000 mirrors similar transitions to democracy from electoral authoritarian regimes in Georgia in 2003, in the Ukraine in 2004, and in the Dominican Republic in 1994, and does not fit with the literature on presidential breakdowns that focuses on the negative impact that presidential crises have on democracy. Consequently, the 2000 Peruvian experience illustrates the need to complement the findings of the literature on presidential breakdowns with the findings of the literature on gray-zone regimes and transitions from electoral authoritarian regimes.

Notes 1. According to the 1993 Constitution, two successive legislatures had to approve an amendment for it to pass. In order to solve the matter quickly, the government used its executive authority to change the legislative calendar. The ongoing Congressional session would end in mid-October after the first affirmative vote, after which the second session would begin allowing for the second affirmative vote (Hamilton 2006, 202). 2. The OAS General Assembly decided not to question the legitimacy of the 2000 elections, choosing instead to send a High-Level Mission to mediate a “national dialogue” on future democratic reform (McClintock 2001, 138). 3. According to U.S. Ambassador Hamilton, six or seven Montecinistas (the turncoats, or transfugas) who took their voting instructions directly from Montesinos) had voted with the government, indicating that the break between Fujimori and Montesinos was not complete at this point (Hamilton 2006, 203). 4. After his arrival in Peru, Montesinos immediately went into hiding. In the days that followed, Fujimori personally led a bizarre police and military manhunt for Montesinos, conducting house-to-house searches. In December 2000, it became known that Montesinos had managed to escape from Peru on October 29, boarding a yacht sailing to the Galapagos Islands. From there he made his way to Costa Rica and Aruba before ending up in Venezuela, where he underwent plastic surgery (Hamilton 2006, 210; Balbi and Palmer 2001, 70). 5. In 2006, Ollanta Humala won the first round of the presidential elections but was defeated by Alan García in the runoff. 6. In this second vote to remove the fujimorista Congressional leadership, the congressmen beholden to Montesinos voted with the opposition (Hamilton 2006, 207).

Peru and the Fujimori Presidential Breakdown

211

7. An autogolpe is defined as a temporary departure from democratic rule (Cameron 1998, 125). Although Peru did not return to full democracy, we have nonetheless used this term. 8. We follow the definition of democracy in Mainwaring, Brinks, and Pérez-Liñán (2001), according to which modern representative democracy has four defining characteristics: universal suffrage, the head of government and the legislature are chosen in open and fair competitive elections, civil liberties are effectively guaranteed, and elected authorities have real power to govern. 9. The right of citizens to initiate a referendum was inscribed in the 1993 Constitution. 10. At a meeting with Fujimori in New York on September 8, 2000, Secretary of State Madeleine Albright had indicated that the United States wanted Montesinos removed from office (Hamilton 2006, 195). 11. It has since been documented that Montesinos had planned to depose Fujimori and make former economy minister Carlos Boloña provisional president (Tanaka 2005, 285–286). 12. The vladivideo was given by Montesinos’ personal secretary and confidante, Matilde Pinchi Pinchi, to an intermediary who posed as an emissary for a group of disgruntled naval officers (Cameron 2006, 275–276). 13. The video-taped revelation provided indisputable proof of corruption. Seeing a bribe actually being paid is more convincing than being told about it (McMillan and Zoido 2004, 90) 14. A series of vladivideos subsequently became public. Taped by Montesinos as he bought off key personnel at all levels of the regime, they showed how the culture of corruption had penetrated almost every elite sector. Electoral authorities, journalists and editors, mayors, politicians, judges, lawyers, and congressmen watched themselves taking bribes on tape. Ultimately all state institutions were discredited, with the exception of the Ombudsman’s office (Taylor 2005, 571). 15. At one point, the military had approached First Vice President Tudela suggesting that it was prepared to remove Fujimori and make Tudela president. Tudela turned down the proposition (Hamilton 2006, 200). 16. Seventy-three percent of the public did not believe that Fujimori’s break with Montesinos was serious (Balbi and Palmer 2001, 69).

This page intentionally left blank

Conclusions: Presidential Breakdowns Revisited* Mariana Llanos and Leiv Marsteintredet

The chapters of this volume demonstrate that Latin American presidential breakdowns have been the outcome of very complex and multifaceted processes. They are situated at the end point of a chain of events, which makes it extremely difficult to weigh which factors are decisive in the exit from power of presidents. Economic crises, coalition breakdowns, intragovernment crises, scandals, policy deadlocks, and social protest have often jointly preceded the early departure of a president. Thus, borrowing from Ragin (1987, 27), presidential breakdowns seem to be caused by the combined effect of a set or a conjunction of causal factors that intersect at a particular moment in time, creating a “perfect storm” that hits a president. Table C.1 below presents an outline of the contributions to this book. In this concluding chapter, we revisit the comparative literature on the causes of presidential breakdowns and discuss how one of the factors highlighted in several chapters—intra-government conflict—may contribute to presidential breakdown. We then examine the implications of presidential breakdown for presidentialism, democracy, and policymaking, which are also briefly reviewed in table C.1.

The Causes of Presidential Breakdowns The comparative literature has tended to emphasize the last link in the chain of events leading to a presidential breakdown, that is, a combination of street challenges and legislative action. This focus prompted a debate on which of these two factors mattered the most (see Pérez-Liñán 2008, for example), leading to the conclusion that it was street protest that is the dominant cause (see introduction). In several of the contributions to this book, the influence of the protest variable is recognized, but the claim that it is relatively more important than Congress was challenged.

Table C.1 Presidential breakdowns: A country-based summary of causes and outcomes Country and Year of breakdown

Causal Factors (Dominant and Others)

Aftermath

Argentina (1989, 2001) (+ 2 caretakers: 2001, 2003)

Institutional: Minority government; midterm electoral defeat. Presidential: Imperial behavior and isolation for de la Rúa; moderate policy preferences for Alfonsín. Others: Economic adversity; lootings Events: Generalized, high-intensity street protests Others: Minority governments; intra-coalition conflicts Institutional: Minority coalition government. Presidential: Extreme policy preferences; imperial behavior first, then confrontation and negotiation Others: Inflation; scandal; declining presidential popularity; social protests

Executive instability Flexibilization of rules (presidential fixed-term; call of early elections)

Bolivia

Brazil

Dominican Republic

Ecuador

Presidential: Incompatibility of authoritarian leadership with democracy Events: Unlawful term extension by electoral fraud Others: international pressure; strong political parties manage to demonstrate fraud Institutional: Limited coalition incentives; adversarial congressional majorities Presidential: Imperial behavior; high cabinet turnover Others: The role of the military; indigenous and popular mobilization; economic crises

Flexibilization of fixed-term rules

No apparent flexibilization of the presidential fixed-term Flexibilization in cabinet politics through changing government coalitions within a presidential term Unilateral strategy together with greater integration between president and congress Stabilization of core democratic principles. Constitutional reforms, few long-term consequences

Executive instability Flexibilization of fixed-term rules Poor economic performance Emergence of political outsiders

Continued

Table C.1 Continued Country and Year of breakdown

Causal Factors (Dominant and Others)

Aftermath

Guatemala (1993)

Causes of self-coup: Minority government; social discontent; neoliberal austerity measures; confrontational style of president Causes of democratic restoration: Actions by the traditional interest groups (the CACIF and the army) Others: Broad civil society alliance; international pressure Institutional: Faction-ridden majority government; president- congress-court conflict; succession rules. Presidential: Strategic mistakes; confrontational leadership. Others: Deteriorating economic situation; scandal; violent street protest; international pressure Institutional: Loss of majority in congress; governing party disintegration. Events: Unlawful term extension; scandals involving president or his close collaborators Others: Popular mobilization against the president

Stabilization of core democratic principles, but: No changes in de facto power relations Delays in peace process No strengthening of representative institutions No gains in democratic accountability and rule of law

Paraguay (1999)

Peru

Venezuela (1993)

Institutional: Erosion of legitimacy and credibility of partyarchy model; centralized and rigid party structures. Presidential: Hostile president-party relations; presidential isolation Others: Scandals; challenges to established parties by new political actors and social protests; neoliberal policies and economic decline;

Flexibilization of presidential rules Electoral weakening of the Colorado Party Shift in the balance of civil-military relations Judicial process against Oviedo remains open

Stabilization of core democratic principles Constitutional reforms including the ban of immediate reelection, and later reforms involving decentralization and local democracy Emergence of another political outsider to the presidency No apparent flexibilization of the presidential fixed-term Collapse of the party system and rise of new political actors Judicial process against Perez remains open

216

Mariana Llanos and Leiv Marsteintredet

According to the authors of this book, legislative action, not popular uprisings, is the main force behind the collapse of elected administrations, whereas street challenges are often only the last factor in a chain of events leading to presidential breakdown. As Nolte argues, Paraguayan demonstrators did not force the political parties to act but rather supported them in their power struggle. In Ecuador, indigenous movements led mass protests and organized opposition against all three deposed presidents, but it is difficult to confirm a direct causal link between this activism and political instability (chapter 4, this volume). In Brazil, public protests were staged after Congress’s investigations began to produce evidence against Collor, and broadened as the evidence against him mounted (chapter 6, this volume). More research is needed about the role of protests and street challenges and their connection to congressional action during successful and unsuccessful challenges to presidents. The type of protest and whether protests express broader and diffuse discontent seem to count. Additionally, it is an open question how the extent to which street demonstrations are spontaneous or orchestrated from above, and the degree of involvement of opposition forces (a point often highlighted about the Argentine breakdowns, but also sometimes about the Venezuelan and Paraguayan cases, as our chapters show), makes a difference. It may also be the case that the “institutions versus streets” debate is misleading, as it ignores the many other (preceding) factors that influence a president’s fate. Three factors motivating Congress and/or the public to seek a president’s deposal stand out in the contributions to this book: the maturity of a democracy, economic performance, and intragovernment politics. It has already been highlighted that the level of maturity of democracy is an underlying factor that may be a necessary condition for a scandal to evolve into impeachment (Pérez-Liñán 2007, chs. 4–5). Furthermore, this factor seems to affect which type of scandal is likely to end a presidency. In hybrid regimes, it appears to be undemocratic behavior, such as a president’s attempt to extend their mandate unlawfully, that signals the end of the presidency. Thus, several of our authors pay attention to the link between presidential breakdown and the consolidation or final transition toward democracy. Indeed, this issue is an integral part of the analyses of the Paraguayan, Guatemalan, Dominican, and Peruvian cases. However, the intervention in these crises of other actors, such as international forces, to protect democracy has been practically absent in comparative studies of presidential breakdowns. Marsteintredet (this volume) shows how the United States and the Organization of the American States (OAS) pressed Balaguer to leave power after the fraudulent elections of 1994 in the Dominican Republic. International players also advised Serrano of the dire consequences of his attempted autogolpe in 1993 (chapter 9, this volume; INCEP 1993; Villagrán de León 1993), and the OAS was present, but not vital, in Peru at the time of Fujimori’s resignation, helping to secure a democratic outcome to the ongoing crisis there. In these cases, international

Conclusions

217

pressure may have helped remove a president on account of the undemocratic behavior that is typical of hybrid regimes. By contrast, international influence in other instances helped to thwart the challenge to a president, as in the case of President Bolaños in Nicaragua in 2004, when former presidents Ortega and Alemán established an “unholy alliance” to (unsuccessfully) orchestrate his ouster. Moreover, various countries in the region have often helped deposed presidents by offering them asylum, and thus helping to relax tense circumstances and potentially lowering the cost for a president to leave power when there is popular pressure for him or her to resign. All this suggests that international pressure may not only affect the outcome of such crises but may also have an impact on the way in which they are resolved (by preventing the repression of street demonstrators, for instance). In short, more research on the impact of international factors is necessary. Concerning structural variables, economic instability, economic development, and economic and socioeconomic performance contribute to presidential breakdown because they trigger protests and weaken presidential popularity. The cases of Hernán Siles Zuazo, Raúl Alfonsín, Carlos Andrés Pérez, and Jamil Mahuad come to mind: they all fell during periods of economic turmoil, and in part due to their own inability to deal with ongoing economic crises. Although the contributors to this book acknowledge contexts of economic malaise that characterize most of the presidential breakdowns analyzed, for the most part they do not refer to the economy as a problem per se, but rather to the way in which economic matters are handled by governments. This adds to the evidence provided by comparative work that is also referenced in our chapters about presidential limitations when dealing with Congress (whether presidential majorities can be built through coalitions or not), government responses to street protest (whether protests are subject to repression or not), and personal behavior (whether presidents or their governments are involved in corruption). In light of the above, we conclude that presidents appear to be at least partially responsible for their own demise. This is obvious in impeachment cases, but we believe that the claim also applies more generally. Considering that instability nowadays manifests itself at the level of government rather than regime, presidential leadership and actions within a given set of political and institutional constraints seem rather under-explored in the literature on presidential breakdowns (chapter 3, this volume). If presidents themselves are part of the problem (Llanos and Margheritis 2006, 81) or, conversely, are partly responsible for surviving in office, presidential strategies, or the style of presidential leadership, must be considered an independent variable that shapes performance in office (Pérez-Liñán 2007, 146).1 There are several examples of how presidential strategy can influence the fate of Latin American presidents in the chapters in this book. Below we present some of these different president-centered independent variables. A striking variable is the relationship between the president and the vice president, which can have unpredictable consequences for presidential

218

Mariana Llanos and Leiv Marsteintredet

stability, as vice presidents are often the first in the line of succession. Linz (1990, 65) argued that it was particularly risky for regime stability if the winning presidential and vice presidential candidates came from different parties, or if the vice president was nominated mainly to balance the ticket, without due consideration for governability and stability. If we take the twelve relevant presidential breakdowns considered here2 (there were five vice presidents belonging to the same party as the president, four with no party affiliation, and three belonging to the other party (or one of the parties) of the government coalition), we find that none of the different political constellations have been exempt from difficulties (although not all deposed presidents have had problems with their vice presidents). The most extreme case in terms of its consequences is that of the Cubas and Argaña presidential duo in Paraguay. Both were Colorados but each belonged to a different, antagonistic, faction. Argaña’s assassination triggered the fall of President Cubas, who, it was believed, was involved in the crime (chapter 8, this volume). Although less dramatic, there are other cases that are significant. In Argentina, the resignation of Vice President álvarez after a conflict with President de la Rúa not only seriously damaged the government’s credibility but also left the office vacant, opening the opportunity for the intervention of (the Peronist dominated) Congress to determine the line of succession, as happened a year later (chapter 3, this volume). In Bolivia, Vice President Paz Zamora had his eyes on the presidency when, with support from Bánzer, he demanded that his party withdraw from the governing coalition, while at the same time he remained in office. Eventually, both president and vice president left together, after calling for early elections. In Ecuador, independent vice president Palacio distanced himself early from President Gutiérrez, and clearly stated that he was ready to take over in case any of the early impeachment attempts in 2004 were successful. Even in the case of President Balaguer in the Dominican Republic, Vice President Morales Troncoso (from the same party) made secret overtures to the U.S. State Department and offered to become president if the United States decided to oust Balaguer (Skol 2009). In short, even though we have found no clear evidence that a specific type of ticket is particularly vulnerable, various cases of presidential breakdown suggest that confrontation within the presidential ticket may shorten a president’s term. On the one hand, since they are next in the line to take over the presidency, vice presidents have little incentive to support the president in case of a government crisis. If the vice president remains too close to the president, he or she runs the risk of falling together with the president, as shown by the fate of the vice presidents of presidents Serrano and Bucaram, Gustavo Espina and Rosalía Arteaga, respectively. On the other hand, by distancing themselves from the president, vice presidents stand a higher chance of taking over the presidency if the president falls, and this behavior can contribute to destabilize the government as well. Third wave presidents and presidential candidates would be wise to bear all this in mind when picking their companion on the presidential ticket.

Conclusions

219

Intra-government problems other than the above-mentioned ones can also have consequences for presidential stability. As presidents have the power to freely appoint and dismiss ministers, presidential cabinets are a reserve of the chief executive. The profile of cabinet members can vary according to the presidential calculus, that is, according to the strategy that presidents adopt to implement their policies (Amorim Neto 2006). We know that presidents who opt to govern alone, refraining from building a “shield” in Congress (PérezLiñán 2007), or by adopting “imperial” (Cox and Morgernstern 2002, 460) or “delegative” (O’Donnell 1994) strategies that bypass Congress, may be tempting fate. On the other hand, the use of unilateral resources in conjunction with a high degree of integration between the legislative and executive branches, mainly achieved through the formation of cabinet coalitions, has proven a successful strategy, as the Brazilian case illustrates (chapter 6, this volume). If coalition building is a source of stability and success, coalition breakdowns and splits in the ruling party and administration may have the opposite effect. They weaken a president’s position in Congress, as well as his popularity and his credibility in various sectors—including, but not exclusive to, the financial sector—by exposing the internal disputes and difficulties that the president has handling them. 3 Mejía Acosta and Polga-Hecimovich (this volume) highlight how coalition rupture caused several presidents to fall in Ecuador; and Serrano’s autogolpe was triggered, among other factors, by the split in the president’s congressional coalition with the National Centre Union (Unión del Centro Nacional, UCN) and the Guatemalan Christian Democratic Party (Partido de Democracia Cristiana Guatemalteca, DCG). However, this is not always the case. In Brazil, it is routine for governments to preside over successive coalitions without breaking down. Again, what seems to count is that the president keeps an eye on how and what coalitions are formed, and the background incentives provided by national institutions (chapter 4, this volume). The next section provides a statistical exercise (albeit provisional and imperfect) to test whether internal conflicts in the administration may be a contributing factor to presidential breakdowns.

