Florida State University Libraries
Electronic Theses, Treatises and Dissertations The Graduate School
2006 Leaving the Past Behind?: a Study of the FMLN and URNG Transitions to Political Parties Michael E. Allison
Follow this and additional works at the FSU Digital Library. For more information, please contact [email protected]
THE FLORIDA STATE UNIVERSITY
COLLEGE OF SOCIAL SCIENCES
LEAVING THE PAST BEHIND? A STUDY OF THE FMLN AND URNG
TRANSITIONS TO POLITICAL PARTIES
By
MICHAEL E. ALLISON
A Dissertation submitted to the Department of Political Science in partial fulfillment of the requirements for the degree of Doctor of Philosophy
Degree Awarded: Spring Semester, 2006
Copyright © 2006 Michael E. Allison All Rights Reserved The members of the Committee approve the dissertation of Michael E. Allison defended on 03/21/2006.
______Paul R. Hensel Professor Directing Dissertation
______Andrew Opel Outside Committee Member
______Damarys Canache Committee Member
______Will H. Moore Committee Member
______Jeffrey K. Staton Committee Member
Approved:
______Dale L. Smith, Chair, Department of Political Science
The Office of Graduate Studies has verified and approved the above named committee members.
ii
To those who have given their lives in the pursuit of social, economic, and political justice in Central America and around the world.
iii ACKNOWLEDGEMENTS
My dissertation would not have been possible without the help of a great number of individuals. First and foremost, I would like to thank the members of my committee: Paul Hensel, Damarys Canache, Will Moore, Jeff Staton, and Andy Opel. Their support and insightful comments were invaluable at each stage of the dissertation process. Their contributions extend well beyond my dissertation. I am especially grateful to Paul Hensel. His mentoring and friendship were extremely important to the realization of this dissertation. I must thank Kűrşan Turan and Steve Shellman for their encouragement and advice all these years. I would also like to acknowledge the support of the faculty and staff at FSU, particularly Dale Smith and Mary Schneider. For encouraging me to pursue graduate studies in Latin American politics, I must thank Ed Dew and Bev Kahn. Special thanks to my family and friends for their emotional and financial support all these years.
Finally, I would like to thank my wife, Marie. This dissertation would not have been possible without your love, support, confidence and patience.
iv TABLE OF CONTENTS
List of Tables………………………………………...…………………………………….... vi
List of Figures……………………………………………………………………………….. viii
List of Maps…………………………………………………………………………………. ix
Abbreviations and Acronyms……………………………………………………………….. x
Abstract……………………………………………………………………………………… xiii
1. INTRODUCTION……………………………………………………………………….. 1
2. EXPLAINING THE SUCCESS OF FORMER INSURGENT GROUPS………………. 12
3. RESEARCH DESIGN…………………………………………………………………… 33
4. INSURGENT GROUPS…………………………………………………………………. 54
5. EVALUATING THE IMPACT OF ORGANIZATIONAL CHARACTERISTICS……. 76
6. EVALUATING THE IMPACT OF ENVIRONMENTAL CONDITIONS…………….. 109
7. CIVIL WAR LEGACY ON FMLN ELECTORAL PERFORMANCE………………… 147
8. GLOBAL INSURGENT TRANSITIONS TO POLITICAL PARTIES………………… 178
9. CONCLUSION………………………………………………………………………….. 233
APPENDICES………………………………………………………………………………. 242
REFERENCES……………………………………………………………………………… 245
BIOGRAPHICAL SKETCH………………………………………………………………... 257
v LIST OF TABLES
4.1 Postwar Presidential Elections in El Salvador (Votes and % of Total Votes)…………. 69
4.2 Postwar Legislative Elections in El Salvador (Votes and % of Total Votes)………….. 70
4.3 Postwar Legislative Elections in El Salvador (Number of and % of Seats)…………… 71
4.4 Postwar Presidential Elections in Guatemala (Votes and % of Total Votes)………….. 72
4.5 Postwar Legislative Elections in Guatemala (Votes and % of Total Votes)…………… 73
4.6 Postwar Legislative Elections in Guatemala (Number of and % of Seats)…………….. 74
4.7 Effective Number of Parties in El Salvador and Guatemala…………………………… 75
6.1 1994 Legislative Election Results for FMLN in El Salvador, Largest Remainder 137 versus D’Hondt…………………………………………………………………………
6.2 1999 Legislative Election Results for URNG in Guatemala, D’Hondt versus Largest 138 Remainder………………………………………………………………………………
6.3 1994 Legislative Election Results for FMLN in El Salvador, Actual and Simulated 139 District Magnitude……………………………………………………………………...
6.4 1999 Legislative Election Results for URNG in Guatemala, Actual and Simulated 140 District Magnitude……………………………………………………………………...
6.5 National Vote in Salvadoran Elections (Percentages)…………………………………. 141
6.6 National Vote in Guatemalan Elections (Percentages)………………………………… 142
6.7 Bipolarization of the Salvadoran Political Party System, Legislative and Executive 143 Elections………………………………………………………………………………...
6.8 Bipolarization of the Guatemalan Political Party System, Legislative and Executive 144 Elections………………………………………………………………………………...
6.9 Growth Rate of Real GDP per Capita, El Salvador……………………………………. 145
6.10 Growth of Rate of Real GDP per Capita, Guatemala…………………………………. 146
7.1 FMLN Performance in 1994 “Elections of the Century”………………………………. 171
vi 7.2 FMLN-Controlled Zones and FMLN Performance in the 1994 Elections…………….. 172
7.3 Distribution of Political Violence and FMLN Performance in the 1994 Elections……. 173
7.4 FMLN Performance in “Ex-Conflict” and “Non-Conflict Zones” in the 1994 174 Elections………………………………………………………………………………...
7.5 Accounting for FMLN Performance in the 1994 Elections……………………………. 175
8.1 Postwar Presidential Elections in Mozambique (Votes and % of Total Votes)………... 225
8.2 Postwar Legislative Elections in Mozambique (Votes and % of Total Votes)………… 226
8.3 Postwar Legislative Elections in Mozambique (Number of and % of Seats)………….. 227
8.4 Postwar Presidential Elections in Colombia (Votes and % of Total Votes)…………… 228
8.5 Postwar Legislative Elections for Colombian Senate (Number of and % of Seats)…… 229
8.6 Postwar Legislative Elections for Colombian Chamber of Representatives (Number of 230 and % of Seats)…………………………………………………………………………
8.7 Postwar Legislative Elections for Colombian Constituent Assembly (Number of and 231 % of Seats)……………………………………………………………………………...
8.8 Global Insurgent Group Electoral Performance………………………………………... 232
vii LIST OF FIGURES
7.1 FMLN Vote in the 1994 Legislative Election………………………………………….. 176
7.2 Distribution of Political Violence and FMLN Performance in the 1994 Election……... 177
viii LIST OF MAPS
4.1 El Salvador, Perry-Castañeda Library Map Collection, 1980………………………….. 67
4.2 Guatemala, Perry-Castañeda Library Map Collection, 2000…………………….……... 68
8.1 Mozambique, Perry-Castañeda Library Map Collection, 1995………………………… 223
8.2 Colombia, Perry Castañeda Library Map Collection, 2001…………………………….. 224
ix LIST OF ACRONYMS AND ABBREVIATIONS
El Salvador
ARENA Alianza Republicana Nacionalista National Republican Alliance
BPR Bloque Popular Revolucionario Popular Revolutionary Bloc
CD Convergencia Democrática Democratic Convergence
CDU Centro Democrático Unido United Democratic Center
CPM Coordinora Político Militar Political-Military Committee
DRU Dirección Revolucionario Unificada Unified Revolutionary Directorate
ERP Ejército Revolucionario del Pueblo People’s Revolutionary Army
FAL Fuerzas Armadas de Liberación Armed Liberation Forces
FAPU Frente de Acción Popular Unificada United Popular Action Front
FARN Fuerzas Armadas de Resistencia Armed Force of National Resistance Nacional
FDR Frente Democrático Revolucionario Democratic Revolutionary Front
FMLN Frente Farabundo Martí para la Farabundo Martí National Liberación Nacional Liberation Front
FPL Fuerzas Populares de Liberación Farabundo Martí Popular Liberation Farabundo Martí Forces
LP-28 Ligas Populares – 28 de Febrero February 28th Popular Leagues
MLP Movimiento de Liberación Popular Popular Liberation Movement
MNR Movimiento Nacional National Revolutionary Movement Revolucionario
MPSC Movimiento Popular Social Popular Social Christian Movement Cristiano
MR-13 Movimiento Revolucionario 13 de November 13th Revolutionary Noviembre Movement
x PAR Partido Acción Renovadora Renovating Action Party
PCS Partido Comunista de El Salvador Communist Party of El Salvador
PDC Partido Demócrata Cristiano Christian Democratic Party
PSD Partido Demócrata Social Social Democratic Party
PRTC Partido Revolucionario de Revolutionary Party of Central Trabajadores Centroamericanos American Workers
RN Resistencia Nacional National Resistance
UCA Universidad Centroamericana “José University of Central America “José Simeón Cañas” Simeón Cañas”
UDN Unión Democrática Nacionalista Democratic Nationalist Union
UNO Unión Nacional Opositora National Opposition Union
Guatemala
AD Alianza Democrática Democratic Alliance
ANN Alianza Nueva Nación New Nation Alliance
ASC Asamblea de Sociedad Civil Assembly of Civil Society
CCPP Comisión Permanente de Permanent Commission of Refugees Refugiados
CNT Confederación Nacional de National Workers Confederation Trabajadores
CUC Comité de Unidad Campesina Committee of Peasant Unity
DIA Desarrollo Integral Auténtico Authentic Integral Development
EGP Ejército Guerrilleros de los Pobres Guerrilla Army of the Poor
FAR Fuerzas Armadas Rebeldes Rebel Armed Forces
FASGUA Federación Autónoma Sindical de Autonomous Trade Union of Guatemala Guatemala
xi FDNG Frente Democrático Nueva Democratic Front for a New Guatemala Guatemala
FRG Frente Republicano Guatemalteco Democratic Revolutionary Front
GAP Grupo de Apoyo Popular Popular Support Group
MDC Movimiento Democrático Civíco Democratic Civic Movement
ORPA Organización del Pueblo en Armas Organization of the People in Arms
PGT Partido Guatemalteco del Trabajo Guatemalan Workers’ Party
PR Partido Revolucionario Revolutionary Party
PSC Partido Social Cristiano Social Christian Party
UD Unión Democrática Democratic Union
UNE Unidad Nacional de la Esperzanza National Union of Hope
UNID Unidad de Izquierda Democrática Democratic Leftist Union
URNG Unidad Revolucionaria Nacional Guatemalan National Revolutionary Guatemalteca Unit
Other
AD M-19 Alianza Democrática M-19 Democratic Alliance April 19th Movement
ANAPO Alianza Nacional Popular National Popular Alliance
FARC Fuerzas Armadas Revolucionarias de Revolutionary Armed Forces of Colombia Colombia
FSLN Frente Sandinista de Liberación Sandinista National Liberation Front Nacional
FRELIMO Frente de Libertaçâo de Moçambique Liberation Front of Mozambique
M-19 Movimiento 19 de Abril April 19th Movement
RENAMO Resistëncia Nacional Moçambicana Mozambique National Resistance
UP Unión Patriótica Patriotic Union
xii ABSTRACT
This project focuses on the aftermath of civil war. Leaving the Past Behind? A Study of the URNG and FMLN Transitions to Political Parties asks a fundamental question related to the transition of insurgent groups to political parties that has not yet been investigated. Namely, how can we account for the success or failure of former insurgent groups as political parties? I argue that certain insurgent groups are more likely to succeed as political parties as a result of a variety of organizational factors (size, prior political party experience, contacts with mass-based organizations, internal unity, control of territory, and minimal violence committed against civilians) and environmental conditions (electoral rules, other political parties, voters, economic crises, and social cleavages).
To test these hypotheses, I combine qualitative and quantitative analyses of two Central American insurgent groups currently undergoing such transitions: the Farabundo Martí National Liberation Front (FMLN) in El Salvador and the Guatemalan National Revolutionary Unit (URNG) in Guatemala. First, I conduct a structured, focused comparison of the FMLN and URNG and their transitions to political parties. Second, I conduct a number of statistical analyses utilizing electoral results, political violence and socio-economic data to determine whether the explanations also hold for within nation analyses. Finally, I pursue the generalizability of the findings through a limited comparison of similar insurgent transitions in Mozambique and Colombia.
Findings from my research suggest that the FMLN drew upon previous experience with electoral politics both prior to and during the war, an extensive urban and rural network of combatants, and a nationwide organization of noncombatants to succeed in postwar electoral politics. On the other hand, the URNG had no prior experience with electoral politics, a weak organizational structure with few noncombatant support personnel, and limited geographic scope. While institutional factors played a marginal role in effecting the electoral performance of each group, it was the FMLN’s experience and organization as an insurgent group that explains its superior performance relative to the URNG.
xiii CHAPTER 1 INTRODUCTION
Over the last several decades, a number of civil wars have ended as a result of negotiated settlements. These negotiated settlements typically involve not only a cessation of armed conflict and the disarmament and demobilization of former combatants, but social, economic, political, and military reforms. The settlements are designed to facilitate the transformation of a system based upon violent political conflict between armed groups to one based upon nonviolent political conflict between competing political parties. Unlike those civil wars that end through military victory, these negotiated settlements do not allow either the government or the insurgent group to claim outright victory and impose its will upon the defeated. Instead, “the issue of determining who governs must be resolved as part of the peace process” (Garber 1998: 2). Elections are frequently used to determine who will rule in these post-civil war societies because they are
supposed to transform a violent conflict into a nonviolent one: ballots take the place of bullets. They are expected to enable the former warring parties to pursue their conflicting ideologies and programs in a peaceful fashion. Elections give all factions an opportunity to present their agendas to the citizens, debate with their opponents, and mobilize public opinion to capture political power. Like other elements of a democratic system, elections contribute to the institutionalization of a conflict resolution mechanism in the body politic (Kumar 1998: 7).
While these are all noble goals of postconflict elections, to date we know very little about either how successful former warring groups have been in adapting to the challenges of nonviolent conflict, or why some groups seem to have performed much better than others. In this dissertation, I ask the following question: how can we account for the success or failure of former insurgent groups as political parties? Previous research on insurgent groups and civil war resolution provides us with a number of possible avenues of research through which to explain the performance of these groups as new political parties. For instance, in an analysis of postconflict Sudan, Zimbabwe, Chad, and
1 Lebanon, Atlas and Licklider (1999) find that violence following negotiated settlements tends not to break out between civil war adversaries, but instead among groups that were once former allies of the insurgency. Coalitions that were once necessary to confront the former regime militarily are not always necessary or manageable once the violence has ceased and that is where we often find violent conflict in the postwar period. But while some postwar periods have been marred by new waves of political violence, there also exists instances in which the former insurgent groups have incorporated themselves into political systems undergoing a process of democratic transition. One of the questions we should then ask relates to whether the unity maintained by insurgent groups also breaks down when making the transition to political party, and how this affects their electoral performance. In a study on Venezuela’s Movement Towards Socialism (MAS), Steve Ellner (1988) traces the history of that group from its roots within the Venezuelan communist party to its insurgent history in the 1960s and finally to its emergence as the third largest political party in the early 1970s. Ellner focuses upon the internal ideological and strategic debate within the Venezuelan left, the MAS’ movement from the left to the right of the political spectrum, and its inability to recruit individuals from outside its core constituents. It is quite possible that the insurgent coalition that was sustained during the civil war broke down in the postwar period as disagreements emerged related to ideology and electoral strategy. The experience of MAS also brings attention to the need for insurgent groups to develop supporters beyond their core constituents if they wish to be competitive as political parties. Both Matthew Shugart and Jeffrey Ryan are more directly concerned with democratic transitions and insurgent groups. Shugart (1992) focuses on the extent to which negotiations to “exit” civil wars reflect how the government and the guerrillas calculate the costs of continued conflict versus the costs of electoral competition. Negotiated settlements to civil wars require that the “rebels’ costs of participation must be lower than their costs of resistance” while the “regime’s costs of toleration must be lower than its costs of suppression” (1992: 126). Although Shugart says he in interested in “the process by which former rebels become significant players within electoral institutions” (1992: 121), he accepts as “relevant” any political party that participates in congressional elections. Shugart’s article provides important insights into the calculations on the parts of the regime and the rebels that might potentially lead to a negotiated settlement, particularly conflict over the electoral rules, but it does not provide much in the way
2 of explaining why the level of electoral success among former insurgent groups varies. Given his concern with political institutions, we have an important area of inquiry to follow. Like Shugart, Ryan (1994) is also interested in the impact of democracy on civil war outcomes. While Shugart focuses on the importance of democratic institutions to negotiated settlements for both the regime and the rebels, Ryan focuses more generally on “how the process of democratic transition affects revolutionary groups” (1994: 27). He argues that revolutionary movements possess a preferred set of outcomes ranging from least preferred (“destruction of movement” and “negotiated surrender”) to most preferred (“negotiated settlement” and “overthrow of the regime”), and that the likelihood of achieving each one of these outcomes is conditioned upon “the scope of the revolutionaries’ coalition, in terms of both size and, perhaps more important, sociopolitical breadth” (29). Revolutionary movements with “narrow” support are unlikely to overcome the challenges of the transition and will most likely be destroyed or surrender. Revolutionary movements with relatively broad-based coalition support prior to the beginning of the state’s transition will be more likely to achieve a negotiated settlement. Ryan’s article is important in two ways. First, it sets out how democratization is likely to affect the outcome of revolutionary movements (overthrow of the regime, negotiated settlement, negotiated surrender, destruction of movement). Second, it expands an argument that a broad-based coalition is necessary not only for the successful overthrow of an authoritarian regime, but also for revolutionary movements to survive the onset of democratization. On the other hand, it is not at all clear how broad-based a coalition is necessary for a revolutionary group to successfully achieve either of its most preferred outcomes. This is an important weakness, because the breadth of the broad-based coalition is Ryan’s critical factor. Furthermore, Ryan does not extend his argument into the postwar period or explore the extent to which a broad-based coalition is useful for the insurgent or revolutionary group as a political party. In another work related to democracy and insurgent groups, Boudon (1997) contrasts the failure of the April 19th Movement in Colombia with the success of the Causa Radical in Venezuela. Prior to their emergence as political parties, both had origins in guerrilla insurgencies. Boudon attributes the success of Causa Radical to its extensive organizational structures, an ideological leader, and economic crisis in Venezuela. The April 19th failure, after initial but limited success, is explained by the opposite conditions – weak organizational structures, the lack of an ideological leader and the lack of economic crisis in Colombia (see also
3 Boudon 2001). Boudon identifies some crucial factors for new party success, but does not really focus on the extent to which each group’s insurgent formation and performance influenced its subsequent electoral performance. Perhaps this is because neither the April 19th Movement nor the Causa Radical had an extensive insurgent history. Vinegrad’s (1998) approach to the success of former insurgent as political parties is most closely related to the present study. In this chapter, Vinegrad focuses upon the transition of the URNG in Guatemala to political party, in comparison to the transitions of the FMLN in El Salvador, the Communist Party of Venezuela, and the April 19th Movement in Colombia. Vinegrad briefly mentions some “general issues in guerrilla transition” such as the group’s origins and history, problems of internal democracy, the role of ideology, wider systemic issues such as security and the strength of party identification. Unfortunately, the analysis is conducted prior to the URNG’s initial electoral performance (leading to excessive optimism) and devotes little attention to the divisions that occurred within the URNG throughout the 1980s (which severely weakened the URNG prior to the Firm and Lasting Agreement), the extent of repression suffered by the URNG, or potential institutional challenges to new party success in Guatemala. I believe these three factors are crucial to understanding the electoral performance of the URNG. Finally, in a book concerned primarily with the role of gender in the Central American revolutionary Left, Luciak (2001) also explores the transitions to political parties experienced by the FMLN, the URNG, the Frente Sandinista de Liberación Nacional (FSLN), or Sandinista National Liberation Front in Nicaragua, and the rightist Resistencia Nicaragüense (RN), or Nicaraguan Resistance, more commonly referred to as the Contras. While Luciak’s focus is on the role of gender, this book is also a very good historical source on these groups as political parties. Though Luciak does not offer an explanation for the relative success or failure of these respective guerrilla groups, he does conclude that the revolutionary Left in Central America needs to continue to democratize internally and to deal more effectively with internal dissent if these parties are to become viable political options. While internal democracy is likely important, it is unclear how critical internal democracy is to explaining the variation in the success of these groups. The performance of the FSLN, FMLN, URNG and RN has varied significantly even though each group has struggled with efforts at internal democracy. In addition, it cannot be argued that the lack of internal democracy in the ruling parties of El Salvador, Guatemala, and Nicaragua has prevented their successful electoral performance.
4 Much of the previous research has addressed some of the issues involved in insurgent transitions to political parties, but with several limitations. First, aside from Boudon (1997), research has not been explicit in conceptualizing what it means for these groups to be successful. There are several ways in which we can discuss the performance of these groups as political parties and it is important that we make explicit just what it is we portend to explain. Are we discussing the group’s success in terms of persistence, winning votes, capturing seats, institutionalization, or policy impact? Second, much of the research tends to focus on one explanation for the success of these new political parties at the expense of others. For example, Shugart (1992) stresses the role of institutions and electoral rules at the expense of the group’s history. Vinegrad (1998), on the other hand, discusses the group’s history and goals, but disregards the importance of institutional explanations. Boudon (1997) addresses the importance of institutions, leaders, and organization. While certainly much more comprehensive, Boudon pays little attention to how the civil war experience conditioned the performance of former insurgent groups as political parties. Again, the performance of these particular groups cannot simply be explained by focusing solely upon a group’s history or the institutional challenges it must overcome. Explaining the success of former insurgent groups as political parties can only be explained through taking into account both organizational factors and environmental conditions.
Theory and Hypotheses In this dissertation, I develop several organizational and environmental explanations of insurgent group success as political parties. I argue that all insurgent groups are not equal and that certain groups are better prepared for the demands of electoral politics than others. Therefore, some are more likely to be successful than are others. Some are much stronger militarily and count on thousands of well-armed combatants. Others may not be as militarily proficient and instead rely more heavily upon an extensive network of non-combatants and/or a political wing. Some insurgent groups are besieged by infighting while others remain relatively unified throughout their existence. Likewise, some insurgent groups are very violent toward civilians while others use force more discriminately. Finally, some insurgent groups maintain a national presence while others are relegated to isolated areas of a country. I argue that larger insurgent groups, those with prior political experience, extensive mass-based contacts, and
5 internal unity, those that control extensive territory, and those that generally do not target civilians are more likely to be successful than insurgent groups with the opposite characteristics – few combatants, minimal political experience, few ties to mass-based organizations, internal divisions, a very localized insurgent presence, and history of targeting civilians. Consequently, the insurgent histories of these groups and their performances during the conflict are likely to be at least as important to explaining their performance as political parties as those issues we typically associate with new political parties. While focusing solely upon former insurgent groups can go a long way towards explaining their performance as political parties, explaining their success or failure would be incomplete without discussing the political system in which they participate. Environmental conditions can inhibit or benefit these groups as political parties. One would expect former insurgent groups to be more successful in parliamentary systems than in presidential systems. If competing in presidential systems, former insurgent groups should tend to be more successful in systems employing majority voting rather than plurality voting. Former insurgent groups are also more likely to be successful when elections for the legislature are determined by proportional representation electoral rules rather than single member district voting, district magnitudes are high, and there are low barriers to entry. The former insurgent group’s success will also vary depending upon how likely voters are to support a new political party. The likelihood that a new party will succeed depends upon whether the voters see a problem with the existing alternatives and are willing to support a new political party. A new political party is also likely to succeed if bipolarization is weak, there is a pool of uncommitted voters, the economy is weak, a political party with a similar platform does not exist, and, if it is an identity-based insurgent group, the social group it seeks to represent comprises a significant percentage of the population. In summary, a former insurgent group is not entirely in control of its own destiny. The characteristics of the insurgent group and its performance during the war are key explanatory factors, but so are the characteristics of the political system in which the new party has to compete with some more likely to inhibit the success of new parties and others more likely to facilitate their success.
6 Research Design I test the hypotheses, briefly outlined above, through qualitative and quantitative comparative analyses of two insurgent groups in Central America. I conduct a most-similar research design. In this most-similar research design, I select one example of an insurgent group that has successfully transformed itself into a minor party and an example of an insurgent group that has successfully transformed itself into a major party. After more than a decade of violence, the Frente Farabundo Martí para la Liberación Nacional (FMLN), or Farabundo Martí National Liberation Front, in El Salvador and the Unidad Revolucionaria Nacional Guatemalteca (URNG), or Guatemalan National Revolutionary Unit, in Guatemala signed negotiated settlements with the government of their respective countries. Since the conclusion of the civil war in El Salvador in 1992, the FMLN has participated in three elections for the presidency and four elections for municipal and legislative office. After ten years, the FMLN has successfully transformed itself into the largest party in the country’s Legislative Assembly, captured the important position of the mayor of the capital in three successive elections, and finished second in three presidential elections. On the other hand, the URNG in Guatemala has not fared nearly as well. Since the end of the civil war in Guatemala in 1996, the URNG has competed in two elections for executive, legislative and municipal offices. After finishing in third place in the presidential and legislative elections in 1999, the URNG fell to sixth place for the presidency and tied for seventh for the legislature in the 2003 elections. As a result, the URNG narrowly avoided losing its legal standing when it received just over four percent of the national vote. In order to lend greater validity to the findings of the case studies, I also conduct statistical analyses of the 1994 election in El Salvador. These are the first elections of the postwar in which the FMLN competed. Using this election, I explore the extent to which the hypotheses developed in Chapter 2 to explain electoral performance across insurgent groups also explain the performance of a single insurgent group across political subunits. We can have greater confidence in the hypothesized relationship if, in addition to explaining the different outcomes of the FMLN and URNG, the relationship also holds for explaining the performance of the FMLN within El Salvador. I am also concerned with the generalizability of the findings. The FMLN and the URNG are not the only insurgent groups to have made the transition from armed opposition group to political party. Therefore, I compare the performance of these two groups to insurgent
7 transitions in Mozambique and Colombia. In Mozambique, the Resistëncia Nacional Moçambicana (RENAMO), or Mozambican National Resistance, signed a negotiated settlement in 1992 and has since gone on to become the second largest party in the country. In Colombia, the Movimiento 19 de Abril (M-19), or April 19th Movement, accepted a government amnesty, disarmed, demobilized, and transformed itself into a strong challenger to the two historically dominant Colombian political parties. I make use of multiple sources of data in order to approach the questions posed through both qualitative and quantitative methods. I utilize include primary and secondary sources published in English and Spanish, fieldwork notes from El Salvador and Guatemala, interviews I conducted in Guatemala and El Salvador, my experiences as an international election observer in El Salvador, and socio-economic and electoral data from both countries.
Significance The purpose of my research is to answer why two organizations, after having successfully negotiated their reinsertion into society as legal actors, have fared so dissimilarly as political parties. I believe that this research is important for a number of reasons. First, civil wars resolved through negotiations are regular occurrences in the world today. In addition to El Salvador and Guatemala, negotiated settlements have ended civil wars in such varied countries as Nicaragua, Northern Ireland, Cambodia, Lebanon, Liberia, and South Africa. While negotiated settlements address a number of social, political, economic and international conditions that contributed to the outbreak of civil war, rarely do they resolve the issues entirely. Instead, most negotiated settlements are meant to contribute to the development of a process by which the former antagonists can work to resolve the problems plaguing the country within a generally accepted institutional framework. Elections, political parties and the democratic process are key elements in this process. So far, though, we do not know to what extent the former civil war adversaries have been successful in the postwar environment in terms of their electoral performance or whether they have contributed to a resolution of any of the issues that originally contributed to the outbreak of violence. My research will probe whether the negotiated settlements have also been able to transform what was a violent conflict between competing armed groups into a nonviolent conflict between competing political parties. If, in the future, we are to convince insurgent groups to shun violence in exchange for the opportunity to
8 compete as political parties, it would be helpful to provide examples of successful transitions. Regardless of whether one is interested in whether the former insurgent group “wins” in the postwar period, one should be concerned with the failure to reintegrate the group into society as a political party. It is quite possible that the failure of the former insurgent group to be successfully reintegrated into political society will lead to higher levels of crime and violence and perhaps even the renewed outbreak of civil war. Second, this research is also significant in that it is one of the first to study the success of new political parties that trace their roots to insurgent groups. While the study of political parties in advanced democracies has a relatively long history, only more recently has there been growing interest in the study of political parties in the democratizing countries of Latin America and post-Communist Eastern Europe. This research will not only increase our knowledge of the factors involved in the successful performance of former insurgent groups, but eventually it will permit us to compare the experiences of all these new political parties. I expect that while many of the factors previously found to influence new party success in general will also apply to these insurgent groups, they will not be sufficient. A useful explanation of insurgent group success as political party must take into consideration the insurgent history of each group. Third, my research will also bring a good deal of closure to the revolutionary periods in El Salvador and Guatemala. During the two conflicts, roughly 300,000 people died and 2,000,000 civilians were turned into internally displaced persons and refugees. The URNG and the FMLN proclaimed that they were fighting for social and economic justice; that they only took up arms when all nonviolent means of participation were forestalled by exclusionary regimes. During the 1980s, both political systems underwent political openings where the election of civilian leaders replaced military rule. Eventually, the end of the Cold War, the inability to attain a military victory, and the shear exhaustion of two societies led to negotiated settlements and the transformation of these two groups into political parties. It is important to all those who died during the conflicts, both civilians and participants, to understand how well these groups have been able to perform as political parties and why. It is also important for those who survived the war and now put their hope of social and economic transformation into the hands of these political parties.
9 Dissertation Outline In Chapter 2, I develop several hypotheses to explain the performance of former insurgent groups as political parties. These hypotheses are based upon organizational characteristics of the insurgent group and the environment in which it must compete as a political party. In Chapter 3, I address research design issues. I discuss what it means for former insurgent groups to succeed as political parties. I discuss a number of ways that previous scholars have approached the study of new party success (persistence, votes, seats, margin of victory/defeat, institutionalization, policy impact, etc.) and introduce a new measure for individual party success. I also discuss how I measure the key variables used to test the hypotheses, including my use of primary and secondary sources, fieldwork, election data, and census data. I discuss the case selection of the FMLN and URNG and the use of the most similar research design. In Chapter 4, I provide the background to the conflicts in El Salvador and Guatemala, the history of each insurgent group, and the electoral performance of each following their incorporation into the political party system. In Chapter 5, I evaluate how well the organizational hypotheses I presented in Chapter 2 and operationalized in Chapter 3 explain the relative performances of the FMLN in El Salvador and the URNG in Guatemala. In Chapter 6, I evaluate how well the environmental hypotheses explain the relative performance of the two new political parties. In Chapter 7, I conduct a statistical analysis of the FMLN’s performance in the 1994 elections for the country’s legislature. After discussing the context of the 1994 elections, I try to explain the geographic distribution of support for the FMLN at the department and municipal level through an analysis of the distribution of political violence and FMLN activity during the war. I provide a separate discussion of research design issues in the chapter. In Chapter 8, I compare the findings from Chapters 5 and 6 on the electoral performances of the FMLN and the URNG to the experiences of RENAMO in Mozambique and the M-19 in Colombia. Finally, in Chapter 9, I present a review of the dissertation and its key findings. My findings suggest that the FMLN drew upon certain experiences both prior to and during the war to succeed as a political party. The URNG, on the other hand, was unable to count on these experiences, and therefore was less successful. Environmental conditions played a marginal role in the electoral performance of each group. I also lay out a future research agenda that
10 includes the continued study of the transitions of the FMLN and URNG to political parties and also extends the analysis to a global study of insurgent transitions to political parties.
11 CHAPTER 2 EXPLAINING THE SUCCESS OF FORMER INSURGENT GROUPS
In many ways, insurgent groups confront obstacles similar to those confronting any new political party. For example, any new party needs to devise a political platform, raise money in order to cover the costs of full-time staff, advertising, and campaign materials, create an organization, and select its leaders and candidates. New political parties also have to overcome legal obstacles including the collection of signatures or the setting up of party offices in a certain number of municipalities. But political parties with roots as insurgent groups are not “new” organizations in the sense that they are starting from scratch. Some have long histories dating back several decades or more. Some trace their histories to former political parties, unions, student organizations, and peasant groups in pre-civil war days. Each insurgent organization is designed with a leadership and organizational structure conducive to conducting an irregular military campaign requiring secrecy, clandestine operations, and vertical command structures. Historically, these groups resorted to violence, relying upon extra-legal channels to raise money (selling drugs or weapons, “war taxes,” kidnapping, foreign donors). These are unlikely to be appropriate avenues of fundraising following the resolution of the war. And while each of the insurgent groups negotiated a political settlement to their respective civil wars, all insurgent groups are not equal. Some were much stronger militarily and counted on a significant number of armed combatants while others may not have been as militarily proficient, instead relying more heavily upon an extensive network of non-combatants. Consequently, the insurgent histories of these groups and their performances during the conflict are likely to be at least as important to explaining their performance as political parties as those issues we typically associate with new political parties. In the following section, I discuss several organizational factors and environmental conditions related to the performance of former insurgents as political parties. In terms of organizational factors, I argue that some insurgent groups are better prepared for the demands of electoral competition and are more likely to succeed than other insurgent groups depending upon both the characteristics of the insurgent organization and its performance during the civil war. First, larger insurgent groups are more likely to be successful than smaller insurgent groups. Second, insurgent groups with a political wing or with a history as a political party previous to
12 taking up armed resistance are more likely to be successful than insurgent groups without a political wing or previous history as a political party. Third, insurgent groups with extensive contacts with mass organizations are more likely to be successful than those with few or no contacts without mass organizations. Fourth, insurgent groups that maintain unity and coherence both during the war, the negotiation process and in the postwar are more likely to be successful than insurgent groups that fracture. Fifth, insurgent groups that control territory during the war are more likely to be successful than groups that are unable to control territory. Finally, I also believe that the manner in which the insurgent groups conduct its military campaign is an important explanation. Insurgent groups that are less violent towards civilians during the war than the government and its forces are more likely to be successful than those groups whose violence outweighs that of the government. In addition to the characteristics of the insurgent group and its performance during the war, we need to take into consideration the environment in which the new political party has to compete. It is easy to imagine that certain environments are more conducive to the success of new political parties than others. In terms of environmental conditions, I develop a series of institutional and voter-based factors expected to increase the likelihood that these new parties will be successful. First, former insurgent groups are more likely to be successful operating in a parliamentary system versus a presidential system. Second, if competing in a presidential system, insurgent groups are more likely to be successful in systems with majority runoff rather than simple majority rules. Third, insurgent groups are more likely to be successful when legislative elections are conducted via proportional representation rather than single member district voting, when district magnitude is high, and when there is a low electoral threshold. Fourth, insurgent groups are more likely to succeed when voters appear willing to support a new electoral alternative. For example, insurgent groups are more likely to be successful when a high degree of voter abstention exists, when bipolarization does not exist, the economy is poor, and there are a number of salient social cleavages dividing the population. On the other hand, insurgent groups are less likely to be successful when there is a history of high voter turnout, low electoral volatility, the economy is performing average to good, and there are no salient ethnic, religious or linguistic cleavages in society. In the remainder of the chapter, I develop the rationale behind each of these hypotheses before moving on to how I intend to test their application to the transition of insurgent groups to
13 political parties in the following chapter. These hypotheses and their application to this dissertation’s two main cases apply to chapters 5 and 6.1
Organizational Characteristics Insurgent Size Insurgent groups vary to the extent that they control national territory, how they manage that territory, the extent of their contacts with legal organizations, their level of popular support maintained during the war, and the number of armed combatants under their command. Some insurgent groups command hundreds of armed combatants while others might count on the support of tens of thousands. In these conflicts, “the total number of active rebels in many wars in which thousands of civilians have been killed (through actions of both governments and rebels) is often in the hundreds or low thousands” (Fearon and Laitin 2003: 80). One can look at the small, but successful revolutionary movements in Cuba and Nicaragua relative to the comparatively larger and better equipped, yet less successful, revolutionary movements in Guatemala and El Salvador. While a “huge” insurgent force is not necessarily essential for the military success of an insurgent group, it would seem likely that those larger groups preparing for electoral competition would be better positioned than smaller insurgent groups for a number of reasons. New political parties require the support of a number of full- and part-time staff that can commit to the building of the new electoral vehicle on a daily basis. In addition, the new political party requires a significant number of individuals above and beyond the paid staff to mount an effective campaign. These new political parties need volunteers to collect signatures, distribute flyers and campaign materials, participate in campaign rallies, and volunteer on election day. The success of the new political party should vary to the extent that the insurgent groups can count on an already mobilized and committed group of followers. Therefore, insurgent groups with a large number of militants should be at an advantage relative to insurgent groups attempting to build a new political party starting with a small number of militants and more likely to be successful.
1 In chapter 7, I explore the extent to which many of the hypotheses developed here help us to understand the performance of a single insurgent group (FMLN) across political sub-national units (departments and municipalities) within a single country (El Salvador). In chapter 8, I explore the extent to which the findings of chapters 5, 6, and 7 contribute to our understanding of the performance of other insurgent group transitions to political parties around the world, specifically in Mozambique (RENAMO) and Colombia (M-19).
14
Hypothesis 1 (Insurgent Size): Larger insurgent groups are more likely to be successful than smaller groups.
Political Party One of the main issues likely to affect the electoral performance of these new political parties pertains to their original formation as an insurgent group. Insurgent groups often evolve from political parties, social movements, student organizations, unions, campesino organizations, religious and ethnic groups. Many emerge from a mixture of these groups. Inspired by some radical ideology or perhaps to escape government repression, these groups have changed tactics to confront their respective governments militarily. During civil wars, insurgent groups might retain contact with these groups or direct new ones to act on their behalf in the legal political arena. After several years of military conflict followed by a negotiated settlement facilitating their transition to a political party, would any of these origins or relationships prove more beneficial to their becoming a successful political party? A group likely to benefit from its prior history is an insurgent group that can trace its roots to a former political party. Such insurgent groups are likely to be much better prepared for electoral competition in the postwar period than those insurgencies that lack such origins. These groups have prior experience, albeit most likely somewhat limited, in selecting candidates, drafting a political platform, and campaigning. These experiences should prove beneficial to the future party upon its “return” to “normal” politics. Student organizations, unions and campesino organizations might possibly possess superior numbers to overcome a lack of electoral experience, but ceteris paribus, insurgents with previous electoral experience should be advantaged relative to those insurgencies lacking such experience. While prior experience as a political party should prove useful in the postwar environment, so should an insurgent group’s relationship with a political party during a conflict. Weinberg (1991) finds that terrorist groups (a category in which he includes the FMLN and the Rebel Armed Forces, one of the groups comprising the URNG) with ties to political parties tended to be more successful (in terms of persistence) than those terrorists groups that lacked such linkages. Weinberg offers a number of possible explanations for why this might be the case, including
15 the experience of having belonged (or of belonging) to a party equipped the terrorists with an array of organizational skills useful in sustaining the violent groups. Or the fact that the longer lasting terrorist organizations were more likely to exhibit relationships with political parties may be symptomatic of their enjoying greater popular support. If the latter is true, the groups involved may have had an easier time in recruiting new members and in obtaining logistical aid from well-wishers (1991: 434).
So it is possible that insurgent groups may be able to count upon their experience as political parties to be more effective as military organizations. While Weinberg (1991) draws the connection from prior political party experience to successful terrorist groups, we see that a similar connection is also likely to be important if and when the terrorist or insurgent group opts (or is permitted) to compete in elections. There exist at least two other ways in which insurgent ties to political parties can contribute to their success both during the insurgent phase and in the post-insurgent phase. First, we can imagine that an insurgent group might create or use a political wing for publicity purposes. A political wing might be used to disseminate propaganda, publicize collective grievances and the goals of the insurgent organization. During a civil war, we can imagine that most if not all of the media is either controlled by the government or, at least, sensitive to how its reporting will play out. The existence of a sympathetic party or political wing should facilitate the process by which the insurgent group can counter government “lies” and “propaganda” and promote an alternative, more favorable, perception of the insurgents. This might contribute to the favorable perception during the war and the postwar period. Second, the existence of a political wing might further help the insurgents if that group actually competes in the electoral process while the insurgent group remains committed to its military cause. The experience that the political wing amasses while competing in elections and the voter loyalties it cultivates should contribute to the future success of the insurgent group as political party.
Hypothesis 2 (Political Party): Insurgent groups with prior political party experience are more likely to be successful than those without prior political party experience.
16 It should be noted that this discussion of the relationship between an insurgent group and a political party combines two different scenarios into a single hypothesis. First, insurgent groups are more likely to be successful if they have prior experience (pre-insurgency) as a political party. And second, insurgent groups are more likely to be successful if, in addition to the military organization, they also possess a political wing. It is possible that these two factors could have distinct effects on insurgent group success, but, a more fundamental distinction is that between those insurgent groups with prior political party contacts and those without, regardless of whether these contacts came before or during the war. Whether an insurgent group with previous experience as a political party and with a political wing is more successful than an insurgent group with only one of these relationships or which of the two scenarios is more important will for now be left to a future project.
Mass-Based Organizations In addition to the insurgent group’s military size and prior political party experience, the group’s relationship with broader sectors of society is likely to influence its electoral success as well. Some insurgent groups develop extensive ties to unions, student organizations, and peasant confederations, while other insurgent groups are unable (or interested) to cultivate relations with other opposition organizations. During periods of civil war, insurgent groups that develop contacts with these other groups can call on them to coordinate opposition to the regime through simultaneous work stoppages, national strikes, civil disobedience, and guerrilla attacks. The development of a broad-based opposition coalition has been found to be a key determinant of revolutionary success (Wickham-Crowley 1992). Shugart (1992: 125) theorizes that “a rebel movement that has focused upon cultivating a mass base may have better resources for extracting sacrifice from its supporters and thereby continuing the war than a rebel movement that lacked such a committed, reliable following.” In addition, those insurgent groups that have developed a mass base may be more inclined to demobilize and then participate in the electoral process, since they will likely to be able to draw upon this mass base for electoral support (1992: 128). Likewise, in an article on how the process of democratization is likely to affect revolutionary movements, Jeffrey Ryan (1994: 41) argues that “the construction of a broad-based coalition extending beyond the core of the revolutionary movement itself” is likely to be the key determinant of whether and how the revolutionaries survive the onset of democratization.
17 Revolutionary organizations that have successfully cultivated a broad-based coalition prior to the onset of democratization are likely to attain a negotiated settlement whereas organizations that have not developed a broad-based coalition prior to democratization are likely to face a choice of either surrender or destruction. In a similar vein, we would expect the development of a broad- based coalition to also contribute to the success of the group as a political party in the post-war period. The existence of a broad based coalition will provide the new political party with a pool of potential voters and also allow the party to overcome defections from disgruntled or disillusioned former militants. Some research has already been conducted in the advanced democracies concerning the relationship a new political party maintained with other previously organized groups, although the results have not always been conclusive. Previous research has hypothesized that “a party that could draw on an existing network of support through unions or other interest groups or which could quickly put together such a network on its own would do extremely well” (Hauss and Rayside 1978: 52). While the authors did not find the existence of a social base critical to the initial success of a new political party, it does appear to be an important factor in their long- term success. Groups with or without a social base have been initially successful as political parties whereas those that have been able to maintain long-term success have tended to do so only with the existence of a social base. In a similar vein, Rose and Mackie (1988: 537) found that new political parties “that drew electoral support from voters already institutionalized into major social organizations” were more successful than those that did not. While there are reasons to expect that insurgent groups with more extensive relations with legal organizations will be better prepared for electoral competition, this is no guarantee of short-term electoral success as Hauss and Rayside found. But all else remaining equal, just as the construction of a broad-based coalition is expected to contribute to revolutionary success, it would appear that insurgent groups with greater contact with mass-based organizations are more likely to be successful as political parties.
Hypothesis 3 (Mass-Based Organizations): Insurgent groups that had extensive contacts with mass-based organizations during the war (unions, student groups, etc.) are more likely to be successful than those without such contacts.