President-Centered Explanations On social change and revolutions, Jon Elster (1989, 164) wrote: “The ability of a regime to resist external pressure depends on its internal health.” But most explanations of presidential breakdown have centered on a government’s relationship with its environment, be it Congress, social movements, or the “streets.” Several chapters in this book have highlighted more president-centered explanations, which focus on style of leadership and internal conflicts in the administration or within the ruling coalition. The following three models in table C.2 (below) attempt to capture the importance of president-centered factors, while controlling for the factors commonly

Table C.2

Models of presidential breakdown

Growth (lagged 1 year) Inflation (natural logarithm) President’s share of seats in lower or single chamber Level of democracy (lagged 1 year) Scandals (dummy) Internal splits Deadlocks Street Protests Constant

Model 1: Presidential breakdown (N = 389)

Model 2: Presidential breakdown (N = 324)

Model 3: Presidential challenges (N = 389)

–.084 (.039) **

–079 (.039)**

–.111*** (.029)

.013 (.100)

.011 (.108)

–.044 (.109)

–.032 (.013) **

-

–.013 (.011)

–.109 (.074)

–.119 (.082)

–.013 (.011)

.937 (.775) .782 (.343)** –1.362 (.886)

.902 (.719) .783 (.367)** 6.205 (1.979)*** 3.149 (2.211) –3.269 (.833)***

1.175 (.535)** .746 (.316) ** –1.559 (.683)**

Notes: Level of significance: *<.1; **<.05; ***<.01. Growth and Inflation based on data from the World Bank (WDI 2008). President’s share of seats based on (Nohlen 2005) and updated with official sources for the most recent elections, measured as a percentage. Level of democracy is based on Polity’s combined authoritarian and democracy index (Jaggers and Gurr 1995). Scandals, internal splits, deadlocks, and street protests are all from the LAWR, coded by Marsteintredet and compiled in a dataset on political conflicts in Latin America for his doctoral thesis on presidential breakdowns in Latin America. Internal splits is a count variable counting for the number of ministerial resignations in a single year. Deadlocks and street protests are level variables, in which the number of deadlocks or street protests (including strikes) directed at the national government is divided by the number of political news reports on a country in a given year as reported in the LAWR. These two latter variables exclude challenges (attempts to remove the president). The coding is inspired by Jones (1995).

10.1057/9780230105812 - Presidential Breakdowns in Latin America, Edited by Mariana Llanos and Leiv Marsteintredet

Conclusions

221

highlighted in the literature on presidential breakdowns, namely, economic, institutional, regime, and contention variables. To do this, we have used the Latin American Weekly Report (LAWR) newsletter, and created a count variable for each time there is an internal conflict in an administration resulting in the resignation of at least one government minister.4 Our assumption is that the more internal splits or conflicts in the administration there are, the more the administration is weakened. The variable varies between 0 and 4, and the extreme value 4 corresponds to Bolivia in 1983, a year without a breakdown. It is important to state here that these are only resignations of ministers that are reported by LAWR as clearly being caused by an internal government conflict. Therefore, we exclude from the count regular reshuffles in the cabinet, ministers who have been forced to resign due to censure by Congress, a common procedure in some periods and places, such as Ecuador (Mejía Acosta 2006), and resignations due to reasons that are presumably not related to internal conflicts (as when ministers, and vice presidents in particular, resign to run for another public office position). The unit for the models are country-years, and we use Rare Events logistic regression since there are very few positive instances compared to negative instances of the dependent variables (King and Zeng 2001b; 2001a). 5 Due to the few cases of breakdown and challenges, the results of our statistical analyses should be interpreted with great caution. Standard errors (in brackets) are robust using countries as clusters (we do not assume that the observations are independent within countries). We only include full semi-democratic and democratic years as coded in the MBP Index for Democracy (Mainwaring, Brinks, and Perez-Liñán 2001) in the period 1980–2005. In all our models, we have excluded the variable of challenges to presidents, or presidential crisis, and opted to run a separate model on challenges (Model 3 below).6 The reason is that, as Model 3 demonstrates a challenge to the president, be it successful or not, either from the streets, Congress, or both, is driven by many of the same factors that provoke the fall of presidents, so it is more appropriate to regard them as a dependent variable. Model 1 uses the presidential breakdowns identified in table C.1 as dependent variable, and shows that when controlling for growth (lagged one year), inflation (natural logarithm), a president’s share of seats in Congress upon election, the level of democracy (lagged one year), and scandals, then internal splits in an administration is significant at 5 percent.7 As expected, economic growth (lagged) is negative and significant at 5 percent, the president’s share of seats is also negative and significant at 5 percent,8 whereas the level of democracy is negative but, as expected, not significant. Furthermore, the scandals variable, although positive, does not reach 10 percent in terms of its level of significance. The latter finding is not surprising since only a few presidents fell as a result of scandals implicating them personally.9 In Model 2 we replaced the president’s share of seats with a direct measure of deadlocks, and also included a direct measure of level of contention in the polity. A deadlock is defined as an open conflict between Congress

222

Mariana Llanos and Leiv Marsteintredet

and the executive, excluding all challenges or attempts to remove a president (see Jones 1995; Negretto 2006). Street protests include any protests, demonstrations, or strikes that present the government with demands, but exclude those calling for the removal of the president. Deadlocks and street protests directed at the government are both measured as level variables to control for the varying coverage in LAWR of different countries in the region: in other words, we divide the number of deadlocks or street protests in a year by the number of political news presented in LAWR about a country in a year.10 In Model 2, all the variables found in Model 1 do not change significantly. Interestingly, Model 2 is partly related to the debate about whether it is institutions or the “street” that matters the most. We note that our deadlock variable is highly significant and positive, whereas the general level of contention in the streets is positive but its level of significance is only 15.4 percent. Institutional factors such as a president’s position in Congress and the level of interinstitutional conflicts seem to play an important role in the breakdown of presidents, and potentially more so than the level of contention in the streets. Together with the information in Model 1 on the president’s share of seats in Congress, our findings again call attention to the factors that Linz associated with the perils of presidentialism but which are now associated with presidential rather than regime breakdown. The relative lack of significance for the street contention variable is perhaps not that surprising, as this has been noted previously (Hochstetler and Edwards 2009). Although street challenges or antigovernment demonstrations clearly increase the risk of presidential breakdown (chapter 2, this volume; Hochstetler 2006), it is more difficult to find a clear association between the general level of contention, evident in the level of antigovernment demonstrations, strikes, piquetes, and other forms of protest, and a presidential breakdown. Finally, in Model 3 we use challenges to presidents from either Congress or the streets as a dichotomous dependent variable, and again find that internal splits are significant (5 percent), and that scandals are also significant at 5 percent. From this, we can conclude that, unsurprisingly, internal splits, scandals, and (negative) growth affect the likelihood of a challenge to presidents. The variable president’s share of seats in Congress when elected, on the other hand, does not seem to explain challenges to a president, but is important to explain challenges that successfully end in breakdowns. In short, internal splits in presidential administrations seem to have both a direct and an indirect effect on presidential survival. They contribute directly when a split in the administration leads to a president’s resignation, either because former party supporters abandon the administration and leave the president with few other options other than to resign; or because the opposition, now strengthened by former members of the governing coalition, can get enough votes in Congress to force the president to leave power early. An internal split has an indirect impact on presidential survival since the opposition may become bolder in its attempts to remove a president when the latter is clearly

Conclusions

223

weakened by the removal or resignation of ministers, or when scandals come out into the open as a result of internal quarrels in the administration. These models thus support many of the findings in the comparative literature on presidential breakdowns, but they also call attention to an omitted variable: presidential leadership and the internal conflicts within presidential administrations. It appears that the political management of crises is relevant. Our internal splits variable is only a crude measure of presidential leadership and the internal erosion of an administration, and we consider that it is worth paying more attention to the internal politics of governments in presidential regimes. However, our analysis also suggests that despite the protection that presidents gain from having fixed electoral terms, they should nonetheless—like their parliamentary counterparts—fear for the survival of their governments if there are internal conflicts and particularly when such conflicts break up governing coalitions.

Presidential Breakdowns and their Aftermath Turning now to the aftermath of presidential breakdown, there are only a few systematic studies about the effects of this phenomenon. Hochstetler and Samuels (2008) find that there are few measurable consequences, although rapid regime re-equilibration is a general outcome. The chapters presented in this volume suggest that presidential breakdowns have implications on three areas: presidentialism, democracy, and policymaking; and that the implications are not uniform across all cases.

Implications for Presidentialism Marsteintredet and Berntzen (2008) take issue with Valenzuela’s (2004) analysis and argue that presidential breakdowns in Latin America have changed how presidentialism works during times of political crisis, mitigating some of the flaws of presidentialism identified by Juan Linz.11 For them, presidential breakdowns demonstrate that presidential regimes can operate in a flexible manner during crises in part by inventing new para-constitutional practices, and, further, that in some instances these practices seem inspired by the flexible solutions that parliamentary regimes find to cope with crisis (namely, early elections, the vote of no-confidence, and congressional elections of the presidential successor). Mejia Acosta’s and Polga-Hecimovich’s chapter support these claims for the case of Ecuador. Similarly, Mustapic (chapter 1, this volume; 2005) suggests that we should reconsider the role of the fixed-term presidential mandate, since it is the impossibility of dissolving Congress (and the dual legitimacy principle) rather than the former that truly distinguishes presidentialism from other systems. Presidential resignations show that Congress is a key to government stability: the capacity to give or withdraw support from presidents puts Congress in a position analogous to that of parliamentary legislatures with their capacity to censure executives.

224

Mariana Llanos and Leiv Marsteintredet

However, Cheibub Figueiredo (chapter 6, this volume) warns against the inclination to generalize about Latin American presidentialism. In her view, the Collor impeachment did not signal a crisis of presidential democracy because, as Marsteintredet and Berntzen (2008, 91) have also maintained, impeachment is a constitutional mechanism—albeit exceptional—that is available for presidential regimes to resolve conflicts between the executive and the legislature, and is one that seems to be working. It is also important to add that there are only three countries in Latin America that have undergone repeated presidential breakdowns. Thus, it is difficult to make region-wide generalizations. In light of the above, we consider it important to retain the picture of regional diversity that is highlighted in the introduction and in the chapter by Cheibub Figueiredo. If impeachments and early resignations were exceptional in some countries, in others they may be here to stay, particularly where presidencies have broken down several times. In such contexts, breakdowns are part of the political memory of the citizens and presidential candidates alike, and an option that remains open regardless of constitutional fixed-term rules. A further element that is typical of presidential systems that this wave of presidential breakdowns has brought to the fore is the role of the vice president and, more generally, of succession rules. Latin American constitutional rules vary considerably regarding the workings of the vice presidency: there are constitutions that do not provide for a vice presidency (Chile and Mexico), and some that provide for more than one (Costa Rica and Peru). The role that vice presidents play when there is no president in office also varies, as there are different succession rules to provide for the termination of a presidential term, including: the appointment of a successor by Congress; calling elections; the adoption of a mixed system; and following a preestablished line of succession (see chapter 1, this volume). In crisis situations, these different dispositions may well shape the actions of legal successors and/ or legislators, thus giving rise to varying political strategies, including challenges to the president either by vice presidents or Congress, as various chapters here show. On the other hand, the increased likelihood of presidential breakdown may lead to changes in the way presidential tickets are built or even promote a change in constitutional succession rules in acknowledgment of previous presidential successions. Our study suggests that we should pay more attention to what happens with succession rules in the future.

Implications for Accountability and Democracy How is accountability between voters, Congress, and the president affected by presidential breakdowns? Pérez-Liñán (2005) observed that, over time, Congress prevails more often than presidents when there are interinstitutional conflicts, which alters our perception of power relations between the two institutions. Marsteintredet (2008) concludes that presidential breakdowns

Conclusions

225

should increase the level of horizontal and vertical accountability, and that they subject presidents to a dual accountability from the streets and Congress. But concerns have also been voiced that this is a failed rather than a virtuous kind of accountability (Schmitter 2004), and that presidential breakdown suggests the prevalence of an “intermittent model” of checks and balances that responds to popular pressure and differs greatly from the institutionalized forms of accountability that characterize well established democracies (Pérez-Liñán 2007, 207). Against this view, the evidence from Brazil (chapter 6, this volume) suggests that Congress functioned as an effective oversight institution, not only during the impeachment of Collor but on many occasions thereafter. Further, as is stressed in other chapters here, the relatively more powerful role of Congress as compared with popular pressure makes it hard to conclude that it is only the Brazilian case that fails to comply with the “intermittent checks and balances” model. In fact, impeachments—as well as votes of confidence in parliamentary systems—are instruments of accountability of last resort: their use is intended to be sporadic and they are therefore complemented by a wide range of other oversight mechanisms that are more appropriate in normal times. This is not to say that horizontal checks on the executive are not “politicized” (Pérez-Liñán 2007, 207): electoral, partisan, and even personal motivations clearly lead legislators to pursue actions against a (discredited) executive. This seemingly optimistic view of accountability is complemented by the immediate and positive democratic effect that presidential breakdowns had in countries with regimes that combine elements of democracy and authoritarianism: Paraguay, Peru, Dominican Republic, and Guatemala were all hybrid regimes when presidential breakdowns occurred there. Several studies published in the early aftermath of presidential breakdowns highlighted the positive effects that the phenomenon had on the consolidation of democracy, including the emergence of a stronger citizenry (Abente-Brun 1999; Hartlyn 1998; Schamis 2002; Villagrán de León 1993). However, the relevant chapters in this volume suggest that we should be more cautious about drawing such conclusions. National and international pressures led to the ouster of presidents on account of their undemocratic behavior and for disrespecting minimal democratic procedures, but with the benefit of hindsight, it is clear that there are many continuities in the way that politics is done, and there has been little progress with strengthening of the rule of law or improving the quality of democracy. In Paraguay, for instance, a high percentage of the population (between 30 and 44 percent) is still open to authoritarian solutions (chapter 8, this volume). It is difficult to arrive at a general conclusion about such a diversity of cases, and we can say that the most important implication for the democratic regimes of Latin America remains separating government instability from regime instability, which, Linz argued, were intrinsically linked under presidential regimes (Linz 1978, 74). That, by itself, is no small feat when

226

Mariana Llanos and Leiv Marsteintredet

the absence of stable democracy was one of the more durable features of the region for decades.

Policy Implications As regards the policy implications of presidential breakdown, Valenzuela (2004) suggests (and the chapters here support) that presidents usually fall as a result of their inability to deal successfully with economic crisis or because they implement unpopular neoliberal reforms. Thus, presidential breakdown is linked to policy switches, be they abrupt implementations of economic reform by decree (Stokes 2001) or more gradual reforms. The first two presidential breakdowns examined here were those of Siles Zuazo in Bolivia and Alfonsín in Argentina. Both occurred after heterodox economic policies promoted by the presidents failed to put an end to economic crisis. Their successors, Paz Estenssoro and Carlos Menem, respectively, implemented broad neoliberal reforms by decree. In his memoirs, Alfonsín (2004, 145) pointed out how his fall facilitated Menem’s radical policy shift. We agree that a government crisis such as a presidential breakdown may facilitate such dramatic policy shifts. Consider the contrast between the introduction of neoliberal reforms in Argentina and Venezuela in 1989: because of the crisis, the Radicals agreed to help Menem pass his reforms in Congress until the new Congress had been installed in December of that year, whereas Pérez’s gran viraje that same year was met with widespread protests—el caracazo.12 Furthermore, there is the breakdown of presidents de la Rúa in Argentina, Carlos Andrés Pérez in Venezuela, Sánchez de Lozada in Bolivia, and Mahuad in Ecuador, which signalled a gradual demise of the neoliberal model. Their successors tried but failed to reverse some neoliberal reforms to satisfy both popular demands and international actors.13 After this short interregnum, Venezuela, Argentina, Bolivia, and Ecuador elected presidents who established a clearer break with the previous economic model.14

Concluding Remarks In this chapter we have emphasized president- or government-centered explanations of presidential breakdown, and we have examined the implications of presidential breakdown for presidentialism, accountability and democracy, and policymaking. As regards explanations of presidential breakdown, it seems clear that government instability is no longer intrinsically linked to democratic regime instability, as Linz argued in 1978. Our focus, then, has been on relations between the president and Congress or popular mobilization, but more specifically on president, or government-centered explanations of presidential breakdown. The chapters in this book and our statistical models strengthen the hypothesis that splits in the governing coalition seem to play an independent and important role in presidential breakdowns both directly and indirectly.

Conclusions

227

As regards the implications for presidentialism, democracy, and policymaking, we conclude, regarding the first, that the role of the vice president needs further attention; regarding the second, that presidential breakdown seems to curtail authoritarian presidents in hybrid regimes, and is preceded in more consolidated democracies by congressional control over presidents who misbehave; and regarding the third, that breakdowns are associated with broad and often dramatic policy shifts. Yet, due to the variety among the cases of presidential breakdowns, and even more so in the region in general, one should be cautious about making general, or overly optimistic, statements about the implications of presidential breakdowns. Perhaps the most important—and encouraging—observation is that, as recent experiences of presidential breakdown show, governments can break down without threatening the survival of democracy. Insofar as this is true, Latin America’s democracies seem to have found a “democratic” way to solve political crises: they have resorted to the early removal from power of elected presidents rather than to military intervention.

Notes * The authors wish to thank Michael E. Álvarez for reading and commenting both this chapter and the introductory chapter. 1. See also Weyland (1993; 2006) for analyses of how neo-populist leadership may explain the falls of presidents Collor de Melo and Fujimori in Brazil and Peru, respectively. 2. We exclude Carlos Mesa because he was a vice president, and Carlos Andrés Pérez, because when he was impeached Venezuela’s constitution did not provide for a vice president (Article 187 states that congress must appoint a caretaker for the remainder of the presidential term in the absence of the president). 3. It is also likely that internal conflicts, or conflicts with Congress, increase the likelihood of scandals implicating members of the administration coming to public light. 4. In other words, our count of internal splits is not affected by whether one, two, or more ministers left at the same time or not. In any case, the event counts as one internal split in the government. 5. The command in Stata for rare event logistic regression is relogit. This type of regression analysis has become quite common in the study of presidential breakdowns (see Negretto 2006, for instance). All calculations are made in Stata/ IC 10.1. 6. To include this variable one should either use selection models (Hochstetler and Edwards 2009) or conceptualize challenges and breakdowns as two different values within the overarching concept of presidential crisis. In the latter case, ordered probit models might be considered (chapter 2, this volume). 7. We also ran separate models using internal splits as a dummy variable (whether present or not in an administration during a year) and found that this did not significantly alter the results.

228

Mariana Llanos and Leiv Marsteintredet

8. We ran separate models using the effective number of parties instead of the president’s share of seats in Congress. This variable remained significant at between .05 and .1, depending on the model. 9. See also Morgenstern, Negri, and Pérez-Liñán (2008). 10. Thus, both variables vary between 0 and 1. We also exclude any country-year with fewer than six political news reports, which explains the drop from 389 to 324 observations in this model. 11. See also Carey (2005). 12. As Corrales (1997) convincingly argues, however, it was not the prior economic crisis that explained the different levels of success in the implementation of neoliberal reforms in these two countries. 13. The successors were Eduardo Duhalde in Argentina, Rafael Caldera in Venezuela, Carlos Mesa in Bolivia, and Lucio Gutiérrez in Ecuador. 14. We refer to Nestor Kirchner in Argentina, Hugo Chávez in Venezuela, Evo Morales in Bolivia, and Rafael Correa in Ecuador.

Epilogue Mariana Llanos and Leiv Marsteintredet

The Breakdown of Zelaya’s Presidency: Honduras in Comparative Perspective On June 28, 2009, when this book had just gone into production, the world was astonished by the breaking news that Honduran President Manuel Zelaya had been deposed and sent into exile. At about 5 am that morning, the military burst into the Honduran presidential palace, seized President Zelaya in his pajamas, bundled him onto a plane and flew him to Costa Rica. According to official communications from the Honduran Supreme Court,1 the military were acting on orders of the court for contentious administrative proceedings (Juzgado de Letras de lo Contencioso Administrativo, JLCA). Later on the same day the military’s actions received the legislature’s consent. The legislature approved what may be a forged letter of resignation (Zelaya publicly denied having signed any such letter), and the line of succession provided for in the constitution was then activated. Because Zelaya’s original vice president, Elvin Santos, had already resigned to run for the presidency in the elections scheduled for November 28, 2009, the president of Congress, Roberto Micheletti, was sworn in as president to serve out the remainder of Zelaya’s term until January 2010. The day Zelaya was deposed, there was meant to be a poll about whether to hold a fourth “election” (the cuarta urna, or fourth urn) concurrently with the upcoming general election. The question the Honduran citizenry would have to pronounce itself about with the cuarta urna would be whether to elect a Constituent Assembly or not. Those supporting the initiative claimed that the aim was to reform the constitution to deepen participatory democracy, but Zelaya’s opponents argued that the constitutional reform was meant to facilitate the president’s reelection. At the time of writing (August 2009), no solution has been found for this presidential crisis. Micheletti is still president, and Zelaya, who has received the backing of all the countries in the hemisphere, is still insisting that he should return to office. Our intention here is to present a short account of the events so far and compare them with the other cases analyzed in this book.

230

Epilogue

The Background to Zelaya’s Ouster2 Zelaya, a businessman and a member of the establishment, won the presidential election in November 2005 as candidate for the Liberal Party (Partido Liberal, PL), a contest that has been described as the dirtiest race in the living memory of Hondurans (NotiCen 2005). He defeated Porfirio Lobo, the candidate for the National Party (Partido Nacional, PN). Zelaya’s party did not win an outright majority in Congress, obtaining 62 out of 128 seats (Taylor-Robinson 2007), but as a result of an alliance with two smaller parties, the Liberals won the important position of the presidency of Congress. Zelaya was perceived as being to the left within the Liberal party, and his policy choices caused friction with his vice president, his own party, as well as the opposition in Congress and the business sector. In August 2008, his relations with the political community—including his own party—began to take a serious turn for the worse when he decided to join the leftist Bolivarian Alternative for the Americas (Alternativa Bolivariana para las Américas, ALBA), the alliance led by Venezuela’s Hugo Chávez. In late 2007, prior to taking this step, Zelaya had led Honduras to join Petrocaribe in order to alleviate the costs of increasing oil prices, having initially failed to circumvent the control of multinational oil companies over the fuel supply to Honduras. Despite early criticism from all sides, Congress ratified the country’s agreement with ALBA with the votes of PL and the small Democratic Unification (Unificación Democrática, UD) party.3 Although the military action of June 28, 2009 took the world by surprise, the fact was that Honduras had been in the midst of an institutional crisis for months. Above all, it was Zelaya’s intention to reform the constitution that alienated the country’s main political parties and institutions. In November 2008, Zelaya launched his reform plan (El Heraldo, November 21, 2008). On March 24, 2009 he called for a referendum to be held in June, to consult voters on whether they wanted a binding referendum held alongside the general election of November 2009 to convene a Constituent Assembly (NotiCen 2009). The President enacted a decree (PCM-005-2009) calling for the referendum, but the Public Ministry immediately declared the decree unconstitutional, a decision confirmed by the courts in late May. Despite this, the President went ahead, issuing a new decree (PCM-020-2009) on May 26. Although it was essentially the same as the previous decree, this one called the referendum a “national survey” (encuesta nacional). The decree was published only on June 25, 2009, which did not give the opposition and the judicial system much time to react (La Gaceta 31, 945, June 25, 2009: 3–4). The PN and the faction of the ruling PL that was loyal to the head of Congress Roberto Micheletti accused Zelaya of using the referendum to perpetuate himself in power and of trampling on democracy. The opposition argued against Zelaya on the basis of constitutional norms, which state that there can be no reelection and that

Epilogue

231

the articles stipulating this cannot be reformed.4 The day before Zelaya’s decree was published, Congress preemptively passed a new law declaring that referenda cannot be held 180 days before or after regularly scheduled elections (Decree 135-2009 of June 24, 2009). Zelaya’s positions also made him very unpopular among the military establishment. On June 24, the president fired General Vásquez Velásquez, the commander of the armed forces, because of his refusal to provide logistical support for the referendum. This led to the resignation of defense minister Edmundo Orellana Mercado, who objected to the president’s action, as well as the resignations (in solidarity) of the commanders in chief of the navy, army, and air force. Finally, on June 25, the Supreme Court declared that General Vásquez Velásquez’s dismissal was illegal and called for his reinstatement. In a series of judicial rulings, the public ministry and the judiciary also declared that the second presidential decree calling for a “national survey” was unconstitutional, and the human rights ombudsman, Ramón Custodio, called for an investigation into the spending of extra-budgetary funds to pay for television and radio advertisements promoting the referendum (LAWR 11-2006). According to Supreme Court documents on this matter, there was an arrest order against Zelaya in place as early as June 26, who was charged with treason, among other crimes. The military arrested Zelaya on the basis of this order, but the decision to send Zelaya out of the country on the morning of June 28 was apparently the army’s. 5 On the day that Zelaya was sent to Costa Rica, Congress confirmed Zelaya’s dismissal by 122 votes,6 and nominated the president of the Congress, Robert Micheletti, as his successor. In short, the events preceding the presidential breakdown show that a broad variety of institutions and politicians acted in legally dubious ways to get rid of Zelaya. Likewise, they demonstrate that the institutional crisis was precipitated by the fact that Zelaya openly defied all the institutions arrayed against him.

Zelaya’s Fall: A Presidential or a Democratic Breakdown? There are good reasons to argue that this is a case of democratic breakdown. It is important to note that, before the military acted against Zelaya, Congress had not used the normal constitutional means to depose the president, namely impeachment. Congress acted ex post on the basis of the charges against Zelaya and in accordance with a presidential letter of resignation of highly questionable origins. Although the Honduran case is probably starker than the other cases of presidential breakdown studied in this volume, both in terms of the deployment of military force and because of

232

Epilogue

the use of “innovative” mechanisms to oust the president, neither of these features is new. As this book shows, such means have been used in other contexts—albeit with local variations. As regards the use of military power, in some cases the role of the armed forces was simply to persuade a president that “the game is over” (Bucaram in Ecuador), or safely to escort a head of state from the presidential palace (Sánchez de Lozada in Bolivia). In other instances, however, military intervention played a central role in ousting a president. In Ecuador, the coupmakers that toppled President Mahuad originally sought to end democracy by installing a civil-military junta, the so-called “government of national salvation” (Zamosc 2007, 13). In contrast to this example and others like it—such as the flagrant coup attempt against Hugo Chávez in Venezuela in 2002—in the case of Honduras the military did not seek to take power. Not only did the armed forces immediately let Congress deal with the new political situation “sans Zelaya,” but they also apparently acted to execute a judicial order to arrest the president. As regards the institutional mechanisms to remove the president, the ouster of Zelaya follows a pattern of a creative reading of the constitution to facilitate actions to remove presidents from office. As the case studies in this volume show, only three out of fourteen presidents have been removed in accordance with constitutional rules of impeachment (Collor de Melo in Brazil, Carlos Andrés Pérez in Venezuela, and Cubas in Paraguay). In every other case, “alternative” rules were followed. In the case of Bucaram in Ecuador in 1997, for instance, Congress declared the president to be “mentally incapacitated”; in several other cases Congress declared that the president had “abandoned” office (Marsteintredet and Berntzen 2008). The Honduran constitution is quite vague about how a president can be removed from power. The article referring to the right of the Congress to decide whether there is proper cause for impeachment (Article 205, 15) ceased to exist in 2002 during the presidency of Maduro. The constitution also stipulates that the Supreme Court must resolve (dirimir) interinstitutional conflicts between Congress and the presidency (Article 316, 2). To further complicate matters, the military has a special constitutionally mandated role to ensure presidential succession (alternabilidad) and, under the order of the Supreme Electoral Tribunal, to guarantee the smooth running of elections (Article 272).7 Clearly, Zelaya’s opponents could make the legal and constitutional case that ousting the president did not require his impeachment. But the abrupt seizure and forced exile of the president by the military undermined the opposition’s strategy. Previous cases of presidential breakdown studied here suggest that if Congress had voted to remove Zelaya before rather than after his arrest (and exile), or if the military had arrested Zelaya without sending him to Costa Rica in his pajamas, things could have been different in Honduras today.8 Our definition of presidential breakdown focuses on the early departure of an elected president without the collapse of democratic rule. Since the

Epilogue

233

Honduran Supreme Court and Congress survived the ouster of Zelaya, power has remained in civilian hands, at least in de jure terms, and since new elections are still scheduled to take place at the end of November, we are inclined to argue that the ouster of Zelaya is a case of presidential breakdown rather than an example of full democratic breakdown. But the case has not come to its logical conclusion as yet, and since we cannot guess the future course of events, we refrain from any final categorization of this case in this epilogue.

The Causes of Zelaya’s Fall in Comparative Perspective Honduras has been a stable although “defective” third-wave democracy based on a century-old two-party system, with strong oligarchic and clientelist elements, so that a select closed group controls political power (Peetz 2009). The country did not suffer the devastating civil wars that beset its neighbors, and so its political society is not as polarized as that of other Central American countries. The presidency has alternated between the PL and the PN, and most presidents have enjoyed majority support in Congress. These two parties have been at the center-right of the political spectrum, a pattern broken by Zelaya in 2006 when he shifted to the left of Micheletti, although he was then still far from taking the leftward turn he opted for later (Alcántara Saez 2009b). Political conflicts take place within the top ranks of the political class, and close observers of Honduran politics say that the attitude of the citizenry to these struggles has been apathetic at best and undemocratic at worst.9 In this context, it is perhaps unsurprising that Zelaya’s ouster was the product of elite action and conflict. In marked contrast to most cases of presidential breakdown in South America, the Honduran presidential ouster took place in the absence of social protest (Hochstetler 2006), or without the participation of “the street.” There was no presidential scandal or economic crisis with sufficient dimensions to generate the public uproar that such conditions created in Brazil and Argentina. Protests and demonstrations for both sides only occurred after Zelaya was no longer in the country. Because of these features, the Honduran case is more similar to the two Central American and Caribbean cases of presidential breakdown: Guatemala in 1993 and the Dominican Republic in 1994. In neither of these instances did pressure from below play a crucial role in the breakdown of the presidency. If we deepen our analysis of the elite conflict that preceded Zelaya’s deposal, it becomes apparent that he was an isolated president. He was running an administration that did not have majority support in Congress, and at the moment of his fall he had lost even the support of his party. In January of 2009, Zelaya lost a battle for the control of the Supreme Court. He quarreled with Congress over the appointment of the justices of the Supreme Court, a prerogative Congress has used to divide the court between the two

234

Epilogue

major parties.10 Zelaya alienated his own party, opposing the presidential ticket of his vice president, Elvin Santos, and supporting Micheletti instead.11 Not two months into his presidency, Zelaya and Santos had aired their differences over a number of issues quite publicly, and the administration had experienced serious rifts and been beset by ministerial resignations (NotiCen 2006). Zelaya also provoked the military, first by insisting that the armed forces carry out his national survey, and then by firing the commander in chief of the forces, General Vásquez Velásquez, when the latter had refused to execute the president’s orders. As various chapters and our conclusions in this volume show, an isolated president is a recurrent feature of presidential breakdowns, for which presidents are partly responsible. Serious differences between presidents and their vice presidents have proved to augur ill for presidential stability; further, when presidents are at odds with both Congress and the judiciary it is unlikely that the president will prevail, as illustrated by the Paraguayan case. Indeed, the fall of Cubas and Zelaya provide the clearest illustration of presidential conflict with these two government institutions: in both cases, judicial and legislative decisions reinforced each other to the detriment of the power of the presidency. In Paraguay, the decision by the Supreme Court to declare unconstitutional the presidential decree freeing General Oviedo was backed by a congressional resolution voted by a substantial majority in a joint session of both houses (see chapter 8 in this volume). Zelaya fought several battles with Congress and the Supreme Court in his drive for constitutional reform, and he opted to carry on alone even after the courts twice rejected his referendum decrees. As in Paraguay, President Zelaya was abandoned by his own party, which eventually voted in favor of his ouster.12

On the Futile International Attempts to Restore Zelaya The televised images of Zelaya in his pajamas in Costa Rica suggested that a traditional coup had occurred. The Honduran case is not the first involving military action, but it is certainly unique when one considers the universal international rejection of the legitimacy of Zelaya’s ouster. From Chávez to Obama, international reactions condemned the golpe unanimously. President Obama stated that the overthrow of Zelaya set a “terrible precedent” for the region (LAWR 02/07-2009). The UN General Assembly immediately adopted a resolution refusing to recognize the new Micheletti government and calling for the “immediate and unconditional” restitution of Zelaya. The General Assembly of the Organization of American States (OAS) delivered an ultimatum to Honduras, calling for the restoration of Zelaya within seventy-two hours on pain of expulsion from the Organization in accordance with the provisions of the Democratic Charter (Honduras responded

Epilogue

235

by withdrawing from the OAS). The ouster also had severe economic consequences for Honduras as several international institutions halted aid programs and loans to the country. The unanimous international response was probably facilitated because international attention was already focused on the region after the earlier attempts to remove Guatemalan President Colom from power for his alleged role in the murder of the lawyer Roberto Rosenberg. Why, despite unanimous condemnation of Zelaya’s ouster, has the international community failed to restore the president to power? The international reaction has differed in this case from the reaction of previous ousters in the region because the international community intervened to restore an ousted president in order to protect democracy. In other cases, international intervention to protect democracy was meant to help depose an autocratic president (as in the Dominican Republic), or to escort the president out of the country following social demands for presidential removal. The international response in favor of Zelaya also came late (there were signs before June 28 that a serious conflict was afoot). Further, the international response has been hindered by the fact that the restoration of the president is seen as synonymous with protecting Honduran democracy. This has made it difficult for local elites to negotiate a solution with the international community, and has paradoxically jeopardized the prospects for a democratic solution. The strong reaction of OAS member states has left little room for negotiation and maneuver. For this reason, the mediation process led by Costa Rican President Oscar Arias was unable to offer any solution other than calling for the immediate return of Zelaya to the presidency. Despite the new prominence of the defense and promotion of democracy in the OAS (Mainwaring and Pérez-Liñán 2005), the Organization has proven to be powerless in this instance. Sadly, it is not just Honduran democracy but also the OAS that is suffering a blow to its prestige. The more cautious attitude adopted by the U.S. administration most recently suggests that the initial reaction may have been exaggerated. Elections are still scheduled for November 29, 2009, but the international community insists that these will not be recognized. The questions still linger: What will the international community do when a new, democratically elected president takes power in Honduras? By that day, the mandate of the Zelaya presidency will have expired anyway. Will Honduras remain ostracized, or will it be allowed to return to the family of Latin American democracies?