18 Unity Another concern of new political parties relevant to former insurgent groups revolves around the extent to which the group can maintain its unity. Schisms can affect an insurgent group at many stages of its existence. During the civil war, differences over strategies and tactics can lead to intra-group conflict that might undermine the goals of the insurgency. The factions that make-up the insurgent group might spend more time debating and fighting each other than they do fighting the government. On the other hand, insurgent groups might not suffer from problems of incoherence during the war when they all share the belief in the military defeat of the government. Instead, problems of unity come to the forefront when the insurgent group must contemplate a shift to nonviolent tactics. Stephen Stedman (1993) has noted how intra-group differences within insurgent coalitions often become exacerbated when groups have to decide between continuing armed opposition to the regime and pursuing a political settlement. It is often the case that individuals and groups on both sides (government and insurgent) of a negotiated settlement do not believe it to be in their best interest to resolve the war through diplomatic negotiations. This might lead to the division and possible breakup of the group as “spoilers” attempt to undermine the peace process (Stedman 1997). Atlas and Licklider (1999: 51) provide an alternative explanation for insurgent group disunity in arguing that
the antagonists on the other side of the civil war divide become the demonized other that helps hold allies together in their wartime coalition. Once a settlement is reached and the fundamental us/them dichotomy begins to break down, the cohesiveness of groups on either side dissipates, and disputes among allies who are now more cognitively aware of their differences and conflicting interests can easily result.
Given these potential differences of opinion within the insurgent (and/or government) coalition, it is not surprising that wartime coalitions break down. Violent conflict has even been known to break out between former allies as a result. In the post war period of Sudan, Zimbabwe, Chad, and Lebanon, Atlas and Licklider (1999) found that conflict within the former rebel coalition led to either a renewal of civil war between the original protagonists (rebel and government) or between the groups that composed the rebel coalition. Of course, violent conflict
19 might not erupt between former insurgent allies, but the defection of its key leaders and supporters are still likely to undermine electoral support for the group in the postwar period. In terms of the literature on political parties, others have also noted the potential negative impact of party disunity. Hauss and Rayside (1978: 52) argue that political parties “suffer when they lose a unified, coherent leadership. Factionalism, and ultimately schisms, can directly lead to the loss of voters and members.” We see that insurgent coalition can fracture at almost any moment: in the midst of civil war, during negotiations for its resolution, and in the postwar period. Schisms might hurt the new political parties in a number of ways including the loss of key leaders, the loss of individuals sympathetic to the party, and the undermining of public trust. If the insurgent group cannot manage differences within its coalition without splintering, how can the public trust them to run the country?
Hypothesis 4 (Unity): Insurgent groups that can maintain unity and coherence during the war, negotiations, and in the postwar are more likely to be successful than insurgent groups that fracture.
Control of Territory Larger insurgent groups are better prepared for the challenges of electoral politics if they not only possess a large army and network of noncombatant supporters, but if that support reaches a greater proportion of the country. Former insurgent groups should be more likely to succeed as political parties if they have won and controlled territory during the war. An insurgent group that maintained territory and developed a certain level of order within the area is more likely to be able to mobilize the population within this area to vote than insurgent groups that controlled only unpopulated or remote areas of the country such as inaccessible mountainous regions. As Kalyvas (1999: 259) discusses, “an insurgent organization which controls a given area (a ‘liberated area’) operates as a counter-sovereign authority, a ‘counter-state’. It provides protection, administers justice, collects taxes, and applies it social program.” In the controlled territory, the insurgent group can gain valuable experience by completing public works projects and working with local organizations to address the needs of the population living within their sphere of control, all critical to its future success as a political party.
20 If, on the other hand, the insurgent group is unable to capture and hold territory, or spread its presence throughout the country, it will be at a disadvantage as it prepares to compete on a national basis as a political party. While even a small insurgent group might be able to inflict damage upon the government or elude destruction, it will be more difficult for such a group to succeed as a political party relative to a group that has a presence throughout a larger area of the country. These organizational experiences should prove advantageous in making the transition and succeeding as political parties.
Hypothesis 5 (Control of Territory): Insurgent groups that control territory during the war are more likely to be successful than groups that are unable to control territory.
Violence But it is certainly possible that the population living under its protection will not look upon the “occupying” forces kindly. Insurgent groups can be very violent towards civilians within their spheres of influence, particularly those individuals they believe sympathetic to the government. While the insurgent group can utilize its experience of having successfully governed controlled zones, or “liberated areas,” “it also enjoys a local monopoly of violence which it uses to punish its enemies and sanction uncooperative behavior, such as the refusal to supply food or pay the ‘revolutionary tax’” (Kalyvas 1999: 25). So beyond the territory controlled by the insurgents, we should expect the relationship between the insurgent group and the general population to be a strong predictor of the insurgents’ future success as a political party. During a civil war, it is often said that one of the goals, perhaps even the primary goal, of both the insurgents and the government is to “win” the loyalty and support of the “masses.” Insurgents and government can vie for the peoples’ support by providing their supporters with benefits at the same time that they deny these goods to those that seek to remain neutral or actively support the opposing group. Unfortunately, not all loyalty or support, though, is won through such “benevolent” means and, instead, is accomplished through coercion. As a result, violence committed against civilians in wartime has grown steadily throughout the twentieth century (Tilly 2002).
21 Neither governments nor insurgents can claim a monopoly on the use of force against civilians during war, as recent history has made quite clear. Violence perpetrated against civilian populations, while a very emotional and highly publicized issue today, is nothing new (Kalyvas 2001). Governments and insurgents from every region of the world and historical period have been known to use violence to deter potential defectors and/or to punish actual defectors. In some wars such as the recent cases in Latin American history, government forces have committed the majority of civilian massacres (Wickham-Crowley 1990). On the other hand, the insurgent army in the Algerian civil war committed most of the civilian massacres during that war (Kalyvas 1999). Human rights violations, and perhaps the massacre of civilian populations, occur in all wars. While the act of massacring civilians might serve some ultimate purpose in the midst of civil war (Kalyvas 1999), it is unlikely to help the insurgent group in the postwar period when it, as a new political party, is attempting to attract electoral support. An insurgent group (or government for that matter) that relied overwhelmingly upon coercion in its pursuit of military victory is unlikely to fare as well as an insurgent group that was able to capture the loyalty and support of the civilian population through less violent means. Therefore, an insurgent group that uses less violence against the civilian population relative to the government is more likely to succeed as a political party relative to an insurgent group that commits a greater proportion of violent acts relative to the government.
Hypothesis 6 (Violence): Insurgent groups that are less violent towards civilians during the war than government forces are more likely to be successful than those insurgent groups whose civilian-directed violence outweighs that of the government.
Discussion In the preceding section, I developed a number of expectations concerning the characteristics of insurgent organizations and their performance in civil wars expected to increase the likelihood that the new political parties will be successful. Larger insurgent groups, those with prior political party experience, more extensive contacts with mass-based organizations, and unity should tend to be more successful than insurgent groups with fewer combatants, fewer ties to political parties and mass-based organizations, and internal divisions.
22 In addition, insurgent groups that control more territory and are less violent towards civilians are more likely to be successful than groups with limited territorial control and an extensive record of civilian-directed violence relative to the government and its associated forces. In the next section, I turn to a number of factors about the electoral environment in which the insurgent group will compete as a political party.
Environmental Conditions While focusing solely upon the characteristics of former insurgent groups can go a long way towards explaining performance as new political parties, explaining their success or failure would be incomplete without considering the influence of the environment in which the group must compete. As I discuss below, some environments are more conducive to facilitating new party success while others tend to hinder the success of new political parties. For example, new political parties should be more likely to succeed when they compete in parliamentary systems rather than presidential systems. But, because many former insurgent groups have competed in presidential systems, including the two I focus on in this dissertation, I also focus on a basic difference within presidential systems that might be expected to impact a new party’s performance: the use of simple plurality or majority runoff elections.2 I also develop hypotheses concerning the legislature such as the electoral rules, district magnitude, and the existence of electoral thresholds. Finally, in addition to the institutional characteristics of the political system, the likelihood of new party success is also going to be effected by whether voters, or potential voters, are likely to throw their support behind a new entrant to the electoral arena. According to Cox (1997: 151-78), new political parties are likely to emerge and succeed when elites perceive that the future party has a high probability of success. In countries undergoing democratic transitions, there are often a high number of new political parties that emerge and are successful at the same time. In new democracies, each new political party is regarded as having an equal probability of success as voters have no solid expectation as to which parties are likely to be successful on election day (Cox 1997: 151-78; Shvetsova 2003). As a result, these new systems are likely to have more political parties relative to more established party systems (Bielasiak 1997; Cox 1997; Reich 2004). While it is easy to imagine that uncertainty is high
2 In Latin America alone, former insurgent groups have competed in presidential systems in Colombia (M-19 and Patriotic Union), Venezuela (MAS), Guatemala (URNG), El Salvador (FMLN), Nicaragua (FSLN, Contras) and Honduras (PUD).
23 during founding elections involving parties with no electoral histories, some argue that this uncertainty even extends to the former communist parties (Kaminski 2002; Shvetsova 2003). While the communist parties have been around for several decades and had thousands, perhaps millions of card carrying members, there was no way to know their true support and, therefore, uncertainty remained high. Over time, the number of political parties and new entrants to the political system should decrease as candidates and voters coordinate on a smaller number of viable political parties (Taagepera and Shugart 1989; Turner 1993; Cox 1997; Duch 2001; Duch and Palmer 2002; Tavits and Annus 2004; Tavits 2004). Because most scholars have accepted that the electoral viability of all potential entrants is equally uncertain, it has allowed those who study the question of new party success to disregard the possibility that some new political parties are more likely to succeed than others given their prior history. Assuming that all new political parties possess an equal probability of success because voters and elites have not had enough time to coordinate on viable electoral candidates allows researchers to focus solely upon institutional factors. As I have discussed regarding the likely success of former insurgent groups, this is not a tenable assumption. In many of the countries in which former insurgent groups are transforming themselves into a political party, the political systems into which they are entering have already experienced a number of electoral contests. While these countries tend not to have extensive histories of completely free and fair electoral competition, there is often critical information we can glean from how the voters reacted. In the following section, I develop a number of institutional-based and voter-based hypotheses that make it more likely that an insurgent group will be successful as a political party.
Electoral Objective An important characteristic of a political system that might be expected to affect the emergence and success of a new political party concerns whether the key electoral objective is the parliamentary seat or the presidency. Generally, it is believed that where parliamentary seats are the objective rather than the presidency, the number of political parties tends to be greater. In those systems where the parliamentary seat is the objective, new political parties have a greater incentive to “go it alone,” because even a single legislative seat can result in the joining of a
24 coalition government. Because a single seat or two can result in the possibility of becoming part of the executive, voters might also be more inclined to vote for a new, most likely small, political party. Therefore, new political parties are more likely to emerge and to succeed in parliamentary systems than presidential systems. On the other hand, the opposite effect occurs in those systems where the main objective is the presidency. Various authors argue that this zero-sum nature of the presidency tends to concentrate the electoral focus on a single office limiting the emergence and success of new political parties (Key 1958; Lipset 1963). Support for the two largest political parties at the presidential level tends to carry over into elections for the legislature thereby directly reducing the number of parties in elections for the executive and indirectly in elections for the legislature. The effect exists regardless of whether legislative elections in presidential systems are conducted under proportional representation or single member district voting (Lijphart 1999: 155). Empirical evidence has not always supported this belief (Hauss and Rayside 1978), although on average it seems likely that any new political party, including those with histories as former insurgent groups, is more likely to experience success in a parliamentary system rather than a presidential system.
Hypothesis 7 (Electoral Objective): Former insurgent groups are more likely to succeed in parliamentary than presidential systems.
Presidential Electoral Rules One of the ways in which presidential systems differ relates to the rules utilized to elect the president. Presidents are typically elected either via plurality rule (the candidate that receives the most votes is declared the winner) or majority rules (the winner must receive 50% of the vote +1). A runoff election involving the top two candidates is frequently used to satisfy this requirement if no candidate receives greater than 50% in the first round. Scholars have argued that the choice between plurality rules for electing the president and majority runoff rules has the effect of increasing or decreasing the number of political parties in a system. Basically, scholars have found that majority runoff requirements are more likely to lead to multiparty systems thereby allowing greater opportunity for new political parties to emerge and succeed.
25 Shugart and Carey (1992) argue that presidential elections employing plurality voting tend to lead to “the formation of a broad coalition behind the front-runner” and “the tendency to coalesce the opposition behind one principal challenger” thus reducing the effective number of political parties to close to two (1992: 209). Similarly, Mark Jones finds that Latin American presidential systems employing a majority electoral formula tend to have greater multiparty systems than those that employ plurality electoral formula (1994). Both Shugart and Carey (1992) and Jones (1994) find that plurality voting for the presidency tends to depress the number of parties whereas majority voting leads to greater multipartism. This is not necessarily the same argument that new parties will be more successful in systems employing plurality voting rather than majority voting. New political parties might find success in systems using plurality voting, but the conditions under which this will likely happen involve a coalition. In majority vote systems, the opportunity for new political parties to succeed will be enhanced by their ability to participate in a first round of voting. Voters are more likely to vote sincerely for their most favored presidential candidate even if the likelihood of victory is minimal or nonexistent, as long as there is a strong likelihood that there will be a second round. The new political party, which is unlikely to attain a fifty-percent majority in the first round, can use the support it receives in the first round to leverage concessions from one of the two larger parties prior to the second round of voting. Therefore, in presidential systems, former insurgent groups are more likely to succeed in those systems using majority runoff than those employing simple plurality electoral rules.
Hypothesis 8 (Presidential Electoral Rules): Former insurgent groups are more likely to succeed in systems with majority presidential runoffs than simple plurality.
Legislative Electoral Rules In addition to the electoral rules for the presidency, the rules for electing the legislature are likely to have a strong impact on the viability of a new political party. At the legislative level, it is believed that new political parties are more likely to emerge in electoral systems where members are elected by proportional representation rather than single member district voting (Duverger 1954; Rae 1967; Harmel and Robertson 1985; Taagepera and Shugart 1989; Mair 1991; Lijphart 1994, 1999; Hug 2001). Single member district voting tends to create a two-
26 party system by overrepresenting the two largest parties, thus making it more difficult for third party challengers to compete effectively. Two processes are at work here. First, voters will tend to vote for those parties that have a clear chance at winning so that there votes are not “wasted” on a smaller political party with little to no chance at winning. Second, political leaders will also tend to gravitate towards those parties that also have the best chance at winning as this will maximize their probability of winning a seat. Typically, this makes it much more difficult for new political parties to break into and succeed in systems employing single member district voting as both political leaders and voters are drawn to the two largest parties with the highest chance at winning the seat. Political systems employing proportional representation, on the other hand, tend to have the goal of preventing the overrepresentation and the underrepresentation of political parties. Under proportional representation, a new political party that captures a small percentage of the vote is more likely to succeed than a party with similar results competing in elections using single member district voting. Therefore, it is not surprising that in elections for the legislature, former insurgent groups are more likely to succeed as new political parties in those systems employing proportional representation rather than single member district voting.
Hypothesis 9 (Legislative Electoral Rules): Former insurgent groups are more likely to succeed under proportional representation electoral rules than under single member district voting.
District Magnitude In addition to electoral rules, district magnitude has also been hypothesized to affect the likely emergence and success of new parties. The district magnitude refers to the number of candidates that are elected in the district. Taagepera and Shugart, in fact, found district magnitude to be the “decisive factor” on the degree of disproportionality and the number of political parties (1989: 112). The impact of increasing district magnitude has opposite effects in plurality-majority and PR systems. Higher district magnitude under plurality-majority formulas leads to greater disproportionality and more favorable conditions for the two largest parties. In proportional systems, a larger district magnitude “results in greater proportionality and more favorable conditions for small parties” (Lijphart 1999: 150), the opposite of the effect of larger
27 district magnitudes in majority or plurality systems. As district magnitude increases, the minimum percentage of the vote needed to win one seat in a district decreases. For example, a party with 10% of the vote in a district with five available seats is unlikely to win a seat. On the other hand, if district magnitude were increase to ten, the party would be guaranteed a seat from the district (Lijphart 1999: 152). Given that the two new parties under consideration in this dissertation participate in legislative elections conducted under proportional representation, I will restrict the hypothesis to the expectations of participation under proportional representation given high versus low district magnitude.
Hypothesis 10 (District Magnitude): Former insurgent groups are more likely to succeed when the district magnitude is high and when there is a nationwide district.
Electoral Threshold In general, systems using proportional representation electoral rules should have a greater number of political parties than systems that do not use PR. Ideally, the percentage of votes a political party receives under this system will closely approximate the percentage of seats it captures in the legislature. But most countries also tend to institute an electoral threshold to limit the proliferation of new political parties. Too many small political parties can make the policy- making process highly inefficient. Parties are thereby required to win a number of seats “in the lower-tier districts and/or a minimum percentage of the total national vote” to retain legal standing and their seats in the legislature (Lijphart 1999: 153; see also Taagepera and Shugart 1989: 133-35). High electoral thresholds should have the effect of preventing a proliferation of new, small political parties whereas lower electoral thresholds should increase the average number of new political parties competing in the system. Former insurgent groups face many obstacles upon initial electoral competition, which makes it likely that electoral thresholds will impact their performance. Therefore, former insurgent groups are less likely to succeed in systems with higher electoral thresholds.
Hypothesis 11 (Electoral Threshold): Former insurgent groups are more likely to succeed in systems with low electoral thresholds.
28 Abstention If the existing political parties are doing a commendable job, it is unlikely that voters will support a new political party. But, on the other hand, if voters do not believe that the system’s current parties are capable of resolving important issues, new political parties are bound to emerge (Hauss and Rayside 1978: 38). This argument pertains more to stable party systems where the emergence and success of new political parties is probably less likely than in newer democracies where voter loyalties are less likely to be so entrenched. Countries emerging from civil war are unlikely to have very stable party systems characterized by a populace fully satisfied with the prevailing electoral choices. In more stable party systems, one of the implications of this is “that there is no pool of available voters in which a new party can build a base of support” (Hauss and Rayside 1978: 38). One way in which we might be able to get a handle on whether voters might support a new political party is by studying the levels of voter abstention in the country. Several elections occur against the backdrop of civil war where insurgent groups often call on their supporters to boycott the vote and convince others to refrain from voting through other coercive means. Under these conditions, low voter participation might be the result of either fear or, perhaps, support for the insurgents. The “true” reasons for low voter turnout in these situations are often quite difficult to discern. Regardless, it appears reasonable that high levels of voter abstention in elections prior to the formation of the new party indicate an obvious pool of potential voters that these new parties can tap into. In that case, high levels of voter abstention should indicate that there is a proportion of the population available to support a new political party.
Hypothesis 12 (Abstention): Former insurgent groups are more likely to succeed in systems with high levels of voter abstention.
Bipolarization It is also argued that the bipolarization that competition for the presidency supports has an “indirect but strong effect on the effective number of parliamentary parties…This tendency is especially strong when the presidential election is decided by plurality instead of majority-runoff (where small parties may want to try their luck in the first round)” (Lijphart 1999: 155). As mentioned previously, presidential systems tend to depress the number of political parties closer
29 to two and make it more difficult for new political parties to succeed. But in new political systems where we expect two political parties to eventually come to dominate, the identity of these two parties is unlikely to be determined immediately. The argument here is similar to the research discussed earlier concerning newly democratizing countries where voters and elites have not yet coordinated on electorally viable parties. Therefore, we would expect new political parties to be more likely to succeed in presidential systems where bipolarization has not yet set it.
Hypothesis 13 (Bipolarization): Former insurgent groups are more likely to succeed in a presidential system when bipolarization has not yet set in.
Economic Conditions An additional factor likely to affect the success of a new political party relates to the state of the nation’s economic health. Poor economic conditions make it more likely that people will punish the incumbent and vote for another political party (Lewis-Beck and Stegmaier 2000; Tucker 2002). This should not only help new political parties, but all other opposition parties not associated with the incumbent party (Lewis 2000). Countries emerging from civil war are often in poor economic shape due to the loss of life, damage to infrastructure, and the over concentration of a state’s resources committed to internal security. Normally, poor economic conditions would lead to the incumbent party being punished by the voters. But former insurgent groups are not necessarily going to be the recipients of transferred votes. If the voting population tends to hold the insurgent group responsible for an economy in crisis, obviously this will not help the new political party’s performance. Therefore, former insurgent group are more likely to succeed when there is an economic crisis [that voters do not blame on the group].
Hypothesis 14 (Economic Conditions): Former insurgent group are more likely to succeed when there is an economic crisis [that voters do not blame on the group].
Social Cleavages In addition to institutional explanations for the success of new political parties, other researchers have focused more on the existence of cleavages in society. Consistent with this line
30 of reasoning, the number of parties in a society is not necessarily determined solely by institutional factors. One must take into account the existence of social cleavages that might be just as important, or perhaps more important, to new party success (Lipset 1963; Lipset and Rokkan 1967; Hauss and Rayside 1978; Kim and Ohn 1992). These cleavages might well include ethnic, linguistic, or religious groups. Therefore, the greater the number of cleavages that exist in a society, the greater the number of political parties will emerge. It appears likely that the electoral fortune of an ethnic, religious, or linguistic based insurgent group will hinge upon the relative composition of society. We can perhaps think of the role of social cleavages in terms of opportunity and willingness. If the insurgent group is made up of an ethnic, religious, or linguistic group that comprises a rather significant percentage of the country’s population, the insurgent group will have a greater opportunity to draw electoral support from this group. For example, an insurgent group formed by members of an aggrieved minority group that represents roughly twenty-five percent of the population is likely to perform better than a similar party representing only five percent of the population. On the other hand, an ethnic-based insurgent group representing an ethnic population comprising less than five percent of the total population is likely to face severe difficulties. This leads to the following hypothesis:
Hypothesis 15 (Social Cleavages): Identity-based insurgent groups are more likely to be successful if they represent a significant percentage of the population.
Existing Parties Now, the simple fact that an insurgent group says that it speaks for a specific population or shares a similar background is no guarantee that the larger population will provide electoral support to the new political party. The likelihood that they will support this group depends upon whether there already exists a successful political party that represents the same population. For example, if an ethnic-based insurgent group opts to compete in a political system where a similar ethnic-based group already exists, the probability of success is much lower than a system where no such party exists. Similarly, if a class-based Marxist-Leninist insurgent group looks to compete in a party system where there already exists a strong left or center-left party, the likelihood of success is lower than a system where these parties do not exist.
31 Hypothesis 16 (Existing Parties): Insurgent groups are more likely to be successful if parties with similar programs do not already exist.
The negative impact of an existing party with a similar platform assumes that the two (or more) political parties do not cooperate with each other. Of course, the existence of a political party with a similar platform might help the new political party if they form an alliance or merge with the new party.
CONCLUSION
In this section, I addressed several factors hypothesized to impact the success of former insurgent groups as new political parties related specifically to the insurgent organization and the environment in which it competes. In sum, larger insurgent groups, those with a political wing or with a history as a political party, extensive relations with mass-based organizations, the ability to capture and control territory, and a less violent history towards civilians than government forces are more to likely succeed as new political parties. Insurgent groups as new political parties are also more likely to succeed in parliamentary systems rather than presidential systems. In presidential systems, new parties are more likely to succeed when elections are determined with majority runoff formula rather than plurality vote. In elections for the legislature, new political parties are more likely to succeed under legislative PR rules with high district magnitudes and low electoral thresholds. As new political parties, insurgent groups should be more successful in systems characterized by high levels of voter abstention, where bipolarization has not yet set in, economic conditions are poor, and there are salient social cleavages in society unrepresented by the existing political parties. In the following chapter, I discuss how I test the hypotheses developed in this chapter.
32 CHAPTER 3 RESEARCH DESIGN
In the previous chapter, I developed a number of hypotheses that explain the electoral performance of former insurgent groups as political parties. In this chapter, I discuss issues related to how I intend to go about testing these hypothesized relationships. First, I discuss the two cases I selected to test these relationships. Second, I discuss how I conceptualize and measure political party success differentiating between failed parties, minor parties, and major parties. Finally, I discuss how I operationalize and measure each independent variable in my analyses.
Case Selection and Methodological Notes Historically, research on the success of new political parties has tended to focus upon the more stable democracies of Western Europe and the United States. In recent decades, research has moved towards explaining the performance of specific parties such as environmental, far- right, and former communist parties. To date, though, there has been little interest in studying the transition of former insurgent groups to political parties. This is surprising given that over the last several years, insurgent groups have disarmed, demobilized, and competed in electoral politics in almost every region of the world. In Central America, in addition to the FMLN in El Salvador and the URNG in Guatemala, insurgent groups have attempted the transition to electoral politics in Nicaragua and Honduras. In Nicaragua, the Sandinista National Liberation Front (FSLN or Sandinistas), following their successful overthrow of the last of the Somoza dynasty, have competed in reasonably free and fair elections since 1984. Following the FSLN rise to power, two anti- Sandinista insurgent groups, commonly known as the Contras, emerged with the intention of removing the Sandinistas from power in Nicaragua. Following the 1990 political settlement, a group of former Contras formed the Party of Nicaraguan Resistance and competed in both local and national level elections. In neighboring Honduras, a number of relatively minor armed opposition groups were defeated by government forces, accepted amnesty, and subsequently formed the Democratic Unification Party in 1994.
33 Insurgent groups have also made the transition to electoral politics in many South American countries. The Movement Towards Socialism was defeated by the Venezuelan military before regrouping as a political party in the 1970s. In Colombia, the Revolutionary Armed Forces of Colombia formed the Patriotic Union and the April 19th Movement formed the Democratic Alliance M-19. Both new political parties competed in the 1990s. In Uruguay, the Tupamaros National Liberation Movement organized as a political party (Movement of Popular Participation) following the return to democratic rule in 1985 and eventually came to power as the lead party in the Broad Front leftist coalition in the 2004 elections. In Africa, insurgents have transformed themselves into political parties in Mozambique (RENAMO), South Africa (African National Congress or ANC), and Angola (the National Union for the Total Independence of Angola or UNITA). Finally, in Europe, the Irish Republican Army (IRA) and its political wing, Sinn Fein, are a good example of a former violent opposition group that is in the process of demobilizing, disarming, and preparing to compete in elections. The transition of former insurgent groups to political parties is a global phenomenon likely to continue well in the future. Recently, the Zapatista Army of National Liberation (EZLN) in Mexico has held internal discussions about the viability of establishing a political party. In Palestine, Hamas recently, and successfully, competed in parliamentary elections for the first time. Finally, the Iraqi government has also begun to reach out to insurgents with the hope that they will disarm, demobilize and participate in the democratic process as political parties. In this dissertation, I focus on two of the insurgent groups in Central America that have made the transition to political party: the Farabundo Martí National Liberation Front (FMLN) in El Salvador and the Guatemalan National Revolutionary Unit (URNG) in Guatemala. The FMLN formed in October 1980 with the merging of five smaller insurgent groups. After more than a decade of violence, the FMLN and the President of El Salvador, Alfredo Cristiani, brought about an end to the civil war with the signing of the Peace Accords at Chapultepec Castle in Mexico City, Mexico on January 16, 1992. In addition to ending the war, the Peace Accords marked the beginning of the transformation of the FMLN into a political party. Since the signing of the Peace Accords, the FMLN has participated in three presidential elections (1994, 1999, and 2004) and four legislative and municipal elections (1994, 1997, 2000, and 2003).
34 While civil war raged in El Salvador, neighboring Guatemala also was experiencing its own violent conflict. In Guatemala, civil war first erupted in 1960 when several insurgent groups protested against an increasingly corrupt and repressive government that was aiding the US-planned invasion of Cuba on its soil. In 1982, four of these groups merged to form the URNG. For the next fourteen years, the URNG battled the Guatemalan military. While not as militarily successful as their counterparts in El Salvador, the URNG was successful enough to convince the Guatemalan government to accept a negotiated solution to the conflict. After several years of negotiations, the URNG and President Alvaro Arzú of Guatemala signed the Firm and Lasting Agreement on December 29, 1996. Like the Peace Accords in El Salvador, the Firm and Lasting Agreement also marked the beginning of the URNG’s transition to political party. Since its incorporation into the political system as a political party, the URNG has competed in presidential, legislative and municipal elections in 1999 and 2003. These two cases share several characteristics. First, several civil war characteristics of each country are quite similar. Each country’s civil war lasted for more than one decade and overlapped the same Cold War period (1979-1992 in El Salvador and 1960-1996 in Guatemala).3 Both wars were also extremely violent resulting in the deaths of more than 200,000 Guatemalans and 75,000 Salvadorans. The estimated number of deaths as a result of each country’s civil war does not do justice to the violence. During both wars, the majority of victims were civilians, not members of the military or the guerrillas. Nor were civilians simply caught in the crossfire; civilians were tortured, raped, and often “disappeared.” Entire villages were destroyed, particularly in Guatemala. Both conflicts were also successfully resolved via negotiated settlements involving heavy international involvement from the United Nations, the United States and other regional actors. Second, while each insurgent group was comprised of socialists, communists, labor leaders and workers, students, campesinos, democrats, and religious workers, they were primarily Marxist-Leninist in nature. Each group was confronted with the challenge of transforming itself from an insurgent group brought together by opposition to the Guatemalan and Salvadoran governments to a political party capable of managing salient internal differences that had been glossed over during the war.
3 These are the commonly accepted starting dates for the respective conflicts. If, instead, we use the founding dates of the FMLN (1980) and the URNG (1982), the time periods are more similar.
35 Third, at the level of the state, a process of democratic transition began in both countries in the midst of civil war. Each country’s democratic transition did not begin with the reinsertion of the insurgents, but several years earlier. El Salvador held elections for a Constituent Assembly in 1982 with presidential elections to follow in 1984. Guatemala also held semi- competitive elections beginning in 1985. The end of the conflicts and the transition of the former insurgents into political parties were defining moments in the transition to democracy, but not the beginning of said transition. Therefore, the performances of these organizations as political parties were not solely based upon their military prowess. Each country had experienced over ten years of elections providing other political parties and voters time to establish some sort of electoral history. Therefore, we can also assess the development of each country’s political party system and its impact on the electoral performance of these new contenders. Fourth, both El Salvador and Guatemala share a similar culture, precluding cultural explanations. Finally, for the researcher, the two groups also provide a useful comparison in that the citizens of both countries speak the same language (although many in Guatemala also speak a variety of indigenous languages) and border each other in Central America, making it relatively easy to conduct fieldwork in both locations. These similarities between the groups, the civil wars, and the political systems allow me to conduct a most-similar research design. In this most-similar research design, I select one example of an insurgent group that has successfully transformed itself into a minor party (URNG) and an example of an insurgent group that has successfully transformed itself into a major party (FMLN). I conduct a structured-focused comparison of the FMLN and URNG. In the previous chapter, I developed a number of organizational and environmental hypotheses to explain the performance of former insurgent groups as political parties. In this chapter, I discuss how I collect data on the same variables for both cases. In chapters 5 and 6, I evaluate all of the variables that are expected to influence the performance of these former insurgent groups, as hypothesized in chapter 2. Chapters 5 and 6 are concerned with testing the hypothesized relationships across insurgent groups. In chapter 7, I am interested in ascertaining whether these hypothesized relationships hold up for within-nation analysis. In choosing the FMLN in El Salvador and the URNG in Guatemala for comparative analysis, I looked to minimize the number of factors that might have played a role in explaining the success of the FMLN relative to the URNG – the role
36 of the international community, each country’s level of political and economic development, cultural effects, and the Marxist-Leninist tendencies of the insurgent groups. Lijphart (1975) and others have noted the value of analyses across sub-national units for testing empirical relationships. While comparing insurgent transitions in Guatemala and El Salvador should limit the number of factors that explain the relative success of new political parties, within-nation analysis should give greater confidence to the relationships. In chapter 7, I conduct statistical analyses of the 1994 Salvadoran election using one-way analysis of variance (ANOVA) and OLS regression. The 1994 election is the first election in which the FMLN competed as a political party. I go into further detail on research design issues for this analysis in chapter 7. I am also interested in the generalizability of these findings to other instances of insurgent transitions to political parties. As mentioned earlier, insurgents have attempted the transition to political party in many other countries, and others are likely to do so for the foreseeable future. Therefore, I compare the performances of the FMLN and URNG to insurgent transitions in other countries in Chapter 8, specifically, the performances of the M-19 in Colombia and RENAMO in Mozambique. I discuss how I go about this analysis in greater detail in that chapter.
Operationalization and Measurement Issues In order to answer my primary research question of why some insurgent groups are more successful than others as political parties, I make use of multiple sources of data. I utilize primary and secondary sources published in English and Spanish and fieldwork notes from El Salvador (February-December 1997, March and May 2004) and Guatemala (April 2004). During these trips to Central America, I conducted a number of semi-structured interviews focused on probing the significance of the armed struggle for preparation as a political party, problems of unity, and relations with other actors in the country. On average, each interview lasted ninety minutes. These interviews were conducted with current and former members of the insurgent groups, insurgent and opposition political parties, media, and political analysts in each country (See Appendix A). My insights were also supplemented by my experience as an international election observer in El Salvador during the 1997 legislative and municipal elections and the 2004 presidential elections. In 1997, I volunteered as an election observer in the town of Apopa in the capital department of San Salvador. In 2004, I was located in Berlín in the department of Usulután. Finally, I also use of a variety of socio-economic and electoral data
37 from each country. The socio-economic data was collected from each country’s national censuses while the electoral data was collected from each country’s electoral commission. In the following section, I go into further detail on the conceptualization and measurement of political party success. Following this discussion of the dependent variable, I discuss the operationalization and measurement of each independent variable.
CONCEPTUALIZING THE SUCCESS OF POST-INSURGENT POLITICAL PARTIES
With the conclusion of a civil war through the transformation of an insurgent group into a political party, how might one evaluate whether these groups have been successful as political parties? On the one hand, some might argue that these groups have already failed, simply by the fact that they have moderated their goals and accepted less than military victory. Revolutionary groups that surrender all hopes for the radical transformation of society required to improve the lives of the country’s impoverished masses, secessionist groups that accept anything less than complete independence, insurgent groups that agree to some form of power-sharing arrangement – each group’s transformation into a political party and its acceptance of the general framework of liberal democratic politics can be seen as evidence of its failure. On the other hand, many authors have pointed out that the great majority of those that join insurgent groups do so because the nonviolent electoral route to power has been denied through fraudulent elections, various exclusionary practices, and intense political and social repression (Goodwin 2001; McClintock 1998; Wood 2000, 2003; Brockett 2005). In many ways, then, the end of wide-scale political violence allows these groups the opportunity to participate in their country’s political, economic and social decision-making processes – the denial of which led them to take up arms in the first place. The manner in which each of these groups has chosen to participate in the postwar period is through the formation of political parties and competition in the electoral process. With that in mind, it is perfectly reasonable and quite appropriate to evaluate the performance of these groups in terms of their performance as new political parties.
38 Successful Political Parties To date, the literature on political parties has not produced a single, generally accepted concept of individual party success. This makes it difficult to systematically evaluate the performance of new political parties. Authors have approached the success of new political parties in terms of their ability to endure or persist (Berrington 1985; Rochon 1985; Rose and Mackie 1988; Kim and Ohn 1992), capture votes (Downs 1957; Hauss and Rayside 1978; Kitschelt 1988, 1989; Iverson 1994; Jackman and Volpert 1996; Boudon 1997; Tavits 2004), capture legislative seats (Sartori 1976; Janda 1980; Rochon 1985; Van Cott 2003), capture both votes and seats (Harmel and Robertson 1985; Müller-Rommel 1985), attain a certain level of institutionalization (Huntington 1968; Dix 1992), influence the design and implementation of policy (Wittman 1973, 1983, 1990; Chappell and Keech 1986) and capture votes and seats while influencing policy (Strom 1990). One might also judge a party’s success by whether it has participated in the government (Harmel and Robertson 1985) or simply possessed “blackmail potential” and could impact the formation of the government (Sartori 1976). The choice of how to conceptualize party success is often driven by how one understands the goals of political parties. Do parties want to win office, maximize votes, influence policy, or some combination of these? Aside from whether the best way to conceptualize success is in terms of votes, seats or policy effect, another issue relates to what level of analysis one should focus on. In parliamentary systems where the distribution of seats in the legislature determines the executive, it is understandable that one focuses upon elections for the legislature. In presidential systems, such as the ones in Central America used in this study, one might also measure the success of new parties at the municipal, legislative and presidential levels. While persistence might be a minimal condition of success and winning seats in the legislature is important, winning the presidency would most likely be the ultimate prize (Linz 1990). In presidential systems, the presidency is clearly the most important position and the indivisibility of the office makes it easy to distinguish between success and failure. It might be unfair, though, to limit success to a party’s ability to win a single office, even if it might be the most important office, considering that only one candidate can win the presidency in any single electoral contest. How successful is a political party if it persists, captures a number of votes and seats in the legislature, but never wins the presidency?
39 Just as scholars disagree as to what the appropriate concept of party success should be, they also disagree on the specific thresholds to be used. Hauss and Rayside (1978), although not explicit with regard to an exact threshold, tend to identify parties with around ten percent of the national vote as successful. In a study of left-libertarian parties in Western democracies, Kitschelt (1989) identifies any party attaining four percent or greater of the vote in at least one national election in the 1980s as significant. Boudon (1997) uses a cutoff of five percent in his study of minor party success involving the AD M-19 in Colombia and Causa R in Venezuela. These studies distinguish “successful” parties capturing at least four, five or ten percent of the national vote from “failed” parties that capture anything less. Obviously, arbitrariness is one of the difficulties is choosing a universal cutoff to divide successful from unsuccessful political parties. Kitschelt acknowledges this problem of selecting “a single and universally applicable criterion for judging when parties are electorally significant” (1989: 10). Given that during any single election there are going to be more parties that lose rather than win, Kitschelt argues “a criterion of electoral relevance should not be too stringent because even a small percentage of the vote may shift the balance of power in an election and determine the chance of government formation” (1989: 10). While the number of votes a party receives can influence the outcome of a single contested seat, it can also affect potential coalitions in a parliamentary system or the existence of a majority or supermajority in the legislatures of presidential systems. We see here that the success of a political party in terms of its vote share is related to how other parties perform in the same election. Instead of employing a threshold based upon the percentage of the national vote, other authors characterize political parties as successful if they surpass a certain number of seats in the legislature regardless of the vote share they received. Sartori (1976) says we should disregard parties with no parliamentary seats. Janda (1980) includes in his analysis legal political parties capturing at least five percent of the seats in the legislature in at least two consecutive elections. Rochon limits his study to all parties “that have gained representation in the Dutch Second Chamber at least once” (1985: 423) thereby making implicit another measure of success, attaining at least one seat in one election. In Van Cott’s (2003) study of ethnic parties in Latin America, she includes parties successful in capturing at least one seat in two consecutive elections at any level of political competition. These studies place the focus of a new party’s success on its ability to place its candidates in office, though they disagree whether success
40 should be measured by capturing one or more seats (regardless of the size of the legislature), five percent of the seats (which will vary according to the size of a country’s legislature; election to election in some cases), or capturing seats in two or more elections. Again, we see a problem in that there is no agreed upon minimum criteria of seats or seats and elections by which a party should be deemed successful. While studying new political parties strictly in terms of capturing a certain percentage of the national vote and/or a number of legislative seats can provide us with some understanding of their success as a political party, this approach leaves much to be desired if we do not understand the context in which the competition occurs. It is unclear what substantive difference exists between a new political party receiving three percent of the legislative seats from another receiving six percent. If the governing party captures a majority of the national vote and automatically forms the government in a parliamentary system or the governing party has a majority (or supermajority) of seats in the legislature of a presidential system, three or six percent of the seats by an opposition party is unlikely to make a significant difference. Worse, when selecting “successful” parties based on the fact that they won a single legislative seat, it treats parties equally regardless of the size of the legislature. Is a party that captures one seat in an assembly of 60 equally successful as a party that captures one seat in a legislature with 150 available seats? At a minimum, it would seem that we should take into consideration not only the number of seats, but that number compared to the performance of other political parties in the same legislature during the same electoral contest.
Effective Number of Parties While many authors have been interested in determining the conditions under which new parties are likely to emerge and succeed, other political party research has focused on the number of parties in a system without concerns about specific parties. This research has been concerned with how one should count the number of parties in a system. In distinguishing between two party systems and multiparty systems, researchers have asked whether or not small parties should count. Blondel (1968) devised a four-category system to take into account the number and relative size of political parties (two-party, two-and-a-half, multiparty with a dominant party, multiparty without a dominant party). Blondel’s classification, while praised for bringing
41 attention to “dominant” and “half” parties, has been criticized for the imprecision of the categories (Lijphart 1999). Sartori (1976: 122-23) argues that parties that do not capture seats in the parliament should not be counted regardless of their share of the vote. On the other hand, not all parties with parliamentary seats should be counted either. Sartori discounts the use of an threshold to separate parties that should count from those that should not and instead argues that only those parties that have “coalition potential” (has the party previously participated in government or do the other parties consider it a potential coalition partner?) or “blackmail potential” (relatively large extreme parties that are unlikely to be seen as coalition partners) should be included. Later, Laakso and Taagepera (1979) developed an index for determining the “effective number of parties” in a system that takes into account both the number of parties as well as their 2 relative sizes. The number (N) of effective parties is equal to 1/∑pi , where pi is the proportion of seats for the ith party. If four parties compete in a system with 100 available seats and each party captures 25% of the seats, the effective number of parties is equal to the total number of parties (4). If, on the other hand, the four political parties are not equal in strength (45, 30, 15, 10 percent of the seats) the effective number of parties (3.08) is fewer than the total number of parties with seats (4). While the effective number of parties provides a useful means of comparing the number of parties across time and across political systems, it has not been used to tell us about specific political parties. As mentioned earlier, this has resulted in research on new political parties using a variety of distinct measures.
Party Success in Context In the more established political party systems, the success of new political parties is a relatively rare phenomenon when compared to the success of parties in newly democratizing countries. For that reason, the ability of a new political party to capture even a small percentage of the vote or seats in more established party systems is highly worthy of scholarship. In newer party systems, the emergence and success of new political parties is likely to be a much more common occurrence. A new political party that captures two or three percent of the seats in a new party system is not as dramatic an occurrence as if the party were competing in the United States. In addition, it is also more likely that in these new systems a new political party will
42 capture a significant percentage of the votes and/or seats, potentially becoming one of the largest parties in the country. In terms of this dissertation, a former insurgent group might command the support of several thousand militants and another ten thousand supporters. It is unlikely that such a group would consider itself successful taking a small fraction of the popular vote or winning two or three seats in the assembly. For this reason, it is important to distinguish between those insurgent groups, or any new party for that matter, that go on to be major political parties from those that become minor political parties. In order to determine the success of these new political parties, I propose to evaluate the success of insurgent groups as new political parties based upon the size of the party’s legislative seats relative to the effective number of parties in the system, as measured by Laasko and Taagepera and discussed above. By this standard, a political party can be considered to be successful within the context of its political system if it qualifies as one of the effective parties in that system. Because the number of effective parties is rarely an integer, fractured numbers are awarded to the nearest integer for purposes of evaluating individual parties. For example, in the first example from the previous section, four political parties each had 25% of the seats in the legislature. In this case the effective number of parties (4) equaled the total number of parties (4). In this case, if the new political party is one of these four, it counts as a major party. There would be no minor parties and those that captured votes, but not seats, would qualify as failed parties. In the second example, there were four political parties of differing size (45, 30, 15, and 10 seats). Because the system-wide effective number of parties is 3.08, the first three parties would qualify as major parties; the remaining parties with seats qualify as minor parties, and those that do not win any seats as failed political parties. If, in another system with 200 seats, there are seven parties with a seat distribution of 75, 60, 27, 21, 12, 4, and 1, the effective number of parties for this system is 3.74. Under this scenario, there are four major parties (75, 60, 27 and 21) and three minor parties (12, 4 and 1). Again, all those that capture a percentage of the vote, but do not gain representation in the country’s legislature, are considered failed parties. Unfortunately, this approach does not resolve all concerns. There remains a sense of arbitrariness in the measure given that I round fractured numbers to the nearest integer. I believe the advantages do outweigh the disadvantages. First, the measure will allow one to consistently compare new party performance across party systems more appropriately than any arbitrary threshold separating successful from failed parties. It does not establish cutoffs of two, five, or
43 ten percent of the votes or seats. Second, any static threshold of votes or seats could produce misleading results across systems. This measure evaluates each group’s performance in the context of the other parties in the system and it allows us to distinguish between parties that receive relatively small shares of seats from those that receive relatively large shares (i.e., a new party with twenty percent of the seats should not automatically be considered to be equally successful as a new party with five percent of the seats).