Final Remarks The Honduran debacle has yet to play itself out. Zelaya’s future, the outcome of the upcoming elections, and the fate of the country’s regime are very uncertain. Indeed, events in Honduras and the recent challenge to President Colom in Guatemala show that the phenomenon of presidential breakdown is

236

Epilogue

still very much a factor in Latin American politics, although the strong international reaction against the ouster and the ensuing isolation of Honduras may prevent presidential breakdowns in the near future. In any case, the Honduran situation demonstrates how fine a line there is between the breakdown of democracy and of a presidency, as Álvarez and Marsteintredet discuss in chapter 2 of this volume. The international community has interpreted the events as a democratic breakdown and has reacted accordingly. Not only is this reaction surprising—presidential crises have, after all, become something of a “tradition” in post third-wave democratic Latin America—but this interpretation may actually have contributed to the powerlessness of external agents. We share the alternative view of one observer who said that we are witnessing the return of old scenarios but in new times of crisis (Alcántara Saez 2009a). It is desirable that events in Honduras help to deepen democracy, as other presidential ousters have done, and encourage higher levels of citizen interest and even participation in politics. Should this turn out to be the case, Zelaya’s ouster may lead to a reform of the Honduran political system. Paradoxically, this is what motivated both Zelaya to call for a referendum and the political elite to oust him. But the first test of democratic progress in Honduras will be the elections of November 2009.

Notes 1. See Comunicado Oficial de la C SJ por Situación Política, at http://www. poderjudicial.gob.hn/ejes/comunicados/, accessed August 28, 2009. 2. This section is based on several issues of NotiCen (of the Latin America Data Base), the Latin American Weekly Report, BBCMundo, and the Economist. 3. President Micheletti actually came out in favor of the ALBA agreement, apparently in exchange for Zelaya’s support for him in the PL primaries for the presidential elections of 2009 (NotiCen 2008). 4. See Articles 239 and 374 of the Constitution. Other articles also refer to the sanctity of the no reelection rule. 5. In an interview with El Faro, legal counsel to the armed forces, Coronel Bayardo, admitted that sending Zelaya abroad was illegal, but justified the decision on the grounds that it prevented bloodshed and further conflict (http://www.elfaro. net/secciones/Noticias/20090629/noticias16_20090629.asp, accessed August 25, 2009). 6. There were sixty-two votes from the PL, fifty-five from the PN, and five from two minor parties. Congress also claimed to have Zelaya’s letter of resignation, and to have evidence that the president was planning to shut Congress down on the day of the referendum. See El Heraldo at http://www.elheraldo.hn/Ediciones/2009/06/29/ Noticias/Zelaya-planificaba-disolver-el-Congreso, accessed August 25, 2009. We have been unable to gain access to the official acts of Congress that day. 7. The constitutional text available at the Political Database of the Americas (http:// pdba.georgetown.edu/) is the 1982 Constitution, including the reforms made until 2005.

Epilogue

237

8. No constitution provides for the exile of an ousted president. But the case of Zelaya is not unique in this respect either: several ousted presidents have gone abroad. President Bucaram was sent to Panama, Fujimori lived for some years in Japan, Sánchez de Lozada was escorted to Brazil, and Pérez of Venezuela has divided his time between the United States and the Dominican Republic. Many exiled presidents have sought to return home as presidents (Fujimori and Gutiérrez are two recent examples), although none immediately after their ouster, or with such fervor or as much international support as Zelaya. 9. Personal communication with Carlos Barrachina Lisón. 10. See, Se Juramentan Corte Suprema en Honduras, at: http://www.nacion.com/ln_ ee/2009/enero/26/mundo1853559.html; Congreso Hondureño elige magistrados y supera crisis, at: http://www.nacion.com/ln_ee/2009/enero/27/mundo1853816. html, both accessed August 27, 2009. 11. In fact, Santos was prevented from running as vice president in the PL primaries, and he presented a surrogate candidate, Mauricio Villeda, instead. When Villeda surprisingly won the primaries, he declared that he would step down to let Santos run, a right Santos won after a Supreme Court ruling in March 2009. 12. In this sense, Honduras and Paraguay are rather unique presidential regimes. A recent study argues that one important difference between presidential and other regime types is that the party of the incumbent in presidential systems very rarely seeks to oust its president (Samuels and Shugart, forthcoming).

This page intentionally left blank

References

Abente Brun, Diego. 1999. “People Power in Paraguay.” Journal of Democracy 10(3): 93–100. ———. 1995. “A Party System in Transition: The Case of Paraguay.” In Building Democratic Institutions. Party Systems in Latin America, ed. Scott Mainwaring and Timothy R. Scully, 298–320. Stanford CA: Stanford University Press. ———. ed. 1993. Paraguay en transición, Caracas: Editorial Nueva Sociedad. Aguilera, Gabriel. 1993. “Guatemala: el carrusel de los golpes.” Revista Fuerzas Armadas y Sociedad 8(2): 1–4. Alcántara Saez, Manuel. 2009a. Centroamérica: viejo escenario en tiempos nuevos de crisis. Política Exterior 131: 1–10. Alcántara Saez, Manuel. 2009b. Un intento de explicación de la crisis de Honduras de 2009 desde la mirada de sus legisladores. Boletín datos de opinión November 2009: 1–4. Alfonsín, Raúl. 2004. Memoria política. Transición a la democracia y derechos humanos. Buenos Aires: Fondo de Cultura Económica. Alston, Lee J., Marcos André Melo, Bernardo Mueller and Carlos Pereira 2006. Political Institutions, Policymaking Processes and Policy Outcomes in Brazil. Washington, DC: Inter-American Development Bank. Altman, David. 2000. “Politics of Coalition Formation and Survival in Multiparty Presidential Democracies: The Case of Uruguay, 1989–1997.” Party Politics 6(3): 259–283. Álvarez, Michael E. 1998. Presidentialism and Parliamentarism: Which Works? Which Lasts? PhD Dissertation. University of Chicago, Chicago. Álvarez, Michael E. and Leiv Marsteintredet. 2007. Presidential Interruptions and Democratic Breakdown in Latin America: Similar Causes, Different Outcomes. Paper presented at the 2nd Joint Workshops of Latin Americanists, GIGA Institute of Latin American Studies, Hamburg, and Department of Comparative Politics at the Bergen University. December 13–14. Hamburg, Germany. Amorim Neto, Otávio. 2006. “The Presidential Calculus Executive Policy Making and Cabinet Formation in the Americas.” Comparative Political Studies 39(4): 415–440. ———. 2002. “Presidential Cabinets and Legislative Cohesion in Brazil.” In Legislative Politics in Latin America, ed. Scott Morgenstern and Benito Nacif, 48–78. Cambridge: Cambridge University Press. Arce, Moisés and Paul T. Bellinger Jr. 2007. “Low-Intensity Democracy Revisited: The Effects of Economic Liberalization on Political Activity in Latin America.” World Politics 60: 97–121.

240

References

Arévalo de León, Bernardo. 1997. Sobre arenas movedizas: sociedad, estado y ejército en Guatemala. Guatemala City: Editorial Serviprensa/FLACSO Guatemala. Atkins, G. Pope. 1981. Arms and Politics in the Dominican Republic. Boulder: Westview Press. Auyero, Javier. 2006. “The Political Makings of the 2001 Lootings in Argentina.” Journal of Latin American Studies 38(2): 241–265. Baer, Werner. 2003. A economia brasileira. São Paulo: Editora Nobel. Balaguer, Joaquín. 1988. Memorias de un cortesano de la ‘era de Trujillo.’ Santo Domingo: Getafe Impresos y Revistas. Balbi, Carmen Rosa and David Scott Palmer. 2001. “ ‘Reinventing’ Democracy in Peru.” Current History 100: 65–72. Banks, Arthur. 2008. Cross-National Time Series Data Archive (CNTS) 1815–2007. hdl:1902.1/11448. Databanks International. Bareiro, Line, 1999. “Crimen y victoria popular ciudadana.” Revista Paraguaya de Sociología 36(105): 227–235. Birnir, Johanna Kirstín and Donna Lee Van Cott. 2007. “Disunity in Diversity: Party System Fragmentation and the Dynamic Effect of Ethnic Heterogeneity on Latin American Legislatures.” Latin American Research Review 42(1): 99–125. Bonvecchi, Alejandro. 2006. “Determinismo y contingencia en las interpretaciones políticas de la crisis argentina.” Revista SAAP 2(3): 509–536. ———. 2004. “La eficacia de las inconsistencias: teoría y práctica del gobiemo de la economía.” In La historia reciente, ed. Marcos Novaro and Vicente Palermo, 75–90. Buenos Aires: Edhasa. Buxton, Julia. 2001. The Failure of Political Reform in Venezuela. Ashgate, Aldershot: Burlington. Caballero, Esteban. 1993. “Los partidos políticos de oposición en la transición en Paraguay: el caso del Partido Liberal Auténtico.” In Paraguay en transición, ed. Diego Abente Brun, 173–188. Caracas: Editorial Nueva Sociedad. Calderón, Ernesto García. 2001. “Peru’s Decade of Living Dangerously.” Journal of Democracy 12(2): 46–58. Calvo, Ernesto. 2007. “The Responsive Legislature: Public Opinion and Law Making in a Highly Disciplined Legislature.” British Journal of Political Science 37(2): 263–280. Calvo, Ernesto and Murillo, María Victoria. 2005. “The New Iron Law of Argentine Politics? Partisanship, Clientelism and Governability in Democratic Argentina.” In Argentine Democracy: The Politics of Institutional Weakness, ed. Steven Levitsky and María Victoria Murillo, 207–298. Philadelphia, PA: Pennsylvania State University Press. Cameron, Maxwell A. 2006. “Endogenous Regime Breakdown: The Vladivideo and the Fall of Peru’s Fujimori.” In The Fujimori Legacy: The Rise of Electoral Authoritarianism in Peru, ed. Julio F. Carrión, 268–293. Philadelphia, PA: Pennsylvania State University Press. ———. 1998. “Self-Coups: Peru, Guatemala, and Russia.” Journal of Democracy 9(1): 125–139. ———. 1997. “Political and Economic Origins of Regime Change in Peru: The Eighteenth Brumaire of Alberto Fujimori.” In The Peruvian Labyrinth: Polity, Society, Economy, ed. Maxwell Cameron and Philip Mauceri, 37–69. Philadelphia, PA: Pennsylvania State University Press.

References

241

Carey, John M. 2005. “Presidential vs. Parliamentary Government.” In Handbook of New Institutional Economics, ed. Claude Ménard and Mary M. Shirley, 91–122. Dordrecht: Springer. ———. 2003. “Transparency Versus Collective Action: Fujimori’s Legacy and the Peruvian Congress.” Comparative Political Studies 36(9): 983–1006. ———. 2002. “Legislatures and Political Accountability.” Harvard Review of Latin America 202: 32–34. Carey, John M. and Matthew Soberg Shugart. 1995. “Incentives to Cultivate a Personal Vote: A Rank Ordering of Electoral Formulas.” Electoral Studies 14(4): 417–439. Carothers, Thomas. 2002. “The End of the Transition Paradigm.” Journal of Democracy 13(1): 5–21. Carrión, Julio F., ed. 2006. The Fujimori Legacy: The Rise of Electoral Authoritarianism in Peru. Philadelphia, PA: Pennsylvania State University Press Carter, April. 2005. Direct Action and Democracy Today. Cambridge: Polity Press. Casaús Arzú, Marta E. 1996. “La pervivencia de las redes familiares en las élites empresariales en Guatemala y su inserción en el nuevo orden mundial.” In Élites, empresarios y estado en centroamérica, ed. Marta E. Casaús Arzú and Teresa García Geraldéz, 37–55. Madrid: Cedeal. CEH/Historical Clarification Commission. 1999. Guatemala: The Memory of Silence. Guatemala City: United Nations Office for Project Services. Charoski, Hernán. 2002. “Honestos y audaces: realizaciones y límites de la política anticorrupción.” In El derrumbe politico en el ocaso de la convertibilidad, ed. Marcos Novaro, 195–251. Buenos Aires: Editorial Norma. Cheibub, José Antonio. 2007. Presidentialism, Parliamentarism, and Democracy. Cambridge: Cambridge University Press. ———. 2002. “Minority Governments, Deadlock Situations, and the Survival of Presidential Democracies.” Comparative Political Studies 35(3): 284–312. Cheibub, José Antonio and Fernando Limongi. 2002. “Democratic Institutions and Regime Survival: Parliamentary and Presidential Democracies Reconsidered.” Annual Review of Political Science 5: 151–179. Cheibub, José Antonio, Adam Przeworski, and Sebastian M. Saiegh. 2004. “Government Coalitions and Legislative Success under Presidentialism and Parliamentarism.” British Journal of Political Science 34: 565–587. CICIG. 2007. Agreement between the United Nations and the State of Guatemala on the Establishment of an International Commission Against Impunity in Guatemala (CICIG), at http://www.un.org/Depts/dpa/CICIG_English.pdf. Collier, David and Steven Levitsky. 1997. “Democracy with Adjectives: Conceptual Innovation in Comparative Research.” World Politics 49(3): 430–451. Conaghan, Catherine M. 2002. “Cashing in on Authoritarianism: Media Collusion in Fujimori’s Peru.” Harvard International Journal of Press-Politics 7(1): 115–125. Cooper, Andrew F. and Thomas Legler. 2001. “The OAS in Peru. A Model for the Future?” Journal of Democracy 12(4): 123–136. Coppedge, Michael. 1994. Strong Parties and Lame Ducks. Presidential Partyarchy and Factionalism in Venezuela. Stanford, CA: Stanford University Press. Corrales, Javier. 1997. “Do Economic Crises Contribute to Economic Reform? Argentina and Venezuela in the 1990s.” Political Science Quarterly 112(4): 617–644. Corte Nacional Electoral. 2007. Boletín Estadístico 7(3) November. La Paz: CNE.

242

References

Costa, José María, and Oscar Ayala Bogarín. 1996. Operación Gedeón. Los secretos de un golpe frustrado. Asunción: Editorial Don Bosco. Cox, Gary and Scott Morgenstern. 2002. “Epilogue: Latin America’s Reactive Assemblies and Proactive Presidents.” In Legislative politics in Latin America, ed. S. Morgenstern and Benito Nacif, 446–469. Cambridge: Cambridge University Press. Crabtree, John. 2001. “The Collapse of Fujimorismo: Authoritarianism and its Limits.” Bulletin of Latin American Research 20(3): 287–303. Dahik, Alberto. 1995. Impeachment Proceedings of Vice-President Alberto Dahik. Quito: Congreso Nacional del Ecuador. Darío Espinal, Flavio. 2001. Constitucionalismo y procesos políticos en la República Dominicana. Santo Domingo: PUCMM. Deheza, Grace Ivana. 1998. “Gobiernos de coalición en el sistema presidencial: América del sur.” In El presidencialismo renovado, ed. Dieter Nohlen and Mario Fernandez, 151–169. Caracas: Nueva Sociedad. Diamond, Larry. 1997. “Consolidating Democracy in the Americas.” Annals of the American Academy of Political and Social Science 550(1): 12–41. Díaz Santana, Juan Bolívar. 1996. Trauma Electoral. Santo Domingo: Ediciones Mograf. Dosal, Paul J. 1995. Power in Transition. The Rise of Guatemala’s Industrial Oligarchy, 1871– 1994. Westport, CO: Praeger. Eaton, Kent. 2005. “Menem and the Governors: Intergovernmental Relations in the 1990s.” In Argentine Democracy: The Politics of Institutional Weakness, ed. Steven Levitsky and María Victoria Murillo, 88–114. Philadelphia, PA: Pennsylvania State University Press. Edwards, Margaret E. 2007. Presidential Survival in South America: Analyzing Institutional, Economic and Social Factors. Paper prepared for the annual meeting of the American Political Science Association, Chicago, IL, Aug 30, at http://www.allacademic.com/meta/p209887_index.html EIU. 2005. Country Report: Bolivia. February. London: Economic Intelligence Unit. ———. 2004. Country Report: Bolivia. January. London: Economist Intelligence Unit. ———. 2004. Country Report: Bolivia. May. London: Economist Intelligence Unit. ———. 2004. Country Report: Bolivia. August. London: Economist Intelligence Unit. Elgie, Robert. 2005. “From Linz to Tsebelis: Three Waves of Presidential/Parliamentary Studies?” Democratization 12(1): 106–122. Ellner, Steve. 1993. “A Tolerance Worn Thin: Corruption in the Age of Austerity.” NACLA Report on the Americas 27(3): 13–16. ———. 2008. Rethinking Venezuelan Politics: Class, Conflict and the Chávez Phenomenon. Boulder, CO: Lynne Rienner Publishers. Elster, Jon. 1989. Nuts and Bolts for the Social Sciences. Cambridge: Cambridge University Press. Escolar, Marcelo, Ernesto Calvo, Natalia Calcagno, and Sandra Minvielle. 2002. “Ultimas imágenes del naufragio: las elecciones de 2001 en Argentina.” Desarrollo Económico 42(165): 25–44. Espinal, Rosario. 2003. “El proceso democrático dominicano: avances, retrocesos y riesgos.” In Democracias en riesgo en América Latina, ed. Robinson Salazar

References

243

Pérez, Eduardo Sandoval Forero, and Dorangélica de la Rocha Almazán, 623–645. Mexico City: Libros. ———. 1999. “Conflictos electorales, reformas políticas y proceso democrático en la república Dominicana.” In La República Dominicana en el umbral del siglo XXI, ed. Ramonia Brea, Rosario Espinal, and Fernando Valerio-Holguín, 281–301. Santo Domingo: Pontificia Universidad Católica Madre y Maestra. Falcoff, Mark, 1998. “Paraguay. The Tail End of the Democratic Revolution.” American Enterprise Institute Latin American Outlook, June 1, at http://www. aei.org/publications/pubID.9084/pub_detail.asp Faúndez, Julio. 1997. “In Defense of Presidentialism: The Case of Chile.” In Presidentialism and Democracy in Latin America, ed. Scott Mainwaring and Matthew Soberg Shugart, 300–320. Cambridge: Cambridge University Press. Faxas, Laura. 2007. El mito roto. Sistema político y movimiento popular en la República Dominicana, 1961–1990. Mexico City: Siglo XXI Editores. Fernández Camacho, Carmen. 2004. “La oposición al autogolpe de Serrano Elías: eficacia de las relaciones públicas políticas.” Ámbitos 11–12: 237–259. Fitch, J. Samuel. 2005. “Post-Transition Coups: Ecuador 2000, an Essay in Honor of Martin Needler.” Journal of Political and Military Sociology 33(1): 39–58. Freidenberg, Flavia. 2008. El sueño frustrado de la gobernabilidad: instituciones, actores y política informal en Ecuador. Barcelona: CIDOB. ———. 2003. Jama, Caleta y Camello. Las estrategias de Abdalá Bucaram y el PRE para ganar elecciones. Quito: Corporación Editora Nacional. Fuentes, Juan Alberto and Maynor Cabrera. 2005. El pacto fiscal de Guatemala: una oportunidad perdida. 17th CEPAL Regional Seminar on Fiscal Policy, Santiago de Chile, March 24–27 at www.eclac.cl/ilpes/noticias/paginas/3/19803/fuentes.doc García Gallegos, B. 2005. “El 20 de abril: presente y pasado de un proyecto militar corporativo.” Íconos: Revista de Ciencias Sociales 23: 95–102. Gargarella, Roberto. 1997. “Recientes reformas constitucionales en América Latina: una primera aproximación.” Desarrollo Económico 36(144): 971–990. Gervasoni, Carlos Hugo. 2003. Debilidad política y crisis financieras: una explicación del fracaso de la Rúa en Argentina. Paper prepared for the meeting of the Latin American Studies Association. Dallas, Texas, March 27–29. Goertz, Gary. 2006. Social Science Concepts: A User’s Guide. Princeton, NJ: Princeton University Press. Graham, John W. 2008. Dominican Republic 94: Stepping Back from the Precipice. Paper presented at the Conference at the Fundación Golobal Democracia y Desarrollo (FUNGLODE), Santo Domingo, November 4. Granovski, Martín. 2001. El divorcio. Buenos Aires: El Ateneo. Guerrero, Miguel. 1999. Al borde del caos: historia oculta de la crisis electoral de 1978. Santo Domingo: Editora Corripio. ———. 1991. Enero de 1962. ¡El despertar dominicano! Santo Domingo: Ediciones Mograf. Hadenius, Axel and Jan Teorell. 2007. “Pathways from Authoritarianism.” Journal of Democracy 18(1): 143–156. Halperin, Morton H. and Kristen Lomasney. 1998. “Guaranteeing Democracy: A Review of the Record.” Journal of Democracy 9(2): 134–147. Hamilton, John R. 2006. “The Fall of Fujimori: A Diplomat’s Perspective.” Fletcher Forum of World Affairs 30(2): 191–212.

244

References

Hartlyn, Jonathan. 1998. The Struggle for Democratic Politics in the Dominican Republic. Chapel Hill, NC: University of North Carolina Press. ———. 1994. “Crisis-ridden Elections (Again) in the Dominican Republic: Neopatrimonialism, Presidentialism, and Weak Electoral Oversight.” Journal of Inter-American Studies and World Affairs 36(4): 91–144. Heston, Alan, Robert Summers, and Bettina Aten. 2006. Penn World Table Version 6.2. Philadelphia, PA: Center for International Comparisons of Production, Income and Prices (CICUP) at the University of Pennsylvania, at http://pwt.econ. upenn.edu/php_site/pwt_index.php Hochstetler, Kathryn. 2008. “Repensando el presidencialismo: desafíos y caídas presidenciales en el Cono Sur.” America Latina Hoy 49: 51–72. ———. 2006. “Rethinking Presidentialism: Challenges and Presidential Falls in South America.” Comparative Politics 38(4): 401–418. Hochstetler, Kathryn and Margaret E. Edwards. 2009. “Failed Presidencies: Identifying and Explaining a South American Anomaly.” Journal of Politics in Latin America 1(2): 31–57. Hochstetler, Kathryn and David Samuels. 2008. Crisis and Rapid Re-equilibration: The Consequences of Presidential Challenge and Failure in Latin America. Paper presented at the conference on New Frontiers on Institutional Research in Latin America, Hamburg, May 5–6. Huber, John. 1996. Rationalizing Parliaments. Cambridge, Cambridge University Press. Honduras, Corte Suprema de Justicia. 2009. Expediente Judicial relación documentada caso José Manuel Zelaya Rosales: Poder Judicial. Human Rights Watch. 1995. Guatemala. Human Rights Developments, www.hrw.org. Human Rights Watch. 1994. Human Rights in Guatemala during President De León Carpio’s First Year. Washington, DC: Human Rights Watch. Immergut, Ellen M. 1992. “The Rules of the Game: The Logic of Health PolicyMaking in France, Switzerland and Sweden.” In Structuring Politics: Historical Institutionalism in Comparative Analysis, ed. Sven Steimo, Kathleen Thelen, and Frank Longstreth, 57–89. Cambridge: Cambridge University Press. INCEP. 1993. “La crisis politico-constitucional de Guatemala. Del golpe de estado de Jorge Serrano a la presidencia constitucional de Ramiro de León Carpio.” In Panorama Centroamericano, INCEP Report 3/93. Ciudad de Guatemala: Instituto Centroamericano de Estudios Políticos. Jaggers, Keith and Ted. R. Gurr. 1995. “Tracking Democracy’s Third Wave with the Polity-III Data.” Journal of Peace Research 32(4): 469–482. Jiménez Polanco, Jacqueline. 1999. Los partidos políticos en la República Dominicana: actividad electoral y desarollo organizativo. Santo Domingo: Editora Centenario. Jonas, Susanne. 2000a. “Democratisation through Peace: The Difficult Case of Guatemala.” Journal of Interamerican Studies and World Affairs 42(4): 3–32. ———. 2000b. Of Centaurs and Doves: Guatemala’s Peace Process. Boulder, CO: Westview Press. ———. 1994. “Guatemala. El problema democrático.” Nueva Sociedad 130: 15–23. Jones, Mark P. 1995. Electoral Laws and the Survival of Presidential Democracies. Notre Dame, ID: University of Notre Dame Press. Jones, Mark P. and Wonjae Hwang. 2005. “Provincial Party Bosses: Keystone of the Argentine Congress.” In Argentine Democracy: The Politics of Institutional Weakness, ed. Steven Levitsky and María Victoria. Murillo, 115–138. Philadelphia, PA: Pennsylvania State University Press.

References

245

Jones, Mark P., Wonjae Hwang, and Pablo Micozzi. 2009. “Government and Opposition in the Argentine Congress, 1989–2007: Understanding Inter-Party Dynamics through Roll Call Vote Analysis.” Journal of Politics in Latin America 1: 67–96. Karl, Terry Lynn. 1995. “The Hybrid Regimes of Central America.” Journal of Democracy 6(3): 72–86. Karl, Terry Lynn and Philippe Schmitter. 1991. “Modes of Transition in Latin America, Southern Europe and Eastern Europe.” International Social Science Journal 128: 269–284. Kasuya, Yuko. 2003. “Weak Institutions and Strong Movements: The Case of President Estrada’s Impeachment and Removal in the Philippines.” In Checking Executive Power. Presidential Impeachments in Comparative Perspectives, ed. Jody C. Baumgartner and Naoko Kada, 45–64. Westport and London: Praeger Publishers. Kenney, Charles D. 2003. “The Death and Rebirth of a Party System, Peru 1978– 2001.” Comparative Political Studies 36(10): 1210–1239. Kim, Young Hun and Donna Bahry. 2008. “Interrupted Presidencies in Third Wave Democracies.” Journal of Politics 70(3): 807–822. King, Gary and Langche Zeng. 2001a. “Explaining Rare Events in International Relations.” International Organization 55(3): 693–715. ———. 2001b. “Logistic Regression in Rare Events Data.” Political Analysis 9(2): 137–163. Kinzo, Maria D’Alva. 1988. Oposição e Autoritarismo: Gênese e Trajetória do MDB, 1966–1979. São Paulo: IDESP/Vértice. Laakso, Markku and Rein Taagepera. 1979. “The ‘Effective’ Number of Parties: A Measure with Application to West Europe.” Comparative Political Studies 12(1): 3–27. Lalander, Rickard. 2004. Suicide of the Elephants: Venezuelan Decentralization between Partyarchy and Chavismo. Helsinki: Renvall Institute for Area and Cultural Studies, University of Helsinki; Stockholm: Institute of Latin American Studies, Stockholm University. Lambert, Peter, 1997. “Assessing the Transition.” In The Transition to Democracy in Paraguay, ed. Peter Lambert and Andrew Nickson, 200–213. Houndmills and London: Macmillan Press. Lambert, Peter and Andrew Nickson, eds. 1997. The Transition to Democracy in Paraguay. Houndmills and London: Macmillan Press. Laserna, Roberto and Miguel Villarroel. 1999. 29 años de conflictos en Bolivia. Cochabamba: CERES Working Paper. Laserna, Roberto, Jesús Ortega, and Douglas Chacón. 2006. Conflictividad sociolaboral en Bolivia: estudios y reflexiones. La Paz: Organization of American States, at www.geocities.com/laserna_r/ConflictosSociolaborales.pdf Lázara, Simón. 1997. El asalto al poder. actores e intereses en la crisis argentina del 89. Buenos Aires: Tiempo de Ideas. Lazarte, Jorge. 2005. Entre los espectros del pasado y las incertidumbres del futuro: política y democracia en Bolivia a principios del siglo XXI. La Paz: Plural. Lazarte, Jorge and Mario N. Pacheco. 1992. Bolivia: economía y sociedad, 1982– 1985. La Paz: CEDLA. Levitsky, Steven. 2005. “Crisis and Renovation: Institutional Weakness and the Transformation of Argentine Peronism, 1983–2003.” Argentine Democracy: The

246

References

Politics of Institutional Weakness, ed. S. Levitsky and M. Murillo, 181–206. Philadelphia, PA: Pennsylvania State University Press. ———. 1999. “Fujimori and Post-Party Politics in Peru.” Journal of Democracy 10(3): 78–92. Levitsky, Steven and Lucan A. Way. 2006. “Linkage versus Leverage: Rethinking the International Dimension of Regime Change.” Comparative Politics 38(4): 379–400. ———. 2005. “International Linkage and Democratization.” Journal of Democracy 16(3): 20–34. ———. 2002. “The Rise of Competitive Authoritarianism.” Journal of Democracy 13(2): 51–65. Levitsky, Steven and Maxwell A. Cameron. 2003. “Democracy without Parties? Political Parties and Regime Change in Fujimori’s Peru.” Latin American Politics and Society 45(3): 1–33. Levitsky, Steven and María Victoria Murillo. 2005. Argentine Democracy: The Politics of Institutional Weakness. Philadelphia, PA: Pennsylvania State University Press. Levy Yeyati, Eduardo and Diego Valenzuela. 2007. La resurrección. Historia de la pos-crisis argentina. Buenos Aires: Sudamericana. Lewis, Colin and Nissa Torrents. 1993. Argentina in the Crisis Years, 1983–1990. London: Institute of Latin American Studies. Linz, Juan J. 2000. Totalitarian and Authoritarian Regimes. Boulder, CO: Lynne Rienner. ———. 1994. “Presidential or Parliamentary Democracy: Does It Make a Difference?” In The Failure of Presidential Democracy, ed. Juan J. Linz and Arturo Valenzuela, 3–87. Baltimore, MD: Johns Hopkins University Press. ———. 1990. “The Perils of Presidentialism.” Journal of Democracy 1(1): 51–69. ———. 1978. The Breakdown of Democratic Regimes: Crisis, Breakdown and Reequilibration. Baltimore, MD: Johns Hopkins University Press. Linz, Juan J. and Alfred Stepan. 1996. Problems of Democratic Transition and Consolidation: Southern Europe, South America, and Post-Communist Europe. Baltimore, MD: Johns Hopkins University Press. Lipset, Seymour Martin. 1959. “Some Social Requisites of Democracy: Economic Development and Political Legitimacy.” American Political Science Review 53(1): 69–105. Lipset, Seymour Martin, Kyoung-Ryung Seong, and John Charles Torres. 1993. “A Comparative Analysis of the Social Requisites of Democracy.” International Social Science Journal 45(2): 154–175. Llanos, Mariana. 2002. Privatisation and Democracy in Argentina: An Analysis of President-Congress Relations. Houndmills and London: Palgrave St Antony’s College Series. Llanos, Mariana and Ana Margheritis. 2006. “Why Do Presidents Fail? Political Leadership and the Argentine Crisis, 1999–2001.” Studies in Comparative International Development 40(4): 77–103. Loewenstein, Karl. 1949. “The Presidency Outside the United States: A Study in Comparative Political Institutions.” Journal of Politics 11(3): 447–496. Lora, Guillermo. 1977. A History of the Bolivian Labor Movement. New York: Cambridge University Press. Lozano, Wilfredo. 2002. Después de los caudillos. Santo Domingo: Editora Manatí.

References

247

Machinea, José Luis. 1993. “Stabilisation under Alfonsín.” In Argentina in the Crisis Years, 1983–1990, ed. Colin Lewis and Nissa Torrents, 124–143. London: Institute of Latin American Studies. Madrid, Raúl L. 2005. “Indigenous Parties and Democracy in Latin America.” Latin American Politics and Society 47(4): 161–79. Mainwaring, Scott. 1993. “Presidentialism, Multipartism, and Democracy: The Difficult Combination.” Comparative Political Studies 26(2): 198–228. ———. 1992. “Transitions to Democracy and Democratic Consolidation: Theoretical and Comparative Issues.” In Issues in Democratic Consolidation: The New South American Democracies in a Comparative Perspective, ed. Scott Mainwaring, Guillermo O’Donnell, and Samuel J. Valenzuela, 294–341. Notre Dame, IN: University of Notre Dame Press. ———. 1990. “Presidentialism in Latin-America.” Latin American Research Review 25(1): 157–179. Mainwaring, Scott and Aníbal Pérez-Liñán. 2005. “Latin American Democratization since 1978: Democratic Transitions, Breakdowns and Erosions.” In The Third Wave of Democratization in Latin America: Advances and Setbacks, ed. Francis Hagopian and Scott P. Mainwaring, 14–59. Cambridge: Cambridge University Press. Mainwaring, Scott, Daniel Brinks, and Aníbal Perez-Liñán. 2001. “Classifying Political Regimes in Latin America, 1945–1999.” Studies in Comparative International Development 36(1): 37–65. Mainwaring, Scott, Aníbal Pérez-Liñán, and Timothy R. Scully. 1995. “Introduction: Party Systems in Latin America.” In Building Democratic Institutions: Party Systems in Latin America, ed. S. Mainwaring and T.R. Scully, 1–36. Stanford, CA: Stanford University Press. Malamud, Andrés. 2006. Social Revolution or Political Takeover? The Argentine Collapse of 2001 Reassessed. Prepared for the 26th Latin American Studies Association Congress, San Juan, Puerto Rico, March 15–18. Maríñez, Pablo A. 1994. Democracia y procesos electorales en República Dominicana. Santo Domingo: Editora Alfa y Omega. Marsteintredet, Leiv. 2008a. “Executive-Legislative Deadlock in the Dominican Republic.” Latin American Politics and Society 50(2): 131–160. ———. 2008b. “Las consecuencias sobre el régimen de las interrupciones presidenciales en América Latina.” América Latina Hoy 49: 31–50. Marsteintredet, Leiv and Einar Berntzen. 2008. “Reducing the Perils of Presidentialism in Latin America through Presidential Interruptions.” Comparative Politics 41(1): 83–101. Martínez, Emma G. 1993. “The Coup All Knew Was Coming.” Revista Envío 144, at http://www.envio.org.ni/articulo/1699. Martini, Carlos and Mar Carlos. 1997. “The Armed Forces.” In The Transition to Democracy in Paraguay, ed. Peter Lambert and Andrew Nickson, 65–71. Houndmills and London: Macmillan Press. Mayorga, René A. 2001. “Presidencialismo parlamentarizado y procesos de decisión en Bolivia.” In Visiones de fin de siglo. Bolivia y América Latina en el siglo XX, ed. Dora Cajías and Seemin Qayum, 291–304. La Paz: IFEA/Coordinadora de Historia. ———, ed. 1987. Democracia a la deriva: Dilemas de la participación social en Bolivia. La Paz: Edobol.