Dependent Variable - Political Party Success As discussed earlier, in presidential systems, one can evaluate the electoral performance of new political parties at the executive, legislative, and municipal level. While persistence might be a minimal condition of success, winning the presidency would most likely be the ultimate prize as the presidency is clearly the most important position. The indivisibility of the office makes it easy to distinguish between success and failure as only one party can win the office. But considering that only one candidate can win the presidency at any single election, I also evaluate the performance of a new political party through both aggregate and proportional vote totals, particularly since winning the presidency might be too much to ask of new parties in the first few elections. In that case, we might view a new political party that increases its vote totals over a series of political contests as being more successful than a new party that loses its electoral share over time, and in a better position to eventually win the presidency. Since the end of each country’s civil war, both insurgent groups have competed in a number of elections for political office. The FMLN has competed in the 1994, 1999 and 2004 presidential elections while the URNG has competed in the 1999 and 2003 presidential elections. My discussion concerning the performance of both groups focuses on the first two presidential elections for two reasons. First, it is easier to compare the two parties’ performances over the same number of elections. Second, and perhaps more importantly, I expect the salience of each group’s insurgent history to weaken over time as an explanation for their electoral performance as new factors come to take precedence, such as how it conducts itself as a political party. In studying the performance of political parties at the legislative level, one also has several options from which to choose. One can investigate the success of a new political party in terms of the percentage of seats or votes it receives, the overall number of seats or votes it wins, or simply whether the party is still in existence. As mentioned previously, measuring political
44 party success in these terms does not provide us with a solid foundation for comparison. Instead, I propose to evaluate the success of these new political parties by taking into account the performance of other political parties and distinguishing between major, minor, and failed political parties. First, I compute the effective number of political parties for the entire party system during the electoral contests under consideration. I compute a score for the effective number of parties following the 1994 and 1997 legislative elections in El Salvador and the 1999 and 2003 legislative elections in Guatemala. Like the discussion of each party’s electoral performance in presidential elections, I focus on the first two elections following the end of each country’s civil 2 war. As mentioned earlier, the number (N) of effective parties is equal to 1/∑pi , where pi is the proportion of seats for the ith party. Then, I rank the political parties according to the percentage of representatives each party has elected to the legislature. In a system where there are four effective parties, the four largest legislative parties are considered major parties. If the effective number of parties is a whole number plus a fraction (e.g., 4.2 or 4.75), I round up for those greater than or equal to 0.5 (4 and 5). Parties that succeed in having their representatives elected to the legislature, but did not finish in the top four or five (in the above examples) are considered minor parties. Finally, those parties that compete in elections, but do not have any of their representatives elected to the legislature, regardless of vote share, are considered failures.
Independent Variables – Explaining the Success of Former Insurgent Groups The hypotheses that I operationalize and measure in the reminder of this chapter relate to the organizational characteristics of the insurgent group and the environment in which it competes as a political party. I hypothesized that an insurgent group’s organizational characteristics are important factors in explaining its future success as a political party. Larger insurgent groups, those with prior political party experience, extensive contacts with mass-based organizations, unity, control of territory and a better record towards civilians relative to government forces are more likely to be successful than smaller insurgent groups, those with little to no experience or relationship with political parties, relatively weak or no relations with mass based organizations, disunity, an inability to control territory and a more violent history towards civilians relative to the government. I also argued that insurgent groups are more likely to succeed if they compete in parliamentary systems. If these new parties must compete in
45 presidential systems, they are more likely to succeed when elections for the executive are determined by majority runoff instead of simple plurality. They are also more likely to succeed when elections for the legislature are determined via proportional representation, district magnitudes are high, a single nationwide district exists, and electoral thresholds are low. Insurgent groups are also more likely to succeed when abstention is high, bipolarization is weak, the economy is in poor shape, and parties with similar platforms do not exist. Finally, an identity-based insurgent group is more likely to succeed if the ethnic, racial, or linguistic people it represents, comprises a large percentage of the population. In the remainder of this section, I discuss how I operationalize and measure the variables needed to test these hypotheses.
Hypothesis 1 (Insurgent Size) The first hypothesis states that larger insurgent groups are more likely to be successful than smaller groups. In order to test this hypothesis, I discuss the number of insurgents each group maintains during the course of the war. Given that wartime insurgent numbers are difficult to come by, and often biased, I rely more heavily upon the total number of insurgents demobilized at the of the war. In many cases, insurgents demobilize with the help of the international community, leaving behind more reliable estimates. The FMLN and URNG are two examples of insurgent groups that have demobilized with the assistance of the international community. While the number of insurgents demobilized at the end of the war might over- or underestimate the level of support the group maintained throughout the course of the war, I believe the size of the insurgent group at the end of the war is a more telling indicator of potential success than either the number at its height or an average throughout the conflict. We still need to take care in interpreting these numbers. The total number of insurgent combatants can be misleading if we do not take into consideration that size of the country’s population. Therefore, in addition to absolute numbers, I interpret the size of the insurgent forces as a ratio to the entire population (insurgents per 100,000 people). This will enable us to compare insurgent group size across groups and countries while taking into consideration their population. The primary source used to identify the total number of insurgents demobilized as part of the peace accords in Guatemala and El Salvador is Luciak (2001), although I also make use of reports from Dunkerley (1985), Halloran (1987), Wickham-Crowley (1992), Pearce
46 (1993), Bracamonte and Spencer (1995), LeBot (1995), McClintock (1998), Spence et al. (1998), and Stanley and Holiday (2002).
Hypothesis 2 (Political Party) The second hypothesis states that insurgent groups with prior political party experience are more likely to be successful than those without prior political party experience. In order to test this hypothesis, I investigate whether the insurgent group’s members or leaders had prior experience in political parties and electoral politics. In addition to whether an insurgent group with a prior political party history is more successful than another insurgent group without such a history, I also investigate whether the insurgent group maintained a political-wing during the civil war. Insurgent groups with prior political party experience or a political-wing are more likely to succeed than insurgent groups with no prior political party experience or political-wing. In the case of the FMLN in El Salvador, I rely on secondary sources (Wickham-Crowley 1992; Dunkerley 1988; Montgomery 1982; Spence, Dye, and Vickers 1994; Baloyra-Herp 1995; Lungo Uclés 1996; Ribera 1996; McClintock 1998; Zamora 2003; Brockett 2005) and a number of interviews I conducted. In the case of Guatemala, I also rely upon secondary sources (Wickham-Crowley 1992; Dunkerley 1988; Luciak 2001; Sichar 1999; Jonas 1991) and a number of interviews I conducted in Guatemala.
Hypothesis 3 (Mass-Based Organizations) The third hypothesis states that insurgent groups that had more extensive contacts with mass-based organizations during the war are more likely to be successful than those without such contacts. In order to test this hypothesis I investigate the relationship between the insurgent movement and mass-based organizations. Mass-based organizations include labor unions, peasant confederations, student groups, and other popularly based interest groups. If the insurgent group and mass organizations coordinated activities in pursuit of a common goal during the war, I would argue that this qualifies as extensive contacts. Insurgent groups with extensive contacts should be more successful as political parties compared to insurgent groups with few or no contacts. In order to test this hypothesis, I also rely upon the secondary sources and interviews identified earlier.
47 Hypothesis 4 (Unity) The fourth hypothesis states that insurgent groups that can maintain unity and coherence during the war, negotiations, and in the postwar period are more likely to be successful than insurgent groups that fracture. In order to test this hypothesis, I look at whether the insurgent group loses key leaders, members, or groups during the conflict, the negotiations, or in the postwar period as a result of internal differences or co-optation by other actors. Looking ahead, insurgents from both the FMLN and URNG left the organizations following the first election in which each party participated to form (in the case of the FMLN) and join pre-existing (in the case of the URNG) parties. I assess the extent to which this fracturing affected the FMLN and URNG by simulating the electoral results that the group would have achieved in the second election (1997 in El Salvador and 2004 in Guatemala) had the original insurgent groups and breakaway parties presented a single slate of candidates. The material used to test this hypothesis also involves interviews and secondary sources as well as election results from the Supreme Electoral Tribunals in each country.
Hypothesis 5 (Control of Territory) The fifth hypothesis states that insurgent groups that controlled territory during the war are more likely to be successful than groups that are unable to control territory. In order to test this hypothesis, I first compare the geographic reach of the FMLN and URNG in capturing and controlling territory. This analysis is performed primarily at the departmental level and makes use of secondary sources. The second test of this hypothesis is the focus of chapter 7 (along with the hypothesis on human rights violations) which I discuss in greater detail in that chapter. In chapter 7, I systematically assess the impact of FMLN controlled zones on the popular vote of the FMLN. As mentioned previously, if capturing and controlling territory is important to explaining the success of former insurgent groups as political parties, then we should also notice a difference in the voting patterns from these areas to the rest of the country. FMLN controlled areas should support the new political party in greater proportion than areas of the country where the FMLN had little presence.4
4 For El Salvador, I code those municipalities where the FMLN prevented elections from being conducted in the first round of the 1984 presidential elections as controlled zones. This is consistent with the approach by McClintock (1998) and includes 59 municipalities. In addition, McClintock supplements these municipalities with others where elections were held, but the mayor was forced to govern in exile because of the presence of the FMLN.
48
Hypothesis 6 (Violence) The sixth hypothesis states that insurgent groups that are less violent toward civilians than are government forces are more likely to be successful than those insurgent groups that are more violent than government forces. In order to test this hypothesis, I discuss the general level of violence at the department and municipal levels during the war using the Truth Commission report from each country. Chapter 7 conducts a secondary test of this hypothesis using data from El Salvador to explore how the geographic distribution of violence throughout El Salvador impacted the FMLN vote in 1994. I make use of a survey by Seligson and McElhinny (1996) to explore the relationship between the violence that occurred in each department and the eventual vote for the insurgent political party.5 In addition, I make use of the National Reconstruction Program’s distinction between municipalities identified as “ex-conflict zones” and “non-conflict zones.” Ex-conflict zones were the areas of the country deemed most seriously affected by the war and in need of reconstruction monies. I discuss both measures in greater detail in chapter 7.
Hypothesis 7 (Electoral Objective) The seventh hypothesis states that former insurgent groups are more likely to succeed in parliamentary than presidential systems. In order to test this hypothesis, I simply look at whether the political system is presidential or parliamentary. In presidential systems, the executive is elected directly by the voters. In parliamentary systems, the executive is not elected directly by the citizens, but by members of the legislature.
Hypothesis 8 (Presidential Electoral Rules) The eighth hypothesis states that former insurgent groups are more likely to succeed in systems with majority presidential runoffs than simple plurality. In order to test this hypothesis, I simply determine whether presidential election rules are determined via simple plurality or majority runoff. Under simple plurality rules, the candidate with the most votes is elected
Unfortunately, when I contacted her, McClintock was unable to locate the original list of municipalities though she did say that it was not very different from simply using the municipalities, where elections were not held. 5 Seligson and McElhinny (1996) published the results of a survey conducted in El Salvador that provides a basic understanding of the geographic reach of the conflict. In the article, the authors present the results of a questionnaire asking respondents whether a family member had been killed as a result of the war.
49 president. Under majority runoff rules, the president-elect must receive 50% of the vote plus 1. If no candidate surpasses 50%, a second round is held between the top two vote getters.
Hypothesis 9 (Legislative Electoral Rules) The ninth hypothesis states that former insurgent groups are more likely to succeed under proportional representation electoral rules than under single member district voting in elections for the country’s legislature. In testing this hypothesis, I perform two tests. First, while elections to determine legislative representation in El Salvador and El Salvador are determined via proportional representation, elections for municipal office are determined via single-member district voting. I compare the electoral performance of each new political party in municipal elections and legislative elections to determine whether the use of proportional representation at the legislative level is significantly different from the outcome of elections determined via single member district voting in municipal elections. Second, I compare the actual electoral performance of each new party in elections for the legislature against simulated results using different electoral formulas. In the case of El Salvador, votes are translated into seats via the largest remainder formula. In Guatemala, votes are translated into seats using the D’hondt method. I simulate the electoral performance of the FMLN and URNG under the different formulas to determine whether the electoral formula had a significant impact on each party’s performance.
Hypothesis 10 (District Magnitude) The tenth hypothesis states that former insurgent groups are more likely to succeed when district magnitudes are high. In testing this hypothesis, one expects that new parties will be more likely to succeed in systems with a high average district magnitude versus those that compete in systems with a low average district magnitude. First, I compare the average district magnitude in El Salvador and Guatemala. The second test involves simulated electoral conditions. Both El Salvador and Guatemala use two-tiered electoral districts, with a number of seats allocated at the department level and a separate list of seats allocated through a national list. One of the advantages of this system is that it combines department and national level representation. There are a number of deputies to represent the entire nation and others to represent the geographic constituents. There are also a number of disadvantages, including the fact that political parties
50 need to compete both locally and nationally. This taxes limited resources, benefiting those large parties with a national presence. Therefore, it might benefit new, smaller parties if there were no national district and instead those seats were reapportioned at the departmental level. In order to test the implication of district magnitude, I simulate electoral results from El Salvador’s 1994 election and Guatemala’s 1999 to determine whether the existence of a national list helped or hurt the FMLN and URNG.
Hypothesis 11 (Electoral Threshold) The eleventh hypothesis states that former insurgent groups are more likely to succeed in systems with low electoral thresholds. In order to test this hypothesis, I examine whether the existence of electoral thresholds has inhibited the success of these former insurgent groups. The data used to test this hypothesis comes from the Supreme Electoral Tribunals in El Salvador and Guatemala.
Hypothesis 12 (Abstention) The twelfth hypothesis states that former insurgent groups are more likely to succeed in systems with high levels of voter abstention. In order to test this hypothesis, I will investigate the overall levels of voter abstention in the elections prior to the first participation of the former insurgent group. High levels of abstention should bode well for a new political party.6 The data used to test this hypothesis comes from the Supreme Electoral Tribunals in El Salvador and Guatemala.
Hypothesis 13 (Bipolarization) The thirteenth hypothesis states that former insurgent groups are more likely to succeed in presidential systems when bipolarization has not yet set in. In order to test this hypothesis, I look at the party system prior to the first electoral competition of the former insurgent groups to determine whether two political parties dominated the vote. The greater the percentage of the
6 In addition, I also expect to find that based upon these statistics, former insurgent groups should perform better in those districts that suffered from the highest level of voter abstention in the previous elections. These would be the districts least committed to the reigning political parties and, arguably, most open to new alternatives. I test and discuss this hypothesis as it relates to the electoral performance of the FMLN in chapter 7.
51 national vote garnered by the two largest parties, the less likely a new party will succeed.7 The data used to test this hypothesis comes from the Supreme Electoral Tribunals in El Salvador and Guatemala.
Hypothesis 14 (Economic Conditions) The fourteenth hypothesis states that former insurgent groups are more likely to succeed when there is an economic crisis (which voters do not blame on the group). In order to test this hypothesis, I compare the economic conditions in El Salvador and Guatemala prior to the first election in which each party competed. In particular, I focus on the period between the signing of the peace accords and first postwar elections. The data used to test this hypothesis comes from the Penn World Tables.
Hypothesis 15 (Social Cleavages) The fifteenth hypothesis states that identity-based insurgent groups are more likely to be successful if they represent a significant percentage of the population. I first identify whether the insurgent group has an identity component – ethnic, religious, or linguistic – and what percentage of the population that identity group comprises. I also look at such factors as whether the leaders of the group and its rank-and-file emerge from this population. I also investigate the relationship that the insurgent group maintained with the larger indigenous population. The data used to test this hypothesis comes from Jonas (1991), the Commission for Historical Clarification’s Guatemala: Memory of Silence (1999), Sichar (1999), Stanley and Holiday (2002), and Brockett (2005),
Hypothesis 16 (Existing Parties) The sixteenth hypothesis states that insurgent groups are more likely to be successful if parties with similar programs do not already exist. In order to test this hypothesis, I explore the relationship between the insurgent party and parties with similar platforms. In many ways, this hypothesis is related to Hypothesis 2, which suggested that insurgent groups with a political
7 In addition, I also expect to find that based upon these statistics, former insurgent groups should perform better in those districts with low levels of bipolarization in the previous elections. These would be the districts least committed to the two largest political parties and, arguably, most open to new alternatives. I test and discuss this hypothesis as it relates to the electoral performance of the FMLN in chapter 7.
52 wing or relationship with an existing political party should be more likely to succeed as a new political party than an insurgent group without these relationships. In this hypothesis, I am more interested in whether the two (or more) are competing for the same votes. In order to test this hypothesis, I combine the vote totals for the insurgent party and the pre-existing party to determine whether the insurgent group would have been more successful without the existence of the other party or whether their existence made no difference at all.8 The data used to test this hypothesis comes from Spence, Dye and Vickers (1994), Sichar (1999), Reding (1998), and the Supreme Electoral Tribunals in El Salvador and Guatemala.
CONCLUSION
Over the past twenty years, a number of insurgent groups have attempted the transition from insurgent group to political party. In this chapter, I discussed how I operationalize the variables and test the hypotheses developed in chapter 2. Before moving to a test of these hypotheses, the following chapter will provide a brief overview of the history of the FMLN and the URNG as insurgent groups and political parties.
8 In addition, I also expect to find that based upon these statistics, former insurgent groups should perform better in those districts with relatively weak support for like-minded political parties based upon results from the previous elections. I test and discuss this hypothesis as it relates to the electoral performance of the FMLN in chapter 7.
53 CHAPTER 4 INSURGENT GROUPS
Frente Farabundo Martí para la Liberación Nacional (FMLN) Throughout most of the twentieth century, an alliance between El Salvador’s armed forces and the country’s economic elite ruled El Salvador. In the early 1970s, a small number of armed opposition groups began to challenge this alliance. In 1970, the Fuerzas Populares de Liberación – Farabundo Martí (FPL), or Popular Forces of Liberation – Farabundo Martí, El Salvador’s oldest guerrilla group, was formed. Many of the FPL’s initial recruits were members of Christian base communities and union activists who had suffered under the repression of the government’s security forces. The FPL supported a prolonged popular war based on the Chinese experience – meaning that they would focus on gaining momentum and support in the countryside before moving to the city. It remained the largest guerrilla group until 1983. The Ejército Revolucionario del Pueblo (ERP), or People’s Revolutionary Party, formed in 1972, two years after the appearance of the FPL. Unlike the other insurgent groups, the ERP consisted of a large middle class and female following. While the FPL supported a prolonged popular war, the ERP believed that is was important for the people to experience the failure of reformism before they would accept and support mass insurrection. By the mid-1980s, the ERP had become the largest of the revolutionary groups in El Salvador, accounting for up to 40% of all armed guerrillas. Another insurgent group, the Fuerzas Armadas de Resistencia Nacional (FARN or RN), or Armed Forces of National Resistance, grew out of a division within the ERP in 1975. Similar to the ERP, though, it pursued a military strategy of urban warfare. The Partido Revolucionario de los Trabajadores Centroamericanos (PRTC, or Revolutionary Party of the Workers of Central America, was formed in 1976 and operated primarily in the capital of San Salvador. The PRTC initially formed because of its dissatisfaction with the strategies and performances of the already existing guerrilla groups. The PRTC also maintained more of a regional outlook. Finally, the Fuerzas Armadas de Liberación (FAL), or Armed Forces of Liberation, formed during the 1977-79 period. The FAL consisted primarily of members from the Partido Comunista de El Salvador (PCS), or Communist Party of El Salvador. The FAL was the last group to surrender the possibility of achieving victory through elections before taking up the cause of armed revolution.
54 The call by these insurgent groups to violent opposition, though, remained largely overshadowed by more moderate, nonviolent opposition efforts. Though many Salvadorans believed change necessary, most clung to the hope that it could occur without resorting to revolutionary violence and through political parties, strikes, and protest marches.9 Growing numbers would become more open to the revolutionary alternative as the decade progressed. Political and social mobilization became more radicalized following fraudulent elections in the 1970s. In 1972, when it became apparent that the candidate of the Unión Nacional Opositora (UNO), or National Opposition Union, José Napoleón Duarte, was going to win the presidential elections, the military intervened and announced it candidate the victor. The 1974 municipal and legislative elections and the 1977 presidential elections were again marred by fraud and violence (Montgomery 1995: 67). With each successive election, an increasing number of protests took to the streets. The government’s response contributed to this radicalization, as official (military and police forces) and unofficial (death squads) security forces responded to this mobilization with even greater levels of repression. Finally, the denial of political and social change through elections and other less confrontational collective action forced increasing numbers of Salvadorans to support a socialist revolution, as they saw “no other way out” (Goodwin 2001). In December 1979, the FPL, RN and PCS established the Coordinadora Político Militar (CPM), or Political-Military Committee, as a first step in the direction of unity for the Salvadoran insurgents. In May 1980, the Salvadoran guerrilla groups began to coordinate their activities in the form of the Dirección Revolucionario Unificada (DRU), or Unified Revolutionary Directorate. In October, these groups formed the Frente Farabundo Martí para la Liberación Nacional (FMLN), or Farabundo Martí National Liberation Front.10 Months later, in January 1981, the FMLN hurriedly launched its “final offensive” in hopes that the Salvadoran people would simultaneously rise up in conjunction with its military offensive and overthrow the government. While the offensive might have been spectacular on its own, relative to the recent success of the Frente Sandinista de Liberación Nacional (FSLN), or Sandinista National Liberation Front, in Nicaragua, it was unsuccessful. The offensive failed as the various insurgent groups did not coordinate with each other and the intended mass insurrection never
9 See Brockett (2005) for a good discussion of “nonviolent” challenges to the Salvadoran state in the 1970s and 1980s. 10 In May 1980, the ERP, FPL, FARN, PRTC and FAL formed the DRU under heavy prodding from Cuba. Shortly thereafter the FARN withdrew from the organization, returning shortly before the unification of the FMLN in October.
55 materialized.11 Following the offensive, the FMLN would pull back from the capital to the rural areas of the country, from where it would fight for the next decade. Following this (first) failed final offensive, the FMLN continued to engage in large-scale attacks against the nation’s military and infrastructure. By the mid-1980s, though, the FMLN had to adjust its strategy because of the Salvadoran military’s increasing use of airpower provided by the United States. The FMLN broke its forces down into “smaller, more mobile and autonomous groups” and moved away from conventional engagements to “intensifying economic sabotage” (Wood 2003: 28). For the next several years, the Salvadoran military and the guerrillas continued fighting although they were involved in sporadic peace talks under the mediation of the Salvadoran Catholic Church. These negotiations failed for several reasons, including the intransigence of each party – the Salvadoran oligarchy and military, the FMLN, and the United States government – toward negotiations. Each believed that it could still secure the military defeat of the other and that anything less was in fact, a defeat. By the late 1980s, changing domestic and international conditions provided an environment that was more conducive to political solution. Domestically, El Salvador had undergone a number of transformations. The FMLN had initially taken up the cause of armed revolution in the late 1970s and early 1980s at a time when an alliance between the military and the economic elite dominated the country. Even with the election of a civilian, José Napoleon Duarte of the Partido Demócrata Cristiano (PDC), or Christian Democratic Party, to the presidency in 1984, it is not entirely an understatement to say that power remained in the hands of the military, the economic elite, and the United States Embassy. While Duarte negotiated with the guerrillas on a number of occasions, his power was weakened by the fact that the three most powerful groups in the country were against negotiations. With the election of Alfredo Cristiani of the Alianza Republicana Nacionalista (ARENA), or Nationalist Republican Alliance, party in 1989, the presidency finally arrived in the hands of an individual that represented the more powerful sectors of the country, albeit the more “moderate” sectors of the conservative economic elite tied to international capital. Unlike Duarte, Cristiani was more likely to be able to pressure both the elites and military into accepting the benefits of a negotiated settlement to the war, if he so desired.
11 There are several reasons for the failure of the offensive, including the fact that the FMLN leadership differed in their understanding of whether the conditions were appropriate for an attack at that moment; the RN did not participate; and the ERP did not share its weapons with the other insurgent groups (McClintock 1998: 54).
56 In November 1989, the FMLN launched its second “final offensive” pushing deep into the heart of the capital, San Salvador, and many of its exclusive neighborhoods. Before the eventual FMLN retreat, it is likely that the parties to the conflict learned several lessons. The FMLN learned that while they could still bring enormous power to bear in any single campaign, military victory was not within sight and the mass insurrection that it had counted upon to complement its military offensive would not materialize, just as it had failed in 1981. For the Salvadoran elite, the offensive brought the war right to its doorstep and disrupted their daily routine when the FMLN was able to move into and occupy the exclusive neighborhoods of Escalón and San Benito. In addition, while the Salvadoran military had been known to bomb poorer neighborhoods, particularly in the countryside, during the FMLN offensive, the military resorted to an aerial bombardment of the capital. The FMLN offensive brought home to many that there was no chance of a quick military victory by either the guerrillas or the Salvadoran military. The FMLN offensive and the military’s response also had international repercussions. During the 1989 offensive, the Atlacatl Battalion, upon orders from the Salvadoran military’s High Command, entered the grounds of the Universidad Centroamericana “José Simeón Cañas” (UCA), or University of Central America “José Simeón Cañas,” and murdered six Jesuit priests, their housekeeper and her daughter. After several years of US-training and several billion dollars in economic and military aid, the Salvadoran military could not defeat the insurgents or prevent them from attacking and occupying exclusive neighborhoods in the capital. In addition, when challenged, the military reverted to bombing civilian neighborhoods and murdering religious figures. The FMLN and the US-backed Salvadoran military had fought to a stalemate, whereby any change in the balance of forces was unlikely in the near future and would most likely require a significant investment (or disinvestment) on the part of international benefactors – the United States, Cuba or the Soviet Union. As a result, pressures for a negotiated settlement to the conflict were brought to bear on each party to the conflict. After eighteen months of negotiations involving the United Nations, the FMLN and the government of El Salvador, a negotiated settlement to the civil war was celebrated at a ceremony at Chapultepec Castle, Mexico City on January 16, 1992. While the Peace Accords addressed social, economic, military, and police reforms, another lasting legacy of the accords has been to further the democratization of the country by facilitating the transformation of the FMLN into a
57 political party. The importance of this event is brought home by the fact that roughly 90% of the Peace Accords addressed the political conditions of the state and the FMLN’s insertion into it as a competitive political party (Zamora 2002). According to the Peace Accords, the legislature was supposed to pass a law decreeing official recognition of the FMLN as a political party. Instead, the FMLN pursued legal recognition as required by the constitution for all other new political parties. The FMLN collected over 3,000 signatures in less than fifteen days (Zamora 2003: 56). Several weeks later, on December 14, 1992, the Supreme Electoral Tribunal officially inscribed the FMLN enabling it to compete for the first time as a political party in the 1994 presidential, legislative, and municipal “elections of the century.”
FMLN as Political Party The FMLN has competed in a number of elections for the presidency (1994, 1999 and 2004) and the Legislative Assembly (1994, 1997, 2000 and 2003) since the conclusion of the war. Table 4.1 presents the electoral results of the country’s presidential elections in these years while Table 4.2 presents the electoral results from the legislative elections. Table 4.3 presents the number and percentage of seats won by all parties in El Salvador’s Legislative Assembly. In the 1994 “elections of the century,” the FMLN captured 21% of the national vote. This translated into twenty-one of the National Assembly’s eighty-four seats and immediately transformed the FMLN into the country’s second largest political party. At the same time, the incumbent party and main adversary of the FMLN, ARENA, won thirty-nine seats in the assembly. After a decade of war, this electoral performance propelled the FMLN into a solid legislative position through which to pursue its political program in the postwar period. At the presidential level, the FMLN entered into a coalition with the center-left Convergencia Democrática (CD), or Democratic Convergence, party. The CD consisted of many members of the Frente Democrático Revolucionario (FDR), or Democratic Revolutionary Front, a political ally of the FMLN during most of the war. In the elections, neither the FMLN-CD candidate, Rubén Zamora, nor the ARENA candidate, Armando Calderón Sol, surpassed the 50% threshold required for a first round victory. This forced a runoff election where Zamora, after attaining 25% in the first round, captured 32% in losing to Calderon Sol.
58 Three years later in the 1997 elections, the FMLN captured twenty-seven seats, which translated into roughly one-third of all assembly seats. Though the FMLN “only” captured six more legislative seats in 1997 than 1994, the results should not be underestimated. During the 1994-1997 legislative session, seven deputies broke ranks with the party lowering its number of deputies to fourteen.12 The 1997 elections saw the FMLN not only survive the loss of a number of its historic leaders and the airing of much of the party’s dirty laundry from the war, but almost double its actual deputies from fourteen to twenty-seven in becoming the second largest party in El Salvador’s Legislative Assembly. During the process of choosing its candidate for the 1999 presidential election, the FMLN again presented itself as a divided party, unable to manage its internal differences and poorly prepared to resolve the country’s problems. The departure of seven deputies in 1994 did not put an end to the divisions within the new political party. After a tainted primary process, Facundo Guardado captured fewer than 30% of the vote for the FMLN in the national election. Though Guardado’s total votes surpassed those captured by Zamora in 1994, the election was a major setback for the FMLN as it was unable to capitalize upon its strong showing in the 1997 mid- term elections. In addition, unlike 1994, the FMLN failed to force a second round of voting as the ARENA candidate, Francisco Flores, attained over 50% of the vote. In 2000, the FMLN rebounded from the previous year’s poor showing in presidential elections. For the third consecutive legislative election, the FMLN increased both its vote and seat share in winning a plurality of seats with thirty-one. The FMLN was followed closely by ARENA, which captured only twenty-nine seats despite having won a slightly higher share of the total national vote relative to the FMLN (36% to 35%). These elections were a remarkable accomplishment, as for the first time and in only its third attempt, the FMLN became the largest party in the country’s unicameral legislature. The FMLN continued its legislative success in the most recent legislative elections of March 2003 retaining the same number of deputies (thirty- one) and increasing its lead over ARENA from two to five seats. The success of the FMLN in the election was somewhat offset by the fact that its percentage of the total vote declined by about 2%.
12 Shortly after the legislature convened in 1994, Joaquín Villalobos, Ana Guadeloupe Martínez, and other members associated with the FARN and the ERP separated from the FMLN because of strategic, tactical and ideological differences related both to the conflict and the party’s new role in the postwar period. Many of these individuals would form the Partido Demócrata (PD), or Democratic Party, and compete in the 1997 elections.
59 The FMLN was poised to mount a serious challenge in the March 2004 presidential elections after a strong showing in the 2003 legislative and municipal elections. Similar to 1999 when the FMLN selected the former comandante Facundo Guardado, the FMLN chose another former comandante, Schafik Handal, as its candidate. Handal’s candidacy was not well received according to national public opinions polls prior to the elections. Unfortunately for the FMLN, the party leadership did not place much faith in polling data. As in 1999, the FMLN was criticized for selecting a candidate directly tied to its revolutionary period and not a candidate with widespread support amongst the general voting population. Whereas ARENA has successfully modernized its image with young presidential candidates having no connections to the civil war era or to right-wing deaths squads of the 1980s, the FMLN continues to struggle with modernizing its image and promoting new leadership, thus leaving itself open to these sorts of attacks. In addition, the continued and obvious financial advantage of the incumbent party, as exemplified in its painting of the country red, white and blue (the tricolors of ARENA), was another decisive factor in the campaign. The campaign was a particularly intense and dirty one, dominated in the print and electronic media with images and references to Handal’s role in the violence of the 1980s and veiled threats from US functionaries as to the potential repercussions of an FMLN victory. Though many had predicted that no candidate would receive a majority in the first round and that a runoff would be needed, Elías Antonio Saca of ARENA soundly defeated Handal in the first round with 57% to Handal’s 36% of the vote. Twelve years after the signing of the Peace Accords, the FMLN performance at the executive level has been decidedly mixed. On the positive side, in three presidential elections, the FMLN has increased its vote totals from just over three hundred thousand in the first round of the 1994 election to over eight hundred thousand in 2004. The FMLN has also successfully increased its share of the national vote from 25% to 29% to 36%, a significant achievement. These performances have solidified the FMLN as the primary alternative to the governing ARENA party, a success experienced by few former insurgent groups. On the negative side, after forcing a second round in the 1994 elections while in coalition with other opposition groups, the FMLN failed to make it to a second round in either 1999 or 2004. While the FMLN vote total and share of the national vote have indeed increased since 1994, the party has been defeated by over 20% in each presidential election. And in the most recent election, while polls indicated that a majority of the population wanted an alternative to ARENA, the FMLN could
60 not convince the majority that they were the alternative. At the legislative level, the success of the FMLN is more clear-cut. After securing the second most seats in the 1994 legislature, the FMLN continued to improve and it is the largest party in the country’s legislature today heading into the 2006 elections.
Unidad Revolucionaria Nacional Guatemalteca (URNG) Guatemala’s civil war was both the longest and bloodiest of the Central American conflicts. From 1960 until 1996 an estimated 200,000 Guatemalans, mostly indigenous Mayans, were killed or “disappeared.” Like the guerrillas in El Salvador, the Guatemalan guerrilla groups fought independently for several years before Cuban prodding encouraged their unification. The Fuerzas Armadas Rebeldes (FAR), or Rebel Armed Forces, began military activities in the early 1960s, making it the oldest guerrilla group in Guatemala. The FAR counted on a maximum of 500 combatants in the mid-1960s, but its military inexperience and poor weapons nearly led to its extinction by the Guatemalan military by the early 1970s. Most of the surviving FAR retreated to the jungle region of El Petén. For the remainder of the decade, the FAR tried to remain out of the Guatemalan military’s crosshairs. Some FAR militants managed to regroup in the late 1970s and, by 1980, had restarted operations in the capital “with a series of kidnappings, assassinations, and bombings” (Brockett 2005:118). However, by the time that the FAR had begun to re-launch its attacks against government personnel and infrastructure, new organizations had surpassed it in terms of effectiveness (Gross 1995: 107). Like El Salvador, the Guatemalan communist party had a long history of opposition to the military regime, but for the most part, it tended to oppose the government through non- revolutionary activity. The Partido Guatemalteca del Trabajo (PGT), or Guatemalan Labor Party, the country's communist party, was legalized in 1951 during Jacobo Arbenz’s democratic government. Following the overthrow of the Arbenz government and imposition of military rule, the now illegal PGT worked mostly with mass-based organizations, particularly those affiliated with the Universidad de San Carlos de Guatemala, or University of San Carlos (Brockett 2005: 117). Those who supported violent confrontation with the regime would join the FAR (Jonas 1991: 136; Brockett 2005: 117-118). By the end of the late 1970s, the PGT/FAR had “formally abandoned guerrilla warfare,” (Gross 1995: 110) though some of its members
61 broke with the organization to for a “new PGT devoted to armed struggle” after government repression increased again in the late 1970s (Brockett 2005: 117-18). Two organizations dominated the armed struggle in the 1970s, “both of which had leadership that had broken years before from FAR and who were determined not to repeat the errors of the 1960s…To these new organizations, revolution would not be ignited by guerrillas confronting the military in the mountains through a Guevarista foquista strategy. Instead, successful revolution would require years of preparation working directly with the country’s poor, especially the vast majority in the countryside” (Brockett 2005: 119). The Organización del Pueblo en Armas (ORPA), or Organization of the People in Arms, began operations in 1971. Throughout the 1970s, ORPA operated in relative secrecy near Lake Atitlán and on the country’s southern coast. ORPA focused on cultivating relationships with the country’s indigenous community and urban labor movements and it “emphasized the need to make broad alliances with progressive middle-class intellectuals and professionals” (Jonas 1991: 138). As a result, ORPA became one of Guatemala’s two strongest guerrilla groups, counting on the support of over 2,000 combatants. A fourth insurgent group, the Ejército Guerrillero de los Pobres (EGP), or Guerrilla Army of the Poor, began activities in 1972. Like the EGP and ORPA, the EGP also formed because of dissatisfaction with the FAR. The EGP focused on the needs of the country’s indigenous population and successfully built a social base amongst these groups in the 1970s. Until the Guatemalan military’s “scorched earth” campaign of the early 1980s, the EGP had become Guatemala’s strongest guerrilla group. After several years of working independently of each other, the FAR, ORPA, EGP, and the PGT began to coordinate their activities as the Unidad Revolucionaria Nacional Guatemalteca (URNG), or Guatemalan National Revolutionary Unit, on February 8, 1982. Over the next decade, the URNG would struggle against the formidable Guatemalan military as well as with itself. The guerrillas were eventually weakened by a series of military defeats and assassinations, the apparent failure of Marxism-Leninism as a viable alternative to the liberal- democratic order, and their failure to develop a broad based opposition to the military, and then civil-military, regimes. After several years of negotiations, the URNG and the government of Alvaro Arzú of the Partido de Avanzado Nacional (PAN), or National Advancement Party, signed the Firm and Lasting Agreement in December 1996, bringing an end to thirty-six years of
62 civil war. The guerrilla groups that formed the URNG in 1982 officially began the legal progress of becoming a political party in June of 1997. The URNG set out to develop a party that was “revolutionary, democratic, and pluralist [in that] it will promote the incorporation to its ranks all those who share progressive, patriotic and democratic ideals” (Sichar 1999: 86).13 The URNG received its formal inscription as a political party in December 1998 and began preparations for the following year’s elections.
URNG as Political Party The URNG has competed in two elections for the presidency and the legislature (1999 and 2003) since its transition to political party. Table 4.4 presents the electoral results of the country’s presidential elections in these years while Table 4.5 presents the electoral results from the legislative elections. Table 4.6 presents the number and percentage of seats won by all parties in the Guatemalan Congress. In 1999, the URNG prepared to form part of the Alianza Nueva Nación (ANN), or New Nation Alliance, along with the Frente Democrático Nueva Guatemala (FDNG), or Democratic Front for a New Guatemala, the Unidad de Izquierda Democrática (UNID), or Democratic Leftist Union, and the Desarrollo Integral Auténtico (DIA), or Authentic Integral Development, party for legislative and presidential elections. The alliance would have allowed the URNG and the Guatemalan left to capitalize on the previous electoral success of the FDNG which had elected six deputies to the Congress in 1995. But the leftist coalition fractured when the FDNG withdrew from the alliance in the midst of the 1999 campaign after complaining that the other members of the ANN were engaged in “discriminatory and undemocratic political maneuvering” (Holiday 2000: 82). In the presidential election, the ANN candidate, Álvaro Colom Caballeros, won just over 12% of the vote and finished in third place, well behind the Frente Republicano Guatemalteca (FRG), or Guatemalan Republican Front, and the PAN. In the congressional elections, the URNG, as part of the ANN, won nine congressional seats in accounting for 8% of the congress’s 113 seats. For many new political parties, 8% of the seats would be quite an accomplishment. According to Rodrigo Asturias (nom de guerre Gaspar Ilom), the URNG was indeed pleased with the results given the difficulties the organization
13 The translation comes from “revolucionario, democrático, y pluralista [en el que] promoverá la incorporación a sus filas de todos aquellos a quienes nutren ideales progresistas, patrióticos y democráticos” (Sichar 1999: 86).
63 confronted in competing as a political party for the first time (Asturias 2004). But given that the majority party led by former dictator Efrain Ríos Montt (FRG) captured sixty-three seats (56%) and the second most successful party (PAN) captured thirty-seven seats (33%), not all shared Asturias’ optimism. David Holiday (2000: 82) argues that the outcome of the elections “nevertheless essentially translated its military defeat into political defeat.” Like the FMLN, the URNG presented one of its historic leaders as its presidential candidate in its second attempt at the presidency after choosing an individual loosely tied to the insurgency in its first attempt. In November 2003, the URNG selected Rodrigo Asturias, the son of Nobel Literature Laureate Miguel Angel Asturias as its candidate. Asturias might have been the URNG candidate in the 1999 elections had his group, ORPA, not been implicated in a kidnapping scheme during the final stages of the peace negotiations. In the eyes of the electorate, this undermined his commitment and that of the party to the peace process and democracy (García 2004). After a disappointing campaign, Asturias finished in sixth place, capturing less than 3% of the national vote. As the URNG’s performance dropped from 12% to 3%, so too its vote total dropped by over 200,000, leaving it with fewer than 70,000 votes. The elections for the country’s legislature in 2003 were similarly disappointing for the URNG. In November, the URNG captured 4% percent of the vote for the legislature and less than 3% for the presidency. In the congressional elections, the URNG only managed to capture two seats and dropped into a tie for seventh. The URNG performance was considerably worse than simply capturing fewer seats relative to 1999, as the total number of seats available had increased from 113 to 158 in the intervening years. On the one hand, the 1999 electoral results were poor for an organization that (1) could trace its roots back nearly forty years, (2) had seventeen years of history together as the URNG, (3) at one time had counted on the support of several hundred thousand Guatemalans, and (4) had been propelled into the national spotlight during several years of negotiations involving the United Nations leading up the signing of the peace accords. On the other hand, the URNG as part of the ANN attained roughly 10% of the vote for the legislature and presidency. Ten percent is very good for a new political party without any previous electoral experience. The electoral results in 2003 were disappointing, though, and require the URNG to reevaluate its role and future in the Guatemalan political process if it is not already too late. The next elections are
64 scheduled for 2008, and it is unclear whether the URNG will be able to reverse its downward trend and become a significant, or even minor, player in Guatemalan politics.
FMLN AND URNG PARTY SUCCESS IN CONTEXT
From its very first electoral contest in 1994, the FMLN established itself as a major political party in El Salvador’s political system. In Guatemala, the URNG has successfully transformed itself into a minor political party. Following the 1994 elections in El Salvador, the effective number of political parties was 3.1 (See Table 4.7). Following the tabulation of electoral results and the commencement of the legislative period, there existed three major (ARENA, FMLN, and PDC), three minor (PCN, MU, and CD), and three failed political parties (MSN, MAC, and MAC). Following the 1997 legislative elections, the effective number of parties in El Salvador was 4.1. There remained four major (ARENA, FMLN, PDC and PCN), five minor (CD, MU, PD, PLD, and PRSC), and two failed political parties (MSN and PL). The FMLN was one of three (1994) and then four (1997) major parties in the country.14 The URNG in Guatemala has not been nearly as successful as the FMLN in El Salvador. Following the 1999 elections, the effective number of parties in Guatemala was 2.35. There existed two major (FRG and PAN), four minor (ANN-URNG, DCG, PLP, and LOV-UD), and seven failed political parties (ARDE, FDNG, UCN, MLN, AD, ARENA, and UN). Prior to the 2003 election, Guatemala changed the size of the congress to 158 members from 113. This helped increase the number of effective parties from 2.35 in 1999 to 4.56 in 2003. Following the 2003 elections, there remained five major (GANA, UNE, FRG, PAN, and PU), three minor (URNG, ANN, and Transparency), and six failed political parties (DCG, UN, DIA, MSPCN, DSP, and UD). The URNG was one of four (1999) and then three (2003) minor parties in the Guatemalan party system. While not a failed political party, the URNG has performed relatively poorly in comparison to both the FMLN and to the other political parties in Guatemala. What explains why the URNG has only been able to transform itself into a minor political party in Guatemala and the FMLN into a major political party in El Salvador? In the next two chapters, I test the
14 The FMN remained a major party following the 2000 and 2003 legislative elections.
65 hypotheses that I developed in the previous chapter to explain the major party success of the FMLN and the minor party success of the URNG.
66
Map 4.1 El Salvador, Perry-Castañeda Library Map Collection, 1980.
67
Map 4.2 Guatemala, Perry-Castañeda Library Map Collection, 2000.