248

References

McCleary, Rachel M. 1999a. Dictating Democracy. Guatemala and the End of Violent Revolution. Gainesville, FL: The University Press of Florida. ———. 1999b. Imponiendo la democracia: Las élites guatemaltecas y el fin del conflicto armado. Guatemala: Artemis Edinter. McClintock, Cynthia. 2006. “Electoral Authoritarian Versus Partially Democratic Regimes: The Case of the Fujimori Government and the 2000 Elections.” In The Fujimori Legacy: the Rise of Electoral Authoritarianism in Peru, ed. J.F. Carrión, 242–267. Philadelphia, PA: Pennsylvania State University Press. ———. 2001. “The OAS in Peru: Room for Improvement.” Journal of Democracy 12(4): 137–140. McCubbins, Mathew and Thomas Schwartz. 1987. “Congressional Oversight Overlooked: Police Patrol versus Fire Alarms.” In Congress: Structure and Policy, ed. Mathew McCubbins and Terry Sullivan, 426–440. Cambridge: Cambridge University Press. McFaul, Michael. 2005. “Transitions from Postcommunism.” Journal of Democracy 16(3): 5–19. McMillan, John and Pablo Zoido. 2004. “How to Subvert Democracy: Montesinos in Peru.” Journal of Economic Perspectives 18(4): 69–92. Medina, Ricardo, 1997. “The Left.” In The Transition to Democracy in Paraguay, ed. Peter Lambert and Andrew Nickson, 72–86. Houndmills and London: Macmillan Press. Mejía Acosta, Andrés. 2009. Informal Coalitions and Policymaking in Latin America. New York, Routledge. ———. 2006. “Crafting Legislative Ghost Coalitions in Ecuador: Informal Institutions and Economic Reform in An Unlikely Case.” In Informal Institutions and Democracy: Lessons from Latin America, ed. Gretchen Helmke and Steve Levitsky, 69–86. Baltimore, MD: Johns Hopkins University Press. ———. 2004. Ghost Coalitions: Economic Reforms. Fragmented Legislatures and Informal Institutions in Ecuador, 1979–2002. PhD. Dissertation. Notre Dame, Indiana. Mejía Acosta, Andrés, María Caridad Araujo, Aníbal Pérez Liñán, and Sebastian M. Saiegh. 2008. “Veto Players, Fickle Institutions, and Low-Quality Policies: The Policymaking Process in Ecuador.” In Political Institutions, Policymaking Processes, and Policy Outcomes, ed. Pablo T. Spiller, Ernesto Stein, and Mariano Tommasi, 243–286. Boston, MA: Harvard University Press. Morales Solá, Joaquín. 2001. El sueño eterno. Ascenso y caída de la Alianza. Buenos Aires: Planeta. ———. 1990. Asalto a la ilusión. Buenos Aires: Planeta. Morales, Juan Antonio. 1993. “Política económica en Bolivia después de la transición a la democracia.” In La política económica de la transición a la democracia, ed. Juan Antonio Morales and Gary McMahon, 97–115. Chile: CIEPLAN. Morgenstern, Scott, Javier Negri, and Aníbal Pérez-Liñán. 2008. “Parliamentary Opposition in Non-Parliamentary Regimes.” Journal of Legislative Studies 12(1–2): 160–189. Morínigo, José Nicolás. 1999. “La plaza, el congreso y la corte: claves de la legitimidad del presidente Gonzalez Macchi.” Revista Paraguaya de Sociología 36(105): 181–225. Moya Pons, Frank. 2000. Manual de historia dominicana. Santo Domingo: Caribbean Publishers.

References

249

Mustapic, Ana María. 2006. “América Latina: las renuncias presidenciales y el papel del Congreso.” Política 47: 55–70. ———. 2005. “Inestabilidad sin colapso, la renuncia de los presidentes: Argentina en el año 2001.” Desarrollo Económico 45(178): 263–280. ———. 2002. “Oscillating Relations: President and Congress in Argentina.” In Legislative Politics in Latin America, ed. Scott Morgenstern and Benito Nacif, 23–47. Cambridge: Cambridge University Press. Mustapic, Ana María and Matteo Goretti. 1993. El congreso de la transición democrática, 1983–1989. Buenos Aires: Instituto Di Tella Working Paper. NDI. 1994. Interim Report on the May 16, 1994 Elections in the Dominican Republic. Washington, DC: National Democratic Institute for International Affairs. NDI/Carter Center. 1990. 1990 Elections in the Dominican Republic: Report of an Observer Delegation. Washington, DC: Carter Center of Emory University and the National Democratic Institute for International Affairs. Negretto, Gabriel L. 2006. “Minority Presidents and Democratic Performance in Latin America.” Latin American Politics and Society 48(3): 63–92. Nickson, Andrew. 1997. “The Wasmosy Government.” In The Transition to Democracy in Paraguay, ed. Peter Lambert and Andrew Nickson, 185–199. Houndmills and London: Macmillan Press. ———. 1989. “The Overthrow of the Stroessner Regime. Re-establishing the Status Quo.” Bulletin of Latin American Research 8(2): 185–209. Nohlen, Dieter, ed. 2005. Elections in the Americas: A Data Handbook: North America, Central America, and the Caribbean. Volumes I and II: Elections Worldwide. Oxford: Oxford University Press. Nolte. 1994. “Paraguay: Eine ‘geschenkte’ Demokratie und ein ‘verschenkter’ Wahlsieg.” Lateinamerika: Analysen-Daten-Dokumentation 11(25–26): 133–142. NotiCen. 2005. Honduran Presidential Election Throws Country into Crisis. Latin American Data Base, January 12, 2005. ———. 2006. Honduran Presidency Unravelling. Latin American Data Base, August 10, 2006. ———. 2008. Honduras’ ALBA Ratification Figures Prominently in Presidential Primaries: A Candidate Comes Around. Latin American Data Base, October 23, 2008. ———. 2009. Honduras’ Lame-Duck President Calls for Constitutional Assembly. Latin America Data Base, March 26, 2009. Novaro, Marcos. 2002. El derrumbe politico en el ocaso de la convertibilidad. Buenos Aires: Grupo Editorial Norma. O’Donnell, Guillermo. 1994. “Delegative Democracy.” Journal of Democracy 5(1): 59–69. ———. 1988. Bureaucratic Authoritarianism. Argentina 1966–1973 in Comparative Perspective. Berkeley, CA: University of California Press. O’Donnell, Guillermo and Philippe C. Schmitter. 1986. Transitions from Authoritarian Rule: Tentative Conclusions about Uncertain Democracies. Baltimore, MD: John Hopkins University Press. OAS. 2003. Informe de la Organización de los Estados Americanos sobre los hechos de febrero del 2003 en Bolivia, http://www.oas.org/OASpage/esp/Documentos/ InfBO-051203.htm. ———. 2002. Report of the Electoral Observation Mission Peru, April 8, 2001. Washington, DC: OAS Unit for the Promotion of Democracy.

250

References

———. 2000. Report: Electoral Observation Mission in Peru: 2000 General Elections. Washington, DC: OAS Unit for the Promotion of Democracy. ———. 1997. Observaciones electorales en República Dominicana 1994–1996. Washington, DC: OAS Unit for the Promotion of Democracy. Olllier, Maria Matilde. 2008. “La institucionalización democratica en el calllejón: la inestsabilidad presidencial en Argentina, 1999–2003.” América Latina Hoy 49: 73–103. Otter, Thomas. 1999. “El Rol de los medios de comunicación en la crisis política del marzo Paraguayo o un ‘marzo’ para los medios.” Serie Enfoques CEPPRO 19: 43–57. Pachano, Simón. 2005. “Ecuador: cuando la inestabilidad se vuelve estable.” Íconos: Revista de Ciencias Sociales 23: 39–46. ———. 1997. “Bucaram ¡fuera! Bucaram ¿fuera? In ¿Y ahora qué . . . ? Una contribución al análisis político-histórico del país” ed. R. Báez, E. Ribadeneira, A. Acosta, F. Muñoz, A. Carrión, W. Paredes, C. Verduga, M. Luna, S. Pachano, M. Donoso Pareja, and R. Rosero Jácome, 229–264. Quito: Eskeletra Editorial. Paige, Jeffery M. 1997. Coffee and Power: Revolution and the Rise of Democracy in Central America. Cambridge, MA: Harvard University Press. Palencia, Mayra P. 2002. El estado para el capital: la arena fiscal, manzana de discordia entre el empresariado y el poder público. Guatemala City: Publicación de la Facultad Latinoamericana de Ciencias Sociales Palermo, Vicente and Marcos Novaro. 1996. Política y poder en el gobierno de Menem. Buenos Aires: Grupo Editorial Norma. Pallares, Amalia. 2006. “Mass Mobilization and Presidential Removal in Ecuador: Entre la Ira y la Esperanza.” LASA Forum 37(1): 22–26. Payne, J. Mark, Daniel Zovatto G., and Mercedes Mateo Díaz. 2007. Democracies in Development: Politics and Reform in Latin America. Washington, DC: InterAmerican Development Bank. Peetz, Peter. 2009. Honduras: Von einem Militärputsch, der keiner sein will. Giga Fokus, No. 7. Pereira, Carlos and Bernardo Mueller. 2004. “The Cost of Governing: Strategic Behavior of the President and Legislators in Brazil’s Budgetary Process.” Comparative Political Studies 37(7): 781–815. Pérez-Liñán, Aníbal. 2008. “Instituciones, coaliciones callejeras e inestabilidad política: perspectivas teóricas sobre las crisis presidenciales.” America Latina Hoy 49: 105–126. ———. 2007. Presidential Impeachment and the New Political Instability in Latin America. Cambridge: Cambridge University Press. ———. 2005. “Democratization and Constitutional Crisis in Presidential Regimes: Toward Congressional Supremacy?” Comparative Political Studies 38(1): 51–74. ———. 2003. “Pugna de poderes y crisis de gobernabilidad: ¿Hacia un nuevo presidencialismo?” Latin American Research Review 38(3): 149–164. ———. 2001. “Crisis presidenciales: gobernabilidad y estabilidad democrática en América Latina, 1950 –1996.” Revista Instituciones y Desarrollo 8–9: 281–298. Pérez Perdomo, Rogelio. 1993. “Crisis política y sistema judicial en Venezuela.” Revista del Centro de Estudios Constitucionales 16: 95–112. Pérez, Carlos Andrés (with Ramón Hernández and Roberto Giusti). 2006. Carlos Andrés Pérez: Memorias proscritas. Caracas : Editorial El Nacional.

References

251

Pérez, Carlos Andrés (with Ramón Hernández and Roberto Giusti). 1993. “Mensaje después del fallo de la Corte Suprema de Justicia que encontró méritos para su enjuiciamiento,” Venezuela Analítica, at http://www.analitica.com/bitblioteca/ cap/csj.asp, accessed November 24, 2008. Political Database of the Americas. 1999. 1990 General Election Results. Georgetown University and the Organization of American States, at http://pdba.georgetown. edu/Elecdata/Guate/guelep90.html, accessed November 23, 2008. Przeworski, Adam. 1992. “The Games of Transition.” In Issues in Democratic Consolidation: The New South American Democracies in a Comparative Perspective, ed. Scott Mainwaring, Guillermo O’Donnell, and Samuel J. Valenzuela, 105–152. Notre Dame, IN: University of Notre Dame Press. ———. 1988. “Democracy as a Contingent Outcome of Conf licts.” In Constitutionalism and Democracy. Studies in Rationality and Social Change, ed. Jon Elster and Rune Slagstad, 59–80. Cambridge: Cambridge University Press. Przeworski, Adam, Michael E. Álvarez, José Antonio Cheibub, and Fernando Limongi. 2000. Democracy and Development: Political Institutions and Well-Being in the World, 1950–1990. Cambridge: Cambridge University Press. Ragin, Charles C. 1987. The Comparative Method: Moving Beyond Qualitative and Quantitative Strategies. Berkeley, CA: University of California Press. Raschke, Joachim. 1988. Soziale Bewegungen: Ein historisch-systematischer Grundriß. Frankfurt am Main: Campus Verlag. Riquelme, Marcial A. 1993. “Desde el stronismo hacia la transición a la democracia: el papel del actor militar.” In Paraguay en transición, ed. Diego Abente Brum, 189–215. Caracas: Editorial Nueva Sociedad. Riquelme, Marcial A. and Jorge G. Riquelme. 1997. “Political Parties.” In The Transition to Democracy in Paraguay, ed. Peter Lambert and Andrew Nickson, 47–64. Houndmills and London: Macmillan Press. Rocabado, Edmundo. 2006. Rebelión. Historia del sindicalismo y los movimientos sociales en Bolivia. Oruro: Latinas Editores. Roett, Riordan. 1989. “Paraguay After Stroessner.” Foreign Affairs 68(2): 124–142. Romero, Aníbal. 1999. Decadencia y crisis de la democracia. ¿A dónde va la democracia venezolana? Caracas: Editorial Panapo. Romero, Salvador. 2003. Geografía electoral de Bolivia. La Paz: Fundemos. Rosada-Granados, Héctor. 1997. Guatemala: el desafío de la paz. Nueva Sociedad 147: 18–26. Sachs, Jeffrey. 1987. “The Bolivian Hyperinflation and Stabilization.” AEA Papers and Proceedings 77(2): 279–283. Sagás, Ernesto. 2000. Race and Politics in the Dominican Republic. Gainesville, FL: University Press of Florida. ———. 1997. “The 1996 Presidential Elections in the Dominican Republic.” Electoral Studies 16(1): 103–107. Samuels, David J. and Matthew S. Shugart. Forthcoming. Presidents, Parties, Prime Ministers: A Framework for Analysis. Cambridge: Cambridge University Press. Sánchez, Omar. 20 08. “Guatemala’s Party Universe: A Case Study of Underinstitutionalization.” Latin American Politics and Society 50(1): 123–151. Sartori, Giovanni. 1997. Comparative Constitutional Engineering. New York: New York University Press. Schamis, Hector E. 2002. “Argentina: Crisis and Democratic Consolidation.” Journal of Democracy 13(2): 81–94.

252

References

Schedler, Andreas. 2002. “The Menu of Manipulation.” Journal of Democracy 13(2): 36–50. Schirmer, Jennifer. 2002. “The Guatemalan Politico-Military Project. Whose Ship of the State?” In Political Armies: The Military and Nation Building in the Age of Democracy, ed. Dirk Krujit and Kees Koonings, 64–89. London: Zed Books. ———. 1998. The Guatemalan Military Project: A Violence Called Democracy. Philadelphia, PA: University of Pennsylvania Press. Schmidt, Gregory D. 2003. “The Presidential and Congressional Elections in Peru.” Electoral Studies 22: 344–351. ———. 2002. “The Presidential Election in Peru, April 2000.” Electoral Studies 21: 346–357. Schmitter, Philippe C. 2004. “The Ambiguous Virtues of Accountability.” Journal of Democracy 15(4): 47–60. Schrank, Andrew. 2003. “Foreign Investors, ‘Flying Geese,’ and the Limits to ExportLed Industrialization in the Dominican Republic.” Theory and Society 32(4): 415–443. Schumpeter, Joseph A. 1952. Capitalism, Socialism and Democracy. London: Allen and Unwin. Shugart, Matthew Soberg and John M. Carey. 1992. Presidents and Assemblies: Constitutional Design and Electoral Dynamics. Cambridge: Cambridge University Press. Shugart, Matthew Soberg, John M. Carey, and Scott Mainwaring. 1997. “Presidentialism and Democracy in Latin America: Rethinking the Terms of the Debate.” In Presidentialism and Democracy in Latin America, ed. Scott Mainwaring and Matthew Soberg Shugart, 394–439. Cambridge: Cambridge University Press. Siaroff, Alan. 2003. “Comparative Presidencies: The Inadequacy of the Presidential, Semi-Presidential and Parliamentary Distinction.” European Journal of Political Research 42(3): 287–312. Sondrol, Paul C. 1992. “The Paraguayan Military in Transition and the Evolution of Civil-Military Relations.” Armed Forces and Society 19(1): 105–122. Spanakos, Anthony Peter. 2000. “The Dominican Republic: A New Way.” In Latin American Politics and Development, ed. Howard J. Wiarda and Harvey F. Kline, 527–541. Boulder, CO: Westview Press. Spence, Jack. 2004. “War and Peace in Central America: Comparing Transitions towards Democracy and Social Equity in Guatemala, El Salvador, and Nicaragua.” Hemisphere Initiatives, November. Spiller, Pablo and Mariano Tomassi. 2007. The Institutional Foundations of Public Policy in Argentina. New York: Cambridge University Press. Stepan, Alfred and Cindy Skach. 1994. “Presidentialism and Parliamentalism in Comparative Perspective.” In The Failure of Presidential Democracy: Comparative Perspectives, ed. Juan J. Linz and Arturo Valenzuela, 119–136. Baltimore, MD: Johns Hopkins University Press. Stokes, Susan C. 2001. Mandates and Democracy: Neoliberalism by Surprise in Latin America. Cambridge: Cambridge University Press. Tanaka, Martín. 2005. “Peru 1980–2000. Chronicle of a Death Foretold? Determinism, Political Decisions, and Open Outcomes.” The Third Wave of Democratization in Latin America. Advances and Setbacks, ed. Frances Hagopian and Scott P. Mainwaring, 261–288. Cambridge: Cambridge University Press.