68 TABLE 4.1 Postwar Presidential Elections in El Salvador (Vote and % of Total Votes)
Political Party 1994 1994 1999 2004 1st Round 2nd Round 1st Round 1st Round
ARENA 651,632 (49.11%) 818,264 (68.21%) 614,268 (51.96%) 1,190,235 (57.71%)
FMLN 331,629 (24.99%)* 378,980 (31.78%) 343,742 (29.05%)** 734,469 (35.68%)
PDC 215,936 (16.27%) - 67,207 (5.68%) 80,592 (3.90%)***
PCN 70,854 (5.34%) - 45,140 (3.82%) 56,289 (2.71%)
MAC 10,901 (0.82%) - - -
MU 31,295 (2.41%) - - -
MSN 13.959 (1.05%) - - -
CDU - - 88,640 (7.5%) ***
LIDER - - 19,269 (1.63%) -
PUNTO - - 4,252 (0.36%) -
Valid Votes 1,326,836 1,197,244 1,223,215 2,061,585
* In 1994, the FMLN presented supported a coalition candidate along with the CD and MNR. ** In 1999, the FMLN presented a joint candidate with the small Social Christian Union (USC). *** The PDC and CDU ran in coalition for the 2004 presidential elections.
69 TABLE 4.2 Postwar Legislative Elections in El Salvador (Votes and % of Total Votes)
Political Party 1994 1997 2000 2003
ARENA 605,775 (45.03%) 396,301 (35.40%) 436,190 (36.04%) 321,685 (32%)
FMLN 287,811 (21.39%) 369,709 (33.02%) 426,298 (35.22%) 339,727 (33.79%)
PDC 240,451 (17.87%) 93,645 (8.36%) 87,078 (7.19%) 73,271 (7.29%)
PCN 83,520 (6.21%) 97,362 (8.70%) 106,804 (8.82%) 131,762 (13.11%)
MAC 12,109 (0.90%) - - -
MU 33,510 (2.49%) 25,244 (2.25%) - -
MSN 12,827 (0.97%) 7,012 (0.63%) - -
CD/CDU 59,843 (4.45%) 39,145 (3.50%) 65,072 (5.38%) 63,530 (6.32%)
MNR 9,431 (0.70%) - - -
PD - 13,533 (1.21%) - -
PL - 2,302 (0.21%) - -
PLD - 35,279 (3.15%) 15,639 (1.29%) -
PRSC - 40,039 (3.58%) - -
PPL - - 4,998 (0.41%) -
PUSC - - 23,338 (1.93%) -
LIDER - - 44,901 (3.71%) -
PAN - - - 10,441 (1.04)
AP - - - 11,673 (1.16)
FC - - - 11,054 (1.10)
PSD - - - 7,349 (0.73)
PMR - - - 19,079 (1.90)
PPR - - - 15,714 (1.56)
Valid Votes 1,345,277 1,119,603 1,210,313 1,005,285
CDU competed in the 2000 and 2003 elections. The PDC-PD captured 39,838 votes in coalition. These votes are counted for individual parties above depending upon which party flag the voter marked.
70 TABLE 4.3 Postwar Legislative Elections in El Salvador (Number of and % of Seats)
Political Party 1994 1997 2000 2003
ARENA 39 (46.4%) 28 (33.3%) 29 (34.5) 27 (32%)
FMLN 21 (25%) 27 (32.1%) 31 (36.9%) 31 (37%)
PDC 18 (21.4%) 10 (10.7%) 6 (7.1%) 4 (5%)
PCN 4 (4.8%) 11 (13.1%) 13 (15.5%) 16 (19%)
MU 1 (1.2%) 1 (1.2%) - -
CD/CDU 1 (1.2%) 2 (2.4%) 3 (3.57%) 5 (6%)
PD - 1 (1.2%) - -
PLD - 2 (2.4%) - -
PRSC - 3 (3.57%) - -
PAN - - 2 (2.4%) -
PSD - - - 1 (1.2%)*
Total Seats 84 84 84 84
*In the 2003 legislative elections, the PSD- PDC-PMR won one seat in coalition.
71 TABLE 4.4 Postwar Presidential Elections in Guatemala (Votes and % of Total Votes)
Political Party 1999 1999 2003 2003 1st Round 2nd Round 1st Round 2nd Round
FRG 1,045,820 (47.72%) 1,173,823 (68.32%) 518,328 (19.31%)
PAN 664,417 (30.32%) 545,151 (31.68%) 224,127 (8.35%)
ANN 270,891 (12.36%)* -
DCG - 42,186 (1.57%)
PLP 67,924 (3.10%)
ARDE 45,470 (2.07%)
FDNG 28,108 (1.28%)
LOV-UD 25,236 (1.15%)
UCN 22,939 (1.05%)
MLN 13,080 (0.60%)
AD 4,929 (0.22%)
ARENA 2,698 (0.12%) - 11,979 (0.45%) UN - - 921,233 (34.33%) GANA - - 707,578 (26.37%) 1,235,219 (54.13%) UNE - - 69,297 (2.58%) 1,046,744 (45.87%) URNG - - 59,774 (2.23%) DIA - - 80,943 (0.03%) PU - - 10,829 (0.40%) MSPCN - - 37,505 (1.40%) DSP -
Valid Votes 2,191,512 1,718,974 2,683,779 2,281,963
In 1999, the URNG competed as a member of the ANN.
72 TABLE 4.5 Postwar Legislative Elections in Guatemala (Votes and % of Total Votes)
Political Party 1999 1999 2003 2003 Department National Department National
FRG 879,839 (43%) 891,429 (42.09%) 522,541 (20.15%) 502,470 (19.69%)
PAN 589,550 (29%) 570,108 (26.92%) 297,126 (11.46%) 278,393 (10.91%)
ANN 231,384 (11.3%)* 233,870 (11.04%)* - -
DCG 71,794 (3.5%) 86,839 (4.10%) 85,876 (3.31%) 82,324 (3.23%)
PLP 91,484 (4.5%) 84,197 (3.98%) - -
ARDE 76,994 (3.8%) 63,824 (3.01%) - -
FDNG 59,336 (2.9%) 60,821 (2.87%) - -
LOV-UD 48,398 (2.4%) 48,184 (2.28%) - -
UCN 40,069 (2%) 42,921 (2.03%) - -
MLN 19,646 (1%) 22,857 (1.08%) - -
AD 6,074 (0.30%) 8,644 (0.41%) - -
ARENA 1,868 (0.09%) 4,178 (0.20%) - -
UN 3,222 (0.16%) - 14,432 (0.56%) 17,478 (0.68%)
UD - - 58,671(2.26%) 55,793(2.19%)
ANN - - 142,674(5.5%) 123,853(4.85%)
Transparencia - - 24,359(0.94%) 27,740(1.09%)
UNE - - 424,418(16.36%) 457,308(17.92%)
PU - - 163,594(6.31%) 157,893(6.19%)
GANA - - 623,505(24.04%) 620,121(24.30%)
DSP - - 30,026(1.16%) 28,425(1.11%)
MSPCN - - 18,383(0.71%) 18,005(0.71%)
URNG - - 104,876(4.04%) 107,276(4.20%)
DIA - - 81,818(3.15%) 75,295(2.95%)
UN-DIA-ANN - - 1,336(0.05%) - Valid Votes 2,047,268 2,117,872 2,593,635 2,552,374 In 1999, the URNG competed as a member of the ANN.
73 TABLE 4.6 Postwar Legislative Elections in Guatemala (Number of and % of Seats)
Political Party 1999 2003
FRG 63 (55.8%) 43 (27.2%)
PAN 37 (32.7% 17 (10.8%)
URNG/ANN 9 (8%) -
DCG 2 (1.8%) -
ARDE - -
PLP 1 (0.01%) -
LOV-UD 1 (0.01%) -
UCN - -
MLN - -
AD - -
ARENA - -
UN - -
UD - 2 (1.3%)
ANN - 6 (3.8%)
Transparencia - -
UNE - 32 (20.3%)
PU - 7 (4.4%)
GANA - 47 (29.8%)
DSP - -
MSPCN - -
URNG - 2 (1.3%)
DIA - -
Total Seats 113 158
74 TABLE 4.7 Effective Number of Parties in El Salvador and Guatemala
Country Legislative Effective Major Parties Minor Parties Failed Parties Term Number of Parties
El Salvador 1994-1997 3.1 ARENA, PCN, MU, CD MAC, MSN, FMLN, PDC MNR
El Salvador 1997-2000 4.1 ARENA, CD, MU, PD, MSN, PL FMLN, PCN, PLD, PRSC PDC
Guatemala 1999-2004 2.35 FRG, PAN URNG, DCG, ARDE, FDNG, PLP, LOV-UD UCN, MLN, AD, ARENA, UN
Guatemala 2004-2009 4.56 GANA, FRG, URNG, DCG, UN, UD, UNE, PAN, PU Transparencia, DSP, MSPCN, ANN DIA
75 CHAPTER 5 EVALUATING THE IMPACT OF ORGANIZATIONAL CHARACTERISTICS
It seems reasonable to expect that some of the most important factors that contribute to the electoral success of a former insurgent group relate to the group’s organizational characteristics and performance as an insurgent organization. New political parties that trace their immediate antecedents to insurgent groups are not equally predisposed to experience electoral success. Insurgent groups vary to the extent that they can call upon prior experience and support in their effort at building a new political party based upon such factors as the number of armed and unarmed supporters, the organization’s territorial reach, and its experience managing that territory. Insurgent groups that have developed more extensive contacts with mass-based opposition groups are also more likely to succeed than insurgent groups with limited support outside the military organization. In this chapter, I evaluate the organizational hypotheses developed in Chapter 2 against the experiences of the FMLN and URNG.
Insurgent Size Hypothesis 1 states that larger insurgent groups are more likely to be successful than smaller groups. An insurgent group that is able to mobilize and direct a large group of committed supporters during its insurgent phase is more likely to succeed as a political party than an insurgent group that is unable to direct a large group of militants.
FMLN Throughout the 1980s, FMLN strength was estimated at 8,000 to 12,000 combatants (Dunkerley 1985; Halloran 1987; Wickham-Crowley 1992; McClintock 1998), ranging from 10,000 to 12,000 during the early 1980s to 8,000 to 10,000 at the end of the decade. One estimate numbered FMLN armed combatants at 14,000 in 1983-1984 (Bracamonte and Spencer 1995: 4), though it does seem out of line with most other estimates.15 Though there exist relatively reliable estimates of FMLN strength for the 1980s (8,000-12,000), it is more
15 Relative to the Salvadoran military, FMLN strength was approximately 9,000 -11,000 armed insurgents while the military counted on 50,845 in 1985 (Bosch 1999: 112).
76 appropriate to measure the size of the insurgent group at the end of the conflict. These are the hard-core supporters with which the FMLN would have to establish its political party. At the end of the country’s civil war, the United Nations Observer Mission in El Salvador (ONUSAL) demobilized 15,009 members of the insurgent group. These numbers included 8,552 combatants, 2,474 wounded noncombatants, and 3,983 political cadres (Luciak 2001: 4).16 8,552 combatants are consistent with estimates from the late 1980s. According to Luciak (2001), FMLN membership might have been ten to fifteen percent greater than the number demobilized, but for various reasons, many did not participate in the official demobilization process.
URNG During the late 1970s and early 1980s, Guatemalan insurgents were thought to have numbered 6,000 (Le Bot 1995: 195) – 7,500 (Perera 1993: 10). This number is significantly higher than the 1960s, when the number of “hard-core guerrilla units was variously estimated at between 200 and 300 men” (Collazo-Davila 1980: 126). After the group reached its greatest strength in the early 1980s, many URNG members were killed, fled into exile, or voluntarily withdrew from the organization in order to participate in a revitalized Guatemalan civil society.17 These defections and deaths included many mid-level cadres, leaving the URNG as an organization of commanders and foot soldiers. As a result, the actual number of armed combatants fell significantly by the mid-1990s. While some believe the total numbers demobilized by the FMLN underestimates its military strength, in the case of the URNG, many believe the demobilization process overestimates its true strength. Some estimates number URNG combatants as low as 1,000-1,500 (Spence et al. 1998), perhaps even as low as 500 (Stanley and Holiday 2002: 447) in the waning years of the conflict. Like measuring FMLN strength, I employ the figures at the moment of demobilization. During its demobilization and transition to political party, the URNG “was revealed to have a relatively small organization and not one with a large group of cadres trained or prepared to be political or electoral organizers” (Spence et al. 1998: 11).18 A UN peacekeeping force oversaw
16 Political cadres were involved in domestic and international political work. 17 Tania Prado Palencia, interview by author, Guatemala City, Guatemala, 19 April 2004. 18 One guerrilla was also concerned about the make-up of the guerrillas at the end of the war. Many of those who joined the guerrillas in the latter years of the war did so out of revenge for the loss of family members. They were not politically conscious like earlier recruits. The changes in the make-up of the organization’s members might have
77 the demobilization and disarmament of 5,753 URNG members, including 2,940 armed combatants and 2,813 international and political support members (Luciak 2001: 24). The URNG demobilized less than 6,000 total personnel fifteen years after it had counted on 6,000- 8,000 armed combatants plus an unknown number of noncombatant personnel.19
Discussion How well does the total number of militants explain the performance of these new political parties? While both the FMLN and URNG were numerically smaller than at earlier points in their respective histories, the FMLN (15,009) still remained significantly larger than the URNG (5,753). In relative terms, the FMLN advantage appears even greater if we compare the ratio of insurgents to the total population in each country. Using El Salvador’s total population according to the 1992 census (5,118,599), the FMLN possessed 293 members per 100,000 Salvadorans at the time of its demobilization. The URNG, on the other hand, demobilized 69 militants for every 100,000 based upon its 1994 census population total of 8,331,874.20 Relative to their respective populations, the FMLN was more than four times larger than the URNG. The FMLN finished second in the 1994 and 1999 presidential and legislative elections. The URNG did not fare as well as the FMLN. The URNG finish third in the 1999 presidential and legislative elections. In 2003, the URNG finished seventh in the legislative election and fifth in the presidential election. The FMLN averaged twenty-seven percent in both presidential and both legislative elections while the URNG averaged seven and one-half percent in both presidential and both legislative elections. These numbers provide strong support for the hypothesis. In terms of the absolute and relative number of militants, the larger insurgent group has been more successful than the smaller insurgent group.
Political Party Another characteristic that seems likely to influence the success of a new political party relates to the experiences upon which it can draw from its pre-insurgency days. One experience relates to whether the group had political party experience prior to its transition from insurgent
had an effect in addition to sheer numbers. Alberto Ramírez Recinos, interview by author, Quetzaltenango, Guatemala, 13 April 2004. 19 Some sympathetic to the guerrilla cause estimate that the guerrillas counted on the support of up to one million active supporters in the late 1970s and early 1980s. 20 The number reduces to 51 per 100,000 if one uses the 2002 census population of 11,237,196.
78 group to political party. Hypothesis 2 states that insurgent groups with prior political party experience are more likely to be successful than those without prior political party experience.
FMLN Given the Salvadoran civil war’s relatively short duration of slightly more than one decade, most FMLN militants had the opportunity to participate in 1970s electoral politics, the 1980s insurgency, and the new political party in the 1990s. Many combatants and guerrilla leaders had extensive histories in political parties prior to the war through their participation in the PDC and the PCS. Both parties had participated in the Salvadoran political system throughout the 1960s and 1970s, winning municipal and legislative elections, though they were prohibited from winning the presidency. Of the five insurgent groups that comprised the FMLN, three (FAL, ERP, RN) had extensive political party experience. The FAL was the armed wing of the PCS. The PCS, which was founded in 1930, had the longest history of any of the groups. However, it was illegal for most of the century and most of its work was accomplished behind the scenes. In 1932, Augustín Farabundo Martí, founder of the PCS, led an indigenous uprising in the country. Over the course of several months, General Maximiliano Hernández Martínez directed the slaughter of 8,000 – 30,000 Salvadoran peasants who had actively participated in the uprising or who might have simply been sympathetic to Martí. Martínez and the Salvadoran government proceeded to extend paramilitary control throughout the country so that similar uprisings would be deterred. Some argue that over the next several decades, the PCS’ presence was only elevated to a higher “reality” when the regime invoked the “communist threat” and not due to any actual threat (Zamora 2003: 41). Fears of global communism allowed the regime to crack down on popular dissent and blame the protests and violence on communist agitators. Given the long history of the PCS and the conditions under which it operated, it developed both direct and indirect political party experience. During the 1950s and 1960s, the PCS twice attempted to form its own political party, the Partido Revolucionario Abril y Mayo, or April and May Revolutionary Party, and the Partido Revolucionario, or Revolutionary Party, but these experiments proved failures. It also tried to “infiltrate” existing parties such as the Partido Acción Renovadora (PAR), or Renovating Action Party (Zamora 2003: 41-43).21 The PCS’
21 The government banned the PAR soon after it captured four seats in the 1966 elections.
79 more recent experience with electoral politics occurred in the late 1960s and 1970s when it operated under the leadership of Schafik Handal during the 1968-1977 period as part of the Unión Democrática Nacionalista (UDN), or Democratic Nationalist Union (Montgomery 1993: 105). The opposition UDN allied the PDC, PCS, and the Movimiento Nacional Revolucionario (MNR), or National Revolutionary Movement. As late as 1979 and at the height of the pre-war repression, the PCS was still committed to the electoral route to power even as other leftists scorned it as reformista after they had thrown their support behind revolution (McClintock 1998: 51-52). Eventually though, the PCS would arrive at the same conclusion as the other four guerrilla organizations and form its own military organization, the FAL. Nevertheless, its experience in electoral politics helped the FMLN during the peace negotiations and in the early years of preparing for the post-war elections.22 Two other insurgent groups that comprised the FMLN also trace their roots back to Salvadoran political parties. The FPL was formed in 1970 when Salvador Cayetano Carpio and a small number of PCS dissidents reached the conclusion that the Salvadoran electoral process was incapable of allowing them to achieve meaningful social, economic and political change. But this was not before they had developed some experience with electoral politics during the national elections of the 1960s. Later that year, other frustrated members of the PCS left, later joining with Christian Youth and New Left groups and dissidents from the PDC to form the ERP in 1972 (Wickham-Crowley 1992: 225; McClintock 1998: 50). After the 1975 killing of one of its members, Roque Dalton, the ERP split, with a number of members leaving to form the FARN. In a way, then, the FARN could also trace its roots to a political party. The PRTC, which began in 1979, was the last guerrilla organization to form. While all the groups shared links to the various student groups around the country, the PRTC did not share a direct relationship with Salvadoran political parties of the 1960s and 1970s. Instead, the PRTC “attracted a cluster of organizations around its attachment to a regional struggle” (Dunkerley 1988: 372). The FAL, ERP and FPL each maintained direct links to Salvadoran political parties of the 1960s and 1970s, particularly the FAL. Many amassed experience in forming electoral coalitions, parliamentary debate procedure, and selecting candidates. The extensive electoral
22 Ana Guadeloupe Martínez, interview by author, San Salvador, El Salvador, 1 May 2004; Tommie Sue Montgomery, interview by author, San Salvador, El Salvador, April-May 2004.
80 history of the groups that comprised the FMLN is consistent with the political party hypothesis as the FMLN has successfully transformed itself into a major political party in El Salvador. Further support for the hypothesis is found by investigating the performance of the individual groups and members within the FMLN. First, the PCS’s armed wing, the FAL, was neither militarily formidable, nor did it maintain strong popular support during the war as “PCS contingents in the FMLN were dwarfed in terms of members, military skill, and a popular base in the countryside by those of the rest of the left” (Dunkerley 1988: 370). Yet, from the peace negotiations forward, the PCS and its leader, Schafik Handal, took on a more dominant leadership role within the party than its overall contribution to the revolution would have suggested. Handal, who had honed his political skills as the leader of the PCS and in successfully negotiating the Peace Accords for the FMLN, used these experiences to capture the leadership within the FMLN in the postwar period.23 For the 1994 elections, Handal was the FMLN candidate for mayor of the capital, San Salvador. The two previous presidents had been mayor immediately preceding their election to the presidency. Though Handal did not win, he slowly worked to established leadership over the party. In 1997, Handal’s PCS and a section of Salvador Sánchez Cerén’s FPL formed an alliance within the FMLN. As a result, Handal assumed greater control over the new political party until his death in January 2006. Following the FMLN’s 1999 loss in the presidential elections, Handal increased his control over the party by forcing Facundo Guardado, former FPL commander and 1999 presidential candidate, from the leadership and eventually from the party. Handal succeeded in elevating the smallest and least representative guerrilla group into the dominant bloc within the political party.24 In addition to the experience of former combatants, I also hypothesized that a former insurgent group would be successful if it had established a political wing or close relationship with a legal political party. The FMLN maintained extensive relations with a “political wing”, the FDR, during the conflict and the early postwar period. In the 1970s and 1980s, not all opposition opted for the cause of armed revolution. In April 1980, some sixty center-left organizations formed the FDR with hopes of uniting the non-revolutionary “left and center-left groups, both political parties, and social organizations” (McClintock 1998: 52-53). In mid-1980,
23 Ana Guadeloupe Martínez, interview by author, San Salvador, El Salvador, 1 May 2004. 24 It might also not be a coincidence that upon the demobilization of the FMLN, the FAL (25%) and FPL (36.1%) were reflected to a much greater extent than the RN (16.5%), ERP (11.6%) and PRTC (10.85) among all demobilized political cadres. The FAL and FPL had the supporters with political experience.
81 FDR leaders traveled Europe and Latin America to bring attention to the plight of the opposition in El Salvador. The FDR was successful in gaining support from a small number of European countries and the Socialist International, as well as the right to establish political offices in Mexico (Montgomery 1995: 111). Months later, several FDR leaders were kidnapped and assassinated by Salvadoran security forces, forcing the surviving leaders to flee into exile. Following the extreme repression of the masses and leadership of the FDR, its more radical members joined the FMLN, leaving only moderates remaining (Ribera 1996: 41-42). The international press and the Salvadoran government frequently referred to the FDR as the political wing of the FMLN, although others describe the relationship more as one of an “uncomfortable alliance” (Ribera 1996: 42). The FMLN and FDR indeed shared common goals such as the reform of the Salvadoran state and the dismantling of its ruthless security forces. The FMLN and FDR also coordinated sending their representatives throughout the world, “aggressively seeking out government ministers and legislators as well as talking to the media and citizens groups on a daily basis” (Montgomery 1995: 114). There was a great deal of cooperation between the two organizations, but there remained critical differences. Ideologically, the FDR was more social democratic than the FMLN, and did not entirely share the idea of a revolutionary society that was favored by many in the FMLN. The FDR was also open to dialogue with the civilian government at a much earlier stage than the FMLN. In 1987, FDR leaders Rubén Zamora and Guillermo Manuel Ungo returned from exile to take advantage of the new political opening. Later that year, the MNR and the Movimiento Popular Social Cristiano (MPSC), or Popular Social Christian Movement, joined with the Partido Demócrata Social (PSD), or Social Democratic Party, to form the CD with the hope of competing in the 1988 legislative and municipal elections. The limited time before the 1988 elections prevented the CD from participating, but it was able to mount a campaign and candidacy for the following year’s presidential elections in 1989. At about the same time, the relationship between the FMLN and FRD also grew increasingly tenuous as FDR leaders criticized the human rights abuses that were committed by the FMLN in the latter half of the 1980s (McClintock 1998: 52-53). While the CD participated in the 1989 presidential elections, the FMLN did little to support its candidates. The 1989 election and criticism of FMLN violence led to the permanent break of the FDR-FMLN alliance that had formed roughly one decade earlier (Baloyra-Herp 1995: 57).
82 Though the “official” alliance between the FDR (and then the CD) and FMLN had been broken, they did continue to cooperate on important issues. In 1991, the CD competed in legislative and municipal elections, capturing a number of seats in the 1991-1994 session of congress. During the legislative term, the CD helped the FMLN by coordinating the negotiations between the insurgents and the executive within the Legislative Assembly to “translate specific accords into legislation” (Spence, Dye and Vickers 1994: 24). The CD also assisted outside the Legislative Assembly by working with the FMLN to negotiate “the many hurdles in adjusting to civilian political life, in particular in dealing with complicated governmental and accords machinery” (Spence, Dye and Vickers 1994: 24). Later, as the FMLN prepared for the 1994 elections, the FMLN chose Rubén Zamora of the CD as the presidential candidate of a leftist coalition. Zamora had experience in running for political office and was well known throughout El Salvador. The FMLN hoped he would be a moderate face on their revolutionary program. In the end, Zamora finished in second place and helped the FMLN force a second round.25 Overall, the FMLN relationship with the FDR helped bring international and domestic attention and much needed political and economic support to the cause of the armed movement. In addition, the new organization of the FDR, the CD, helped the FMLN during peace negotiations and in directing legislative matters to prepare for the “arrival” of the FMLN. Finally, the CD provided useful experience and even the presidential candidate of the left at the time of the FMLN’s insertion into electoral politics. This relationship that the FMLN maintained with a political wing also provides support for the political party hypothesis. The FMLN used the prior electoral experience of its political wing in order to make a smooth transition to political party and to transform itself into a major political party in El Salvador.
URNG The opposite relationship between political party experience and success exists in Guatemala. The FAR, ORPA, and the EGP that formed the URNG could all trace their roots to the original FAR guerrilla group of the 1960s. The PGT traced its roots directly to the Communist Party of Guatemala. Former guerrillas argued that the URNG did not benefit from
25 I discuss the relationship of the FMLN and CD as political competitors in the next chapter of the dissertation.
83 experiences that any of its members had with prior electoral competition.26 The reasons that the insurgents provided basically revolved around the long duration of the war. In 1954, the CIA- orchestrated overthrow of the democratically elected Jacobo Arbenz government, radicalized many Guatemalans and crushed their “illusions about peaceful, legal and reformist methods” (Dunkerley 1988: 429). The conflict in Guatemala would last for over three decades, with civilian rule and electoral competition only returning in the mid-1980s – over thirty years since the country’s last free and fair elections. Insurgents that might have had electoral experience from the democratic experiment in the 1940s and 1950s had passed away or were no longer active in the insurgency. This meant that at the time of establishing a political party and preparing for electoral competition, the URNG could not count on any members’ prior political party experience. The URNG was also inhibited by its inability to establish a political wing. During the conflict, the URNG was sometimes hurt by the lack of such a political wing. For example, in January 1982, the URNG issued a public pronouncement of its revolutionary platform. But in many ways, the proclamation fell on deaf ears, as the URNG lacked the political wing to publicize it – a noted contrast from the ties between the FMLN and the FDR (Dunkerley 1988: 491-92). It was not until the mid-1990s that what could be interpreted as a political wing of the guerrillas was established in Guatemala. In 1994, the Frente Amplio para la Dignificación de Guatemala, or Broad Front for a Dignified Guatemala, and the Movimiento Cívico Democrático, or Democratic Civic Movement, formed the FDNG. The latter would exit, though, before the 1995 elections. Like the relationship of the FDR to the FMLN in El Salvador, the relationship between the FDNG and URNG is not altogether clear. The URNG encouraged the formation of the FDNG in 1994, but most URNG members were not associated with it, nor was the FDNG overly supportive of the insurgents (Sichar 1999: 80). In 1995, the FDNG competed in national elections. For the first time, the URNG established a two-week ceasefire during the elections and called on Guatemalans to vote (McCleary 1997: 138). In the elections, the FDNG finished in third place, with six of the eighty congressional seats. Following the end of the civil war, the FDNG and URNG did not always see eye-to-eye. The URNG made it clear that it did not want to integrate into the FDNG, since the FDNG did not
26 Rodrigo Asturias, interview by author, Guatemala City, Guatemala, 2 April 2004; Alberto Ramírez Recinos, interview by author, Quetzaltenango, Guatemala, 13 April 2004; Juan Francisco García, interview by author, Guatemala City, Guatemala, 14 April 2004.
84 share its revolutionary ideals (Sichar 1999: 86). However, the URNG remained interested in working with the FDNG. Some members of the FDNG had been members of the guerrillas in the 1970s and early 19809s, so it made sense that there would be a desire to form an electoral alliance. In addition, the URNG needed to cooperate with other leftists groups, because at the time of its demobilization it was wholly unprepared for the demands of a transition to political party. The left eventually formed a coalition to compete in the 1999 elections involving the FDNG, URNG and other progressive groups. But the FDNG did not see the coalition as one built on mutual respect. The URNG was not interested in selecting candidates who might be considered “qualified” for elective office, such as those that were serving in the current Congress, as members of the FDNG legislative bloc (Sichar n.d).27 Instead, the URNG was more interested in selecting candidates based upon their contribution to the insurgent organization.28 The FDNG was eventually expelled from the left coalition, the ANN, months before the 1999 elections. The lack of prior political party experience and the failure of the URNG support the political party hypothesis. URNG militants had no prior experience with electoral politics. The insurgency was weakened during the war by its failure to develop a political wing (or establish relations with an existing political party) that could publicize the organization’s principles and engage in fundraising activities. This weakness continued into the postwar period as some insurgents lamented the lack of knowledge in building a political party and running an electoral campaign.29
Discussion The prior electoral experience of the FMLN and the lack of similar experiences within the URNG support the hypothesis. As seen before, the experience of the FMLN guerrilla groups in electoral politics, the ascension of Handal and the PCS to the party’s leadership, and its relationship with the FDR support the importance of prior political party experience. In Guatemala, the lack of prior political party experience on the part of the URNG also lends support for the hypothesis. None of the Guatemalan guerrilla groups had relevant experience
27 Sichar argues that the disconnect between the URNG and the FDNG within the ANN coalition was caused by the FAR specifically. 28 Tania Prado Palencia, interview by author, Guatemala City, Guatemala, 19 April 2004. 29 Rodrigo Asturias, interview by author, Guatemala City, Guatemala, 2 April 2004
85 with party politics. The lack of a political wing to raise funds and to increase national and international awareness of the insurgents’ goals and accomplishments also hindered its performance. The URNG only established relations with the FDNG shortly before the war ended, could not hold a coalition together through a single electoral campaign, and failed to capitalize on the experience and the seats that the FDNG had gained in participating in the 1995 elections.
Mass-Based Organizations Hypothesis 3 postulates that insurgent groups that had extensive contacts with mass- based organizations during the war (unions, student groups, peasant confederations, etc.) are more likely to be successful than those without such contacts.
FMLN Each individual Salvadoran guerrilla group established contacts with mass-based organizations in the 1970s “that could operate aboveground legally, enlisting the participation of protesters who were not yet prepared to embrace armed struggle. Through this participation, it was expected, many would be radicalized, and eventually a number of those might become revolutionary militants themselves” (Brockett 2005: 87). The amount of attention that should be provided to these political organizations was a frequent source of conflict within the military organizations. The first organization to seriously dedicate attention to mass-based organizing was the RN. During the mid-1970s differences had emerged within the ERP concerning how much attention the organization should provide to the political relative to the military struggle. The ERP had pursued a strategy entirely focused upon the military struggle. Following the murder of an ERP member and the nation’s most famous poet, Roque Dalton, by the ERP leadership, those erpistas who shared Dalton’s commitment to political organizing broke away and formed the RN. The RN was critical of the ERP’s exclusive reliance upon militarism at the expense of other forms of opposition, and set out to “break the traditional division between armed vanguard and mass organization through the establishment of an alliance with Frente de Acción Popular Unificada (FAPU), set up in 1974” (Dunkerley 1988: 371).30 The RN was successful in organizing the urban working class.
30 FAPU stands for the United Popular Action Front.
86 Differences of opinion over interpreting the national reality quickly resulted in a division within FAPU. In 1975, two large unions (teachers and campesinos) left FAPU and formed the core of the Bloque Popular Revolucionario (BPR), or Popular Revolutionary Bloc, an even larger mass-based organization. The BPR established ties with another guerrilla organization, the FPL. The FPL had been involved in the mass political struggle from the start, believing that a political struggle was needed to accompany a military struggle (Byrne 1996: 36). By the end of the 1970s, the BPR had become the “largest and most socially representative left-wing organization in the country” with estimates ranging from 60,000 (Dunkerley 1988: 372) to 80,000-100,000 members (Byrne 1996: 36). The PCS also maintained good relations with mass- based organizations. Throughout the 1960s and 1970s, the PCS cultivated support among the country’s trade unions. The PCS had not given up the possibility of radical reform without the need for revolutionary violence, and thus established strong relations with mass-based, non- violent organizations. The ERP, after initially shunning relations with mass organizations resulting in the establishment of the RN, eventually helped to establish its own organization – the Ligas Populares – 28 de Febrero (LP-28), or February 28th Popular Leagues. The PRTC also established a mass-based organization, the Movimiento de Liberación Popular (MLP), or Popular Liberation Movement, in 1979. Like the PRTC itself, though, the MLP was small in size (Dunkerley 1988: 372). Prior to the outbreak of revolutionary violence in 1979, “the left had established a clear set of linkages between politico-military bodies and mass-organizations, which possessed a combined membership of at least 250,000 and had carved out a major role in public life” (Dunkerley 1988: 372). At the end of the decade, these organizations were bringing the regime to the point of crisis following a series of strikes, demonstrations and protest marches. The Salvadoran government responded by unleashing repression against these mass-based organizations, killing a significant amount of the nonviolent left and pushing many of its survivors into the ranks of the rural insurgency. For the next few years, mass-organizations declined as the focus moved to revolutionary warfare. In the early 1980s, the insurgent left concentrated primarily on overthrowing the Salvadoran government through military means. The FMLN was inspired by the success of the Sandinista revolution in neighboring Nicaragua in July 1979. In January 1981, the FMLN
87 launched a similar insurrection, although, unlike the Sandinista insurrection, this effort failed. For the next few years, the FMLN continued to prepare the country for the overthrow of the regime through military means. FMLN work with the masses focused on engaging them in the armed struggle as combatants, not as a complementary force. At the time, there was little attention to the political struggle, unlike the 1970s, because of the FMLN belief that military victory was a realistic short-term possibility. Following the increasing use of air power by the Salvadoran military and the election of Duarte in 1984, the FMLN began to refocus its efforts on the political struggle and the mass-based organizations that had been decimated in the late 1970s and early 1980s. During the second half of the 1980s, protest movements again filled the streets of San Salvador and other Salvadoran cities. Labor and students groups again organized relatively openly, many of them still directed by the FMLN. Many of the remaining protesters were simply anti-Duarte and the PDC. These protesters had been instrumental in electing Duarte in 1984, but had turned against him after the economy worsened, he failed to bring an end to the war, and his government was perceived as one of the most corrupt in memory. In 1992, over one hundred thousand people, many of them from these labor and student groups, took to the streets to celebrate the negotiated end to the war. They would be important to the urban success of the FMLN. This was clearly different than the Guatemalan experience, where only about 20,000 turned out in the capital to celebrate the end to the conflict in a country almost three times as large as El Salvador. After the war, problems developed between the FMLN and mass-based movements as each attempted to adapt to new postwar challenges.31 Some of the challenges concerned mass- based groups attempting to become more independent of the FMLN by working with other political parties, but effects on the electoral performance of the FMLN appear to have been marginal.
URNG During the 1960s, 1970s, and 1980s, each time the Guatemalan government’s repression intensified against non-revolutionary movements, those who became radicalized found protection in the Guatemalan guerrilla groups. “When repression rapidly escalated in 1980,
31 Sonia Baires, interview by author, San Salvador, El Salvador, 30 April 2004.
88 intensifying grievances and foreclosing peaceful protest, the guerrillas provided the vehicle by which thousands could take up arms against the murderous military regime” (Brockett 2005: 117). But the relationship between insurgents and mass movements in Guatemala remained much weaker than the relationships that had developed between the insurgents and mass movements in El Salvador. As a result, the URNG could count on fewer noncombatant supporters in the postwar period. During the 1970s, the Guatemalan guerrilla groups sought to radicalize students, unions, and other mass-based organizations in support of revolution. The PGT cultivated a strong base throughout the country’s university system, “especially among faculty, including some high level administrators” (Brockett 2005: 105). By the end of the decade, the PGT had developed a strong relationship in the Federación Autónoma Sindical de Guatemala (FASGUA), or the Autonomous Trade Union Federation of Guatemala (Brockett 2005: 112). For most of the 1960s and 1970s, the PGT had pursued nonrevolutionary confrontation, which allowed it to concentrate on relations with other nonrevolutionary mass-based organizations. Two other insurgent groups also sought to cultivate relations with these mass-based groups. By the mid-1970s, the FAR, in addition to the PGT, was making its presence felt in the Confederación Nacional de Trabajadores (CNT), or National Workers Confederation (Brockett 2005: 112). The Comité de Unidad Campesina (CUC), or Committee of Peasant Unity, formed in May 1978 “as a broad rural workers’ union that sought no recognition from the state and possessed no formal leadership structure. Organized largely underground and through local meetings, the CUC effectively became a popular extension of the guerrilla, and with a declared commitment to prolonged people’s war it stood particularly close to the EGP” (Dunkerley 1988: 478). While the CUC and the guerrillas, particularly the EGP, shared common goals of furthering the plight of the poor in Guatemala, their exact relationship is not exactly clear. Some disagree that the CUC was a “popular extension of the guerrilla.” But, at a minimum, it is relatively clear that the CUC acted as a vehicle for Guatemalan peasants to enter the guerrilla ranks (Wickham-Crowley 1992: 132; Brockett 2005: 140) even if they were not one and the same. According to Ryan (1994: 35), “No political parties or professional or popular organizations of significance publicly affiliated themselves with the URNG, and there is little evidence of covert support. Undoubtedly, the systematic and brutal repression of the military
89 regime made such support hazardous in the extreme.” This is true, but partially misleading. In the early 1980s, the Guatemalan military decapitated the leadership of the country’s mass-based organizations and incurred deep losses upon the insurgent organizations. In 1985, the Guatemalan military turned over power, at least in name, to a civilian elected government. Almost immediately, mass-organizing was again seen in the urban areas of the country. Many of the re-emerging movements’ leaders were former insurgents, or at least insurgent sympathizers. They had left the insurgents after concluding that military victory was impossible and that they had more to gain by finding space within the system created by the new civilian government. Some of these individuals would go on to form the core of the FDNG in 1995, while the URNG was left with its former commanders and its core fighters for the next decade.32 Some have argued that the URNG might have gained more through a prolonged period of negotiations than its military strength would have suggested (Vinegrad 1998). The agreement between the URNG and Guatemalan government included concessions by the military and government that would not have been expected in the mid 1980s. It is quite possible that in the long run, though, the process hurt them. During the negotiations, the URNG and Arzú worked on a number of groundbreaking social and economic reforms. The Asamblea de Sociedad Civil (ASC), or Assembly of Civil Society, an organization comprising a number of civil society organizations (women’s groups, indigenous groups, etc), was responsible for developing the groundwork for agreements on women’s rights, indigenous rights, and education. The URNG adopted many of the proposals as their own, but was criticized for its unwillingness to debate and negotiate with civil society the various components of the peace agreement (Vinegrad 1998: 219). Eventually, many key provisions dealing with the indigenous in Guatemala were defeated in the May 1999 referendum from the lack of popular support. So, while the URNG and the Arzú government did accomplish a great deal during the negotiation process, the reforms were voted down. There was a distancing between the leadership of the URNG and its base and civil society.33 At the same time, the URNG continued to bleed combatants who came down from the mountains to seek medical attention or were frustrated by the commandancia’s negotiations. The URNG also continued to lose support in the countryside, as the economic stress upon their rural
32 Tania Prado Palencia, interview by author, Guatemala City, Guatemala, 19 April 2004. 33 Prado Palencia, interview by author, Guatemala City, Guatemala, 19 April 2004.
90 supporters did not appear to have an end in sight.34 The protracted negotiations, all in an effort to make the agreement as comprehensive as possible, had come at a great price for the URNG. The distancing between the URNG and mass-based organizations continued following the end of the war, as the URNG failed to identify how it as a political party could lead the mass movements that had developed with the democratic opening. As discussed in the previous section regarding the URNG’s experience with party politics, differences between the URNG and the FDNG within the ANN resulted in the FDNG being forced out of the coalition along with key members and sectors of Guatemalan civil society. In June 2003, the URNG rejected a call by leftists, academics and directors of Guatemala’s social organizations to unite for the 2003 elections. Over the next several months, the relationship between the URNG and mass-based organizations continued to deteriorate (Velia Jaramillo 2003).
Discussion In the cases of the FMLN and URNG, each insurgency’s relationship with mass-based groups supports the hypothesis. The FMLN had more extensive relations with political, social and economic organizations and has been more successful as a political party, while the URNG did not develop strong relations with similar organizations and has been less successful as a political party. In the postwar period, the FMLN and URNG were in different situations with respect to mass-based opposition groups. In El Salvador, many mass-based organizations had become intimately tied to the revolutionary organization during the conflict. In the post-war period, many mass-based organizations looked to develop a greater sense of autonomy from the insurgent organizations, perhaps with the option of working with other political parties. While many members of the organization were still loyal to the FMLN, the organizations sought independence. In Guatemala, relations between the URNG and mass-based organizations were very different. In 1995, many mass-based organizations participated in the electoral process through the formation and electoral competition of the FDNG. When the URNG entered electoral politics following the peace process, the URNG looked to co-opt the mass-movements for their own utility. This has led to several difficulties over the years, including the fracturing of the ANN as many members of the mass-based movements were not revolutionary and had no desire to align with the URNG. The URNG’s attempt to transform itself into a successful
34 Alberto Ramírez Recinos, interview by author, Quetzaltenango, Guatemala, 13 April 2004.
91 political party was hurt by its inability to lead frentes populares or frentes amplios during the negotiations for ending the war and immediately in the postwar period prior to the 1999 elections.
Unity Hypothesis 4 stated that insurgent groups that maintain unity and coherence during the war, negotiations, and in the postwar are more likely to be successful than insurgent groups that fracture. How have the FMLN and URNG been affected by internal divisions?
FMLN During the war in El Salvador, the different guerrilla organizations that comprised the FMLN overcame their strategic and ideological differences, more often than not, to coordinate their opposition to the regime. While the five groups officially merged to form the FMLN in 1980, “each one of the organizations maintained its own statutes, conserved its own organic structure, procured its own financing and independently elected its leaders. The militants maintained discipline to their organization and not to the FMLN” (Zamora 2003: 52).35 Though the agreement among the five organizations on the primacy of armed revolution kept them sufficiently unified throughout the war, postwar challenges threatened and eventually destroyed this unity. The FMLN had to focus its attention towards the selection of candidates and the drafting of a political platform, the possibility of alliances, and the type of opposition political party it wished to become. This was an obvious challenge for any party on the left in the post- Cold War period, perhaps more so for one simultaneously making a transition from a predominantly military to a predominantly political organization. Although there were concerns surrounding the possible fragmentation of the party in the months leading up to the “elections of the century,” the FMLN did not suffer its first serious defection until the elections had passed and the legislature had convened. During the war, the FMLN’s General Command was made up of the commander of each of the five guerrilla organizations. These five commanders, plus two additional members from each organization, made up the fifteen-person Political Commission. The General Command and the Political Commission were the two highest-ranking decision-making bodies in the
35 Author’s translation.
92 FMLN. Three commissioners represented each group regardless of the membership or military strength of the individual group. While the five organizations were not always in complete agreement on strategy and tactics, each group followed the decisions of the General Command. During the war, the FMLN did not have significant problems with people or groups abandoning the armed struggle as most still believed the FMLN could “win.” Following the 1992 Peace Accords, FMLN unity broke down in part because of the internal decision-making process. The FMLN voted to reform the party structures, removing the General Command, and elevating the Political Commission to the primary decision-making body. The Political Commission needed ten of the fifteen members to agree to the election of a coordinator, but the commission was unable to satisfy the two-thirds requirement. This failure resulted in the coordinator position remaining vacant for several months following the accords and an eventual change in the rules (Luciak 2001: 97). In 1994, the FMLN reformed the coordinator selection process and the decision-making process more generally, so that it only required three-fifths of the members to approve a coordinator. This rule change disadvantaged two of the FMLN’s groups, the ERP and RN, who tended to share similar views on the future of the party that differed drastically from the FPL, FAL, and PRTC. With a two-thirds majority needed on certain party issues, the ERP-RN alliance could effectively block the decisions of the remaining three groups. The ERP-RN strength was greatly diminished following the adoption of the more easily attainable 60% threshold. One of the key issues where the ERP-RN stance was defeated involved the selection of the FMLN presidential candidate for the 1994 elections. The ERP-RN believed that the FMLN should establish some sort of alliance with the PDC and choose someone like the social- democratic Abraham Rodriguez as its presidential candidate. The ERP-RN believed that a center-right, reformist candidate, would be strategically the best option for continuing the transformation of the country that had begun under the Peace Accords.36 From this perspective, it would be helpful for the FMLN to establish an alliance with moderates in the business and political communities that were also opposed to the policies of the ruling ARENA party. The FAL, FPL and PRTC disagreed with the selection of a center-right candidate and instead preferred Rubén Zamora of the CD, an organization and individual with which the FMLN had cooperated during the course of war. In the end, the FAL, FPL and PRTC outvoted the ERP and
36 Ana Guadeloupe Martínez, interview by author, San Salvador, El Salvador, 1 May 2004.
93 RN three-to-two and Zamora was selected to represent the left. It was at this point, with the selection of its candidate, that the ERP and RN realized that they had lost the election.37 They also believed that the experiences proved that there was little room for debate and consensus within the FMLN. Following the selection of Zamora, the ERP and RN failed to fully support the FMLN candidate, which might have hindered his performance in the 1994 election (Luciak 2001: 100). Divisions among the former insurgents continued into the 1994-1997 legislative period. Former guerrilla leaders Joaquin Villalobos and Ana Guadalupe Martínez, along with several members of the RN and ERP, broke from the FMLN. FMLN deputies associated with the RN and ERP left the party after reaching an accord with the ruling party. In return for supporting ARENA’s controversial budget, the ERP-RN deputies would receive two positions on the legislative council in the Legislative Assembly. According to Martínez, the division was caused by differences over the FMLN’s program, political strategy, and the future direction of the country, many of the same issues that had been involved in the earlier selection of the left’s presidential candidate.38 The RN and ERP believed that they lost out on all three issues and would continue to lose on future decisions because they would always be outvoted by the FAL, PRTC, and FPL. For the 1997 elections, Villalobos, Martínez and others formed and competed as the PD. While the PD maintained seven deputies in the 1994-1997 Assembly, these deputies were elected as FMLN candidates in the 1994 elections. After these deputies broke away from the FMLN and formed their own political party, they were only successful in getting one deputy elected in the March 1997 elections. Having attained less than one percent of the national vote, the PD lost legal standing as a political party under Salvadoran electoral rules.39 While the PD performed poorly in the 1997 elections, it experienced short-term success in capturing two positions in the legislative council and helping to pass the governing party’s controversial budget during the 1994-1997 legislative session. In the long run, this alliance with the governing party and its supporters, individuals whom the FMLN had just spent 12-years attempting to overthrow, appeared to have alienated many of the traditional supporters of the FMLN and contributed to
37 Ana Guadeloupe Martínez, interview by author, San Salvador, El Salvador, 1 May 2004. 38 Ana Guadeloupe Martínez, interview by author, San Salvador, El Salvador, 1 May 2004. 39 It subsequently merged with the CD, the Popular Labor Party (PPL), the Movement of Christian Democratic Unification (MUDC), and the Faith and Hope (FE) Party to form the United Democratic Center (CDU) in 1998.
94 the failure of the PD in the 1997 elections.40 As to the impact on the FMLN, the division was very public and involved the airing of dirty secrets from the war. The break also involved claims and counterclaims of treason and cooperation with the US Central Intelligence Agency. While this public fighting was quite messy for the FMLN, the electoral impact was negligible. The PD captured less than one percent of the vote in the next elections, while the FMLN became the second largest party in the country’s legislature. While the FMLN had suffered a serious division following the first election in which it participated, the breakaway faction was unable to take a sizeable portion of the FMLN with it or capture a significant percentage of the voting population. The withdrawal or defection of a group of dissident FMLN personnel could have helped the party if this removed a source of friction among the remaining efemelistas and allowed them to present a unified party. This has not been the case as ten years after it first competed in an election the FMLN continues to have highly publicized dissent leading to the ouster of key leaders. The inability of the FMLN to deal with internal dissent and the subsequent withdrawal or expulsion of former members has led to questions concerning the FMLN’s commitment to democracy and its ability to work with other political parties and organized groups in order to move the country forward. These divisions may have weakened the FMLN, but they have by no means crippled the party.
URNG In Guatemala, one of the conclusions of the Commission for Historical Clarification (CEH) is that “for the greater part of the confrontation, the cohesion of the Guatemalan insurgency revolved around the idea of the need for, and the primacy of, armed struggle as the only option” (1999: 21). This adherence to the belief in the necessity of armed struggle is similar to the experience of the FMLN in El Salvador. While the belief in the need for armed struggle was important in brining the left together, so was the role of Fidel Castro and Cuba. The Cubans
40 In April 2002, the FMLN again suffered an internal schism leading to the breakaway of a prominent group of efemelistas from the revolutionary period. Facundo Guardado (the FMLN’s presidential candidate in the 1999 presidential elections) along with other founding members of the FMLN left to form the Renovating Movement (MR). The MR emerged out of differences between some of the more orthodox leaders of the FMLN and the more moderate leaders. The MR defines itself as a party of the “democratic left, social-democratic and heir to the traditions of the popular organizations of the last thirty years.” Though the new party was made up of a number of the FMLN’s key historical figures, the MR attained less than 2% of the national vote in the March 2003 elections and ceased to exist as a political party. Like the PD, the MR failed to surpass the 3% threshold required by Salvadoran electoral laws to remain a political party.
95 offered assistance to the four insurgent groups under the conditions that they work more closely together.41 Unlike the FMLN, though, it is unclear whether the URNG ever attained the same level of unity. Two major events during the 1980s fractured URNG cohesion. First, the overwhelming response of the Guatemalan military convinced many insurgents that revolutionary success was unlikely, if not impossible. As discussed later in this chapter, the military struck ferociously in the early 1980s at the same time that the URNG looked to unite. Dunkerley claims “this failure to maintain an authentic unity must be attributed in large measure to the adverse military situation in which the left was placed within a few months of the URNG’s formation although it cannot be divorced from the distinct background of its component organizations” (1988: 481). In 1984, Mario Payeras led a breakaway faction of the EGP away from the URNG. Following the military’s scorched-earth policy of the early 1980s, Payeras and others believed that there was no room for discussion within the revolutionary movement over the correct response to the military’s increased repression and the political opening with the return to electoral politics in 1985 (Vinegrad 1998: 217). The second major event was the beginning of a return to civilian rule in Guatemala with the election of Christian Democrat Vinicio Cerezo in 1985. With the beginning of the return to civilian rule, there were discussions among the insurgents as to whether there was a significant opportunity for them within the new environment. Many withdrew from the insurgency to participate in a revitalized civil society.42 Many mid-level URNG cadres left during this period. These defections during the war hurt the URNG’s negotiating leverage and weakened its prospects for electoral competition.43
“Hay una verdad aquí: la guerra unifica, porque en la Guerra es todo más sencillo, todo mundo puede distinguir al enemigo. Pero en la paz muchas veces no se sabe distinguir al enemigo, y eso generó diferencias esenciales entre nosotros” (Jorge Soto quote from Velia Jaramillo 2003).
41 Juan Francisco García, interview by author, Guatemala City, Guatemala, 14 April 2004. 42 Others left to attend to medical concerns, from sheer exhaustion, or because of differences with the leadership over the direction of the organization. 43 Tania Prado Palencia, interview by author, Guatemala City, Guatemala, 19 April 2004; Alberto Ramírez Recinos, interview by author, Quetzaltenango, Guatemala, 13 April 2004.
96 This fracturing of the URNG in Guatemala that had begun in the 1980s continued after the elections in 1999. Many Guatemalans on the left held Rolando Morán of the EGP (aka Ricardo Ramírez) in high esteem and believed he was the one most capable of keeping the revolutionary left united and bridging the gap with the non-revolutionary left.44 Unfortunately for the URNG, Morán died in 1998, leaving behind a divided left. The next in line to Morán would likely have been Rodrigo Asturias, but during the latter stages of the peace process he and his party were implicated in a kidnapping scandal. Asturias was forced to remove himself from the public face of the URNG. Jorge Soto (aka Pablo Monsanto), leader of the FAR, became the leader of the URNG upon Ramírez’s death. Soto and the FAR belonged to the most outdated and Marxist of the four organizations comprising the URNG (Sichar n.d.). After losing out for the position of the Secretary General to Rodrigo Asturias (ORPA) in August 2001, Soto questioned the relationship that the URNG had developed with the country’s ruling party. The governing FRG was also the party of Ríos Montt, one of the men most responsible for the mass killing of the early 1980s. Soto resigned from the party in May 2002. Less than two hundred members of the URNG resigned alongside Soto, many of which followed Soto into the ANN.45 Similarly, Arnoldo Noriega resigned from the URNG in August 2001 after becoming disillusioned with the direction of the new political party (Latin American Regional Reports, September 25, 2001: 3). Noriega was a leader in the EGP. He had been involved in the peace process and had served on the URNG’s national executive for a number of years. Another high profile break within the URNG involved one former comandante defecting to the “enemy.” Pedro Palma Lau (aka Comandante Pancho of ORPA) took a position in the government and ran for Congress in 2003 as a member of Ríos Montt’s FRG. Rios Montt had come to power in the early 1980s through a military coup and had overseen some of the most brutal years of the repression. In the 2003 elections, former guerrillas were thus associated with the URNG, ANN, FDNG, and FRG. In addition, the ANN coalition’s presidential candidate from the 1999 election ran as the candidate for the center-right Unidad Nacional de la Esperanza (UNE), or National
44 Tania Prado Palencia, interview by author, Guatemala City, Guatemala, 19 April 2004; Juan Francisco García, interview by author, Guatemala City, Guatemala, 14 April 2004. 45 Soto’s subsequent move to control the ANN also led many to leave that party in order to find a home in the FDNG.
97 Union of Hope. According to Rodrigo Asturias, the withdrawal of many former combatants from the URNG was to be expected as the organization made the transition to political party. Unfortunately for the URNG, it did not possess enough noncombatant support to survive the loss of these members and it proved unable to recruit significant numbers of new supporters. These divisions led to disillusionment among former combatants and confusion among the general public.46 As a result, the former insurgent group’s electoral support declined precipitously in 2003.
Discussion While both the URNG and the FMLN have suffered internal schisms in the postwar period, the debilitation has been much greater for the URNG. Throughout the 1980s and 1990s, the URNG lost important members to civil society, other political parties and even high profile defections to the government. These events undermined the URNG’s unity during the war and as a political party. The divisions within the URNG also made it difficult for voters to follow as former combatants ran for office on the party lists from the extreme left to the extreme right. They also weakened the URNG’s argument that it was a party distinct from those already in existence. A larger party might have been able to overcome these challenges, but the URNG was relatively weak both militarily and popularly, making it extremely difficult to overcome the obstacles. In El Salvador, while ideological and strategic differences also existed and sometimes resulted in deadly outcomes, the FMLN remained unified during the war. Most of its leaders and combatants believed that the FMLN was capable of military victory and saw little benefit to leaving the insurgent group. It was not until after the negotiations that eventually led to the Peace Accords and after the FMLN occupied its seats in the Legislative Assembly that the divisions within the FMLN became irreparable. The evidence provides mixed support for the hypothesis. On the one hand, both insurgent groups suffered a breakdown in unity, yet it does not appear that both groups suffered to the same extent. New political parties often fail because of a breakdown in unity. In the case of the smaller insurgent group, the URNG, it has struggled to remain a minor political party with all of these divisions. The larger of the insurgent groups, the FMLN, also suffered severe divisions. While the divisions might have undermined its chances of winning the presidency or
46 Alberto Ramírez Recinos, interview by author, Quetzaltenango, Guatemala, 13 April 2004.
98 having a larger majority in the legislature, the FMLN was sufficiently large to remain a major political party. It continues to be the second strongest party in El Salvador while the URNG fights for its political survival.
Control of Territory Hypothesis 5 states that insurgent groups that controlled territory during the war are more likely to be successful than groups that are unable to control territory. A group that has already established a nationwide presence and organization as an insurgent group will be much better prepared to utilize that national organization for electoral competition. In this section, I evaluate the hypothesis at the departmental level. Those insurgent groups that have a nationwide presence and control greater national territory are more likely to succeed than insurgent groups that do not.47
FMLN In January 1981, the FMLN launched its “final offensive” to bring down the government of El Salvador. The FMLN planned for a nationwide insurrection to accompany its military attack. The offensive failed to achieve its goals as the newly formed insurgent group had difficulty coordinating its military attack and the civilian insurrection never materialized. Following the failed offensive, the FMLN dug in for a long drawn out war. It sought to capture national territory in order to dismantle “the existing state’s infrastructure of local government offices, jails, telephone exchanges, police stations, military garrisons, and counterinsurgency military and civil defense facilities” (Alvarez 1988: 79). As the FMLN would dismantle the state’s institutions, it would simultaneously establish “alternative governing authority” based upon “widespread” and “voluntary” campesino support (Wood 2003). A new society would emerge in these FMLN zones based upon “cooperation” (McClintock 1998: 74) and “solidarity” (Alvarez 1988: 87) thus preparing them for the day that the FMLN would capture state power. In FMLN controlled zones, there existed “numerous collective initiatives, including collective agricultural production, and a strong sense of community among the FMLN militants” (McClintock 1998 74). Collective agricultural production served the needs of the guerrilla army
47 If the ability to control territory as an insurgent group is critical to the success of the new political party, we would also expect the party to perform better in those areas where they maintained a presence, than in those areas where they did not. I evaluate this corollary in Chapter 7 using municipal level electoral data.
99 as well as the needs of those who could not adequately produce for themselves (Pearce 1984). Another collective initiative in the zones under FMLN control involved the development of popular education programs. These programs “established close links between education and other practices and goals not normally considered part of it – community organization, political struggle, and social transformation” (Hammond 1998: 5). The popular education programs were designed to serve the needs of the revolutionaries and prepare the citizens for the day that the FMLN came to power. The political, social, and economic benefits of living in FMLN controlled zones should have benefited the FMLN during its transition to political party. During the war, the FMLN was said to control areas in Chalatenango and Morazán, with other positions in Cabañas, the central zone of San Vicente, and surrounding the Guazapa volcano in San Salvador (Alvarez 1988: 85; Byrne 1996).48 Its presence also extended into the departments of San Miguel, La Unión and Usulután. The FMLN established a significant military presence in eight of the country’s fourteen departments and had developed a national structure to conduct a military campaign that was easy to “superimpose” on the country in carrying out a political campaign.49 The national presence of the FMLN as an insurgent group and its success as a major political party provide support for the territory hypothesis.
URNG During the early years of the war in Guatemala, Zacapa, Izabal, El Progreso, Alta Verapaz, and Chiquimula were considered “guerrilla strongholds” (Wickham-Crowley 1992). The Guatemalan insurgents attempted to establish alternative authority in these zones in order to compete with the Guatemalan military. One of the initial guerrilla groups, the Movimiento Revolucionario 13 de Noviembre (MR-13), or the November 13th Revolutionary Movement, created “peasant committees” to be left behind after the guerrillas had withdrawn from an area of the country; such groups were primarily active in Izabal and parts of the capital. These committees were selected by the village inhabitants and had the goal of challenging “the real authority in each village with the military commissioners and the auxiliary mayors, impart justice ‘outside the framework of bourgeois justice,’ coordinate efforts to repair the damage the army provokes in its excursions into the zone, and economically sustains the families of comrades
48 Alvarez cites an undated document, El Poder Popular en El Salvador, by J. Ventura(e) as the source of FMLN activity. 49 Ana Guadeloupe Martínez, interview by author, San Salvador, El Salvador, 1 May 2004.
100 incorporated into the guerrilla struggle or killed by the army” (Galeano as cited in Collazo- Davila 1980: 126-27). Unfortunately for the guerrillas, in a scenario that would be repeated over the next twenty years, between November 1966 and March 1967 the Guatemalan military killed roughly 8,000 civilians – in the process wiping out, for the moment, the guerrilla base of support. The offensive also decimated the guerrillas, although they did re-emerge in the 1970s. In the late 1970s and early 1980s, the guerrilla presence was no longer centered in Zacapa, Izabal, El Progreso, Alta Verapaz, and Chiquimula, but in the departments of San Marcos, El Quiché, and Huehuetenango (Dunkerley 1988: 473-74). But, again, as the guerrillas sought to establish a new civilian base among the region’s indigenous, the military struck ferociously through its scorched-earth campaign. As Dunkerley (1988) writes
The inability of the URNG to sustain military operations outside its core areas of support and, more importantly still, its failure to provide protection where it possessed a longstanding base, proved to be disastrous...[T]this strategic failure manifested several similar features, the most pronounced of which was the vulnerability of the civilian population to retribution from the army. Although the rebel combatants conducted a measured retreat in reasonable military order, all of the groups later recognized that the precipitate offensive of September 1981 to May 1982 permitted the army to separate them from a large sector of their support (491).
By the mid to late 1980s, the military’s offensive had forced the guerrillas back to their historic “core zones of support” – northern Huehuetenango and Quiché, San Marcos and Petén – along the Mexican border (Dunkerley 1988: 497). Unlike the FMLN, the Guatemalan guerrillas were unable to control and manage territory for extended periods of time. In addition, many of the villages where the URNG used “armed propaganda” were severely repressed by the Guatemalan military. Given that the URNG’s presence placed these communities in danger, there are concerns that the territorial presence of the URNG might have a negative effect on its electoral support in the postwar as a political party. The URNG was too weak to protect the village when confronted by the military leaving the communities to fend for themselves.
101 Discussion I hypothesized that insurgent groups that are able to capture and control territory are more likely to succeed as political parties than those that are not. The evidence from the FMLN in El Salvador and the URNG tend to support the hypothesis. During the Salvadoran conflict, the FMLN established a significant presence in eight of the country’s fourteen departments, with two of these departments (Chalatenango and Morazán) under their control for nearly the entire period of the war. The FMLN also set out to prepare the people living within these areas for life after the war. At the point of becoming a political party, the FMLN had little difficultly in superimposing its military and nonmilitary structures into party structures throughout the nation. On the other hand, in Guatemala, the URNG struggled to gain a presence outside of its core bases of support. During each decade when the guerrillas looked to extend their presence, the Guatemalan state struck back ferociously, forcing the URNG back into the highlands and its core base of support. As a result, the URNG could only count a significant presence in four of the country’s twenty-two departments. To put the territorial presence of each group in a little more perspective, the FMLN received less than ten percent of the vote in only one department (out of fourteen) in its first election. The URNG, on the other hand, received less than ten percent in nine departments (out of twenty-two). These results are consistent with the FMLN’s performance as a major party and the URNG’s performance as a minor party.
Human Rights Violations Hypothesis 6 addresses the relationship between the insurgent group and the civilian population. I hypothesized that insurgent groups that are less violent towards civilians during the war than government forces are more likely to be successful than those insurgent groups whose civilian directed violence outweighs that of the government. Both insurgent groups and government security forces can be very violent towards civilians, particularly an individual believed to be sympathetic to the “enemy”. Therefore, we should expect the manner in which the war is fought to be a strong predictor of the insurgents’ future performance as a political party.
102 FMLN As part of the Peace Accords that brought about an end to the Salvadoran civil war, both sides agreed to the establishment of a commission to investigate human rights violations committed during the conflict. Among its conclusion, the United Nations Truth Commission Report determined that repeated “serious acts of violence” had been committed by the FMLN in roughly 5% of the cases, while the state and related security forces were identified as the perpetrators roughly 85% of the time.50 Given the preponderance of human rights violations committed by government forces against the civilian population, one would expect the FMLN to fare very favorably as a political party. On the other hand, it is quite possible that the crimes or violations committed by the FMLN were not accounted for in the Truth Commission Report. While there is no credible evidence that the FMLN targeted civilians on a scale comparable to the Salvadoran military, the FMLN frequently targeted the country’s infrastructure – electricity grids, bridges, dams, water stations, highway chokepoints, etc. The attacks by the FMLN against the country’s infrastructure represented an attempt to undermine support for the government by proving that it could not provide for the general population and that billions of dollars in US economic aid was wasted. It is quite possible, though, that FMLN attacks against the country’s infrastructure had the opposite effect of what had been envisioned. Instead of encouraging civilians to support the insurgents, who were capable of “protecting” them and “providing” such services, the FMLN could have pushed many “independent” civilians to support the government by making life unbearable for civilians, many of which had no strong allegiance to either side. In addition to the attacks upon the civilian infrastructure, the FMLN, particularly the ERP, sometimes resorted to forced recruitment. The forced recruitment of soldiers most likely resulted in further alienation from the population. Therefore, both the forced recruitment of civilians and attacks upon the country’s infrastructure could have had a damaging effect upon support for the guerrillas in the war-torn areas of the country, more so than would be discerned simply by looking at the Truth Commission Report. On the whole, the FMLN should not have been disadvantaged when competing as a political party. The FMLN did not have a reputation for committing large-scale violence
50 The remaining 10% of the human rights violations noted by the Truth Commission were either committed by individuals or unknown assailants.
103 towards the civilian population, whereas the forces aligned with the government did. In addition, while the FMLN did resort to forced recruitment at times, it was not very widespread.51
URNG Guatemala’s civil war was one of the longest and most violent in Latin America. Like El Salvador, a commission was established to investigate human rights abuses committed during the conflict. The Commission for Historical Clarification (CEH) collected information on 42,275 victims of the Guatemalan conflict, including 23,671 arbitrary executions and 6,159 forced disappearances.52 The CEH concluded that over 200,000 Guatemalans were killed, mostly members of the unarmed indigenous population, over the thirty-six years of the war. In terms of the human rights violations committed against the population, the CEH found the state’s security forces and related paramilitary groups responsible for 93% of all violations while the insurgents were responsible for 3% (CEH 1999).53 The violations included 669 massacres (“collective killing of the defenceless population”) and 440 villages completely destroyed. The CEH concluded “that the violence was fundamentally directed by the State against the excluded, the poor and above all, the Mayan people, as well as those who fought for justice and greater social equality” (CEH 1999: 1). The types of violations included torture, forced disappearances, murder, and rape, among others. While the Guatemalan military and the country’s security forces committed the majority of all violations, the Guatemalan guerrillas were not without fault. Insurgents regularly engaged in acts of economic terrorism and sabotage. In urban and rural areas, the insurgents kidnapped and killed businessmen and foreign diplomats, both to spread terror throughout the population and to raise funds to advance their revolutionary goals (Collazo-Davila 1980: 109-113). Collazo-Davila argues that while the guerrillas of the 1960s were more “selective” in their use of terrorism, their actions in the 1970s and beyond had the goals of “attempting to provoke government counterterror” by engaging in “armed propaganda” (1980: 124-125). Many targets
51 To further evaluate the effect of the violence during the country’s civil war on the FMLN’s electoral fortunes, one might also want to consider how well the FMLN performed in those areas most severely affected by the violence relative to those areas where the war’s impact was less severe. One would expect that those areas of the country most affected by the violence would favor the FMLN because of the overwhelming imbalance of human rights violations committed by the military and state security forces. I explore this line of reasoning in Chapter 7. 52 http://shr.org/guatemala/ceh/report/english/conc1.html 53 The remaining 4% of the human rights violations noted by the CEH were either committed by individuals or unknown assailants.
104 of the government’s scorched-earth program of the early 1980s were towns where the guerrillas had engaged in armed propaganda. Typically, the guerrillas would seal off an area so that residents would be required to listen to what the insurgents had to say. The guerrillas would then hold a meeting and lecture the villagers “about such things as exploitation, the people’s army, security measures, and so forth” (Collazo-Davila 1980: 122). Given that the guerrillas could not confront the military head on, they retreated before the military advanced, leaving the village to fend for itself. As a result, the CEH partially blamed the guerrillas for these deaths. The guerrillas knew what the reaction by the state to their presence would be, and did little to protect the villagers that they had put in danger. Following the end of the war, the CEH found that the guerrillas had committed 3% of all violations. In a separate study, Sichar (2000) found that the URNG had committed fifty-one massacres relative to over 1000 committed by the government. It is clear from both these studies, though, that the crimes committed by the guerrillas were in no way comparable to the Guatemalan state. It is quite possible that Guatemalan voters did not punish the URNG for crimes that it committed directly against the civilian population during the war, but it did hold them responsible for leaving them to suffer under the wrath of the country’s security forces.
Discussion In terms of the human rights violations committed during the conflicts in El Salvador and Guatemala, both the FMLN and URNG were clearly responsible for fewer abuses than the respective state security forces. In this manner, one cannot make the argument that the electoral performances of the FMLN or the URNG as political parties were largely determined by their responsibility in committing violent acts against the larger population. While the URNG and FMLN were clearly responsible for fewer abuses than the respective state security forces, that does not mean that they were entirely benign when it came to the civilians with whom they came into contact. The FMLN and the URNG are both known to have engaged in acts of economic sabotage in hopes of undermining the government. What might explain the relatively poor electoral performance of the URNG was its inability to protect its supporters, or potential supporters; something that the FMLN did a much better job of accomplishing after the initial years of the war. As a result, the inability of the URNG to protect its potential electoral
105 supporters and the relative success of the FMLN in protecting its potential electoral supporters might explain the minor party success of the URNG and the major party success of the FMLN.
OVERALL IMPACT OF ORGANIZATIONAL CHARACTERISTICS
How well do organizational factors account for the relative performance of the FMLN in El Salvador and the URNG in Guatemala? Overall, the organizational characteristics and experiences of the FMLN and URNG provide a solid basis for explaining each party’s performance in the postwar period as a political party. In terms of the number of combatants each insurgent group can count on at the time of demobilization, the evidence supports the hypothesized relationship. The FMLN possessed a greater number of militants both in absolute and relative terms in comparison with the URNG. The larger insurgent group was successful while the smaller insurgent group was less successful. Evidence concerning an insurgent group’s political party experience also supports the hypothesized relationship. The FMLN had extensive ties with political parties prior to the outbreak of war, individually and as groups. The FMLN also maintained a strong relationship with a political wing, the FDR. Even though their relationship became strained during the latter years of the conflict, they were still able to cooperate in the selection of a common presidential candidate in 1994. The URNG, on the other hand, did not maintain a relationship with a political wing during the war. As a result, the URNG suffered in its inability to promote its vision of an alternative society to a larger domestic and international audience. The closest relationship the URNG developed with a political wing occurred shortly before its insertion into electoral politics. The URNG and FDNG maintained a strained relationship from the beginning and failed to maintain an alliance for even a single election. As a consequence, the URNG failed to capitalize on the experience that the FDNG had gained during the 1995 electoral campaign and in the 1995-1999 Congress. These relationships also support the political party hypothesis. Hypothesis three states the insurgent groups that can maintain their unity will be more likely to succeed than those groups that fracture. The finding here is somewhat mixed. Both insurgent groups suffered the loss of key members, but with a number of differences. First, unlike the FMLN, the URNG was hurt by the loss of key members prior to the signing of the peace agreement. Former insurgents left the party during the war to join opposition political
106 parties, to direct civil society groups, and at least one even joined the party of the individual most responsible for the scorched-earth campaign of the 1980s. By the first election in which the URNG competed, former guerrillas were represented in political parties from across the political spectrum. It is quite possible that former guerrillas participating in a number of parties in 1999 might have confused or dismayed voters, and did little to help the URNG prospects. Second, the URNG was much smaller organizationally than the FMLN, and the internal divisions might have been too much for a party of this size to overcome. While the FMLN also suffered a breakdown in unity, the first group to leave did so after the FMLN had established itself as the country’s primary opposition party following the 1994 elections. The FMLN had the advantage of being the second largest party in a presidential system favorable to the two largest parties.54 In terms of territorial control, the evidence also supports the hypothesis. During the conflict in El Salvador, the FMLN maintained a strong presence in eight of the country’s fourteen departments. In 1989, during the FMLN’s second “final offensive,” it coordinated attacks throughout the entire country, including all of its major cities. While the FMLN could not hold all the territory, it showed that as the country’s civil war was winding down, the Salvadoran military could not dislodge the FMLN from its strongholds in Morazán and Chalatenango and that the FMLN could coordinate attacks throughout the country. As a result, the FMLN was much better prepared for the challenges of electoral competition because it had a presence in most of the country. On the other hand, by the end of the war in Guatemala, the URNG only had a presence in four of the country’s twenty-two departments. It was unable to capture and control significant territory during the war, as the military onslaught always forced the guerrilla retreat. As the URNG attempted the transition, it had little presence in most of the country’s departments and municipalities. Finally, I also hypothesized that an insurgent group that is less violent towards civilians relative to the government should be more successful than those that are more violent relative to their respective governments. Both the FMLN and URNG were responsible for many fewer human rights violations than their respective opponents, yet each group fared quite differently as political parties. Neither the FMLN nor the URNG should have been disadvantaged because of its violent nature.
54 During the period under investigation, the FMLN successfully weathered the breakaway of one dissident faction. Following the 1999 election (the last election covered in the study), the FMLN also survived the dismissal of Facundo Guardado (2002) and the resignation of other party faithful (2005).
107 The FMLN appears to have been successful as a political party because it drew upon a significant number of former combatants, a strong relationship with mass-based organizations, and a military presence throughout large portion of the country. The FMLN was also helped by the prior experiences its member could bring to bear upon the new political party as well those of its “political wing” (political party). The URNG’s transition to political party was hindered because of its relatively small number of combatants and noncombatants and its limited territorial presence. In addition, the URNG was unable to draw upon any prior experience with electoral politics and failed to take advantage of the possible help that allying with the FDNG might have provided. These findings fully support the hypotheses. The two remaining hypotheses are less conclusive. Both insurgents suffered from schisms in the postwar period, yet only the URNG appears to have been significantly effected. Of course, it is possible that the FMLN would have won the presidency or a large majority in the legislature if it had not been for the disunity. Finally, there is less to be said concerning the role of violence and the performance of these two groups. The Guatemalan and Salvadoran states were responsible for the majority of human rights violations committed against the civilian population during each country’s civil wars. It is possible that the municipal level analyses in Chapter 7 will shed some light on this relationship. While some insurgent groups are better prepared for electoral competition given their organizational characteristics and civil war performance, the success of these groups as new political parties is not entirely of its own making. Electoral environments in which political parties compete can be more or less favorable to the success of new political parties. This is the subject of the next chapter.
108 CHAPTER 6 EVALUATING THE IMPACT OF ENVIRONMENTAL CONDITIONS
Organizational factors are important, but one must also take into consideration the environment in which these new political parties must compete. Some environments are generally considered more favorable to the success of new political parties than others. In this chapter, I evaluate how important the environment in which the FMLN and the URNG participated in was in explaining the performances of each insurgent group as a new political party.
Electoral Objective Hypothesis 7 suggested that former insurgent groups are more likely to succeed in parliamentary than presidential systems. Presidential systems tend to depress the number of parties in a system closer to two, making it more difficult for a new political party or third party challenger to succeed.
FMLN Like most Latin American nations, the president dominates the Salvadoran political system. In 1983, El Salvador adopted a new constitution establishing three independent branches of government: the executive, legislature, and judiciary.55 The president and vice- president are elected by popular vote for a five-year term without the possibility of reelection.
URNG Guatemala also has a presidential system where the president and vice-president are elected by popular vote without the possibility of reelection. The Guatemalan system differs from the Salvadoran one in that elections in Guatemala are held every four years, not five, and the elections are held concurrently with those of the legislature.
55 Though the presidency was established as the most powerful organ of government in the 1980s, it is not unreasonable to say that the independence and power was not really established prior to the Cristiani Administration (1989-1994). During the 1980s, the US embassy and Salvadoran military were at least on par with the president.
109 Discussion With regards to the FMLN and URNG, there is little one can say regarding the likelihood of new party success in presidential versus parliamentary systems. While both political parties have competed in presidential systems, the FMLN has gone on to become a major political party in the postwar period and the URNG has become a minor political party. At a minimum, one can say that competing in a parliamentary system is not a necessary condition for new party success; new political parties do emerge and succeed in presidential systems, as evidenced by the performance of the FMLN. While, in general, competing in presidential systems might be a negative for new political parties, it does not appear to be a strong explanation for the varied performance of the FMLN and URNG. As mentioned earlier, previous research has also identified a number of ways in which presidential systems differ such as the rules used to elect the president and members of the legislature. These factors might provide better explanations for the performance of these two parties.
Presidential Electoral Rules Hypothesis 8 suggested that former insurgent groups are more likely to succeed in systems with majority presidential runoffs than simple plurality rules for electing the president. If no candidate receives greater than fifty percent of the vote in the first round, a second round is held involving the candidates that received the two greatest vote totals in the first round. On the other hand, if a single party or coalition attains greater than 50% of the vote, it is declared the winner. One of the implications is that new political parties can try their luck in the first round of a presidential election. If no candidate captures greater than 50%, smaller political parties can attempt to work out an alliance with one of the two larger political parties that go on to compete in a second round. Under this scenario, voters might be more inclined to vote for a new political party even if its probability of winning is very low. Voters can hope that a second round is necessary and their preferred political party can leverage the support captured in the first round of voting into policy consideration and/or positions in the new government through negotiations with the two candidates that advanced to the second round.
110 FMLN In El Salvador, the president must be elected with a majority of the vote. As a result, elections are determined via majority runoff. As mentioned earlier, although the FMLN has competed in three presidential elections (1994, 1999, and 2004), I am only focusing on the first two elections. During the 1994 elections, the FMLN formed a coalition of the left involving the CD (the party from which the candidate emerged) and the MNR. This alliance prior to the first round helped propel the FMLN into an eventually unsuccessful second round runoff against the ARENA candidate. Given that the FMLN finished second, I do not believe that forcing the selection of the president into a second round of voting did anything to improve the overall success of the FMLN. The FMLN would have remained the second largest party regardless of whether a second round was held. In addition, none of the political parties that lost in the first round supported the FMLN candidate in the runoff. Prior to the 1999 elections, the FMLN and other left and center-left actors – including the Centro Democrático Unido (CDU), or United Democratic Center – failed to coordinate a common front. The FMLN presented Facundo Guardado, an ex-guerrilla commander, as its candidate. The alliance’s presidential candidate in 1994, Rubén Zamora, was the candidate for the CDU. One of the consequences of the former allied party presenting its own candidate was that the election was decided in the first round. An alliance would likely have forced a second round of voting since the winning candidate received slightly more than 50% of the vote while the FMLN ticket received 29% and the CDU ticket received less than 8%. On the other hand, it is doubtful that the outcome of a runoff election would have been any different.
URNG The Guatemalan constitution also requires that the president be elected with a majority of the vote. As discussed previously in Chapter 4, the URNG attempted to establish an electoral alliance for the 1999 elections along with three other left and center-left political parties -the FDNG, DIA, and UNID. However, in the months prior to the election, the FDNG withdrew from the alliance, leaving the URNG and ANN in a much weaker position. In the election, the alliance’s presidential candidate finished in third place with 12% of the national vote. Considering that neither of the two largest parties captured a majority of the vote necessary to win in the first round, the URNG was in a key position to help determine the outcome of the
111 second round. In the end, though, the ANN did not support either candidate and had no influence on the outcome of the second round. Unfortunately for the URNG, during the period between the 1999 and 2003 elections, it continued to have problems maintaining its party unity. As discussed in the previous chapter, differences of opinion existed among the former guerrilla commanders that eventually resulted in Jorge Soto’s resignation from the party. The URNG remained distant from civil society and many of its former combatants. By the time of the 2003 elections, the URNG was in a significantly weaker position, whereby the electoral rules for the presidency would have negligible effect on its overall performance. In the elections, the URNG captured 3% of the vote and finished in sixth place. Like 1999, the URNG was unable to negotiate an alliance with either of the two parties that advanced to the second round of the presidential elections.
Discussion Elections for the presidency in both El Salvador and Guatemala are determined by majority runoff. The existence of majority runoff elections is a condition that is supposed to work in favor of new political parties. For these two cases, it is not apparent that majority runoff elections instead of simple plurality elections had a significant impact upon either party’s performance and is therefore not a good explanation for their different performances. The FMLN would likely have remained the second strongest party regardless of the rules for electing the president. In the case of the URNG, it was unable to translate its third place finish into any significant accomplishments. It is always possible that the URNG would have done significantly worse under simple plurality, but, in explaining the varied performances of the FMLN and URNG, the rules for electing the president do very little.
Legislative Electoral Rules In addition to the electoral rules for the presidency, the rules for electing the legislature are likely to have a strong impact on the viability of a new political party. Hypothesis 9 suggested that former insurgent groups are more likely to succeed under proportional representation electoral rules than under single member district voting. How important are legislative electoral rules to explaining the performance of the FMLN and URNG?
112 FMLN In El Salvador, elections for the Legislative Assembly are conducted via proportional representation. The use of proportional representation should have helped the FMLN more than a system using single member district voting. In 1994, the FMLN captured a total of twenty-one seats (sixteen by department vote and five additional ones as a result of a nationwide district). This performance was good for 25% of all available legislative seats, making the FMLN the second largest party in the Legislative Assembly. In the 1997 elections, the FMLN increased its share of seats in the Legislative Assembly from twenty-one to twenty-seven with 33% of the national vote. We can also see the benefit of proportional representation to the FMLN by comparing these legislative results with the results of the FMLN in elections for the municipal council. Municipal council elections are determined using single member district electoral rules. It provides a good estimation of how well the FMLN would have performed had it competed under the alternative rules.56 In 1994, the FMLN captured 25% of the legislative seats with 21.39% of the national vote. In municipal council elections, the FMLN captured fifteen out of 262 (5.7%) municipal offices with roughly the same percentage of the vote as it received in the legislative elections (20.33%). For purposes of comparison, let us assume that the Salvadoran legislature also contained 262 seats, with one from every municipal district in the country. If legislative elections were conducted under single member districting voting similar to municipal council elections, the FMLN would not have been nearly as successful as it turned out to be with proportional representation, as the same percentage of the vote would have resulted in only about 6% of the seats. While comparing the FMLN performance against the hypothetical example of single member districts is a little far-fetched, a change in proportional representation from the current electoral formula to another is less so. Table 6.1 presents the seat allocation for the FMLN following the 1994 legislative elections under the existing largest remainders system where it captured twenty-one seats. I simulated the electoral results for the FMLN under conditions where its votes are translated into seats using the D’Hondt formula (which is the formula utilized in Guatemala). If FMLN votes had been translated into seats using the D’Hondt formula, the
56 Obviously, the strategic interaction between votes and parties under a different set of electoral rules would have led to somewhat inexact results than the ones I am using here.
113 party would have gained two fewer seats at the district level while the number of seats at the national level would have remained unchanged, leaving it with only nineteen. Again, we can see that the electoral rules under which the FMLN competes have been highly favorable.
URNG Like the elections for the Salvadoran Legislative Assembly, elections for the Guatemalan Congress use proportional representation. Under this system, the URNG captured nine total seats in the 1999 elections with 11% of the national vote. Nine seats accounted for 8% of the congress’ 113 seats. In 2003, the URNG received about 4% of the national vote in capturing 1.3% of all available seats.
Like El Salvador, elections for municipal council are determined via single member district voting. Therefore, we can get a reasonable approximation for how well the URNG would have performed had it competed under single member district voting for the legislature. Out of the 331 municipal council elections in 1999, the URNG’s alliance won fourteen. Hypothetically, if elections were determined using single-member district voting and 331 seats, the URNG performance would have been much worse, garnering only about 4%.
I also simulated the outcome of the 1999 elections for the URNG had the votes been translated into seats under an alternative formula for translating votes into seats. Table 6.2 presents the electoral results for the URNG in the 1999 elections under both D’Hondt and largest remainder electoral formulas. Under the present D’Hondt system, the URNG captured a total of nine seats. Seven were captured on district lists and two additional seats via the national list. Simulating the URNG’s performance under the largest remainder formula, the URNG would have won an additional four seats in the 1999 elections for a total of thirteen, picking up additional seats in Chimaltenango, Escuintla, Totonicapan, and Izabal. While thirteen would still leave the URNG in third place, an increase of four seats would have been 44% greater than their actual result.
Discussion As it relates to Guatemala and El Salvador, both the FMLN and URNG operate under relatively favorable conditions in elections for the legislature, more so for the FMLN. Each party competes under proportional representation electoral rules rather than the more inhibiting
114 single-member district rules. Both parties would likely have performed significantly worse under SMD rules. In terms of translating votes to seats, the FMLN operates in a more favorable environment. In the 1994 elections, the FMLN captured two more seats under the largest remainders formula than it would have had El Salvador used the D’Hondt method. The URNG might have been helped had it competed under the largest remainder formula where it would have captured four more seats than under the existing D’Hondt formula used in Guatemala including the 1999 elections. Simulating the electoral results of each party under a different type of electoral formula in its first elections does not cause the successful FMLN to fail or the unsuccessful URNG to succeed. On the other hand, the FMLN would have captured about 10% fewer seats under D’Hondt and the URNG would have captured 44% more seats using the largest remainder formula. Though the electoral formula did not determine success or failure, the URNG would have been significantly helped with a change in the electoral rules. Given that some viewed the URNG’s performance with nine seats or 8% of the total seats as a failure, thirteen seats or 11.5% of the seats might have provided the URNG with a greater opportunity to build on this performance in 2003.
District Magnitude Hypothesis 10 suggested that former insurgent groups are more likely to succeed when the district magnitude is high and when there is a nationwide district. Under proportional representation, as district magnitude increases, the cost of winning a seat decreases. Therefore, new political parties should benefit from competing in proportional representation systems that have high district magnitude. In addition to district level representation, both El Salvador and Guatemala elect a national slate of representatives from a single nationwide list. A potential advantage of a nationwide district is that it provides smaller political parties with the opportunity to win seats under more proportional electoral conditions. Nationwide districts are supposed to be more proportional than the smaller departmental districts, offering new political parties a greater opportunity to win seats from this district. Instead of utilizing a nationwide district to make up for the disproportionality of the geographic-based districts, a country might redistribute the seats among the geographic-based districts and get rid of the nationwide districts.
115 FMLN In El Salvador, sixty-four legislative seats are distributed among fourteen departments and another twenty from a single nationwide district. Voters cast a single ballot for their geographic-based district and the national district. The average district magnitude for El Salvador’s 14 departments is 4.57, while the district magnitude for the national district is simply the total number of seats available, 20. The average district magnitude is constant in El Salvador as there has not been a change in the number of deputies elected since the 1988 elections. In El Salvador, I simulate the impact that removing the national district would have had on the FMLN’s performance in 1994. I redistribute the twenty nationwide seats throughout the fourteen departments proportionally according to the departmental vote relative to its percentage of the national total. For instance, the 421,206 votes from the department of San Salvador accounted for 31% of the total national vote in 1994. Therefore, with eighty-four total seats available, San Salvador would now maintain twenty-six seats (84x0.31=26). After rerunning the analyses, it does not appear as if district magnitude significantly impacted the FMLN. As seen in Table 6.3, under the largest remainder formula with eighty-four seats distributed proportionally throughout the country’s fourteen departments, the FMLN picks up one additional vote compared to the current system.
URNG In Guatemala, 113 seats were distributed in the 1999 elections among twenty-three districts (twenty-two departments in addition to the capital district). Of the 113 legislative seats, ninety-one were distributed in proportion to district-level vote totals while the remaining twenty- two seats were distributed based upon each party’s electoral results on the national list. The average district magnitude by department in the 1999 elections is 3.96, and twenty-two for the national district. In 2003, the number of seats increased to 158. Of the 158 congressional seats, 127 seats were distributed among the twenty-two departments and the capital district and the remaining thirty-one seats were distributed according to the results of the vote for the national list. In 2003 the average district magnitude by department in the 2003 elections is 5.52 and thirty-one at the national level. In Guatemala, I also simulate the results of the first elections in which the URNG competed as if there were no national list and those seats were redistributed among the twenty-
116 two departments and capital district. Table 6.4 illustrates that the URNG would have gained two additional seats in the 1999 elections if no additional nationwide district existed.
Discussion In terms of the importance of district magnitude, both El Salvador and Guatemala combine departmental districts and a single nationwide district using proportional representation electoral rules for elections to the legislature. In the first election of each postwar period, the greater district magnitude in El Salvador and the success of the FMLN combined with the lower district magnitude in Guatemala and the poorer performance of the URNG does provide support for the hypothesized relationship between district magnitude and new political parties. As seen in Table 6.3, the FMLN picked up 20% of the available seats on the national list (5) while in Table 6.4 the URNG picked up 9% of the available seats (2). One way in which one can assess the impact of the nationwide district is to rerun the electoral results in both countries without the nationwide district and with those seats redistributed among the other districts. The results are not particularly significant. The FMLN would have picked up one additional seat while the URNG would have picked up two. These results suggest that district magnitude is not a key explanation for the parties’ divergent performance.