References

253

Taylor, Lewis. 2005. “From Fujimori to Toledo: The 2001 Elections and the Vicissitudes of Democratic Government in Peru.” Government and Opposition 40(4): 565–596. ———. 2001. “Alberto Fujimori’s Peripeteia: From ‘Re-Election’ to Regime Collapse.” Revista Europea de Estudios Latinoamericanos y del Caribe 70: 3–24. Taylor-Robinson, Michelle M. 2007. “Presidential and Congressional Elections in Honduras, November 2005.” Electoral Studies 26(2): 521–524. Taylor-Robinson, Michelle M. 2001. “Old Parties and New Democracies: Do They Bring out the Best in One Another?” Party Politics 7(5): 581–604. Torre, Juan Carlos. 2005. “Citizens versus Political Class: The Crisis of Partisan Representation.”In Argentine Democracy: The Politics of Institutional Weakness, ed. Steve Levitsky and María Victoria Murillo, 165–180. Philadelphia, PA: Pennsylvania State University Press. ———. 1993. “Conflict and Cooperation in Governing the Economic Emergency: The Alfonsín Years.” In Argentina in the Crisis Years, 1983–1990, ed. Colin Lewis and Nissa Torrents, 73–89. London: Institute of Latin American Studies. Torres-Rivas, Edelberto. 2006 “Guatemala: desarrollo, democracia y los acuerdos de paz.” Revista Ciencias Sociales 3(6): 11–48. ———. 2001. “Reviews: Centaurs and Doves: Guatemala’s Peace Process by Susanne Jonas; and Dictating Democracy: Guatemala and the End of the Violent Revolution by Rachel M. McCleary.” American Political Science Review 95(1): 238–239. Train, Kenneth. 2003. Discrete Choice Methods with Simulation. Cambridge: Cambridge University Press. Tsebelis, George. 2002. Veto Players. New York: Russel Sage. ———. 1995. “Decision-making in Political Systems: Veto Players in Presidentialism, Multicameralism, and Pluripartism.” British Journal of Political Science 25: 289–325. Urbina, Andrés Benavente. 2003. “Corrupción y estatismo: el caso de América Latina.” Revista Enfoques 1(1): 155–167. Valdéz, Fernando J. 2003. El ocaso de un liderazgo: las elites empresariales tras un nuevo protagonismo. Guatemala City: FLACSO/ Editorial de Ciencias Sociales. Valenzuela, Arturo. 2008. “Presidencias latinoamericanas interrumpidas.” América Latina Hoy (49): 15–30. ———. 2004. “Latin American Presidencies Interrupted.” Journal of Democracy 15(4): 5–19. ———. 1997. “Paraguay: The Coup that Didn’t Happen.” Journal of Democracy 8(1): 43–55. ———. 1994. “Party Politics and the Crisis of Presidentialism in Chile: A Proposal for a Parliamentary Form of Government.” In The Failure of Presidential Democracy: The Case of Latin America, ed. Juan J. Linz and Arturo Valenzuela, 91–150. Baltimore, MD: Johns Hopkins University Press. ———. 1993. “Latin America: Presidentialism in Crisis.” Journal of Democracy 4(4): 3–16. Van Cott, Donna Lee. 2005. “Building Inclusive Democracies: Indigenous Peoples and Ethnic Minorities in Latin America.” Democratization 12(5): 825–842. ———. 2003. “Institutional Change and Ethnic Parties in Latin America.” Latin American Politics and Society 45(2): 1–39. Vega, Bernardo. 1991. Kennedy y los Trujillo. Santo Domingo: Fundación Cultural Dominicana.

254

References

Villagrán de León, Francisco. 1993. “Thwarting the Guatemalan Coup.” Journal of Democracy 4(4): 117–124. WDI. 2008. World Development Indicators. Washington, DC: World Bank Group. Wedge, Bryant. 1969. “Case Study of Student Political Violence: Brazil 1964 and the Dominican Republic 1965.” World Politics 21(2): 183–206. Weyland, Kurt. 2006. “The Rise and Decline of Fujimori’s Neo-populist Leadership.” In The Fujimori Legacy: The Rise of Electoral Authoritarianism in Peru, ed. J. Carrión, 13–38. Philadelphia, PA: Pennsylvania State University Press. ———. 1993. “The Rise and Fall of President Collor and Its Impact on Brazilian Democracy.” Journal of Interamerican Studies and World Affairs 35(1): 1–37. Whitehead, Laurence, ed. 1996. The International Dimensions of Democratization: Europe and the Americas. New York: Oxford University Press. Whitehead, Laurence, Guillermo O’Donnell, and Philippe C. Schmitter, eds. 1986. Transitions from Authoritarian Rule: Prospects for Democracy. Baltimore, MD: Johns Hopkins University Press. Wiarda, Howard J. 1998. “The Dilemmas of Democracy in the Dominican Republic: A Paradigm for all of Latin America.” In Assessing Democracy in Latin America: A Tribute to Russell H. Fitzgibbon, ed. Philip Kelly, 185–196. Boulder, CO: Westview Press. Wood, Elisabeth Jean. 2000. Forging Democracy From Below: Insurgent Transitions in South Africa and El Salvador. Cambridge: Cambridge University Press. Yashar, Deborah J. 1998. “Contesting Citizenship: Indigenous Movements and Democracy in Latin America.” Comparative Politics 31(1): 23–42. Yuste, Juan Carlos and Hugo Valiente. 1996. “Fuerzas Armadas.” In Derechos Humanos en Paraguay, ed. SERPAJ (Servicio Paz y Justicia) Paraguay, 345–382. Asuncion: SERPAJ-PY. Zagorski, Paul W. 2003. “Democratic Breakdown in Paraguay and Venezuela: The Shape of Things to Come for Latin America?” Armed Forces and Society 30(1): 87–116. Zamosc, León. 2007. The Indian Movement and Political Democracy in Ecuador. Latin American Politics and Society 49(3): 1–34.

Contributors

Michael E. Álvarez is Associate Professor in the Department of Comparative Politics at the University of Bergen. His research focuses on political regimes and democratization. Einar Berntzen is Associate Professor of Comparative Politics at the University of Bergen, Norway. He is the author of numerous chapters and articles on Latin American and European politics, and on comparative methods. Miguel A. Buitrago is a Doctoral Fellow and Associate Researcher at the GIGA Institute of Latin American Studies in Hamburg, Germany. He has a Masters degree in Economics from the American University in Washington, DC. His current research focuses on decentralization, democratization, presidentialism, and regionalism. Maren Christensen Bjune has a Masters/MSc in Latin American Politics from the Institute for the Study of the Americas (ISA), University of London, and is currently working as a college lecturer and for Save the Children in Bergen, Norway. Her research has focused on (neo-) Pentecostalism in Guatemala. Argelina Cheibub Figueiredo is Professor of Political Science at the Rio de Janeiro University Research Institute (IUPERJ), and Senior Researcher at the Brazilian Planning and Analysis Centre (CEBRAP). She has published articles and books on executive-legislative relations and coalition government in Brazil. Tor-Einar Holvik Skinlo holds a Masters degree in Comparative Politics from the University of Bergen, and is a former university lecturer. He is currently an Advisor to the County Municipality of Sogn og Fjordane. Rickard Lalander has a Ph.D. in Latin American Studies, is a Research Fellow at the Institute of Latin American Studies at Stockholm University, and is Associate Researcher at the Renvall Institute at the University of Helsinki, and at the Universidad Andina Simón Bolívar, in Quito, Ecuador. Mariana Llanos has a D.Phil in Politics (Oxford, 1999), and is a Senior Research Fellow at the GIGA German Institute of Global and Area Studies,

256

Contributors

Latin American Studies, in Hamburg, Germany. Her works include several chapters and articles on Argentine politics and comparative analyses of Latin American political institutions. Leiv Marsteintredet is a Ph.D student at the Department of Comparative Politics at the University of Bergen, Norway. His thesis deals with presidential breakdowns in Latin America. He has published articles on presidential breakdowns and politics in the Dominican Republic. Andrés Mejía Acosta is a Governance Research Fellow at the Institute of Development Studies (IDS), University of Sussex. He has published a book on informal coalitions and policymaking in Latin America (2009), and is currently working on a project on the politics of natural resource management in the Andes. Ana María Mustapic is Associate Professor at the Department of Political Science and International Studies of the Torcuato Di Tella University in Buenos Aires. Her main areas of research include congress, political parties, and electoral systems. She is currently finishing a book on the micro foundations of party politics in Argentina. Detlef Nolte has a D.Phil from the University of Mannheim, and is Professor of Political Science, International Relations, and Latin American Studies at the University of Hamburg. He is Vice President of the GIGA German Institute of Global and Area Studies and Director of the GIGA Institute of Latin American Studies in Hamburg. Stina Petersen has a Masters degree in Comparative Politics from the University of Bergen. She currently works for the Norwegian Program for Development, Research, and Education (NUFU) as a Higher Executive Officer, and was formerly Junior Researcher at the Christian Michelsen Institute in Bergen. Her previous research focused on the peace process in Guatemala. John Polga-Hecimovich is a Ph.D student at the University of Pittsburgh. He has a Masters degree from the Simón Bolívar Andean University in Quito, Ecuador. His research interests include presidential instability in Latin America and coalition-building in presidential systems.

Index

21 January Patriotic Society Party (Partido Sociedad Patriótica 21 de Enero, PSP), 83 A Agency of Deposit Guarantees (Agencia de Garantía de Depósito, AGD), 82 Agrarian Reform Law, 77 Alarcón, Fabián, 82, 85 Alfarista Radical Front (Frente Radical Alfarista, FRA), 85 Alfonsín, Raul causes and aftermath of presidential breakdown, 67–69, 214, 217 de la Rúa and, 64 early exit from office, 8, 20 Menem and, 57 mid-term elections and, 58–62 policy implications of breakdown, 226 presidential breakdown, 4, 9, 55 resignation, 19, 21–22 Alliance, the, 62–64, 70 Álvarez, Carlos “Chacho,” 31, 62–63, 68, 70 Álvarez, Michael A., 5, 6, 8, 9, 33–49, 236 Angeloz, Eduardo, 59–60, 69 approval ratings, 58, 205, 208 Argaña, Luis Maria, 150–161, 218 Arias Cárdenas, Francisco, 135, 141 Arteaga, Rosalía, 26, 218 Arzú, Alvaro, 177 Austral Plan, 58–59

Authentic Radical Liberal Party (Partido Liberal Radical Auténtico, PLRA), 149, 160 authoritarianism, 12, 35, 49, 191–192, 197–198, 201–202, 207, 209–210, 225 autogolpe, 12, 36, 43, 165–172, 174, 176–178, 197, 199, 201, 202, 209, 216, 219 B Babbit, Harriet, 189 Balaguer, Joaquín, 181–193 Blanco and, 185–187 early exit from office, 4, 20 election, 21–23 electoral fraud and, 187 factors influencing presidential breakdown, 184–185 fall as presidential breakdown, 12, 190–192 fall of, 182–184 features of presidential breakdown, 36 historic falls of, 189–190 implications of fall, 192–193 international pressure on, 187–189, 216, 218 shortened term, 199 bargaining, political, 75–76, 80, 84, 125, 157

258

Index

Berntzen, Einar, 1, 12, 75, 112, 124, 126, 197–210 Black February, 101 Black Friday, 132 Blanco, Jorge, 184–186, 187 Bolivian Communist Party (Partido Comunista de Bolivia, PCB), 95–96, 103–104 Bosch, Juan, 23, 182, 186, 188, 193 Brazilian Communist Party (Partido Comunista do Brasil, PCdoB), 123 breakdowns, presidential civil society and, 42–43 compared to democratic breakdowns, 37–39 economic factors and, 41–42 explained, 5–8 factors influencing presidential survival, 40–44 institutional factors and, 44–48 overview, 3–5, 34–37 regime characteristics and, 43–44 Bucaram, Abdalá, 4, 8, 10, 20, 24–26, 74, 76, 81–84, 86, 189, 218, 232 C cabinets, turnover of, 79, 80, 214 Caldera, Rafael, 132, 135, 139–142, 143, 145 Campíns, Luís Herrera, 132, 136 Capital Markets law, 77 caracazo, 129, 134–135, 138, 142, 144, 226 caretaker governments, 4, 9, 56, 57, 65–67, 74, 84, 214, 227 Catholic Church, 22–23, 35, 96, 105, 156, 169, 173, 183 Causa R., 134–135, 137, 141–142, 143, 144, 145 Cavallo, Domingo, 63–65, 70 Central Electoral Board (Junta Central Electoral, JCE), 181, 183, 189–190

Cerezo, Vinicio, 167 Chamorro, Violeta, 137, 140 Chávez, Hugo, 134–135, 140, 141–142, 143, 209, 230, 232, 234 checks and balances, 11, 77, 111–112, 121, 125, 225 Cheibub, José Antônio, 39, 49 Cheibub Figueiredo, Argelina, 10, 111–126, 224 Christian Democratic Committee of Independent Electoral Politics (Comité de Organización Política Electoral Independiente, COPEI), 11, 130–138, 140–143, 144 Civic Solidarity Union (Unión Cívica Solidaria, UCS), 97, 98 civil society, 10, 38, 42–43, 46, 48–49, 90–106, 159, 166, 170, 172–173, 174, 177–178, 186, 203, 215 role of, 34, 93–94 social protests and, 104–105 working definitions of, 92–93 Civil Society Assembly (Asamblea de la Sociedad Civil, ACS), 177 class, 66, 84, 87, 105, 120, 131–132 coalition formation, 74–76, 78 coca, production of, 93, 97–98, 105 Cold War, 3, 33–34, 38, 42, 86 Collor de Melo, Fernando, 4, 8, 10–11, 20, 24, 37, 111, 136, 143, 189, 232 Congress and, 115–121 rise and fall, 113–115 Colorado Party (Partido Colorado, PC), 11, 24, 148–157, 159, 218 Common Southern Market (MERCOSUR), 151, 158 Concentration of Popular Forces (Concentración de Fuerzas Populares, CFP), 85

Index Confederation of Indigenous Nationalities of Ecuador (Confederación de Nacionalidades Indígenas del Ecuador, CONAIE), 87 Confederation of Peasant Workers of Bolivia (Confederación Sindical Única de Trabajadores Campesinos de Bolivia, CSUTCB), 95, 98 Confederation of Private Entrepreneurs in Bolivia (Confederación de Empresarios Privados en Bolivia, CEPB), 100 Confederation of Workers of Venezuela (Confederación de Trabajadores de Venezuela, CTV), 131–132, 134 Congress, 5–7, 9, 10, 11, 58, 87, 103, 112, 121–125, 172, 184, 213 conflicts between president and, 2, 5, 24, 25, 36, 45, 55, 60, 96, 103, 135, 153–155, 165, 168, 200 dissolution of, 5, 18, 26–30, 149, 169, 176, 199, 223 early exits and, 17–32 Congressional performance during crises, 24–26, 73, 75, 81, 213 Congressional support, presidents and, 18, 63, 68, 103, 115, 186 Constitutional Guardians, 166, 177, 178 constitutional reform, 12, 22, 28, 31, 56, 78, 82, 88, 99, 106, 135, 137, 176, 181–182, 193, 199, 215 presidential breakdown and, 191, 214 Convertibility Plan, 21, 55, 62–69, 70, 71 Coordinating Committee of Agricultural, Commercial,

259

Industrial, and Financial Associations (Comité Coordinador de Asociaciones Agrícolas, Comerciales, Industriales y Financieras de Guatemala, CACIF), 166–168, 170–178 Correa, Rafael, 88, 228 corruption Balaguer and, 192 Bucaram and, 82 Carlos Andrés Pérez and, 133, 136–140, 142–143 Collor de Melo and, 113–115, 119, 122, 125 Dahik and, 78 de la Rúa and, 63, 68 Ecuador and, 84–88 Fujimori and, 199, 202, 205–206, 208–209 impeachment and, 10–11 presidential breakdowns and, 57, 217 Sanchez de Lozada and, 97 Serrano and, 28 Serrano Elías and, 165–169, 172–176 Cossio, Mario, 101 Coups. see also autogolpe Carlos Andrés Pérez and, 129, 134–135, 137–139, 141–143 Cubas and, 148–151, 153, 157–158 institutional, 64 Mahuad and, 85–86 presidential breakdowns and, 3, 18, 33, 35–36 presidential resignations and, 28–29 Serrano Elías and, 11 Siles Zuazo and, 95–96 Courts Constitutional Court (Corte Constitucional, CC), 167, 169, 173, 176