Electoral Thresholds Hypothesis 11 suggested that former insurgent groups are more likely to succeed in systems with low electoral thresholds. It is expected that a low electoral threshold makes it easier for new political parties to succeed.
FMLN According to the electoral laws in El Salvador, a political party that does not attain three percent of the national vote (or six percent in coalition) loses its legal status. The party maintains any seats it won in the assembly, but must re-apply for recognition as a political party. In order for the party to retain legal standing it must collect a number of signatures equivalent to three percent off all votes cast in the most recent presidential election (Article 159). The existence of a three percent electoral threshold has not directly factored into FMLN performance, as the party has attained at least 20% of the vote in each election in which it has participated.
117
URNG The electoral code (article 93b) in Guatemala establishes a four percent threshold for a party to capture a seat from the national list. If the party successfully attains representation in the Congress by winning at least one of the district level seats, the four percent requirement is waived. As a member of the ANN in 1999 the URNG easily cleared the four percent hurdle with 11% of the legislative vote and 12% of the presidential vote. The URNG did come close to losing its legal standing following the results of the 2003 election, where it received 4.04% of the departmental vote in legislative elections. The URNG captured two seats with 4.2% of the national list vote in legislative elections.
Discussion As of yet, the existence of an electoral threshold has not affected either group, although it is quite likely to be a challenge for the URNG in the 2008 election. Neither the FMLN nor the URNG has failed as a political party by missing the electoral threshold required by each country’s political system. In El Salvador, the FMLN has established itself among the strongest parties in the country and was never in danger of being disqualified for failing to surpass the required threshold. The URNG, on the other hand, while it has not fallen below the electoral threshold, is teetering on the brink of disqualification after narrowly surpassing the electoral threshold in the 2004 presidential elections by hundredths of a percentage point. This result has ensured that it will survive to compete in at least one more election, but offers no guarantees for future elections after that.
Voter Attitudes Hypothesis 12 suggested that former insurgent groups are more likely to succeed in systems with high levels of voter absenteeism. A new political party should be more likely to succeed if there is a large percentage of the population without ties to existing political parties, than if most voters have already voted for an existing party.
118 FMLN In El Salvador in the 1970s, political violence escalated following the fraudulent 1972 and 1977 presidential elections. When elections returned in 1982, the design of the electoral process focused, though not always successfully, on preventing fraud and other irregularities (Baloyra-Herp 1995: 49). The process was not designed to bring as many people into the political process as possible. The cumbersome registration and voting process, combined with the violence associated with the civil war and with the FMLN’s call to boycott the elections, kept voter participation levels low. Baloyra-Herp (1995:59) characterizes the elections of the early 1980s as having “had decent levels of popular participation” relative to the “lower levels” of the 1988-1991 elections. Table 6.5 presents voter turnout figures from El Salvador from the 1982 Constituent Assembly elections until the 1994 postwar elections. While there are several explanations for the decline in voter participation into the early 1990s, the lower levels do seem to suggest that there existed a pool of potential voters for a new political party to tap. A new political party, such as the FMLN, should have had the opportunity to mobilize citizens who dropped out of the electoral process after the initial elections of the early 1980s, had heeded the call to boycott the elections, or simply had never participated in the process. There does not seem to have been a shortage of potential voters for the FMLN as a new political party in 1994.
URNG In Guatemala, there was a significant pool of potential voters from which the URNG could draw voters. Table 6.6 presents voter turnout figures for Guatemala from 1950 – present. The average percentage of registered voters who have voted in elections since 1950 is 52%. Another measure of the lack of participation in the political process of Guatemalans is that the average percentage of people voting as a percentage of the total voting age population is even lower at 27.51%. Upon the return of competitive elections in 1985, 70% of registered voters turned out. These levels had not been seen since the CIA-backed coup to remove the democratically elected government in 1954. As the civil war dragged on for another decade, voter turnout dropped significantly. As in El Salvador, the initial enthusiasm and participation of the population following the return of civilian government was followed by a sharp drop-off in voter participation. 69% of all
119 registered voters turned out to vote in 1985 when a civilian was elected president for the first time in over thirty years. The percentage of all registered voters who turned out dropped to 56.4% in 1990 and 46.7% in 1995. Two of the more likely causes of the drop in voter participation were the inability of the civilian government to provide for economic (jobs) or physical (end the war) safety of ordinary Guatemalans. The country has had one of the lower voter participation rates in Latin America (Spence 2004), and a new political party should have had plenty of opportunity to mobilize support among a population that had once turned out or that had never turned out at all.
Discussion In both Guatemala and El Salvador, voter absenteeism has historically been high, so there is little evidence that the citizens were content with their previous electoral choices. As Spence (2004:81) notes, “Since the first round of elections in the late 1980s both Guatemala and, until 2004, El Salvador had been plagued with low voter turnout, among the lowest in Latin America.” Therefore, both the FMLN and the URNG were entering under relatively favorable political party systems in terms of voter turnout – a significant proportion of the population had remained outside electoral politics, a population the new political parties might be able to reach. Of course, a history of low voter turnout might hurt these new political parties. If the FMLN and URNG capitalized on the high levels of voter absenteeism that existed prior to their first electoral competition, we should also have witnessed a positive change in voter turnout with their initial entry. In fact, that is exactly what we see in both cases. In elections for the Salvadoran Legislative Assembly in 1991, 1,153,013 total votes were cast. In the 1994 elections, 1,453,299 were cast, an increase of 300,286 additional votes. In the case of El Salvador, the FMLN appears to have benefited from the historically high level of voter absenteeism and the increase in voter turnout in 1994 given that the FMLN received slightly less than 300,000 votes at 287,811. In Guatemala, less than half the registered voters turned out in 1995 at 47% (See Table 6.6). In 1999, voter turnout increased to 54%, 7% higher than 1995. In the 1999 elections, the URNG and ANN captured eleven percent of the total vote. It appears that both the FMLN and URNG benefited from the historically high levels of voter absenteeism and the increased voter turnout in the first election in which they participated.
120
Bipolarization Hypothesis 13 suggested that former insurgent groups are more likely to succeed in a presidential system when bipolarization has not yet set in. In general, two major political parties tend to emerge in presidential systems, making it more difficult for third party challengers to succeed if these two parties have already been well established.
FMLN Elections occurred throughout the 1960s and 1970s. Competitive elections, though, only began with the 1982 Constituent Assembly and 1984 presidential elections. While these elections were competitive in the sense that there existed a range of electoral alternatives along a somewhat limited ideological spectrum (center/center-right to extreme right), they were still flawed – neither left nor center-left parties could legally compete; the US government intervened in the political process, and; voting was conducted in the midst of a civil war. The FMLN was threatening voting centers, while the state’s security forces were threatening people who could not prove that they had voted. These were less than ideal voting conditions. Despite these flaws, the elections of the 1980s offer useful information about popular support for the top two parties relevant to this hypothesis. During the 1980s, there were two main political parties in El Salvador, the PDC and ARENA. During the three elections for the Legislative Assembly prior to the incorporation of the FMLN, ARENA and the PDC received 84% (1985), 83% (1988), and 82% (1991) of the national vote. The two parties also dominated the voting in the 1984 (73.5%) and 1989 (90.3%) presidential elections. The fact that the two largest parties had attained a such a high percentage of the national vote in each election since the return to competitive party politics in El Salvador should have worked against the FMLN relative to the experience of the URNG in Guatemala. This unfavorable condition was counterbalanced somewhat by the fact that one of the two largest parties, the PDC, had been losing votes and overall support. From 1982’s election for a Constituent Assembly to the 1985 legislative election, the PDC averaged 47% of the vote. From 1988 to 1991, the PDC only averaged 33%. The PDC was suffering for a number of reasons. The PDC failed in its campaign promise to bring an end to the war. The party and its leader, José Napoleon Duarte, had been involved in several corruption scandals. The PDC’s association
121 with the United States might also have hurt its credibility and electoral performance. Therefore, the FMLN entered at a relatively favorable point in time where bipolarization had not yet stabilized and, it appeared, the system was actually moving away from it with the decline of the PDC. In 1994, the PDC average vote fell to 18% across the presidential, legislative, and municipal elections. It seems that the party most hurt by the performance of the FMLN was the PDC.
URNG In Guatemala, the URNG first participated in the 1999 elections. During the prior three elections, five different parties had finished in either first or second place. Unlike the 80% plus vote for the top two parties in El Salvador, the two largest Guatemalan parties only captured 52% (1985), 43% (1990), and 59% (1995) of the national vote in recent elections. In addition, the three largest political parties to emerge out of the 1995 elections (PAN, FRG, FDNG) were all relatively new to the Guatemalan political scene. Given the frequent turnover in the top two positions and the extremely low percentage of their combined vote, bipolarization is not a strong explanation for the URNG failure. No two political parties dominated the Guatemalan political system. In fact, the low vote total for the two largest parties and Guatemalan voters’ apparent willingness to support new alternatives boded extremely well for the URNG. If anything, the situation in Guatemala was much more favorable for the URNG than the situation in El Salvador was for the FMLN.
Discussion The electoral conditions in terms of bipolarization were much more favorable to the URNG than the FMLN. The two largest parties in Guatemala were unable to secure more than 60% combined in any single election. In addition, the top two parties had never been the same. In El Salvador, the two largest parties secured greater than 80% of the vote combined in each election. The performance of the two main political parties in each country prior to the incorporation of the insurgents should have provided a more favorable environment for the URNG than the FMLN. Given the nature of the vote for the two main political parties in each country and the performance of the former insurgent groups, it does not appear as if bipolarization is a strong explanation for insurgent group performance as new political parties.
122 The relevance of bipolarization might be minimal in the case of these new political parties and many other groups that share insurgent histories. The idea behind bipolarization is that presidential political systems tend to reduce the number of parties in the system until it reaches some number close to two. Once the populace has coordinated on the top two parties, third party challengers are severely disadvantaged. In the case of former insurgent groups becoming political parties, we are often talking about systems with limited electoral history. In these two countries, elections had only been reinstituted for ten years before the inclusion of the armed groups. The short time period was unlikely to have been enough time for voters and elites to agree on the two most important parties. Perhaps more relevant for these groups, leftist and center-left parties were prohibited in both countries until shortly before the inclusion of these insurgent groups in the electoral process. The bipolarization that had begun to form in El Salvador involved center-right and far- right parties; no center-left or leftist party was allowed to participate until the 1989 presidential and 1991 legislative elections. The same could be said for the URNG; center-right and far-right political parties dominated the political system and leftist political parties were not allowed to participate until 1995. In both cases, systematic exclusion of leftist parties meant that 1980s vote totals did not measure bipolarization across the political spectrum. Once leftist parties were allowed to participate, it would not be surprising to see rapid change in previous voting patterns in the first few elections. After several elections that feature this full range of parties, voting patterns should be expected to stabilize and bipolarization should have its hypothesized effect on new parties.
Economic Conditions Hypothesis 14 suggested that former insurgent groups are more likely to succeed when there is an economic crisis [that voters do not blame on the group].
FMLN El Salvador experienced relatively robust growth rates in the early 1990s after suffering negative growth in the 1980s (See Table 6.9). In 1989, Alfredo Cristiani of the ARENA party was elected president, partly as a result of the PDC’s poor management of the economy. Despite
123 whatever ties ARENA’s founders had to the country’s death squads, the party was able to take credit both for bringing an end to the civil war and overseeing the country’s economic revitalization. During the early 1990s, the Salvadoran economy benefited from the end of the conflict, international aid associated with the Peace Accords and the continued growth in remittances from Salvadorans living abroad. In the two years prior to the first election in which the FMLN participated (1992-1993), the growth rate of real gross domestic product (per capita) grew by 3.97% and 4.86%. The growth rate of GDP per capita also grew 3.38% during the 1994 election year. Though El Salvador still remained poor by most standards, the economy was growing at its fastest rate in over fifteen years just as the FMLN was entering the electoral system. This factor could not have worked in favor of the FMLN.
URNG In Guatemala, the economy was not growing quite as rapidly at the moment of the URNG’s insertion as the Salvadoran economy was at the time the FMLN insertion. During the 1980s, Guatemala “saw the worst economic crisis since the 1930s, with massive social consequences. By 1990, according to U.N. and Guatemalan government figures (Jonas 1991, 177-78), nearly 90 percent of the population lived below the poverty line (and more than 70 percent in extreme poverty – that is, unable to afford a basic minimum diet)” (Jonas 1995: 30). The low growth rates in Guatemala continued into the 1990s and into the 1999 election year even with funds flowing in from the international donor community. The economic conditions in Guatemala prior to the 1999 elections were highly favorable to the success of a new political party (See Table 6.10). In 1997 (0.12%) and 1998 (-0.87%), the two years prior to the election, Guatemala experienced very low rates of economic growth, with 1998 actually being negative.
Discussion While both El Salvador and Guatemala suffered from high rates of unemployment, poverty, and inequality (and continue to this day), the economic situation in Guatemala was much more favorable to the potential success of the URNG. Economic growth fluctuated around zero in Guatemala (even though it “jumped” to a positive rate of 1.39% in 1999) while the Salvadoran economy grew quite well. In 1995, Alvaro Arzú of the National Advancement Party led the first round of presidential voting with thirty-seven percent. Arzú eventually won the
124 second round with fifty-one percent. In 1999, the incumbent party’s vote share fell to thirty percent. The poor economic conditions might have contributed to the electoral success of opposition parties, just not enough to make the difference for the URNG between a minor party and a major party.
Social Cleavages Hypothesis 15 suggested that identity-based insurgent groups are more likely to be successful if they represent a significant percentage of the population.
FMLN In the case of the FMLN, there is no relevant connection between the insurgent group and an ethnic population. The last census to ask about the indigenous background of Salvadorans occurred in 1930. In the early twentieth century, indigenous people were estimated to have comprised roughly 6% of the population, but by the late 1970s, as revolutionary violence engulfed the country, the indigenous made up only 1% of all Salvadorans (U.S. Library of Congress). In addition, the FMLN had no identity-based component to its organization or goals.
URNG During the 1960s, the early failure of the insurgent forces was often blamed on the inability of the insurgents to attend to the specific needs of the country’s indigenous majority (Jonas 1991: 69). The government responded to guerrilla activity by murdering thousands of indigenous peasants living in suspected guerrilla zones, regardless of their connection to the insurgent forces. When the insurgents reorganized in the 1970s, they purposefully adopted a more indigenous character; this was particularly true for the EGP and ORPA. Jonas (1991: 133- 35) identifies several factors contributing to the increase in indigenous support for the guerrillas: increasing economic and ethnic marginalization of the indigenous population, the denial of legal and political recourse to pursue justice, religious experiences, the “crisis of the counterrevolutionary state,” and the concerted effort by the revolutionaries to develop a base among the indigenous population. By the late 1970s, indigenous support for the insurgents was at an all-time high. The guerrillas had succeeded in making inroads among the Mayan population and related social
125 movements (Stanley and Holliday 2002). As the 1980s began, though, the military and the other security forces targeted the indigenous communities of Guatemala mercilessly. During the bloodiest years of the war, the Guatemalan military conducted a scorched-earth campaign in order to eradicate the popular support base of the insurgents. From 1981-1983, 440 villages were destroyed; up to 150,000 Guatemalans were killed, mostly Mayan Indians though some ladinos (a westernized Spanish-speaking Latin American) were also among the casualties; and roughly an additional 1.5 million were internally displaced or sent fleeing the country. Though the Guatemalan military was clearly responsible for the atrocities, a number of Mayans and even the Commission for Historical Clarification’s Guatemala: Memory of Silence (1999) found fault with the guerrillas for their campaign of “armed propaganda.” After occupying a town to “gain supporters or demonstrate their strength,” the guerrillas withdrew prior to the arrival of the military. In many instances, civilians were subsequently killed and entire villages even destroyed because of the dangerous situation that the guerillas had placed them in. As a result of the URNG failure to protect its sympathizers and supporters, the remaining communities developed a “broad sense of abandonment, deception and rejection in these sectors” (page 7 of the online CEH). While the military difficulties of the insurgency in the 1960s led to a self-criticism within the guerrilla groups of the role of the indigenous in the trajectory of the revolution, difficulties from the increased state repression in the late 1970s and early 1980s also had an important affect on the relationship between the indigenous and the URNG. Over the next several years, the indigenous community began to organize on its own, becoming much more autonomous as a result.57 The new social movements of the 1980s reflected indigenous concerns “through specifically Maya-identified organizations…Divided into twenty-two linguistic groups and, within each, further by more local loyalties, the emergence of a pan-Maya identity and movement in the 1980s is undoubtedly one of the developments of this decade of the greatest significance for the future of Guatemala” (Brockett 2005: 226). For the 1993 elections, indigenous groups formed comités civícos (civic committees) instead of participating within the more “traditional” political parties (Jonas 1994: 39). In the postwar period, the URNG and FDNG division might also have damaged URNG support among the country’s indigenous population. During discussions for selecting the
57 Prado Palencia, interview by author, Guatemala City, Guatemala, 19 April 2004.
126 presidential and vice-presidential ticket for the 1999 election, the FDNG had wanted to nominate the Mayan activist Rosalina Tuyuc for vice-president, but the URNG and other members of the ANN were not interested in supporting her candidacy. The nomination process not only led to increasing tension between the URNG and FDNG, but tension between the URNG and the larger Mayan vote (Sichar 1999: 83-84). The growing tendency of indigenous communities to participate on their own, as well as the URNG’s break with the FDNG and its resulting image as one of insensitivity to indigenous concerns undermined the URNG’s ability to pursue its wartime objectives and its successful transition to political party.58 Though the URNG was never able to capture the support of the wider indigenous community during the 1980s and its 1990s, the majority of its armed combatants and popular support did emerge from the indigenous communities of Guatemala. The indigenous comprised 82% of the URNG’s demobilized combatants and roughly 50% of its political cadres (Luciak 2001: 58). The best ties that the URNG had to the Guatemalan people were in the heavily indigenous departments of Huehuetenango, Quetzaltenango and El Quiché. The URNG won one seat in each of these departments in 1999. Were if not for the support that the URNG maintained amongst the indigenous population, the URNG would have likely ceased to exist as a political party.59 Even if the entire indigenous population did not support the URNG, there was a significant pool of voters from which the URNG could build a political party as, according to the 2002 Guatemalan census, forty-one percent of all Guatemalans self-identify themselves as indigenous.
Discussion How does the experience of the URNG relate to the social cleavages hypothesis? According to the 2002 census, forty-one percent of all Guatemalans self-identify themselves as indigenous. The URNG had an opportunity to represent the indigenous population as a political party, but as discussed above, it had difficultly translating the available support into actual support. The URNG has had to compete for the support of the indigenous population against other purely ethnic based political parties and social groups that developed autonomously over the final decade of the war. The URNG was unable to capitalize on the extreme marginalization
58 Tania Prado Palencia, interview by author, Guatemala City, Guatemala, 19 April 2004. 59 Prado Palencia, interview by author, Guatemala City, Guatemala, 19 April 2004.
127 of the country’s indigenous, although there is a strong likelihood that the URNG would have already ceased to exist as a political party if it could not count on any indigenous support.60 Therefore, the experience of the URNG provides limited support for the social cleavage hypothesis.
Existing Political Parties Hypothesis 16 suggested that former insurgent groups are more likely to succeed if political parties with similar political platforms do not exist. If a political party with a similar platform to what the new political party is proposing already exists, the likelihood that the new political party succeeds is diminished. This hypothesis assumes that the new political party is not cooperating with the previously existing party. In effect, they are competing for the same voters. This does not always have to be the case, though. The existence of a political party with shared positions might work to benefit the new political party if it forms an alliance or merges with the pre-existing party. The electoral experience and voter loyalty from the existing political party might work to improve the prospects for the former insurgent group as a new political party if it can cooperate with the “old” political party.
FMLN In El Salvador, insurgents took up arms against the regime at a time when they believed that social and political change via the ballot box was impossible. The electoral process, when it functioned, was fraudulent and designed so that the opposition could not win. But while civil war raged, political conditions on the ground changed. In 1982, elections were held for a Constituent Assembly and in 1984 for the presidency. While these elections did mark an initial political opening of the system, leftist groups still were not allowed to participate. As the civil war continued, political conditions also changed. In 1987, exiled members of the FDR returned to El Salvador with the intention of taking advantage of the political opening. These individuals
60 The URNG is not a great test of this hypothesis. The URNG was never an ethnic-based party. The majority of its combatants were from the indigenous communities of Guatemala as was the majority of its popular support. Over the years, the Guatemalan guerrilla groups grew more concerned with the plight of the indigenous population requiring a more nuanced interpretation of the national reality than a class-based analysis could provide. As a result, the URNG platform during the war and in the postwar contributed greater attention to the indigenous issue than most parties, but, in the end, the URNG was never a purely ethnic-based party.
128 joined a coalition of progressive Christian and social democratic parties to form the CD. The CD proceeded to compete in the 1989 presidential and 1991 legislative and municipal elections. In the 1989 elections, the CD captured 3% and finished in fourth place with Guillermo Manuel Ungo as its presidential candidate. In the 1991 elections, the CD captured eight (out of eighty- four) seats with 12% of the nationwide vote. In the 1994 elections, the FMLN, CD and the MNR competed in coalition to support Rubén Zamora for president. As mentioned earlier, while the left was able to coordinate a single presidential and vice-presidential ticket, they were unable to attain a similar consensus on candidates for legislative and municipal offices. In elections for the Legislative Assembly, the FMLN and CD presented their own lists of candidates.61 According to Spence, Dye and Vickers (1994: 25), a number of possibilities might explain the inability of the FMLN and CD to work more closely. These include the desire of the CD to retain a singular identity, the FMLN view that the CD desired too much candidate representation given its popular support, and the CD view that the FMLN desired to maintain “hegemony” over the entire process (25). Therefore, each party had its own reason for running independently. To what extent did the alliance at the executive level and the failed alliance at the legislative level help or hurt the FMLN? In elections for the presidency, it is not clear that the alliance made much of a difference. In the first round, the FMLN-CD-MNR alliance captured 24.9% against ARENA’s 49%. It is possible that if the left had failed to coordinate their candidate in 1994, potential voters might have become disenfranchised, leading to a first round victory for ARENA. On the other hand, it is always possible that all opposition parties running individual candidates would still have captured a total of 51% forcing a second round regardless. In elections for the Legislative Assembly, it is somewhat easier, though by no means certain, to quantify what effect an alliance of the left would have accomplished. In presenting three slates of candidates for the assembly, the FMLN (21), CD (1), and MNR (0) won a combined twenty- two seats. In order to determine whether the left as a single party would have performed better, I add the votes from the FMLN, CD and MNR and rerun the analyses. If the individual votes for the FMLN, CD and MNR candidates had gone toward a single slate of candidates for the
61 In elections for municipal councils, while the FMLN presented candidates in 240 municipalities and the CD in 115 (out of 262), the FMLN and CD only presented a single candidate in 54 races (Spence, Dye and Vickers 1994: 25).
129 assembly, similar to their alliance for the presidency, the left would have captured 25 total seats, an increase of 14%.62 The problem with investigating the impact of an alliance lies with those aspects that cannot be measured in at least three ways, two of which are discussed by Spence, Dye and Vickers (1994: 25). Spence et al. detail concerns that they had with the actual ballot and the perception of the FMLN. First, voters are supposed to place an “X” over the flag and acronym of the party of their choice on their ballot. In the case of the coalition presidential ticket, a voter for the coalition would have to mark the coalition’s flag with the three acronyms for the FMLN, CD and MNR below. In legislative and municipal elections, voters again had to place an “X” over the flag and acronym of the party of their choice. In elections for the legislature, the parties presented their own list, while in elections for municipal office, the parties most often ran independently. There remains a concern that the left lost votes in the 1994 legislative and municipal elections as a result of individuals marking the flag for either two or three of the parties on the legislative and municipal ballots. As a result, those ballots would be void. In a country with a high rate of illiteracy, this is a distinct possibility. In the end, it is not clear how much of a problem this was. If this were the case, we would likely have seen a greater number of impugned and null votes at the municipal and legislative levels relative to the executive. In addition, the vote for the coalition candidate for the presidency should be greater than the combined votes of the parties at the other levels. This is not what we see.63 A second concern involves the FMLN’s, and the left in general, need to appear as though it was capable of managing the country (Spence et al. 1994: 25). After working closely together during the war, it is impossible to measure the extent to which the inability of the left to coordinate their electoral campaign in the postwar diminished their credibility amongst the wider population. If the radical left (FMLN) was not willing to bargain and negotiate with the
62 The increase reduces to two seats if we factor in that the left as separate parties should have received 23 in the elections. In 1994, CD failed to register its candidates for the national list on time meaning that the only seats it could occupy would come from departmental lists. 63 In first round of elections for the presidency 75,818 votes were impugned or null. At the legislative level 81,390 votes were null or impugned. And in municipal elections, 77,755 were impugned or null. This would fit the expectations of those concerned with ballot problems. On the other hand, the vote totals for the coalition versus the individual party votes at the legislative and municipal levels are mixed. The three parties combined captured 357,085 votes for legislative, 329,392 for municipal, and 331,629 for presidential elections. Again, if ballots were a problem we would likely have seen a greater number of votes for the coalition presidential candidates (where there would less likely have been a problem) relative to the other levels. Instead, the presidential ticket captured more votes than the municipal elections and less than the assembly total.
130 moderate left (CD and MNR), how were Salvadorans going to believe that they would be able to bargain and negotiate with other domestic and international actors? Finally, while the FMLN, CD and MNR were able to agree on a unified candidate for the presidency, as discussed in the unity hypothesis, divisions within the FMLN were exacerbated during the process. The eventual support by the FMLN for the Zamora candidacy contributed to the withdrawal of members from two of the FMLN’s five guerrilla groups (ERP and RN). The attempt and inability of the left to work together might have had a greater impact on the FMLN beyond the lost opportunity in the 1994 elections. It led directly to the breakaway of a contingent from the FMLN that has had more long-term effects. In the 1997 legislative elections, the CD and FMLN did not form an electoral alliance for the most part. As a result, the CD suffered the loss of roughly one-third of the electoral support it received in 1994, declining by over 20,000 votes. After capturing one seat with 4.11% of the legislative vote in 1994, the CD only received 3.33% and two seats. Unlike 1994, the FMLN and CD presented their own candidates for the 1999 presidential elections. Rubén Zamora, the coalition candidate in 1994, captured 7.5% of the vote for the CD. Facundo Guardado of the FMLN captured 29%. Francisco Flores of ARENA captured 52% to secure a first round victory. In strictly numerical terms, the inability of the left and center-left to present a unified front might have cost it an opportunity to contest a second round, but little else.
URNG In 1984, Guatemala held elections for a Constituent Assembly and the following year for the presidency. For the next decade, Guatemala conducted a number of elections that, while not outright fraudulent, were quite limited in terms of the issues open to debate and the ideological range of parties participating in the electoral process. Even if elections were relatively free and fair, the military still retained veto power over the civilian-elected president. In addition, the left was prohibited from competing. This changed in 1995 when the FDNG formed and competed in the national elections. As mentioned previously, the FDNG was comprised of a coalition of progressive social groups (human rights, women’s, and indigenous organizations), some of whose members were sympathetic to the revolutionary cause and had been encouraged by the URNG to form a party. The 1995 elections were a success for the party, which had registered less than two months prior to the elections. The FDNG successfully captured six seats with 10%
131 of the vote, making it the third largest political party in the country. Three of the FDNG’s six congressional deputies were popular women from Guatemalan civil society.64 The results provided it with an opportunity to cultivate those represented by these progressive organizations as long-term supporters prior to the insertion of the guerrillas. Even before the transition of the URNG, though, there were mixed signals about the future direction of the FDNG. From 1995 to 1999, the FDNG had increased its membership from 7,554 to 14,402 members, twice the number that the URNG could count on as of 1998 (Sichar 1999: 83). On the other hand, in the 1998 municipal elections, the FDNG won only one of the thirty municipal elections being contested.65 There were a number of small leftist parties in addition to the FDNG that entered the political landscape in the mid-1990s. UNID, was formed by the Grupos de Apoyo Popular (GAP), or Popular Support Group, Unión Democrática (UD), or Democratic Union, and the Comisión Permanente de Refugiados (CCPP), or Permanent Commission of Refugees.66 The GAP and UD withdrew from the FDNG after FDNG leaders made the decision to align with the Partido Revolucionario (PR), or Revolutionary Party, a party with ties to many of the country’s former right-wing military rulers (Reding 1998: 20). UNID was concerned because one of the original objectives of the FDNG was to avoid aligning itself with any of the traditional Guatemalan political parties (Sichar 1999: 79). UNID formed part of the ANN coalition in the 1999 elections. The Alianza Democrática (AD), or Democratic Alliance, is another political party that shares political goals comparable to the URNG. The AD was formed in August 1997 as the successor party to the Movimiento Democrático Civíco (MDC), or Democratic Civic Movement, and the Partido Social Cristiano (PSC), or Social Christian Party (Reding 1998: 20). The AD chose not to form the ANN. The DIA is another leftist party founded in 1994 that was more of a network of nongovernmental organizations than it was a political party (Sichar 1999: 78). On February 12, 1999, the DIA, FDNG, UNID, and the URNG established the ANN. Each group would have brought something unique to the coalition. UNID had a strong following
64 The FDNG was represented by Nineth Montenegro, Rosalina Tuyuc, and Amílcar Méndez. Montenegro was president of an organization created by relatives of the disappeared, GAM. Tuyuc was president of the National Coordinating Committee of Guatemalan Widows (Coordinora Nacional de Viudas Guatemaltecas, CONAVIGUA). Finally, Méndez was the director of the Runujel Junam Council of Ethnic Communities (CERJ). 65 The FDNG won the municipal council election in Tajumulco, San Marcos. 66 UD had captured 2.55% of the presidential vote in 1995 elections. The UD was formed in 1993 as a popular, progressive party. However, the UD sought alliances with parties from across the political spectrum including the extreme right.
132 among intellectuals and artists, particularly in the urban areas (Reding 1998: 20). The DIA maintained a network of social organizations focused upon citizen participation (Sichar 1999: 78). The FDNG had a strong following with indigenous and women’s-based organizations as well as among the greater indigenous population. Finally, the URNG also had a strong following among the indigenous community as well as the publicity engendered by the multiyear peace process. But none of the organizations had anything approaching an extensive electoral history. Therefore, the URNG would not have been disadvantaged by the existence of a leftist, or center- left, political party with significant electoral support as no such party was allowed to participated prior to the 1995 elections. During the transition to political party, the URNG called for an alliance of the left to contest the upcoming elections. Unfortunately, the alliance fractured. As mentioned earlier, the FDNG left the ANN in the months leading up to the 1999 elections. In the elections, the FDNG and the ANN ran separate slates of candidates at each level of office. To what extent did the collapse of the alliance influence the elections? By most indicators, the fracturing had very little impact at the presidential level, but a significant impact on legislative elections. In the first round of the presidential elections, the ANN won 12.4% and the FDNG 1.28%. For congressional elections, the ANN captured nine seats with 8% of the vote while the FDNG failed to win a seat, capturing only 2.87% of the national vote. If the FDNG had remained a member of the ANN coalition for the 1999 elections, it is possible that the larger coalition would have picked up an addition five seats, finishing with fourteen.67 This would have been a 55% increase.
Discussion How do the experiences of the FMLN and URNG/ANN relate to the hypothesis? In the first place, neither the FMLN nor the URNG was hurt by the existence of preexisting leftist parties. Though leftist parties existed in both countries prior to the transition of the insurgents, in neither case had the existing party competed in more than one presidential or legislative electoral competition. In terms of an electoral alliance, both the URNG and FMLN entered into alliance with preexisting parties. For the FMLN, it does not appear that the alliance with the CD had a
67 The ANN with the FDNG would have picked up additional seats in Chimaltenango, Sololá, Alta Verapaz, Petén and on the national list.
133 significant impact upon the electoral fortunes of the new party. In Guatemala, the URNG/ANN did suffer as a result of the inability to forge an election to the tune of a loss of 55% potential seats. The experiences of the two groups provide little support for the hypothesis. It is possible that the limited electoral history of the FDNG and CD were key factors. The experience of having participated in a single election was not enough to significantly influence the fortunes of the new parties.68
OVERALL IMPACT OF ENVIRONMENTAL CONDITIONS
Although there do exist a number of differences between the political systems and electoral rules between the two countries, the similarities generally outweigh the differences. Overall, the institutional differences do little to explain FMLN and URNG performance (Hypotheses 7-11). Whereas we might expect a new political party to have a higher likelihood of succeeding in a parliamentary system versus a presidential system, both parties have had to compete under similar institutional arrangements. In terms of the rules for electing the president, El Salvador and Guatemala both employ majority runoff electoral rules for the presidency whereby if no candidate receives 50% of the vote plus one, a runoff is held several weeks later between the two candidates with the highest votes in the first round. In El Salvador, the FMLN successfully advanced to a second round in 1994, but since then, the presidency has been determined in the first round of voting. In Guatemala, the URNG has not played a direct role in the second round of the elections in either 1999 or 2004. It is possible that without the 12% received by the URNG in 1999, one of the two largest parties might have secured a first round victory instead of advancing to a runoff. The ability of the URNG to influence the electoral outcome in 1999 should have elevated the party to a more prominent position in Guatemalan politics and improved its standing for the 2003 elections. Instead, during the inter-election period, the URNG came under criticism both from within the party and its supporters for its
68 An interesting finding does emerge from the analysis that is not directly related to the hypothesis. In both cases, the former insurgent groups severely undermined support for the preexisting leftist parties. In Guatemala, the FDNG lost its legal standing as a political party following the 1999 elections and in El Salvador, the CD lost its standing after failing to attain 3% in the 2000 legislative elections. It might be that existing parties with a similar ideological background undermine the likelihood that a former insurgent group will succeed as a political party, but that the former insurgent group is more likely to undermine the support for the preexisting electoral alternatives.
134 close relationship to the governing party, the FRG. So even with the relatively favorable electoral rules and performance in its first electoral competition, the URNG failed to take advantage of the situation. Therefore, the executive level characteristics provide no leverage in explaining the relative success of the FMLN to the URNG. At the legislative level, we see some important differences that help explain the varied outcomes of the two parties. Each similarly possesses a single legislative chamber elected using some form of proportional representation. The proportional representation system is more favorable than single-member district voting. The URNG is at a slight disadvantage given that Guatemalan employs the D’Hondt formula to translate votes into seats while the FMLN competes in a system using the largest remainder formula. If the FMLN and URNG had to compete under the other’s method, the URNG would have been more successful while the FMLN would have been somewhat less successful. The differences would not be staggering, but they are important. I also hypothesized that an insurgent group is more likely to succeed in proportional representation systems with high district magnitude. The evidence in support of this hypothesis is also weak. Guatemala and El Salvador combine two-tiered district voting. In El Salvador, voters cast a single ballot for the party of their choice; this single ballot is then applied to candidate lists at both the department (14) and national level (1). In Guatemala, voters cast separate ballots for the department (22 plus one for the capital district) and the nation (1). El Salvador had a higher district magnitude relative to Guatemala in 1999, but in the 2003 elections, district magnitude favored the URNG relative to the FMLN. In simulating the results had each country participated in a system without a national district, both would have benefited. The FMLN would have picked up one additional seat while the URNG would pick up two. Again, we do not see district magnitude having a substantial impact on the two parties. Typically, we would expect low electoral thresholds to benefit new political parties because it makes it easier for new political parties, which tend to be small, to survive at least one round of elections. In both Guatemala and El Salvador, the new political parties have had to contend with average electoral thresholds (3% in El Salvador and 4% in Guatemala). Neither party has failed to surpass the threshold and is not relevant to explaining the varied performance of the two political parties.
135 Voter based variables (Hypotheses 12-14) show greater variation between the two cases, but, whereas the conditions appeared to be more favorable to the URNG, it was the FMLN that was relatively more successful. In Guatemala, high levels of absenteeism, low rates of bipolarization, and poor economic conditions relative to El Salvador, each should have worked in favor of the URNG. Finally, factors related to social cleavages (Hypothesis 15) and other political parties (Hypothesis 16) provide minimal support for the divergent performances of the FMLN and URNG. The majority of URNG members and support came from the indigenous communities of Guatemala. But, as explained above, the URNG was unable to capitalize on wide-scale support among the indigenous population beginning with the violent crackdown of the state against the people in the early 1980s and continuing with conflicts over representation and indigenous rights’ issues in the 1990s. Both the FMLN and URNG should have been somewhat hindered by the existence of other leftist political parties, but that does not appear to have been the case. In the end it appears as if the insurgent parties undermined support for the preexisting parties. Overall, the FMLN and the URNG competed under environmental conditions relatively favorable to new political parties. Both parties competed in more unfavorable presidential systems, but the rules for electing the president in each country is the more favorable majority runoff formula. Candidates were elected to the legislature using highly favorable proportional representation electoral rules combined with low barriers to entry and reasonable geographic and nationwide districts. There also appears to have been a significant pool of voters in each country ready to support a new electoral alternative. Therefore, each political party’s performance in its respective system was not hindered by unfavorable environmental conditions. Environmental conditions cannot explain the varied performance of the transition of these two insurgent groups to political parties. In the next chapter, I explore how well the hypotheses developed in chapter 2 and tested against the FMLN and URNG in chapter 5 and 6 help to explain the performance of a single insurgent group within its political system.
136 TABLE 6.1 1994 Legislative Election Results for FMLN in El Salvador, Largest Remainder versus D’Hondt
Department Votes Seats under LR Seats under D’Hondt (Actual) (Simulated)
Cuscatlán 8,899 1 1 Chalatenango 9,492 1 0 San Vicente 8,491 1 1 Cabañas 3,680 0 0 Morazán 5,670 1 0 Usulután 11,440 1 1 La Unión 20,946 0 0 San Salvador 124,196 5 6 San Miguel 12,575 1 1 La Paz 10,906 1 0 La Libertad 33.288 1 1 Sonsonate 17,715 1 1 Santa Ana 28,685 1 2 Ahuachapán 9,633 1 0 Deputies (District) - 16 14 Deputies (National) - 5 5 Totals 287,811 21 19
137 TABLE 6.2 1999 Legislative Election Results for URNG in Guatemala, D’Hondt versus Largest Remainder
Department Votes Votes Seats under Seats under LR National List Departmental List D’Hondt (Actual) (Simulated)
Central District 33,381 39,754 1 1 Guatemala 14,800 13,980 1 1 Sacatepequez 4,448 5,055 0 0 Chimaltenango 11,898 12,311 0 1 El Progreso 808 765 0 0 Escuintla 12,186 11,643 0 1 Santa Rosa 5,344 5,727 0 0 Solola 13,649 13,205 0 0 Totonicapan 5,851 5,533 0 1 Quetzaltenango 17,652 17,221 1 1 Suchitepequez 9,775 9,909 0 0 Retalhuleu 10,094 10,987 0 0 San Marcos 13,778 13,724 1 1 Huehuetenango 12,268 10,781 1 1 Quiche 14,441 13,967 1 1 Baja Verapaz 3,391 3,381 0 0 Alta Verapaz 18,500 18,201 1 1 Peten 12,317 12,252 0 0 Izabal 6,998 6,582 0 1 Zacapa 1,361 1,643 0 0 Chiquimula 2,042 1,798 0 0 Jalapa 1,101 - 0 0 Jutiapa 2,784 2,965 0 0 Deputies (District) - - 7 11 Deputies (National) - - 2 2 Totals 233,870 231,384 9 13
138 TABLE 6.3 1994 Legislative Election Results for FMLN in El Salvador, Actual and Simulated District Magnitude
Department Total Seats Seats under LR Total Seats Total Seats w/o Available (Actual) Available National District (Simulated) (Simulated)
Cuscatlán 3 1 3 1 Chalatenango 3 1 3 1 San Vicente 3 1 2 1 Cabañas 3 0 2 2 Morazán 3 1 2 0 Usulután 4 1 4 1 La Unión 3 0 3 0 San Salvador 16 5 26 8 San Miguel 5 1 5 1 La Paz 3 1 4 1 La Libertad 5 1 9 2 Sonsonate 4 1 7 1 Santa Ana 6 1 8 2 Ahuachapán 3 1 4 1 Deputies (District) 64 16 84 22 Deputies (National) 20 5 - - Totals 84 21 84 22
139 TABLE 6.4 1999 Legislative Election Results for URNG in Guatemala, Actual and Simulated District Magnitude
Department Total Seats Seats under Total Seats Seats w/o Available D’Hondt Available National District (Actual) (Simulated) (Simulated) Central District 10 1 19 2 Guatemala 12 1 10 1 Sacatepequez 2 0 2 0 Chimaltenango 3 0 4 1 El Progreso 1 0 2 0 Escuintla 4 0 6 0 Santa Rosa 3 0 4 0 Solola 2 0 3 1 Totonicapan 3 0 3 0 Quetzaltenango 6 1 7 1 Suchitepequez 3 0 5 0 Retalhuleu 2 0 3 0 San Marcos 8 1 7 1 Huehuetenango 7 1 7 1 Quiche 5 1 5 1 Baja Verapaz 1 0 2 0 Alta Verapaz 6 1 5 1 Peten 2 0 3 1 Izabal 3 0 3 0 Zacapa 1 0 3 0 Chiquimula 2 0 3 0 Jalapa 2 0 2 0 Jutiapa 3 0 5 0 Deputies (District) 91 7 113 11 Deputies (National) 22 2 - - Totals 113 9 113 11
140 TABLE 6.5 National Vote in Salvadoran Elections (Percentages)
Year Election Vote/Registered Votera
1982 C.A. 63.6% 1984 Presidential 56.3 1984 Presidential Runoff 58.3 1985 Legislative 42 1988 Legislative 59 1988 Municipal 50.6 1989 Presidential 39.9 1991 Legislative 44.7 1991 Municipal 44.1 1994 Presidential 50 1994 Legislative 51.5 1994 Municipal 51.3 1994 Presidential Runoff 44.2 Average 50.4
a Vote / Registered Voter is based upon the total number of Salvadorans found on the registry for each election.
I derived the Vote / Registered Voter (Turnout) from Baloyra-Herp’s (1995: 50) reporting of voter turnout and total registered voters. Baloyra-Herp reported both figures in millions so the percentages are approximations.
141 TABLE 6.6 National Vote in Guatemalan Elections (Percentages)
Year Vote/Registered Votera Vote/VAPb
1950 71.5% 30.4% 1954 69.5 32.5 1958 66.8 28.3 1959 44.6 18.0 1961 44.5 19.0 1966 55.0 23.7 1970 53.3 25.9 1974 46.4 25.5 1978 40.0 22.9 1982 45.6 30.6 1985 69.3 49.9 1990 56.4 14.1 1994 21.6 14.5 1995 46.7 33.3 1999 53.8 44.0 2003c 46.7 - Average 52% 27.5
aVote / Registered Voters is simply the percentage of Guatemalans that voted divided by the number of Guatemalans inscribed in the Electoral Registry. These figures come from “Por qué no votan los Guatemaltecos?” bVote / VAP is simply the percentage of Guatemalans that voted divided by the total voting age population (18 and over). These figures come from “Por qué no votan los Guatemaltecos?” cThis percentage comes from the Guatemalan Supreme Electoral Tribunal (http://www.tse.org.gt/).
142
TABLE 6.7 Bipolarization of the Salvadoran Political Party System, Legislative and Executive Elections
Year Election 1st Place 2nd Place Total
1984 Presidential PDC ARENA 74%
1985 Legislative PDC ARENA 84%
1988 Legislative ARENA PDC 83%
1989 Presidential ARENA PDC 90%
1991 Legislative ARENA PDC 82%
143 TABLE 6.8 Bipolarization of the Guatemalan Political Party System, Legislative and Executive Elections
Year Election 1st Place 2nd Place Total
1985 Legislative DCG UCN 56%
1985 Presidential DCG UCN 52%
1990 Legislative UCN DCG 59%
1990 Presidential UCN MAS 43%
1995 Legislative PAN FRG 80%
1995 Presidential PAN FRG 59%
1999 Legislative FRG PAN 90%
1999 Presidential FRG PAN 78%
144 TABLE 6.9 Growth Rate of Real GDP per Capita, El Salvador
Year Growth (%) 1980 -11.85 1981 -10.18 1982 -6.57 1983 +1.55 1984 +1.18 1985 +0.89 1986 -1.71 1987 +0.85 1988 +1.11 1989 -1.97 1990 -0.83 1991 +0.94 1992 +3.97 1993 +4.86 1994 +3.38 1995 +4.24 1996 +1.34 1997 +1.94 1998 +0.27 1999 +2.14 2000 +0.30
Source: Alan Heston, Robert Summers and Bettina Aten, Penn World Table Version 6.1, Center for International Comparisons at the University of Pennsylvania (CICUP), October 2002.
145 TABLE 6.10 Growth Rate of Real GDP per Capita, Guatemala
Year Growth (%) 1980 +3.70 1981 -1.58 1982 -3.52 1983 -3.06 1984 -2.44 1985 -1.82 1986 -1.21 1987 -1.67 1988 +1.58 1989 +1.36 1990 +0.56 1991 +0.37 1992 -0.62 1993 +1.73 1994 +1.29 1995 +2.00 1996 +1.77 1997 +0.12 1998 -0.87 1999 +1.39
Source: Alan Heston, Robert Summers and Bettina Aten, Penn World Table Version 6.1, Center for International Comparisons at the University of Pennsylvania (CICUP), October 2002.
146 CHAPTER 7 CIVIL WAR LEGACY ON FMLN ELECTORAL PERFORMANCE
To this point in the dissertation, I have focused on how well the hypotheses developed in chapter 2 account for the electoral performance of insurgent groups in El Salvador and Guatemala. As mentioned earlier, though, I am also interested in determining the extent to which the hypotheses developed to explain success across insurgent group transitions also contribute to our understanding of the new political party’s performance across political sub- national units. In this chapter, I explore the extent to which the hypotheses developed in chapter 2 contribute to explaining the performance of the FMLN in the 1994 “elections of the century.”69 While these findings alone will not prove or refute the hypotheses developed earlier, I believe they will provide valuable insight into evaluating the theory’s overall usefulness.
Elections of the Century During the two years between the signing of the Peace Accords and the holding of elections in El Salvador, great uncertainty surrounded the likely electoral performance of the FMLN. How would a former insurgent group perform when forced to compete at the ballot box? On the one hand, prospects for the FMLN looked favorable. The FMLN was probably the largest and militarily most formidable insurgent group in the Western Hemisphere. It had successfully fought a US-trained, funded and supplied Salvadoran military to a stalemate under geographic conditions highly unfavorable to a successful protracted insurgent conflict. The FMLN also maintained extensive relations with unions, students’ and teachers’ groups, and farmers’ associations, which could provide party militants and voters. On the other hand, there were reasons to suspect that the FMLN would struggle as a political party. The FMLN was an insurgent army organized to fight a military, not an electoral, campaign. In addition, while elections had returned to El Salvador in the midst of the civil war as part of the US-sponsored counterinsurgency program, the FMLN had rejected these “demonstration elections” (Herman
69These elections have been referred to as the “elections of the century” because, in addition to being the first election in the postwar period, it was the first time since the adoption of the 1983 constitution that the election for each level of government was held simultaneously. Elections for the presidency in El Salvador are held every five years while those for the legislature and municipal office every three and will not be held simultaneously again until 2009.
147 and Brodhead 1984) as illegitimate tools of yanquí imperialism. Perhaps more detrimental to the political future of the FMLN, as the war progressed, the FMLN increasingly
attacked the elections in areas of the country not under their military control. They took away ID cards from voters, an implicit threat. They refused to declare a cease fire on election day, and some military actions were close enough to polling places so as to give the impression of an attack on the elections, though no voters were killed (Spence and Vickers 1994: 6).
Finally, the FMLN assassinated opposition mayors, forcibly recruited civilians to fight on its behalf, kidnapped wealthy businessmen for ransom, and engaged in widespread acts of economic terrorism. While the Salvadoran government and military were not without fault, it remained unclear how an organization with such a history could be successful at the ballot box. As is seen in Table 7.1, though, the FMLN performed relatively well in the 1994 elections. In municipal elections, the FMLN won 15 municipalities out of 262 nationwide (5.7%).70 In legislative elections, the FMLN became the country’s second largest party by winning 21 out of a possible 84 seats (25%). Unlike the single member district (SMD) voting used to elect the municipality’s mayor and councils, membership in the Legislative Assembly is determined via proportional representation. And in elections for the presidency, the FMLN formed a coalition along with the CD and MNR parties in support of Rubén Zamora. The coalition captured 25% of the national vote. This forced a second round of voting under a majority runoff formula, since no candidate secured an outright majority in the first round. In the second round, the coalition candidate captured 32% of the vote, but was defeated by the ARENA candidate, Armando Calderón Sol, who captured 68%. While the FMLN did not capture the majority of the Salvadoran vote, it was obvious that they had significant popular support, more than the US and Salvadoran governments had recognized. As might have been expected, while the FMLN captured 20-25% of the vote across the three elections, its popular support was not evenly distributed throughout the entire country. The FMLN received greater support in some departments and much less in others. As seen in Figure 7.1, FMLN support in legislative elections ranged from low levels in La Unión (7%) and
70 The FMLN won thirteen municipal elections by itself and two in an alliance with the CD.
148 Cabañas (12%) to higher levels in San Salvador (30%) and San Vicente (23%). The primary interest of this chapter is to explain this geographic distribution of popular support for the FMLN as reflected in the 1994 elections. In the remainder of the chapter, I assess the legacy of the Salvadoran civil war on the performance of the FMLN. I argue that two key factors related to the legacy of the war explain this distribution of FMLN support. First, the FMLN should have performed better in those areas of the country in which it maintained a strong presence during the war than in those areas where it did not. During the conflict, the FMLN laid the groundwork for a national electoral campaign through political organizing in support of their revolutionary goals in those areas of the country that it controlled. This is an additional test of the territorial control hypothesis developed and tested earlier to explain insurgent group success as a new political party. Second, FMLN support should have been helped by the extreme reaction of the state’s security forces against the civilian population. During the conflict, the Salvadoran military and its related security forces committed the majority of human rights violations against civilians as part of its counterinsurgency campaign. El Salvador’s post war truth commission found that the Salvadoran security forces, and paramilitary and deaths squads associated with them, were responsible for 85% of the human rights violations committed during the conflict, while the guerrillas were only responsible for roughly 5%.71 Several authors have also found that one of the strongest motivations for people joining the guerrillas was the intense and brutal violence that the state and its security forces perpetrated against these civilians. As a result, I argue that those areas of the country most severely affected by the war should support the FMLN in greater proportion relative to those areas where the violence was less pronounced. This is an additional test of the violence hypothesis developed and tested in earlier chapters. I test these hypotheses through one-way analysis of variance (ANOVA) and ordinary least squares regression (OLS) using political violence, electoral, and socio-economic data at both the departmental and municipal level for the 1994 elections in El Salvador. The results suggest that the FMLN as a political party was helped more by the state’s violently disproportionate response than by its ability to hold territory during the war. Finally, I conclude by relating these results to the findings from chapters 5 and 6.
71 Again, the remaining 10% of the violations were either committed by individuals not directly tied to the war or they could not be attributed to any particular assailant.
149
EXPLAINING FMLN ELECTORAL PERFORMANCE
Control of Territory Previously, I argued that insurgent groups that won and controlled territory during a civil war should be more likely to succeed as political parties than insurgent groups that failed to capture and control territory. An insurgent group that captures and controls territory, develops a certain level of order within that territory, and provides benefits to those living within the territory is more likely to be able to mobilize the population within this area in order to support it as a political party than an insurgent group that controls only unpopulated or remote areas of the country. As Kalyvas (1999: 259) discusses, “an insurgent organization which controls a given area (a ‘liberated area’) operates as a counter-sovereign authority, a ‘counter-state’. It provides protection, administers justice, collects taxes, and applies its social program.” In the controlled territory, the insurgent group can gain valuable experience by completing public works projects and working with local organizations to address the needs of the population living within its sphere of control, all critical to its future success as a political party. These organizational experiences should prove advantageous in making the transition to and succeeding as political parties. On the other hand, if the insurgent group is unable to capture and hold territory throughout the country, it will be at a disadvantage when preparing to compete as a national political party. While even a small insurgent group might be able to inflict damage upon government forces or elude destruction, it will be more difficult for such a group to succeed as a political party relative to a group that has a presence throughout a larger area of the country. Obviously, success does not come simply from holding territory. It is certainly possible that the population living under insurgent protection will not look upon occupying forces with appreciation. Insurgent groups can be very violent toward civilians within liberated zones, particularly toward those suspected to be sympathetic to or actively aiding the government. It is also possible that the government will identify these areas as free-fire zones, where anyone found within these areas is considered a guerrilla or guerrilla collaborator. While the insurgent group can utilize its experience of having successfully governed controlled zones, or liberated areas, “it also enjoys a local monopoly of violence which it uses to punish its enemies and sanction
150 uncooperative behavior, such as the refusal to supply food or pay the ‘revolutionary tax’” (Kalyvas 1999: 25). An insurgent group that relies almost exclusively upon coercion to control territory is unlikely to be successful in the postwar electoral period. So beyond the territory controlled by the insurgents, we should expect the relationship between the insurgent group and the population living within its zone of control to be a strong component explaining an insurgent group’s performance as a political party. As mentioned in chapter 5, the FMLN sought to capture national territory in order to dismantle “the existing state’s infrastructure of local government offices, jails, telephone exchanges, police stations, military garrisons, and counterinsurgency military and civil defense facilities” (Alvarez 1988: 79). As the FMLN dismantled the state’s institutions, it simultaneously established “alternative governing authority” based upon “widespread” and “voluntary” campesino support (Wood 2003). The FMLN was intent on creating a new society based upon “cooperation” (McClintock 1998: 74) and “solidarity” (Alvarez 1988: 87). The political, social, and economic benefits of living in FMLN controlled zones should have resulted in an increased level of electoral support relative to areas of the country where the FMLN had little control or military presence. To the contrary, several sources have portrayed less than ideal conditions that existed within FMLN zones. As a result, perhaps a strong FMLN presence should not have had a positive impact on its postwar electoral support if it ruled these areas through fear and intimidation. Bracamonte and Spencer (1995: 8) argue that while the FMLN initially relied upon ideological education in its recruitment efforts, the FMLN increasingly relied upon forced recruitment of those living in war zones as it continued to suffer heavy casualties. Bosch also argues that following the failed offensive of 1981, “terrorism in the countryside would replace combat in major cities” (1999: 109). Instead of ruling with the voluntary participation of the people, the FMLN controlled these zones through terrorizing its inhabitants into submission. Likewise, the United Nations Truth Commission on El Salvador details a number of incidents of forced recruitment by the FMLN, particularly by the ERP. Given this evidence, we would expect lower levels of electoral support in FMLN controlled zones during the 1994 elections relative to the rest of the country, or simply low levels across both zones. Therefore, we have two competing expectations. First, FMLN support should be greater in controlled zones than in other areas because of the voluntary support provided by the citizens
151 in these areas and the political work of the guerrillas in establishing alternative governing authority in these zones. Second, FMLN support in controlled zones may have been coerced and attained through terror and fear; this should produce a negative relationship between FMLN controlled zones and its electoral support. Overall, I believe the first proposition has greater merit. There has been much written about FMLN attempts to transform the areas under their control to a more just society based upon cooperation and solidarity (Alvarez 1988; Hammond 1998; Luciak 2001; McClintock 1998; Wood 2000, 2003). While the FMLN was guilty of assassinating civilian government officials and engaging in sporadic forced recruitment, there is no credible evidence that the FMLN participated in large-scale attacks against the civilian population. One might make the argument that had the FMLN captured state power, it would have ruled with an iron fist and through the use of terrorism, but there is little systematic evidence that during the war the FMLN controlled the population within its sphere of influence through the large-scale use or threat of violence. A positive relationship would support the first set of arguments, while a negative relationship would support the second.
Hypothesis 1 (Territory): FMLN support should be greater in areas of the country that it controlled during the war than in areas that it did not control.
Political Violence During a civil war, one of the goals, perhaps even the primary goal, of insurgents and government, is to “win” the loyalty and support of the “people.” Insurgents and governments can vie for the people’s support by providing their supporters with benefits at the same time that they deny these goods to those that seek to remain neutral or actively support the opposing group. Unfortunately, not all loyalty or support is won through such “benevolent” means. Instead, support is accomplished through coercion. As recent history has made quite clear, neither governments nor insurgents can claim a monopoly on the use of force against civilians during war. Violence perpetrated against civilian populations, while perhaps a more emotional and highly publicized issue today, is nothing new (Kalyvas 2001). Governments and insurgents from every region of the world and historical period have been known to use violence against civilian populations in order to deter potential defectors and/or punish actual defectors. In some wars such as the recent cases in Latin American history, government forces have committed the
152 majority of civilian massacres (Wickham-Crowley 1990). Civilians have suffered at the hands of government forces disproportionately relative to the insurgents in Guatemala, El Salvador, Honduras, Chile, Argentina, Brazil, and Nicaragua during the Somoza dynasty. On the other hand, insurgents have also targeted civilians equally to the government, or perhaps more so, in Peru (Shining Path or Sendero Luminoso), Nicaragua (the Nicaraguan Resistance or Contras), and Colombia (Armed Forces of Colombia and the United Self-Defense Forces of Colombia). Unfortunately, human rights violations, and perhaps the massacre of civilian populations, occur in all wars. While the act of massacring civilians might serve some strategic purpose in the midst of civil war (Kalyvas 1999), it is unlikely to help an insurgent group in the postwar period when, as a new political party, it must attract electoral support. An insurgent group (or government for that matter) that relied overwhelmingly upon coercion in its pursuit of military victory is unlikely to fare as well as an insurgent group that was able to capture the loyalty and support of the civilian population through less violent means. While most Latin American insurgent groups did not target civilians on a scale comparable to their respective governments, this does not mean that insurgent violence did not negatively impact the civilian population. Insurgent groups must also be concerned with how their use of violence affects the civilian population. Insurgents do not always utilize conventional, army-to-army warfare to take power. Instead, they often attempt to undermine the government’s popular support by disrupting its ability to manage the economy and/or political structures. As a result, insurgents often target the government’s political (ministry offices, party/political headquarters, the electoral process) and economic (roads, bridges, water and utility plants) structures. How did the role of violence play out in El Salvador’s conflict? Following the end of violent activities in El Salvador, the United Nations Truth Commission Report determined that the majority of violence inflicted upon the civilian population was a direct result of the actions of the state and its security forces. Government violations included rape, torture, and forced disappearances of not only individuals connected to the insurgents, but religious, opposition, and labor leaders and civilians from all walks of life. Many guerrilla recruits and supporters often cited government repression as the reason for supporting or joining the FMLN (McClintock 1998; Wood 2003). The UN Truth Commission Report also faulted the FMLN for forced disappearances, forced recruitment and extrajudicial killings of public officials. McClintock (1998: 60) estimates
153 that the FMLN killed about forty civilians each year throughout the mid-1980s. It is quite possible, though, that the crimes or violations committed by the FMLN were not accounted for in the Truth Commission Report. As a result, the report by the Truth Commission might have underestimated the manner in which guerrilla violence affected ordinary Salvadorans. While there is no credible evidence that the FMLN targeted civilians on a scale comparable to the Salvadoran military, the FMLN frequently targeted the country’s infrastructure – electricity grids, bridges, dams, water stations, highway chokepoints, etc – and civilians aligned with the government. FMLN attacks against the country’s infrastructure were attempts to undermine support for the government by proving that the government could not provide for the general population and that the billions of dollars in US economic aid was wasted. During the first half of 1980 alone, Bosch estimates that the FMLN “committed 3,140 acts of violence, including arson, assault, assassination, and the destruction of bridges, electric power towers, and private business establishments” (1999: 60). Some argue that this violence was part of a grand strategy of the FMLN to increase recruitment levels and eventually bring down the government:
[The FMLN] plan was insidious. The destruction of the economy and the infrastructure would have a profound impact on the social conditions of the population. The people would blame their new social problems on the government. Mass organizations would publicly and visibly protest against the government in an attempt to radicalize the masses. The newly radicalized masses would provide recruits for the urban guerrillas and the mass organizations to perpetuate more protest, destruction, and sabotage. The FMLN had created a strategy that would grow and feed on itself, a functional, perpetual motion machine (Bracamonte and Spencer 1995: 27).
It is quite possible that FMLN attacks upon the infrastructure had the opposite effect than what the FMLN had envisioned. Instead of encouraging civilians to support the insurgents, the FMLN could have made life so unbearable that their actions pushed many “independent” civilians to support the government. Similarly, Bracamonte and Spencer argue that increasing terrorist tactics employed by the FMLN as the 1980s progressed, particularly in the urban areas, brought some additional support to the guerrillas, but not nearly enough to compensate for those that were pushed away from the more moderate PDC to the more hard line ARENA party, as
154 seen in the strengthening of ARENA and the weakening of the PDC in the 1989 presidential elections (1995: 33). What is the expected relationship between civil war violence and FMLN group support? Again, like the territorial control hypothesis, we have two competing expectations. First, we would expect the FMLN to perform better in those areas where the conflict occurred relative to those areas less affected by the war given the disproportionate violence committed by the government against the civilian population. Second, the forced recruitment of civilians, assassination of government officials, and attacks upon the country’s infrastructure could have had a negative effect upon guerrilla support in war-torn areas of the country. Given the preponderance of evidence concerning the overwhelming violence of the state, I believe the first expectation is more plausible.
Hypothesis 2 (Violence): FMLN support should be greater in areas of the country that were more severely affected by the civil war violence than in areas that were less affected.
Control Variables While territorial control and civil war violence are expected to be the key variables in explaining the geographic distribution of the FMLN vote, there are also a number of political factors for which we can control, including voter abstention, bipolarization, and the performance of other opposition political parties.
Abstention In general, one would expect that when the existing political parties are doing a commendable job, voters will tend not to support new political alternatives. But, on the other hand, if voters do not believe that the system’s current parties are capable of resolving important issues, new political parties are bound to emerge (Hauss and Rayside 1978: 38). One way in which we can get a handle on whether citizens are satisfied with the existing political parties and whether voters might support a new political party is by looking at levels of voter abstention. High voter abstention rates mean that there might be a significant portion of the population that
155 is not completely satisfied with the status quo and would be more open to supporting a new political party. Two of the initial triggers of the escalation of political violence in El Salvador were the fraudulent presidential elections of 1972 and 1977. In response, the 1980s electoral process was designed to prevent fraud and other irregularities (Baloyra-Herp 1995: 49). The result was a cumbersome registration and voting process that, combined with civil war violence and the FMLN call to boycott the elections, kept voter participation levels low. Baloyra-Herp (1995:59) characterizes the elections of the early 1980s as having “had decent levels of popular participation” relative to the “lower levels” that were characteristic of the 1988-1991 elections. While there are several explanations for the decline in voter participation into the early 1990s, the lower levels do seem to suggest that there existed a pool of potential voters for this new political party to tap. A new political party, such as the FMLN, would have the opportunity to mobilize citizens who dropped out of the electoral process after the initial elections of the early 1980s, had heeded the call to boycott the elections, or simply had never participated in the process. There does not seem to have been a shortage of potential voters for the FMLN as a new political party in 1994. If a high level of voter abstention at the national level indicates that there is significant support available for a new political party, we should also expect this relationship to exist at lower levels; in this case, at the municipal level. Therefore, I expect the FMLN to perform better in those areas of the country with high levels of voter abstention in the most recent election prior to the insertion of the FMLN as a political party.
Hypothesis 3 (Abstention): FMLN support should be greater in areas with high levels of voter abstention.
Bipolarization Some have also argued that the bipolarization that competition for the presidency supports has an “indirect but strong effect on the effective number of parliamentary parties” (Lijphart 1999: 155). Presidential systems tend to depress the number of political parties closer to two making it more difficult for a third party, or new political party, to succeed. But in new
156 political systems where we expect two political parties to eventually come to dominate, the identity of these two parties is unlikely to be determined immediately. Therefore, we expect a new political party such as the FMLN to be more successful in areas of the country where bipolarization is weak. This means that the FMLN should perform better in municipalities where the two largest parties capture the lowest combined percentage of the vote. El Salvador conducted elections throughout the 1960s and 1970s, but the semblance of competitive elections only began with the 1982 Constituent Assembly and 1984 presidential elections. These elections were competitive in the sense that there existed a range of electoral alternatives along a somewhat limited ideological spectrum (center/center-right to extreme right), but they were less than fully competitive in that neither left nor center-left parties were legal, the US government intervened in the political process, and voting was done in the midst of a civil war. While these elections were not conducted under ideal voting conditions, they do permit us to look in a very basic way at the preferences of those voters that did turn out to vote. During the three elections for the Legislative Assembly prior to the incorporation of the FMLN, the two largest parties combined, ARENA and the PDC, received 84% (1985), 83% (1988), and 82% (1991) of the national vote. The fact that the two largest parties had attained over 70% of the national vote in each election since the return to competitive party politics should have worked against the FMLN. On the other hand, this unfavorable condition might be countered by the fact that there was still 20-30% of the vote uncommitted to the two largest parties, particularly in the most recent election. In addition, one of the two largest parties, the PDC, had been losing votes and the general support of the Salvadoran people as evidence by its performance in the 1989 presidential elections and the 1991 legislative elections. The PDC had proven unable to bring an end to the war and had been involved in several corruption scandals. Therefore, the FMLN entered at a relatively favorable point in time where bipolarization had not yet stabilized and, it appeared, the system might actually have been moving away from it with the decline of the PDC. An additional favorable condition for the FMLN was that the bipolarization that had begun to form involved center-right (PDC) and far-right (ARENA) parties. The FMLN might have benefited from the fact that no center-left or left party had built up an electoral history or significant following in the country. The left was only first allowed to participate with the 1991 legislative and municipal elections. Therefore, we expect the FMLN to
157 perform better in those areas where large percentages of the population did not support the PDC and ARENA.
Hypothesis 4 (Bipolarization): FMLN support should be greater in areas where bipolarization is weak.
Opposition Parties New, or potential, political parties also need to be concerned with the presence of other opposition political parties. It is possible that while one group takes up the cause of armed conflict, another opposition group might establish itself as a legitimate opposition political party within the political system. These “legitimate” opposition parties might undermine support for the insurgent group as a political party. The existence of a political party with similar ideological underpinnings and electoral support might make it more difficult for a new political party, such as one that has its roots as a former insurgent group, to be successful. For example, the existence of a left or center-left legal opposition political party should make it more difficult for a Marxist-Leninist insurgent group such as the FMLN to become successful. Likewise, an ethnic-based insurgent group might have a more difficult transition to political party if a similarly ethnic-based party already exists. In El Salvador, insurgents took up arms against the regime at a time when they believed that social and political reform enacted through change at the ballot box was impossible. The electoral process, when it functioned, was fraudulent and designed so that the opposition could not win. By 1979, civil war had broken out. But while civil war raged for the next decade, political conditions changed. Constituent Assembly elections were held in 1982 and elections for the presidency in 1984. While these elections did mark an initial political opening of the system, though, leftist groups still were not allowed to participate. This changed in 1987 when exiled members of the FDR returned to El Salvador with the intention of taking advantage of the new political space. The FDR was a center-left organization that shared many of the same criticisms of the Salvadoran social, economic and political situation levied by the guerrillas, although it did not share the FMLN’s advocacy of violent revolution. Prior to the 1994 elections in which the FMLN first competed, the FDR and a coalition of progressive Christian and social democratic parties formed the CD and competed in the 1989 presidential and 1991 legislative
158 and municipal elections. The CD had a five year head start on building up an electoral support base in El Salvador. In the 1994 elections, while the FMLN and CD supported a single presidential candidate, each party presented its own slate of candidates for the Legislative Assembly. I expect that where the CD has performed well, the FMLN will not perform as well.
Hypothesis 5 (Opposition Parties): FMLN support should be greater in areas where political parties with similar political platforms are weak.
Other Control Variables Ideally, one would prefer to control for a number of socio-economic factors at the municipal level. One would expect a leftist party such as the FMLN to perform better in areas of the country where such conditions as unemployment and poverty rates are high. Unfortunately, socio-economic data is only available at the department level, not for individual municipalities, beginning in 1998. In the 1992 census, population is available at the municipal level, but socio- economic data is aggregated at the departmental level and is not useful for statistical analysis.
RESEARCH DESIGN
In the following analyses, I set out to explain the geographic distribution of FMLN electoral support. Since the signing of the Peace Accords in 1992, the FMLN has competed in three presidential elections (1994, 1999 and 2004) and four legislative and municipal elections (1994, 1997, 2000, and 2003). The 1994 elections of the century marked a symbolic end to the war as well as the beginning of the FMLN’s life as a political party. It is also the election where the most immediate impact of the war on the FMLN performance should exist and the only election with which this analysis is concerned. The unit of analysis is FMLN electoral performance. I evaluate the FMLN’s performance at both the department and municipal level. Politically, El Salvador is divided into fourteen departments, roughly the equivalent of US states. These fourteen departments are further divided into 262 municipalities, roughly the equivalent of US counties. Given the small number of departments, statistical analyses using analysis of variance (ANOVA) and multivariate regression are confined to electoral data at the municipal level.
159
Variable Operationalization Electoral Performance In this paper, the dependent variable used to measure electoral performance is the percentage of the vote captured by the FMLN in elections for the Legislative Assembly in the March 1994 elections of the century. I analyze electoral data from the Supreme Electoral Tribunal (Salvadoran electoral commission) at both the department level and the municipal level. I use the vote for the FMLN in legislative elections because the FMLN supported a coalition candidate in elections for the presidency and a number of municipal offices.
FMLN Controlled Territory Hypothesis 1 suggested that the FMLN would perform better in those areas of the country that it controlled during the war relative to areas it did not. Obviously, this is a difficult hypothesis to test. The Salvadoran government contended that any serious offensive on its part would dislocate the FMLN from areas in which it had concentrated and, therefore, these could not be considered FMLN-controlled areas. To the contrary, others maintain that the FMLN controlled “the zones north of the Salvadoran departments of San Salvador, Chalatenango, Cabañas and Morazán, and in the central zone of San Vicente – Volcán Chinchontepec” (Alvarez 1988: 85).72 During the war, FMLN presence also extended into the departments of San Miguel, La Unión and Usulután. Given that the FMLN had a relatively strong presence in these eight departments, we might expect that its electoral support in the 1994 elections should have been higher there than in the other six departments, which had little history of FMLN activity. It is also possible that because the FMLN did not control any department in its entirety, the impact of FMLN presence might not be apparent in voter tallies at the department level. Therefore, I also explore FMLN control at a more localized level. In order to systematically assess the impact of FMLN controlled zones on the popular vote of the FMLN, I code FMLN-controlled zones as those municipalities where the FMLN prevented elections from being conducted in the first round of the 1984 presidential elections. All other zones are coded zero. In 1984, the war prevented 59 of the country’s 262
72 Alvarez cites an undated document, El Poder Popular en El Salvador, by J. Ventura(e) as the source of FMLN activity.
160 municipalities from holding elections. Employing municipalities where votes were not cast in 1984 provides a useful, albeit rough, indicator of FMLN controlled territory. This approach is relatively consistent with McClintock (1998)73 and with reports that the FMLN controlled 15- 25% of the country in the early 1980s. There are a number of concerns with operationalizing FMLN controlled zones in this manner. While these are important concerns, they should not preclude systematic analysis. First, typically municipalities did not conduct elections for two distinct reasons. One reason is that the FMLN controlled certain zones and would not allow elections to be conducted. These are exactly the municipalities I am interested in. Second, the FMLN military strength was significant enough in other areas to thwart the electoral process even though the zone was not directly under FMLN control. Therefore, I am collapsing these two categories (FMLN control and FMLN presence) into a single category. Unfortunately, I know of no way to systematically divide these two categories where we would then be able to further distinguish between FMLN controlled territory and territory it simply threatened. A second problem is that the FMLN frequently controlled municipalities for weeks, months or years. The FMLN did not always permanently control these areas. Simply occupying the municipality for a number of days close to the date of the elections might have been sufficient to prevent them from being held and, hence, having the municipality qualify as a FMLN controlled zone. The ability of the FMLN to control territory and establish some sort of alternative governing structure capable of influencing its vote in the 1994 election is much more likely to have been a process of months and years, not days or weeks. On the other hand, including some municipalities in the analyses that were only influenced over a short period of time should prove an even harder test of the hypothesis, as we are including a number of municipalities with limited long-term FMLN presence.
Political Violence Hypothesis 2 addresses the relationship between the insurgent group and the population. I hypothesized that the FMLN was more likely to succeed in those areas of the country that were
73 In addition, McClintock supplements these municipalities with others where elections were held, but the mayor was forced to govern in exile because of the presence of the FMLN. Unfortunately, when contacted, McClintock was unable to locate the original list of municipalities though she did say that it was not very different from simply using the municipalities where elections were not held.
161 more severely affected by the violence than those that were not, given that the government was more responsible for the violence committed against the civilian population than the guerrillas. One way in which we can probe the impact of the legacy of the civil war on the electoral fortunes of the FMLN is by comparing the electoral results of the FMLN against the geographic distribution of the violence. Seligson and McElhinny (1996) conducted a survey whereby respondents were asked whether a family member had died as a result of the war. At the high end, over 70% of the respondents in Cuscatlán said that at least one family member died as a result of the war while at the low end less than 20% in Ahuachapán claimed to have lost a family member as a result of the war. At one level, this allows a basic assessment of FMLN performance relative to those departments where civil war victims originated. A second way in which we can probe the violence hypothesis is by assessing whether war-torn areas of the country were more likely to support the FMLN than areas that were spared the bulk of the fighting. The United States Agency for International Development sponsored a program during the peace negotiations, the National Reconstruction Program (NRP), that set out to identify those areas of the country that had been most severely affected by the civil war. The goal was to identify the municipalities that were most in need of reconstruction funds and development projects. Out of the country’s 262 municipalities, the NRP identified 115 “ex- conflict” zones and 137 “non-conflict” zones. During the initial coding of the municipalities most severely affected by the civil war, ten municipalities were not coded because they were too dangerous. In the following analyses, I have coded these ten as conflict zones.74 Comparing “ex-conflict” and “non-conflict” zones can provide insight into the electoral support provided to the FMLN by those municipalities most seriously affected by the war (“ex- conflict zones”) from those least affected (“non-conflict zones”). This study allows us to look at the FMLN performance at the municipal level relative to where most of the violence during the civil war took place. Given that several authors have argued that the FMLN provided various services to the Salvadoran population and that FMLN support was widespread and voluntary, one expects ex-conflict zones to have a positive impact on the FMLN vote in the 1994 elections.
74 I ran the analyses with and without these ten municipalities as ex-conflict zones. There were no differences in the results.
162
Control Variables The data on voter abstention, bipolarization, and strength of other left parties come from the TSE in El Salvador. I measure each of these variables at the municipal level. High voter abstention in the 1991 elections should be positively related to the FMLN vote in 1994. Abstention is measured by dividing the number of total votes from the 1991 legislative elections at the municipal level and dividing that number by the total voting age population for that municipality as reported in the 1992 census. I then subtract this number from one hundred to come up with the level of abstention for the municipality. Second, I expect that a high combined vote for the two largest parties (bipolarization) will be negatively related to FMLN performance. I measure the existence of bipolarization by combining the vote percentage of the two largest political parties, ARENA and the PDC, in the 1991 elections. I also expect that the high levels of support for the CD should be negatively related to the FMLN vote in 1994. I measure the strength of left parties using the CD’s vote percentage in the 1991 elections. Finally, I also control for population and poverty. I use municipal level population numbers from the 1992 census.75 Unfortunately, as mentioned earlier, poverty rates are only available at the department level beginning with a yearly survey of homes conducted by the Ministry of the Economy in El Salvador in 1998.76 These rates come from four years after the election. Therefore I am less confident in their ability to explain FMLN performance. On the other hand, it is unlikely that these rates changed dramatically in the four years and at such an uneven rate across departments that the OLS results would be biased. I code total poverty rates as the combined percentage of people in each department living in relative and extreme poverty.
EMPIRICAL ANALYSES
In this section, I perform tests of the hypotheses at both the department and municipal levels. First, I discuss the evidence for or against the two main hypotheses using one-way analysis of variance (ANOVA). Second, I test for the independent effects of the two main variables when controlling for other factors using multivariate regression.
75 Censo de Población y IV de Vivienda,El Salvador 1992 76 Ministerio de Economía, Dirección General de Estadística y Censos. Encuesta de Hogares de Propositos Multiples, 1998.
163
FMLN Electoral Performance Controlled Zones The first hypothesis states that the FMLN should be more successful in FMLN-controlled zones than in uncontrolled zones. As discussed earlier, while the FMLN maintained its presence in eight of the country’s fourteen departments, its presence was strongest in the departments of Morazán and Chalatenango, as well as parts of San Salvador. Returning to Table 7.1 and focusing on legislative results, the FMLN captured 30% of the vote in San Salvador, by far its best performance in any single department. In Chalatenango, the FMLN captured 19% of the vote, its fifth best performance. These results, in addition to the FMLN’s performance in San Vicente (23%) and La Libertad (22%) where it also maintained strong presences during the war, provide relative support for the hypothesis. On the other hand, Morazán was a FMLN stronghold, yet the party only captured 16% of the vote, its eleventh best showing out of fourteen. In addition, Santa Ana, a department with little FMLN military activity or outright signs of support for the insurgents during the civil war, supported the FMLN with 22%. This was good for the fourth greatest level of FMLN support at the department level. At the departmental level, then, it does not appear as if there is a very strong association between FMLN insurgent activity and its electoral support in the 1994 elections. How does FMLN insurgent activity relate to its electoral performance when focusing on the municipal level? Table 7.2 presents a comparison of FMLN support in the 1994 legislative elections between FMLN-controlled zones of El Salvador and all others, using one-way analysis of variance (ANOVA). In the 1984 elections, the FMLN prevented elections from being held in 59 municipalities.77 The difference between FMLN support in these controlled zones from the rest of the country’s municipalities is statistically significant (F=17.625, 1 d.f., p < .0001). FMLN support in formerly controlled zones averaged 23% while FMLN support in the country’s other municipalities averaged 15%. FMLN support is indeed much greater in FMLN controlled zones versus all others.
77 San Luis de la Reina in San Miguel drops from all analyses, though, because there is no available electoral data from 1994. San Luis would have counted as an FMLN controlled zone and conflict zone.
164 Political Violence The second main hypothesis postulates that FMLN support in conflict-prone areas of the country should be higher than in those areas of the country that did not experience similar levels of political violence because of the overwhelming violence committed by the Salvadoran security forces against civilians relative to the insurgents. Table 7.3 presents a list of El Salvador’s fourteen departments along with the percentage of people in the department responding that they had lost a family member as a result of the war in the Seligson and McElhinny study (1996) and the FMLN vote in the 1994 legislative elections. At the department level, two stand out – La Unión and Cuscatlán. La Unión ranked as the seventh most conflict prone department with 38% of its residents having had a family member killed as a result of the war. Given that six departments were more conflict prone and seven were less, we would have expected FMLN support in La Unión close to the median. Instead, less than 7% of the department’s votes were cast for the FMLN making it the department that provided the least support for the FMLN. In Cuscatlán, on the other hand, over 70% of its residents reported having a family member killed as a result of the war. This was far and above, the most for any single department. Given the hypothesis, we would have expected FMLN support to be greater than the mid-range of all departments – yet Cuscatlán finished 6th most. Figure 7.2 plots the vote for the FMLN by department against the percentage of the department’s population responding that they had a relative killed as a result of the war. Again, if we divide the seven most severely affected departments from the seven least affected, there does not appear to be much of a relationship.78 But perhaps the significance of the legacy of the war on FMLN performance is stronger at the municipal level than can be discerned at the department level. Table 7.4 presents a comparison of voting in non-conflict and ex-conflict zones for the FMLN in the 1994 legislative elections using one-way analysis of variance. As one can see, the FMLN averaged 21% of the vote in ex-conflict zones and only 13% in non-conflict zones. On average, the FMLN received about 60% more votes in zones that were severely affected by the war than those that were not. The difference between zones is also statistically significant (F=28.167, 1 d.f., p<.0001). These
78 One possible explanation, of course, is that the FMLN was responsible for more of the deaths in these areas than the government. Unfortunately, at this point, I am unable to distinguish between the two.
165 results support the hypothesis. Those areas of the country more affected by the war in El Salvador showed greater support for the FMLN than areas less affected by the conflict.
Controlled Zones and Political Violence I also test the effects of these two primary independent variables while controlling for a number of political conditions using ordinary least squares (OLS) regression with robust standard errors. The political variables measure the level of abstention, strength of other left parties, strength of bipolarization, and population. Again, while including a number of socioeconomic variables would be ideal, they are not available for El Salvador at the municipal level during the time period under consideration. I do report results using these 1998 poverty rates in Model B. Table 7.5 presents the results of the multivariate regression models using robust standard errors. Overall, the first model (A) is a good fit (R2 = .61) with several independent variables reaching conventionally accepted levels of statistical significance. Interestingly enough, though, one of the main independent variables is not statistically significant. The first hypothesis expects FMLN control over an area to be positively related to its electoral performance. Though the coefficient for FMLN control is in the predicted direction, it is not statistically significant (p < .154). FMLN control during the war is not a systematic predictor of postwar electoral performance when controlling for the effects of the other variables in the model. The second hypothesis concerns the geographic distribution of civil war violence. I expect conflict zones to be positively related to FMLN electoral support. The coefficient is in the predicted direction and statistically significant (p < .01). This finding is consistent with those who argue that the violence perpetrated by the state resulted in many Salvadorans providing support to the FMLN. Their support appears to have continued into the postwar period. One control variable was also statistically significant. A negative relationship was hypothesized between CD performance in the 1991 elections and FMLN electoral performance in 1994. The coefficient is both positive and statistically significant (p < .001). I had predicted that the relationship of a strong left party would negatively impact the performance of the FMLN, but this has proven not to be the case. One possible reason is that the CD had only participated in two previous elections. The limited electoral history of the CD did not allow for the party to develop a committed group of voters. Alternatively, this could indicate that the
166 people in these areas were more inclined to support leftist ideals, so they were a better ideological fit for the FMLN than areas where the CD had weak support. Lastly, the effects of bipolarization (p < .317), abstention (p < .168), and population (p < .658) are not statistically significant. Model B presents the results of multivariate regression using robust standard errors, with the addition of poverty rates at the department level. Controlling for the level of poverty at the department level does not change the model significantly. The variables for conflict zones and opposition strength remain statistically significant while all other variables again fail to reach accepted levels of statistical significance. Finally, the variable for poverty is statistically significant, although its direction is negative. As poverty increases across departments, support for the FMLN tends to decrease. One possible explanation for this unexpected relationship is that the ruling ARENA party employed its extensive patronage network to heavily influence the vote in the poorer areas of the country.
DISCUSSION
Following a decade of political violence and a negotiated settlement to the Salvadoran civil war, the FMLN competed for the first time as a political party in the 1994 “elections of the century”. In the elections, the FMLN secured roughly 20-25% of the vote. The spatial distribution of the vote varied from a low of 6% in the department of La Unión to over 30% in the capital department of San Salvador. In this paper, I tested two main variables expected to explain the geographic support for the FMLN in the 1994 election. The results are somewhat mixed. At the departmental level, we see a weak relationship between FMLN controlled zones and the distribution of political violence and the FMLN support in the 1994 elections. The FMLN performed well in some departments where it had a strong presence (San Salvador, Chalatenango, and San Vicente), but also in some where its presence was relatively weak (Santa Ana). Likewise, FMLN support was strong in departments with both high (Chalatenango, Cuscatlán and San Vicente) and low (Santa Ana and La Libertad) levels of political violence. In order to analyze the legacy of the Salvadoran civil war on the electoral performance of the FMLN, I also conducted a number of statistical analyses at the municipal level. I ran one-
167 way analyses of variance to determine whether there existed a statistical relationship between support for the FMLN in controlled zones and in non-controlled zones and in ex-conflict and non-conflict zones. In these analyses, FMLN support was indeed greater in controlled zones (relative to non-controlled zones) and ex-conflict zones (relative to non-conflict zones). These findings support the hypotheses. What conclusions can we draw from the preceding analysis? On the one hand, the FMLN did perform better in zones it controlled during the war than in those it did not. This supports the finding that the ability of the FMLN to develop alternative authoritative structures during the civil war had positive implications for the FMLN in the postwar period. Many scholars cite the importance of civilian support in FMLN controlled zones and these zones do tend to favor the FMLN more so than non-controlled zones. On the other hand, even in formerly controlled zones, the FMLN managed less than one-quarter of the total vote. FMLN support was perhaps “widespread” and “voluntary,” but it was nowhere near universal. Most residents continued to support other political alternatives. Furthermore, the effect of this variable weakens when one controls for whether the area was an ex-conflict zone and the electoral history of the area, suggesting that FMLN control is not a particularly good predictor of FMLN support. The legacy of violence in a given area has a more consistent impact on FMLN support. FMLN support is greater in ex-conflict zones than in non-conflict zones, both using one-way analysis of variance and when controlling for FMLN control and a number of factors using multivariate regression. Only the coefficient for ex-conflict zones remains statistically significant; the effect of FMLN control during the war no longer reaches statistical significance. Finally, the only control variable that consistently reached statistical significance is strength of opposition parties. The FMLN performed well in those areas where other leftist parties had done well previously. It appears that there was sufficient electoral support in these areas to support another party of the left. The performance of the top largest parties (bipolarization) and the history of voting (abstention) in the most recent election prior to the incorporation of the FMLN do not contribute to understanding how the new political party performed.
168 CONCLUSIONS AND IMPLICATIONS
Previously, I argued that the FMLN’s success in capturing and controlling territory throughout El Salvador better prepared it for the challenges of electoral competition, relative to the experience of the URNG in Guatemala. The relationship between FMLN-controlled territory and the party’s electoral performance in the 1994 election is mixed. ANOVA results support the hypothesized relationship, but when controlling for other factors using OLS, FMLN-control is no longer statistically significant. The second main relationship in this chapter relates to the level of political violence. Previously, I found that both the FMLN in El Salvador and the URNG in Guatemala were less responsible for the violence perpetrated against the civilian population in both countries, though, their electoral performances were quite different. In this chapter’s municipal analyses of the FMLN vote, conflict zones tended to vote for the FMLN more so than non-conflict zones. Given these findings, governments should be aware that violence against the civilian population in the name of fighting a guerrilla insurgency might have long-term effects on political support once the fighting has ended. For insurgents, these results suggest that they may see more long-term gains by bringing the conflict to more of the country (particularly if the government responds by punishing the civilian population) than by trying to transform the areas they control into revolutionary societies. In terms of the control variables, only opposition strength appears to be helpful in explaining the performance of the FMLN as a new political party. In those municipalities where the left performed well in 1991, the FMLN did well in 1994. The other controls – voter abstention and bipolarization – did not have any systematic impact on FMLN performance. For insurgent groups contemplating the transition to political party, it would appear that they should seriously consider the performance of like-minded political parties prior to making their decision. In addition, new political parties may be able to benefit more from cultivating voters in those areas where previous opposition groups have already shown promise, than by venturing out into areas that appear to be ripe to support a new political party (that have not previously shown much sympathy for similar ideologies). In the beginning of this chapter, I mentioned that I was interested in how well the hypotheses developed in chapter 2 and tested in chapters 5 and 6 could help us to understand the
169 electoral performance of the FMLN in the first election in which it participated. But the FMLN and URNG are not the only two insurgent groups that have attempted the transition to political party in recent years. How well does the theory help to explain some of the other transitions that have taken place? That is a question I pursue in the next chapter.