260

Index

Courts—Continued Constitutional Tribunal, 84, 90, 202–203, 208 Electoral Tribunal, 77, 82–83, 167, 232 Supreme Court 11, 25, 65–66, 77–78, 81, 84, 96, 101, 105, 119, 137–140, 150, 152–154, 157–160, 168–169, 175–176, 187, 229, 231–234 Supreme Electoral Tribunal (Tribunal Supremo Electoral, TSE), 167, 169 Cubas, Raúl, 4, 6, 8, 11, 20, 24, 143, 147–148, 150–160, 218, 232, 234 D Dahik, Alberto, 77–78, 82 Daló, Cristóbal Fernández, 138 de la Rúa, Fernando, 4, 9, 19, 24–25, 55–57, 62–65, 66, 67, 68, 158, 218, 226 de León Carpio, Ramiro, 28, 169, 173, 174–177 decentralization, 130, 133–134, 137, 139, 142–143 defections, 78–80, 120, 199, 207 del Carmen Ariza, José, 189 Democratic Action (Acción Democrática, AD), 11, 24, 129–143 democratic breakdowns, 1, 6, 9, 26–27, 33–49, 190–192, 209, 231–233, 236 Democratic Left (Izquierda Democrática, ID) Party, 83, 84 Democratic Popular Unity (Unión Democrática y Popular, UDP), 23, 95 demonstrations, anti-government, 156 deregulation, 60, 77, 133 Differential Exchange Rates Regime (Régimen de Cambios

Diferenciales, RECADI), 136–137 Dirceu, Jose, 122 Dominican Liberation Party (Partido de la Liberación Dominicana, PLD), 22–23, 183, 186–189, 193 Dominican Revolutionary Party (Partido Revolucionario Dominicano, PRD), 22–23, 181, 183, 185–190, 193 Duhalde, Eduardo, 9, 19, 20, 21–22, 66–67 Durán Ballén, Sixto, 76–77, 81, 86 E early elections, 20, 21, 23, 29, 96, 107, 126, 199–200, 202, 206, 218, 223 economic development, 35, 41–42, 217 economic elite, 120, 167 economic indicators, 59, 63 effectiveness, 35–36, 44 efficacy, 35, 62 election fraud, 22, 141, 149–151, 181–189, 191–193, 197–198, 202–204, 216 electoral authoritarian regimes, 12, 191, 193, 197–198, 200–208 Electoral Tribunal. see courts Electric Sector Regime Law, 81 Elite Settlement, 11, 165, 172–175 erosion, 35, 73–76, 78–84 Escobar Salom, Ramón, 137 Espina Salguero, Gustavo, 26, 172, 174, 218 exceptionalism, 130 exit modalities, 18 exits from crisis parliamentary, 24–26 presidential resignations, 19–23, 26–29

Index F factionalism, 30, 86, 130–133, 167, 185 federalism, 67 Federation of Chambers and Associations of Commerce and Production (Federación de Cámaras y Asociaciones de Comercio y Producción, FEDECAMARAS), 131 Fermín, Claudio, 139 Fernández, Leonel, 193 Fernández Daló, Cristóbal, 138 Fernández Meijide, Graciela, 69 Fitch, Samuel, 85–86 flexibility, 27, 73, 147 flexibilization, 56, 75, 111–113, 122, 124, 174, 214–215 foreign currency sales, 60 fraud, electoral. see election fraud Fujimori, Alberto, 4, 12, 19, 20, 36, 170, 189, 197–210, 216 G Galdo, Andrés, 137 Gaviria, César, 152 general strikes, 38, 42–43, 45, 48, 134, 152, 155 ghost coalitions, 76–78, 84, 88 Goertz, Gary, 37 González Macchi, Luís Ángel, 157 Goulart, João, 36 “great turnaround, the” (el gran viraje), 133–134 growth, economic, 35, 38, 45–48, 55, 62, 86, 184–187, 221 Guatemalan Christian Democratic Party (Democracia Cristiana Guatemalteca, DCG), 167, 168, 172, 176, 219 Guatemalan National Revolutionary Union (Unidad Revolucionaria Nacional Guatemalteca, URNG), 173, 176–177

261

Guatemalan Republican Front (Frente Republicano Guatemalteco, FRG), 172 Gutiérrez, Lucio, 4, 10, 20, 24–25, 74, 76, 83–84, 86–87, 218 H Hadenius, Axel, 202 Hartlyn, Jonathan, 187 Hausman, Ricardo, 133 Historical Revolutionary Nationalist Movement (Movimiento Nacionalista Revolucionario Histórico, MNRH), 95, 96 Hochstetler, Kathryn, 7, 8, 42, 94, 106, 121, 158, 191, 204, 223 Humala, Ollanta, 199, 207 human rights ombudsman (Procuraduría de Derechos Humanos, PDH), 167, 175 Hydrocarbons Law, 77 hyperinflation, 21–22, 37, 48, 55, 61–62, 95, 101, 105, 135 hyper-presidentialism, 209 I impeachment, 6, 8, 10–11, 17–18, 24 Impuesto a la Circulación de Capitales, ICC, 82 incapacity, 81–82, 96, 200, 207 independent election, 18, 28, 125 indicators, economic performance. see economic indicators indigenous peoples, 10, 73–74, 83, 85–87, 98, 171, 173, 216 inflation, 6, 8, 35, 38, 42, 45–48, 58–60, 114, 122, 184, 186–187, 221. see also hyperinflation instability, 74–76, 85–87, 101–104 coalition breakdowns and, 103–104 economic policy and, 101

262

Index

instability—Continued minority share of seats an, 101–103 Instancia, 171–173, 176 institutional design, 36, 38, 49, 67 institutional reforms, 10, 73–74, 76, 78–80, 149 Institutional Renewal Party of National Action (Partido Renovador Institucional de Acción Nacional, PRIAN), 84 International Monetary Fund (IMF), 65, 67, 96–97, 129, 141, 170, 184, 186 J Jaramillo, Fidel, 82 Jefferson, Roberto, 122 judicial independence, 84 Justicialist Party (Partido Justicialista, PJ), 22. see also Peronism K Kouri, Luis Alberto, 198 L Laíno, Domingo, 149 Law of Authentic Interpretation of the Constitution, 202 Ledezma, Antonio, 139 Left Nationalist Revolutionary Movement (Movimiento Nacionalista Revolucionario de Izquierda, MNRI), 23, 95 legislative shields, 6, 11, 37, 83–84, 147, 157, 184–185, 199, 207 legitimacy, 5, 35–36, 55, 57, 69, 94, 142, 158, 167, 171–173, 175, 197–198, 204–206, 223, 234 Lepage, Octavio, 138 Linz, Juan J., 1–2, 5–6, 7, 8, 26–28, 49, 57, 94, 166, 218, 222, 223, 225, 226

Llanos, Mariana, 1–13, 55–71, 213–228, 217, 229–237 Lula da Silva, Luís Inácio, 114, 122–123 Lusinchi, Jaime, 132–133, 136, 140 M Madrid, Raúl L, 87 Mahuad, Jamil, 4, 10, 24–25, 36, 43, 74, 82–83, 85–86, 217, 226, 232 Mainwaring, Scott, 18, 19, 35, 39, 43 Marsteintredet, Leiv, 1–12, 33–49, 75, 112, 126, 181–193, 213–227 Matos, Cecilia, 138, 142 Medina, Doria, 100 Medina, Pablo, 135 Mejía, Hipólito, 193 Mejía Acosta, Andrés, 10, 73–88, 219, 223 Menem, Carlos, 8, 21, 57, 59–65, 67, 69, 135, 156, 226 Mesa, Carlos, 4, 10, 20, 24–26, 91–92, 98–101, 103, 104–106 midterm elections, 10, 55–58, 59–65, 184, 193, 214 Alfonsín and, 59–62 de la Rúa and, 64–65 presidential survival and, 56–58 military intervention, 36, 42–43, 85, 88, 94, 115, 227, 232 military rule, 111, 116, 167–168 minority governments, 5, 7, 10, 23, 36–37, 56–57, 67–68, 74, 91, 101, 185 mobilization, civil society and mobilization, political. see also social protests, streets, the, 9, 10, 25, 36–39, 42, 45, 48–49, 56, 68, 74, 83, 85–87, 94, 98–99, 134, 154–158, 160, 170, 188, 200, 226

Index

263

Montesinos, Vladimiro, 198–199, 201–202, 204–209 Mora Padilla, Indira Josefina, 142 Morales, Evo, 97–98, 100–101, 103, 218 Movement toward Socialism (Movimiento al Socialismo, MAS), 97–101, 103, 105, 134–135, 138, 141–142, 165 Multisectorial Forum (Foro Multisectorial), 170–173 Mustapic, Ana María, 17–30

135, 141, 165, 169, 184–185, 201, 226 New Republican Force (Nueva Fuerza Republicana, NFR), 97–100, 103–104 Noboa, Álvaro, 82, 84 Noboa, Gustavo, 25, 83, 85 no-confidence votes, 27, 73, 126, 223 Nohlen, Dieter, 39, 41 Nolte, Detlef, 11, 147–161 “notables strategy,” 137–138

N Naim, Moises, 133 National Advancement Party (Partido de Avanzada Nacional, PAN), 172, 177 National Centrist Union (Unión del Centro Nacional, UCN), 168, 172, 219 National Democratic Action (Acción Democrática Nacional, ADN), 23, 95–97, 103 National Democratic Institute (NDI), 183, 187 National Election Board (Junta Nacional Electoral, JNE), 203, 209 National Office of Electoral Procedures (Oficina Nacional de Procedimientos Electorales, ONPE), 203–204, 209 National Unity Front (Frente de Unidad Nacional, FUN), 100 National Unity of Ethical Colorados (Unidad Nacional de Colorados Éticos, UNACE), 152, 159 Nationalist Democratic Action (Acción Democrática Nacionalista, ADN), 23, 95–97, 103 Negretto, Gabriel L., 7, 37 neoliberalism, 8, 48, 55, 60, 62, 94, 97, 101, 129–130, 133,

O opposition parties, 23, 74–77, 80, 84, 88, 99, 103, 105, 112, 114, 120–122, 135, 153–156, 186 Organization of American States (OAS), 34, 35, 152, 171, 187, 189, 192, 199, 201, 204, 206, 209, 216, 234–235 Oviedo, Lino, 147, 150–160, 234 P Pachakutik Plurinational Unity Movement (Movimiento Unidad Plurinacional Pachakutik, MUPP), 83–84 Pact for Democracy (Pacto por la Democracia), 22, 181 “pacted democracy” arrangement, 96–97 Palacio, Alfredo, 25, 76, 84–86, 218 Paniagua, Valentin, 199, 200, 208–209 Paredes, José Luís, 100 Party of Dominican Liberation (Partido de la Liberación Dominicana, PLD), 22–23, 183, 186–189, 193 party switching, 79–80, 88 partyarchy (partidocracia), 11, 129–143

264

Index

Paz Estensoro, Victor, 8, 93, 101, 226 Paz Zamora, Jaime, 23, 95, 100, 103–104, 218 Peña Gómez, José Francisco, 23, 181, 183, 193 People’s Electoral Movement (Movimiento Electoral del Pueblo, MEP), 132, 135 Pérez, Carlos Andrés (CAP), 4, 6, 8, 11, 20, 24, 129–143, 217, 226 impeachment, 136–140 return of, 133–136 Pérez de Cuellar, Javier, 208 Pérez Perdomo, Rogelio, 139–140 Pérez-Liñán, Aníbal, 2, 6–7, 18, 19, 35–37, 39, 43, 55–58, 73, 75, 81–82, 83, 86, 94, 115, 140, 157, 191–192, 200, 221, 224 Peronism, 22, 55, 57–62, 64–69 Peronist Party. see Justicialist Party (Partido Justicialista, PJ) Pietri, Arturo Uslar, 137, 139 Piñerúa, Luís, 132 Pinochet, Augusto, 28 polarization, political, 5, 36, 38, 42, 82, 168, 233 Popular Action Party (Partido de Acción Popular, PAP), 199, 208 Popular Democracy (Democracia Popular, DP), 82, 85 pork, budgets and, 68, 75–76, 78 presidential breakdowns, 3–8 Presidential Military Guard (Estado Mayor Presidencial, EMP), 167, 170 presidentialism, 1–2, 5–6, 9, 17–30, 36–37, 49, 56, 111, 113, 124, 126, 200, 209, 213, 222–224, 226–227

privatization, 59–60, 62, 83, 93, 97, 101, 114, 133, 170 public opinion, 123, 130, 140, 154 Public Sector Budgets Law, 77 Pugliese, Juan Carlos, 60 Punto Fijo Pact (1958), 131, 134 purgables, 28 Q Quispe, Felipe, 98 R Radical Civic Union (Unión Cívica Radical, UCR), 22, 55, 59–70 Radicalism. see Radical Civic Union radicalization, 49 Rangel, José Vicente, 137 Real Plan, 123 “recalcitrant” legislatures, 57, 60, 112 regime institutionalization, 148 Republican Union Party (Partido Unión Republicana, PUR), 77 Requena, Mario, 98 resignations, presidential. see also exits from power, 2, 9, 12, 19–30, 55, 59, 61, 66–69, 73, 82–83, 91, 94, 96, 98, 100, 103, 105, 155, 157, 172, 186–190, 198, 200, 205–207, 216, 218, 222–224, 229, 231 Revolutionary Armed Forces of Colombia (Fuerzas Armadas Revolucionarias de Colombia, FARC), 204 Revolutionary Bolivarian Movement 200 (Movimiento Bolivariano Revolucionario, MBR-200), 135 Revolutionary Left Movement (Movimiento de Izquierda Revolucionaria, MIR), 23, 95–100, 103–104, 132 Reyes Villa, Manfred, 97–98

Index

265

Rodríguez, Andrés, 148–150, 190 Rodríguez, Eduardo, 101 Rodríguez, Miguel, 133 Rodríguez Saa, Adolfo, 9, 19, 20, 21, 25, 56, 57, 66 Roldosista Party (Partido Roldosista Ecuatoriano, PRE), 81

Stroessner, Gustavo, 28, 148–151, 159 Superior Institute of Administration Studies (Instituto de Estudios Superiores de Administración, IESA), 133 Supreme Court. see courts

S Salom, Escovar, 137–138 Sánchez de Lozada, Gonzalo, 4, 10, 20, 24–26, 91–92, 96–99, 100, 101, 103–106, 158, 226, 232 semi-democracy, 19, 38, 43, 221 Serrano Elías, Jorge, 4, 11, 20, 24–26, 28, 165–178, 216, 218, 219 “signatures scandal,” 203 Siles Zuazo, Hernán, 4, 8, 10, 20, 21–23, 91–92, 94–96, 101–105, 217, 226 Skol, Michael, 188–189 Social Christian party (Partido Social Cristiano, PSC), 77–78, 81–85 Social Christian Reformist Party (Partido Reformista Social Cristiano, PRSC), 22–23, 181, 183, 186, 193 social protests, 10, 55–56, 87, 91–106 Solares, Jaime, 100 Solidarity Front (Frente País Solidario, FREPASO), 62–64, 66 Sourrouille, Juan, 58–59, 60 Spring Plan, 60 State Modernization Law, 77 “steam roller” (aplanadora) coalition, 82–83 streets, the. see also mobilization, social protests, 6–7, 213–224, 233

T taxation, 60, 64, 77, 81–82, 97–99, 133, 141 Tello, Benitez, 137 Teorell, Jan, 202 term limits, 74, 76, 78 third wave of democratization, 1, 3, 18–19, 27, 34, 38, 41, 55, 75, 101, 148, 218, 233, 236 Toledo, Alejandro, 200, 203–204, 209 Trujillo Molina, Rafael Leonidas, 182, 190, 192 Tudela, Francisco, 199–200, 207 U Ucero, Alfaro, 139 Unified Revolutionary Directorate (Dirección Revolucionaria Unificada, DRU), 96 unilateralism, presidents and, 59, 63, 67–68, 112, 116, 219 V Vaca Diez, Hormando, 100 Value-Added Tax (VAT), 81, 82, 141 Velásquez, Ramón J., 134, 138, 140–141 Velásquez, Vásquez, 231, 234 Venezuelan Communist Party (Partido Comunista Venezolano, PCV), 131–132 Voluntary Civilian Self-Defencse Committees (Patrullas de Autodefensa Civil, PAC), 175

266

Index

W Wasmosy, Juan Carlos, 150 Weyland, Kurt, 36, 86, 121 Workers’ Party (Partido dos Trabalhadores, PT), 114, 122–123 World Development Indicators (WDI) database, 39, 46

Y Youths for Democracy movement (Jóvenes por la Democracia, JD), 155–156 Z Zelaya, Manuel, 229–236