170 TABLE 7.1 FMLN Performance in 1994 “Elections of the Century”
Department FMLN FMLN FMLN Coalition Municipal Legislative Executive
San Salvador 115,740 (27.7%) 124,196 (29.5%) 134,993 (33.8%)
Santa Ana 27,808 (20.8) 28,685 (21.5) 33,462 (25)
San Miguel 12,327 (14.5) 12,575 (15) 15,703 (18.6)
La Libertad 31,464 (20.6) 33,288 (22) 39,074 (25.7)
Usulután 10,958 (15.2) 11,440 (16) 15,151 (21)
Sonsonate 16,616 (15.7) 17,715 (16.7) 22,701 (21.4)
La Unión 2,744 (6) 3,141 (6.8) 4,669 (10)
La Paz 10,315 (15) 10,906 (16) 12,409 (18.2)
Chalatenango 10,121 (20) 9,492 (18.9) 10,531 (21)
Cuscatlán 8,215 (16.9) 8,899 (18.3) 11,561 (23.7)
Ahuachapán 9,530 (15.8) 9,633 (16) 11,015 (18.2)
Morazán 5,929 (16.4) 5,670 (15.8) 6,031 (16.7)
San Vicente 7,907 (21.1) 8,491 (22.7) 9,388 (25.1)
Cabañas 3,824 (12.5) 3,680 (12.1) 4,941 (16.2)
Overall, the FMLN captured 15 municipal elections (13 alone and 2 in coalition with the CD) with 20.33% of the total valid vote (273,498 out of 1,345,454). 15 municipal elections account for 5.7% (15 out of 262) off all municipal councils. The FMLN captured 21 out of 84 (25%) of the seats in the Legislative Assembly with 21.39% of the vote. Finally, the FMLN-CD alliance captured 24.99% of the vote for the presidency in the first round.
171 TABLE 7.2 FMLN-Controlled Zones and FMLN Performance in the 1994 Elections
Zones FMLN Performance Number of Zones Mean (S.D.)
FMLN-Controlled Zones 22.85 (18.81) 58
Rest of Country 15.28 (9.39) 203
Total 16.97 (12.49) 261
F= 17.625 (1df, p < .0001)
172 TABLE 7.3 Distribution of Political Violence and FMLN Performance in the 1994 Elections
Department % with Victims FMLN Performance
Cuscatlán 72.5% 18.3% Chalatenango 54.5 18.9 San Vicente 47.9 22.7 Cabañas 46.7 12.1 Morazán 44.4 15.8 Usulután 42.0 15.9 La Unión 37.6 6.8 San Salvador 32.6 29.5 San Miguel 31.4 15 La Paz 29.2 16 La Libertad 28.3 21.9 Sonsonate 28.6 16.7 Santa Ana 19.2 21.5 Ahuachapán 17.1 16 National Average 34% 21.39%
The FMLN vote is based on the 1994 legislative elections % with victims reflects the percentage of respondents in the department that claim to have had a family member killed as a result of the Salvadoran civil war. The figures are cited in Seligson and McElhinny (1996) Bold departments must indicate strong FMLN presence.
173 TABLE 7.4 FMLN Performance in “Ex-Conflict” and “Non-Conflict Zones” in the 1994 Elections
Zones FMLN Performance Number of Zones Mean (S.D.)
Ex-Conflict Zones 21.07 (15.59) 124
Non-Conflict Zones 13.25 (7.03) 137
Total 16.97 (12.50) 261
F= 28.167 (1df, p < .0001)
All municipalities where elections were not held in the first round of the 1984 presidential election are coded as FMLN controlled zones.
174 TABLE 7.5 Accounting for FMLN Performance in the 1994 Elections
Variable Model A Model B Coefficient Coefficient (Robust S.E.) (Robust S.E.)a
FMLN-Controlled Zone 2.01 (1.66) 3.00 (1.98)
Conflict Zone 3.12 (1.13)** 3.42 (1.39)*
Bipolarization 0.13 (0.13) 0.11 (0.10)
Opposition Strength 0.94 (0.11)*** 0.92 (.0.08)***
Abstention -10.94 (7.91) -10.74 (9.07)
Poverty - -0.13 (0.05)*
Population -0.000005 (0.00) -000005 (0.00)
Constant 5.30 (13.28) 12.88 (9.30)
N = 251 N=251 R2 = .61 R2 = .63 F (5, 245) = 30.65 F (7, 13) = 76.25 Prob > F = 0.0001 Prob > F = 0.0000
*p < .05, **p < .01, ***p < .001 a Adjusted for clustering on department
175 Usulutan
Sonsonate
San Vicente
Santa Ana
San Salvador
San Miguel
Morazan artment
p La Union De La Paz
La Libertad
Cuscatlan
Chalatenango
Cabanas
Ahuachapan
0 5 10 15 20 25 30 35 Percentage
FIGURE 7.1 FMLN Vote in 1994 Legislative Election.
176 $ Cuscatlan
70.0
60.0
d $ Chaletenango
e
l
l
i K
50.0 r e $ San Vicente
b $ Cabanas m
e $ Morazan M
$
y Usulutan l
i 40.0 m
a $ La Union
F
% $ San Salvador $ San Miguel 30.0 $ $ La Paz Sonsonate $ La Libertad
20.0 $ Santa Ana $ Ahuachapan
10 15 20 25 30 1994 FMLN Leg Vote
FIGURE 7.2 Distribution of Political Violence and FMLN Performance in the 1994 Legislative Elections.
177 CHAPTER 8 GLOBAL INSURGENT TRANSITIONS TO POLITICAL PARTIES
Insurgent group transitions to political parties are not unique to El Salvador and Guatemala. As mentioned previously, insurgent groups have also attempted transitions to electoral politics in Nicaragua, Honduras, Venezuela, Colombia, Uruguay, Mozambique, South Africa and Angola, among others. It is also probable that some of today’s insurgent groups will make the transition to electoral politics at some point in the future. In this chapter, I consider whether the factors that I have found to explain the divergent performances of the FMLN and URNG are generalizable to other cases of insurgent transitions to political parties. I chose to compare the transition of the FMLN and URNG to the Mozambican National Resistance (RENAMO) in Mozambique and the April 19th Movement (M-19) in Colombia. Both insurgent groups also concluded peace agreements with their respective government, transformed themselves into political parties, and competed in at least two elections following the end of their conflicts.
ACCOUNTING FOR THE PERFORMANCE OF NEW POLITICAL PARTIES
Resistëncia Nacional Moçambicana (RENAMO) In 1974, a military coup in Portugal overthrew the country’s government. One year later, the new government granted independence to its colonial possession, Mozambique. Upon independence, the anti-Portuguese independence movement that had formed in 1962, the Frente de Libertaçâo de Moçambique (FRELIMO), or Front for the Liberation of Mozambique, assumed power and set out to establish a one-party state. At the same time that Mozambique was achieving its independence, neighboring minority-ruled white Rhodesia was confronting a black insurgency led by the Zimbabwe Africa National Union (ZANU) and its military wing, the Zimbabwe African National Liberation Army (ZANLA). The Rhodesian government responded to FRELIMO's support for ZANU by forming the Resistëncia Nacional Moçambicana (RENAMO), or Mozambican National Resistance. Originally, RENAMO was organized to carry out attacks against ZANU and FRELIMO bases inside Mozambique in order to punish FRELIMO for its support of the rebels in Rhodesia (Turner et al. 1998: 153-154; Waterhouse
178 1996: 12). According to Waterhouse (1996: 12), these attacks were primarily carried out by a small number of Mozambicans and involved “setting villages aflame, destroying clinics, shops, and railways, and vandalising commerce.” In 1980, Rhodesia gained its independence, renaming itself Zimbabwe. The new Zimbabwean ZANU-led government had no further need to support RENAMO. At this time, the South African military undertook support for RENAMO. As in Rhodesia, a white minority government ruled South Africa and had little interest in a black Marxist regime on its border. South Africa “funded, trained, and equipped” RENAMO in order to destabilize the Frelimo government by targeting “roads, railways, power plants, and other economic installations” in addition to “all social symbols of FRELIMO: State-run schools, health centres, and local administration, and the people who staffed them” (Waterhouse 1996: 12). In 1984, South African support for RENAMO took more covert forms following the signing of the Nkomati friendship Accord. In return for South Africa ending its support for RENAMO, the FRELIMO government promised to stop the African National Congress (ANC) from continuing to work out of Mozambique.79 While covert support from South Africa would continue following the accord, RENAMO began to become “more dependent on its own resources” (Waterhouse 1996: 13). At the same time, its reputation as an externally supported insurgency remained intact.80 Similar to events in El Salvador, domestic and international factors converged to end the Mozambican war by means of a political settlement. Militarily, the civil war was at its most destructive in the mid-1980s as RENAMO achieved “significant advances in the central provinces, splitting the country into government-occupied and rebel-occupied territories” (Turner et al. 1998: 154). By the end of the decade, though, “it was obvious that no military victory was possible” as neither side appeared capable of militarily defeating the other (Waterhouse 1996: 14). Each side’s ability to maintain its armies was also deteriorating (Weinstein 2002: 148). In addition to the military stalemate, political and economic reforms were also underway. In 1984, FRELIMO began to align itself with the West, gaining membership in the International Monetary Fund (IMF) and the World Bank. In 1987, FRELIMO adopted an economic stabilization program encouraged by the IMF. By 1990, FRELIMO had formally rejected
79 At the time, the ANC was trying to overthrow the apartheid government in South Africa. 80 Rhodesia and South Africa were not the only international benefactors for RENAMO. Its anti-communist rhetoric attracted support from Malawi, Kenya, and the United States.
179 Marxism, introduced and adopted a liberal constitution, declared Mozambique a democratic state and called for competitive multiparty elections (Waterhouse 1996: 23; Turner et al. 1998: 154). FRELIMO’s economic and political reforms made it difficult for RENAMO to continue to fundraise in the West based upon its anti-communist crusade.81 The Soviet Union and Eastern European nations cut their support for FRELIMO and they turned their attention towards internal problems (Turner et al. 1998: 154; Waterhouse 1996: 14). Finally, with the decline of apartheid, South Africa's support for RENAMO also declined (Waterhouse 1996: 14). Following Foreign Minister Joaquim Chissano’s succession of President Samora Machel in 1987, Mozambique began its path to peace. Initially, Zimbabwe, Kenya and Malawi were involved in trying to mediate an end to the war, but the peace process really gained momentum following the involvement of the Catholic Church and the Italian government in July 1990. After two years of negotiations, RENAMO and FRELIMO concluded the General Peace Agreement in Rome on October 4, 1992. This agreement included a cease-fire, the demobilization of both armies and the formation of a new unified national army, a new electoral law, and the holding of presidential and parliamentary elections within one year. The agreement also called for the United Nations to verify and monitor the implementation of military, political, electoral, and humanitarian accords through the United Nations Operations in Mozambique (UNOMOZ). For the next two years, RENAMO transformed itself into a political party and prepared for the country’s first multiparty elections.
RENAMO as Political Party Since its transition to political party, RENAMO has competed in presidential and legislative elections in 1994, 1999, and 2004.82 Table 8.1 presents the results from the 1994 and 1999 presidential elections in Mozambique. Table 8.2 and Table 8.3 present each party’s vote totals and seat totals from the 1994 and 1999 Assembly of the Republic elections. In October 1994, Mozambique held its first postwar multiparty election. The campaign lasted from September 22nd until October 24th and functioned fairly smoothly (Waterhouse 1996: 24; Turner et al. 1998: 159-160), although there were difficulties both during the campaign and election weekend. For example, during the campaign, elections workers, the police assigned to guard the
81 In addition, RENAMO was discredited internationally when Robert Gersony of the US State Department released a refugee-based report finding RENAMO directly responsible for 100,000 civilian deaths. 82 In 1999 and 2004, RENAMO competed as the lead party in the RENAMO-Electoral Union coalition.
180 polls, and demobilized soldiers each threatened to disrupt the electoral process if their complaints were not addressed (Turner et al. 1998: 160). The International Centre Against Censorship also complained that it was difficult to move to and from RENAMO-controlled areas of the country and that the people living within these areas appeared to be living in fear (Waterhouse 1996: 24). The U.S. State Department also claimed that FRELIMO used its position and influence to raise illicit campaign contributions from unwilling local businessmen (Turner et al. 1998: 160). Finally, the night before the elections, the entire electoral process was thrown into doubt. RENAMO’s presidential candidate, Afonso Dhlakama, reiterated concerns he had raised during the campaign about FRELIMO planning fraud to win the election. He announced that he would be withdrawing from the contest. The legitimacy of the elections and the viability of the peace process depended on the participation of both RENAMO and FRELIMO. In order to head off disaster, the international community pressured Dhlakama to remain in the race and “after the Commission for Supervision and Control provided a written guarantee that it would closely monitor the elections, and an additional $1 million incentive was added to the Renamo Trust Fund, early on 28 October Renamo agreed to continue with the process” (Turner et al. 1998: 161). The parties also agreed to extend the voting one day to October 29th. Even with Dhlakama’s threat to boycott the election and the confusion that accompanied his flip-flopping, over 87% of all eligible voters turned out.83 Twelve parties presented candidates for the presidency. RENAMO’s candidate, Afonso Dhlakama, finished second with 34% of the vote to FRELIMO’s candidate, Joacquim Chisano, who finished first with 53%. None of the ten remaining candidates captured more than three percent of the vote. In elections for the 250-seat Assembly of the Republic, twelve political parties and two coalitions competed, but only three won seats. Effectively, RENAMO and FRELIMO split the country’s ten provinces. RENAMO won majorities in the central provinces of Manica, Nampula, Sofala, Tete and Zambézia, and FRELIMO won in the north and the south (Waterhouse 1996: 24; Turner et al. 1998: 162).84 The five provinces won by RENAMO were the areas it was most active during the war (Turner et al. 1998: 162). FRELIMO won 129 seats with 44%, RENAMO won 112 seats with 38%, and the Democratic Union (UD) won nine seats with 5% of the vote. The UD’s votes and seats are generally considered a mistake because the placement of its symbol on the
83 The CNE registered 6.3 million (Turner et al. 1998: 159) – 6.4 million (Waterhouse 1996: 23) of the 7.8 million eligible voters. 84 Zambézia and Nampula are the two most populous provinces in the country and were won by RENAMO.
181 legislative ballot corresponded to FRELIMO’s placement on the presidential ballot (Waterhouse 1996: 25; Turner et al. 1998: 162). Many believe that the UD’s votes were intended for FRELIMO. They also point to the fact that the UD’s share of the vote fell to a more reasonable 1.5% in the 1999 election. In the 1999 elections, RENAMO also finished second. Turnout fell from 87% in 1994 to just below 60%. In elections for the presidency, RENAMO led a coalition of eleven parties (RENAMO Electoral Union or RENAMO-UE) behind Afonso Dhlakama to 48% of the vote. This fell short of the incumbent Chissano’s 52%. Initially, Dhlakama screamed fraud, although eventually, RENAMO accepted the results. In elections for the Assembly of the Republic, FRELIMO won 133 seats with 49% of the vote and RENAMO-UE won 117 seats with 39% of the vote. None of the remaining ten political parties attained over five percent of the vote. Therefore, none were eligible to occupy a seat in the legislature. After two elections in Mozambique, RENAMO and FRELIMO have effectively transformed the country into a two-party system with RENAMO successfully transforming itself from the country’s leading insurgent group to its leading opposition political party. In each election, RENAMO finished second. RENAMO's most successful performance in elections for the executive came in 1999, when it came close to forcing a second round presidential election. In elections for the country’s legislature, the Assembly of the Republic, RENAMO has done very well. In the 250-member assembly, RENAMO captured 112 (1994) and 117 (1999) seats making it the second largest political party in the country.
Movimiento 19 de Abril (M-19) During the second half of the twentieth century, a number of insurgent groups took up arms against the Colombian government. In 1970, the Alianza Nacional Popular (ANAPO), or National Popular Alliance, supported former dictator General Gustavo Rojas Pinilla for president. ANAPO was an alliance of several left-wing groups that believed the time was ripe to break Colombia’s two-party political system comprised of the Liberal Party and the Conservative Party. On the day of the vote, preliminary results suggested a likely ANAPO victory. But before victory for ANAPO and Rojas Pinilla, the vote count was interrupted. When counting resumed, ANAPO had lost by a slim margin. Many ANAPO supporters called the elections fraudulent and took to the streets in protest. The protests were relatively calm and short-lived because Rojas
182 Pinilla accepted the results and called for calm among his supporters. Some of those that remained convinced that the electoral outcome had been rigged and did not accept the results, then set out to form a new revolutionary movement. They joined with some disgruntled members of the country’s other insurgent groups and named the new organization after the date of the stolen election, the Movimiento 19 de Abril (M-19), or April 19th Movement. Many consider the M-19 to have been unlike other guerrilla groups in Colombia for a number of reasons. First, its central issue was democracy. Although Marxist thought influenced the thinking of the organization, the M-19 strove to break the two-party stranglehold on Colombian politics. They claimed that they were interested in reforming democracy, not replacing it. Second, the organization also said that it wanted to represent the poor and marginalized in the hopes of further democratizing the country. The unique aspect was that, as an insurgent group, the M-19 stole milk, food and medicine and then distributed it to the urban poor (Dudley 2004: 49). They portrayed themselves as a modern day Robin Hood organization. Finally, the M-19 was also different because it focused on the urban areas of the country. Other insurgent groups, such as the Fuerzas Armadas Revolucionarias de Colombia (FARC), or Revolutionary Armed Forces of Colombia, concentrated on the rural areas of Colombia. After fifteen years of guerrilla warfare, the M-19 was not very successful militarily. However, it did gain domestic and international notoriety because of its high-profile campaigns: the M-19 stole one of Simon Bolivar’s swords from a museum; it held more than one dozen ambassadors hostage at the Dominican Republic Embassy in 1980; and it occupied the National Palace of Justice in November 1985, in the midst of peace talks with the government. In this last action, the military shunned negotiations and laid siege to the palace. Over 40 militants from the M-19 and over ten Supreme Court Justices were killed in the retaking of the court house. In the assault, the M-19 lost five of its top leaders and much of its public support (Gross 1995: 57). In the late 1980s, the M-19 came to believe that it could accomplish more through electoral politics than it could by continuing its guerrilla campaign. A 1985 poll, taken prior to the fiasco at the Palace of Justice, had found that the M-19 would have captured 36.7 percent of the vote if elections were held that year (Dudley 2004: 55). Before disarming, polls also reported that the M-19 had the most support among all guerrilla groups (Boudon 2001: 76). These were both encouraging signals for the M-19. In September 1988, the M-19 began negotiations with the Colombian government. In March 1989, the government pardoned the M-19 in return for its
183 demobilization and disarmament. M-19 militants also received training in small-business management and a small stipend to be paid over six months (Brooke 1990). On November 2, 1989, the M-19 and Colombian President Virgilio Barco signed the Political Pact for Peace and Democracy. The M-19 then prepared for the March Constituent Assembly elections and the May presidential elections.
M-19 as Political Party Since its transition to political party, the M-19 has participated in the 1990 Constituent Assembly election, the 1990, 1994, 1998 and 2002 presidential elections, and the 1990, 1991, 1994, 1998, and 2002 legislative elections. Shortly after its demobilization, the M-19 joined with a number of leftist opposition groups in Colombia to form the Alianza Democrática M-19 (AD M-19), or Democratic Alliance M-19. Table 8.4 presents the results from the 1990 and 1994 presidential elections. Tables 8.5 and 8.6 present the vote totals and seat totals from the 1990, 1991, and 1994 elections to the bicameral Colombian legislature. Table 8.7 presents the seat totals for the 1990 Constituent Assembly election. In the March 1990 legislative elections, the AD M-19 won one seat with 0.35% of the vote in the 199-member Chamber of Representatives and no seats in the 114-member Senate. According to Boudon (2001: 77), the AD M-19’s performance was due in great measure to its inability to register many candidates in such a short window of opportunity and not necessarily a sign of its lack of popular support. As the AD M-19 began preparing for the May presidential elections, the AD M-19 suffered a serious setback when its presidential candidate, Carlos Pizarro, was shot and killed on April 26, 1990. The AD M-19 replaced Pizarro with another guerrilla leader, Antonio Navarro Wolff. Navarro captured 12.5% of the vote, finishing just ahead of the fourth place candidate from the Social Conservative Party who finished with 12.15%. In December 1990, the AD M-19 participated in elections for a Constituent Assembly. In these elections, the AD M-19 captured nineteen of the seventy-member assembly with 27% of the votes. In the assembly, the AD M-19 played a very important role in revising the country’s constitution. New elections were called for October 1991, both to pass the constitutional reforms and to elect a new legislature. At this time, it appeared as though the AD M-19 was ready to challenge the Liberal and Conservative stranglehold on Colombian politics.
184 Unfortunately for the AD M-19, it was unable to capitalize on its previous strong showings. On October 27, 1991, the AD M-19 captured nine seats (8.8%) in the new 102-seat Senate with 9.4% of the vote. It also captured fifteen seats (9.3%) in the 161-seat Chamber of Representatives with 10% of the vote. In the 1994 presidential elections, the AD M-19 and Navarro again finished in third place, but the party’s share of the vote fell below 4%. The AD M-19’s total votes fell by more than 500,000, or 70%, even though more than one million votes were cast compared to the 1990 presidential election. Likewise, the AD M-19 fared poorly in legislative elections. The AD M-19 failed to capture a seat in the Senate and only managed one seat (0.6%) in the Chamber of Representatives. After initial success in the early 1990s and sporadic victories in the latter half of the decade, the AD M-19 had failed to break the two-party dominance of the Colombian system, though it remained a minor political party.
Discussion Table 8.8 presents the political party performance of RENAMO, the AD M-19, the FMLN, and the URNG in terms of minor and major party success. Following the political settlements that led to the transformations of RENAMO and the M-19 into political parties, RENAMO has gone on to become a major political party and the M-19 has become a minor political party as the AD M-19. In Mozambique, RENAMO finished second in each presidential and legislative election since its incorporation into electoral politics. Following the 1994 election for the Assembly of the Republic, the effective number of parties was 2.14. There existed two major parties (FRELIMO and RENAMO), one minor party (the Democratic Union), eleven failed parties, and one failed coalition. In the 1999 elections, the effective number of parties was 1.99. There were two major parties (FRELIMO and RENAMO) and eleven failed parties. RENAMO had successfully transformed itself into the number two party in a two-party race. In Colombia, the AD M-19 finished in third place in the 1990 and 1994 presidential elections. In the 1990 Senate elections, the effective number of parties was 2.24. The Liberal and Conservative parties finished as major political parties, while the AD M-19, because it failed to win a seat, finished a failed party. In elections for the Chamber of Representatives, the effective number of parties was 2.20. The Liberal and Conservative parties are major political
185 parties and the AD M-19 with one seat, is a minor party. The 1990 election was held immediately following the peace agreement and in generally not regarded to be representative of the AD M-19’s support. The M-19 had its greatest success when it finished with the most seats in the 1990 Constituent Assembly elections. For the assembly, the effective number of parties was 4.083 with the AD M-19, National Salvation Movement, Liberal Party and Conservative Party as major political parties. Following the 1991 Senate elections, the effective number of parties was 3.06. The three major political parties were the Liberal Party, Conservative Party, and the AD M-19. Sixteen political parties qualified as minor political parties. Following the Chamber of Representatives elections, the effective number of parties was 3.03 with the Liberal Party, Conservative Party, and the AD M-19 again qualifying as major political parties. Sixteen parties also qualified as minor political parties. Finally, the effective number of parties for the Senate following the 1994 elections was 2.79 with the Liberal Party, Conservative Party and New Democratic Force as major political parties. The AD M-19 became a minor political party with one seat. And for the Chamber of Representatives, the effective number of parties was 2.75 with the Liberal Party, Conservative Party and the National Conservative Movement as major political parties and the AD M-19 among twenty-two minor political parties. In terms of the four insurgent transitions to political parties that have been analyzed in this dissertation, three groups transformed themselves immediately into major political parties (the FMLN, RENAMO, and the AD M-19), through, only two remained so (the FMLN and RENAMO). The AD M-19 quickly became a minor party in Colombian politics. The fourth party, the URNG, transformed itself into and remains a minor political party.
Size Hypothesis 1 states that larger insurgent groups are more likely to be successful than smaller groups. Previous analyses of the FMLN and URNG insurgent groups supported this hypothesis. The larger insurgent group, the FMLN, became a major political party and the insurgent group with fewer combatants, the URNG, became a minor political party.
186 RENAMO RENAMO primarily consisted of disgruntled Mozambicans and some Portuguese settlers following the establishment of one-party rule by the Marxist-FRELIMO party. Though RENAMO was primarily comprised of Mozambicans, its popular support was limited and its funding and training by external patrons gave it a less than fully Mozambican image (Weinstein 2002: 148). Slowly, RENAMO took on a more indigenous character, both in image and reality. Throughout the 1980s, RENAMO membership grew from two recruitment sources. First, its ranks increased because many Mozambicans were disenchanted with the new government and its destructive Marxist policies. FRELIMO alienated many Mozambicans by undermining traditional authorities and church leaders and forcing many peasants onto collectivized farms (Waterhouse 1996: 12; Turner et al. 1998: 154). A second membership source involved forced recruitment. Some claim that this was the primary method of recruitment for RENAMO as well as FRELIMO forces (Weinstein 2002: 145). There is also evidence that, of those captured and forced into fighting by RENAMO, some voluntarily stayed with the organization (Manning 1998: 170). Regardless, from 1984 (Weinstein 2002: 148) until the end of the war (Alden 1995), it is estimated that RENAMO counted on 20,000 combatants. While there are no exact numbers on how many soldiers RENAMO had at the time of its demobilization, 20,000 is still the general estimate (Alden 1995). At the time of its demobilization in 1992, RENAMO counted on 132 combatants per 100,000 Mozambicans.85
M-19 In the late 1970s, the M-19 was still relatively small and insignificant. It only began to increase “exponentially” at the end of the decade (Dudley 2004: 50). The original members of the M-19 came from ANAPO, existing guerrilla groups, and the Colombian Communist Party. Later, recruits joined because they were attracted to the M-19’s urban focus and Robin Hood image. Radu & Tismaneanu (1990: 168) place the number of M-19 combatants from 800 to 5,000 during its armed confrontation with the Colombian military and government. Gross (1995: 55) estimates that the M-19 never counted greater than 2,000 combatants. It is unclear whether the M-19 ever had anywhere close to 5,000 combatants, but it appears safe to say that its membership declined following its attack on the Palace of Justice in 1985. At the time of its
85 According to the CIA World Fact Book, Mozambique’s population stood at 15,113,300 in 1992.
187 reincorporation into society, the M-19 counted on roughly 800 (Carrigan 1993: 31) to 1,200 (Coleman 1990).86 Relative to the country’s population, the M-19 counted on 2.5 to 3.75 combatants per 100,000 Colombians.87 In the late 1980s, the M-19’s fellow insurgents, the FARC, counted on 4,000 to 7,000 combatants (Gross 1995). Relative to the country’s population, the FARC counted on 13 to 22 combatants per 100,000 Colombians.
Discussion Like the experiences of the URNG and FMLN, the experiences of RENAMO and the M- 19 are consistent with the size hypothesis. RENAMO, as a major political party, counted on 132 combatants per 100,000 Mozambicans, while the AD M-10, a minor political party, had counted on 2.5-3.75 combatants per 100,000 Colombians. Overall, the two major political parties, of RENAMO and the FMLN, counted on 132 and 293 members per 100,000 people, respectively. The two minor political parties, the M-19 and URNG, only counted on 2.5-3.75 and 69 members per 100,000.
Political Party Hypothesis 2 states that insurgent groups with prior political party experience are more likely to be successful than those without prior political party experience. The experiences of the FMLN and URNG support the political party hypothesis. The FMLN had experience with party politics prior to the war and relatively good relations with a political wing during it. The FMLN easily transformed itself into a major political party. In Guatemala, the URNG did not count on prior political party experience and only developed what some consider a political wing in the latter years of the conflict. Partly as a result, the URNG only experienced minor party success in the postwar as a political party.
RENAMO As mentioned previously, RENAMO’s initial recruits came from Mozambicans disenchanted with FRELIMO’s policies. Upon independence, FRELIMO established a one- party state that would rule for roughly the next twenty years. Therefore, there was no
86 Brooke (April 14, 1990) reports about 1,000 combatants and while Navarro cites 881 (Brooke May 31, 1990). 87 According to the CIA World Factbook, Colombia’s population stood at 31,945,165 in 1989.
188 opportunity for RENAMO’s members to have gained experience with electoral politics prior to the first time they occupied seats in the 1994 legislature. While prior experience with party politics was not a factor contributing to the performance of RENAMO as a political party, the development of a political wing during the war most likely was. For roughly the first decade of its existence, RENAMO paid little attention to the development of any political structures, only to the development of its military structures. Only in the early 1980s, after South Africa took control of the organization from Rhodesia, did RENAMO begin to develop its political structures (Hall and Young 1997: 131). For the most part, RENAMO’s initial focus was on improving its image at home and abroad, rather than on actually building a political organization. RENAMO’s leaders, including Dhlakama, traveled to West Germany, France, and Portugal seeking support. RENAMO eventually established overseas offices in Lisbon, Washington, D.C., Heidelberg, and Nairobi. Following a number of trips to these foreign capitals, RENAMO was convinced to “restructure politically with the aim of transforming Renamo into a political party which, together with political figures and organisations already existing in exile, could form a single Front” (Hall and Young 1997: 131). RENAMO formed a National Council in 1982 with Afonso Dhlakama as president, Orlando Cristina as secretary-general and a number of political exiles (Hall and Young 1997: 132). But even then, RENAMO’s political structures within Mozambique remained weak. RENAMO continued to develop its political structures following the Nkomati Accords in 1984. It was following the accords that RENAMO began to concentrate on recruiting individuals that would enable it to compete as a political party (Manning 1998: 162-163; 175). RENAMO recruited secondary students with promises of foreign scholarships in hopes that they would eventually form the political core of RENAMO following a peace agreement.88 Hall and Young (1997: 171) agree that RENAMO primarily maintained military structures until the mid to late 1980s, as it dedicated minimal time to “building political structures in RENAMO held areas, or at political teaching within the military itself…The reality was that Renamo retained all the features of an army.” In addition to the individuals that staffed RENAMO’s international offices and to its student recruits, RENAMO prepared to transform itself into a political party by training political “commissars” and political “delegates” (Manning 1998: 177). Initially, the
88 These scholarships never materialized and became a source of friction following the peace agreement. I discuss this issue in more detail while discussing unity in RENAMO.
189 commissars were responsible for informing RENAMO combatants about its political project and negotiations with FRELIMO; at a later point, another group of commissars explained the same issues to the civilian population (Manning 1998: 177).89 In June 1989, RENAMO held its First Congress in Gorongosa. Manning (1998: 180- 181) claims that it was at this congress where RENAMO moved “closer to something resembling a coherent opposition movement that would be capable of negotiating effectively with Frelimo and with potential external supporters” (Manning 1998: 181). Political preparations continued through 1990 as RENAMO president Dhlakama “began to receive a substantial amount of political coaching and ‘image enhancement’ at the hands of private foreign advisers” and “he assigned certain Renamo officials with international experience to give ‘political education classes’ to all heads of departments and other important officials” (Manning 1998: 182). RENAMO had been preparing for party politics for the last several years of the conflict, but it was still relatively unprepared for the challenges that awaited it (Waterhouse 1996: 23). There was no one in Mozambique with experience in electoral politics. To assist the new political party, the international community set up the “UN Trust Fund for the Implementation of the Peace Process in Mozambique,” more commonly referred to as the “RENAMO Trust Fund,” in May 1993.90 The international community believed that the success of the peace process depended on RENAMO successfully transforming itself into a political party and competing in the 1994 elections (Turner et al. 1998: 156-157). International donors contributed about $18 million to this fund, which RENAMO used for “offices, houses for party representatives who had to move to Maputo, office equipment, cars, trusted ‘intellectuals’ who could fill the new political positions and sufficient patronage resources both to compensate them and to buy the cooperation of party members who were being eased out” (Manning 1998: 189). Turner et al. (1998: 174) argue that “without the Renamo Trust Fund, Renamo could not have competed in the elections” and that “this transformation was key to the success of the peace process.”
89 The delegates were “supposed to be trusted people from among the local population, chosen by the people. Their job was to keep people informed about the progress of the war, to organize meetings, to explain RENAMO’s idea of how things should be in Mozambique” (Manning 1998: 179). Manning (1998: 180) has doubts about whether these commissars and delegates were ever actually chosen by “the people” or “served as an effective liaison between RENAMO and the people.” 90This fund was separate from the one established to help all opposition political parties. The European Union and its member countries, the United States, South Africa, and Namibia contributed $17 million to the RENAMO Trust Fund (Waterhouse 1996: 23). The international community contributed an additional $1 million to entice RENAMO to return to the electoral process after it withdrew the day before the 1994 elections were to begin.
190 How important was prior political party experience to the performance of RENAMO? On the one hand, RENAMO had no prior experience with electoral politics that would help it during its transition to political party. On the other hand, RENAMO had begun to build up its political structures in the mid-1980s both at home and abroad. These developments appear to have been useful, though not nearly enough, during the peace process and its transformation into a political party. The RENAMO trust fund and international technical assistance also contributed significantly to RENAMO’s success as a political party.
M-19 Many of the M-19’s initial recruits came from ANAPO. As mentioned previously, ANAPO competed in the 1970 election when its presidential candidate, General Gustavo Rojas Pinilla, allegedly lost due to outright fraud. Jaime Bateman, a disillusioned member of the FARC, sought out several ANAPO members in hopes of constructing a new insurgent group that would be comprised of both a military wing and a political wing (Carrigan 1993: 78). Bateman hoped that this new insurgent group would capitalize on the mass following of ANAPO, a mass following that the existing insurgent groups lacked. Therefore, many of the M-19’s initial recruits had been active in political work and electoral politics. While the M-19’s membership increased during the second half of the 1970s, it suffered a severe setback following its theft of weapons from the army barracks in Bogotá on New Year’s Eve, 1979. Following the attack, the Colombian military captured or killed most of the M-19’s top political and military leaders and recovered the stolen weapons. During the 1980s, M-19 political work was hindered by the dirty war being carried out against those critical of the government. It is unclear how important the experience of the M-19 was to its minor success as a political party. On the one hand, some in the M-19 had accumulated experience in ANAPO twenty years earlier. It is unclear, though, how many surviving members remained in 1989. In addition, the M-19 did not have a political wing that operated inside or outside Colombia.
Discussion The experiences of RENAMO and the AD M-19 provide mixed support for the political party hypothesis. On the one hand, the less successful political party, the AD M-19, could trace
191 its roots to a party that had previously competed in elections. The more successful political party, RENAMO, did not have prior experience with electoral politics. On the other hand, RENAMO focused on building up its political structures so that it would be better prepared for the challenges of party politics. The M-19, while it had set out to cultivate a strong political wing, saw many of its political cadres killed prior to its transition to political party. RENAMO definitely appears to have better prepared itself for a successful transition than the M-19. Overall, the experience of the four insurgent groups and party politics is mixed. The two successful insurgent groups, the FMLN and RENAMO, had prior experience in electoral politics during the war (FMLN) or maintained a relationship with a political wing during the conflict (FMLN and RENAMO). These experiences support the political party hypothesis. For the two minor political parties, those with prior electoral experience were no longer around at the time of their reincorporation into the political system. Therefore, while both had connections to prior political movements, these connections were no longer relevant when they participated in postwar elections. In addition, one group maintained a weak relationship with a political wing as its war drew to a close (URNG) and one did not have any relationship (M-19).
Mass-Based Organizations Hypothesis 3 states that insurgent groups that had extensive contacts with mass-based organizations during the war (unions, student groups, etc.) are more likely to be successful than those without such contacts. The experiences of the FMLN and URNG support the mass-based organizations hypothesis. In El Salvador, the FMLN maintained extensive relations with mass- based organizations on its way to becoming a major political party. In Guatemala, the URNG had poor relations with mass-based organizations and only achieved minor party success
RENAMO RENAMO does not appear to have had any contacts with what one would call mass- based organizations. In the 1970s and the early 1980s, RENAMO was primarily a military organization. Beginning in the mid-1980s, RENAMO made a concerted effort to develop a party program and political structures both to contribute to its wartime activities as well as to prepare itself for the challenges of a postwar situation. While it established political fronts overseas and began to train individuals in Mozambique, RENAMO does not appear to have reached out to any
192 organized social groups within the country. RENAMO was an organization that was “eminently flexible, unconstrained as it was by the demands of any pre-existing social or political movement” (Manning 1998: 178). RENAMO relied more on support from people who were disenchanted with FRELIMO than on any organized political groups.
M-19 The M-19 was the only Colombian insurgent group that focused on cultivating the country’s urban support. In the early 1970s, M-19 founder Jaime Bateman had concluded that Colombia needed a new revolutionary alternative. For this new alternative to be successful, it needed to combine the military might of a guerrilla army with the popular support and numbers of Colombia’s popular sectors (Carrigan 1993: 77). In the late 1970s, it appeared as if the M-19 would be very successful. The M-19 was growing rapidly and attracting a wide base of support. The M-19 counted university students, professionals, artists and members of the poor among its primary supporters in the urban areas (Sharpless 1992: 200). Unions, student organizations, and peasant groups were intensifying their demonstrations against the Colombian government (Dudley 2004: 50). The M-19 conducted much of its political work through FIRMES, an organization designed to unite Colombia’s left under a single banner, until a change in strategy in 1981 refocused the M-19 on a “rural militarist path” (Goméz 1996). The M-19’s goals of creating a mass-based movement to work alongside its military organization had failed and FIRMES disappeared from the scene in 1981 (Goméz 1996). While the M-19 continued to count on urban-based support, it does not appear to have had organized support from mass-based organizations. In the postwar, the M-19 also had difficulties establishing relations with mass-based organizations. According to Navarro, “When we came down from the hills we left behind part of the social base we had: the population that wanted the total defeat of the establishment. But we opened up to a different social base, which supports our proposals for change but believes that some sort of agreement is possible with the more modernising sectors of the system” (Latin American Weekly Report March 28, 1991). While Navarro was optimistic about this new support base, it is unclear how successful the organization was in cultivating this support. The new AD M-19 was comprised of over 120 political groups, but it “never established any real
193 linkage with the segment of Colombian society that had supported the M-19 as a guerrilla group and with the 1 million voters who backed the party in the December 1990 elections” (Boudon 2001: 86). The AD M-19 was wholly unprepared for managing such a large coalition and “little effort was made to create some kind of umbrella organization that would have more fully integrated them” (Boudon 2001: 77, 87).
Discussion The experiences of RENAMO and the M-19 do not support the mass-based organizations hypothesis. RENAMO had few connections with mass-based organizations in Mozambique, but still transformed itself into a major political party. The M-19 in Colombia set out to establish strong relationships with mass-based movements, but failed to do so. Even though most of its support came from the urban sectors of the country, the M-19 had a difficult time coordinating a common front among these mass-based movements. While the M-19 still maintained better relations with mass-based movements than RENAMO, they were less successful electorally. Overall, the relationship between insurgent groups and mass-based organizations and the electoral performance of new political parties is inconclusive. In chapter 5, I found that the FMLN’s relationship with mass-based organizations during El Salvador’s civil war contributed to its transformation into a major political party in the post war period while the URNG’s weak relationship with mass-based organizations resulted in its success as a minor political party. Of the two successful insurgent groups that have transformed themselves into political parties, only the FMLN had any significant relationship with these mass-based organizations. In addition, the two minor political parties, while not as strong as those of the FMLN, had stronger connections with mass-based organizations than RENAMO, yet they were less successful as political parties.
Unity Hypothesis 4 states that insurgent groups that can maintain unity and coherence during the war, negotiations, and in the postwar are more likely to be successful than insurgent groups that fracture. Both the FMLN and the URNG had difficulties keeping their insurgent coalitions together. In Guatemala, the URNG became a minor political party because it lost a number of combatants during the war, the negotiation phase, and in the postwar. The FMLN only began to lose important members in the postwar after it had established itself as a major political party.
194
RENAMO RENAMO confronted a small number of internal challenges as an insurgent group, during the peace process, and while competing as a political party. As an insurgent group, RENAMO’s biggest threat to its internal cohesion came during the 1979-1980 period. In 1979, there were indications that RENAMO was on the verge of defeat following the death of its leader, Andre Matsangaíssa, and, in a separate incident, its base was overrun by government troops (Hall and Young 1997: 120). Following Matsangaíssa’s death, a power struggle ensued among RENAMO’s leadership that “culminated in a shoot-out at Chisumbanje in southern Zimbabwe (June 1980) with the supporters of Afonso Dhlakama emerging victorious” (Hall and Young 1997: 120). Under Dhlakama’s leadership and the contribution of the South African military, RENAMO survived and even strengthened over the next decade. When RENAMO began to develop its political structures abroad in the mid-1980s, a new threat emerged as tensions intensified between its supporters living abroad and those fighting within Mozambique. RENAMO’s political wings helped attract international attention to its cause, but they also contributed to its characterization within the country as an organization dominated by “disgruntled former Portuguese settlers” (Manning 1998: 180). For years, RENAMO had been trying to dispel this myth. RENAMO then moved to replace these representatives with others more closely aligned to the insurgent organization within Mozambique. A similar problem occurred when Dhlakama chose to provide “political education classes” to mid- and high-ranking officials in RENAMO. The trainers brought in from outside the country were of relatively low rank within the organization, leading to conflicts with the students. The students were upset because they tended to be of mid- and high-rank (Manning 1998: 182). Unity problems also arose during the peace process. First, RENAMO had been preparing itself for its transition to political party for a few years before the peace process started, but even at the moment of negotiations with FRELIMO, RENAMO was still unprepared organizationally. Dhlakama again sought to have RENAMO bring in trained personnel to overcome the organization’s new challenges. RENAMO brought into the open personnel from within its “small urban clandestine branches, or on the recommendation of these members” (Manning 1998: 185). These new recruits “were unfamiliar to most ‘bush’ members of RENAMO, who
195 suspected them, correctly in many cases, of using the party only to gain potentially lucrative financial positions” (Manning 1998: 188). Although tensions existed, they did not lead to defections from the organization. Finally, RENAMO also confronted problems in the postwar period. In the late 1980s, RENAMO recruited students in order to prepare for the challenges of party politics, luring 100- 200 students into the organization with promises of foreign scholarships. In reality, though, “none of them were sent on scholarships abroad, and many interrupted their studies in their final stages only to find themselves in a remote RENAMO outpost with no possibility of resuming their academic careers” (Manning 1998: 182-183). RENAMO’s failure to fulfill its promise to the students caused problems for the party, particularly in 1995. A group of students attacked RENAMO representatives on their way to the party’s first national conference following the 1994 elections (Manning 1998: 183). While many students remained with RENAMO and some even assumed leadership positions within the party (four were elected to the legislature), these events brought to light yet another fault line within the new political party. RENAMO has overcome a number of internal challenges at each stage of its existence in becoming the second largest party in a two party race. This seems to have been the case for a number of reasons. First, as an insurgent organization, RENAMO was large enough to absorb the loss of a small amount of personnel. Second, the internal difficulties might not have been salient enough to result in the “losers” of these conflicts abandoning the organization. Finally, both during the conflict and in the postwar, FRELIMO and RENAMO were the only viable alternatives. FRELIMO established a one-party Marxist state following Mozambique’s independence from Portugal. For many, the only viable opposition existed with RENAMO. In the postwar period, FRELIMO and RENAMO designed an electoral system that complicated third party alternatives Outside of RENAMO and FRELIMO, only one other party has ever had its members elected to the legislature and, by most accounts, the votes for this party were unwarranted.
M-19 There have been a number of times throughout the M-19’s history when its unity has been tested. An early challenge occurred following the first peace agreement signed between the M-19 and President Belisario Betancur. In 1982, President Betancur amnestied Colombia’s
196 imprisoned guerrillas in the hopes of kick starting a peace process. Before peace talks concluded, the M-19 leader who had initiated talks with the government, Jaime Bateman, died in a plane crash. Shortly after the peace agreement took affect, the M-19 began to reconsider its decision as M-19 militants and supporters were being killed in a dirty war carried out by right- wing opponents of the peace. The M-19 also claimed that Betancur was slow in bringing about the reforms necessary for a sustainable peace (Carrigan 1993: 81). In December 1984, the remaining M-19 leadership met in Havana to consider whether they should return to war because the government had failed to live up to its promises. According to Carrigan (1993: 82), the debate “raged long and inconclusively” with “the M-19 leadership split down the middle” between the political and the military wings. The political wing “voted to give the peace process more time” while the military wing “sought an immediate return to war.” In June 1985, the M- 19 made its decision to return to war. Although a decision to break the peace and return to war led to intense debates within the M-19, it did not lead to defections from the organization. The M-19 has also confronted challenges to its internal unity as a political party. The M- 19 elites who assumed the direction of the party following Carlos Pizarro’s death in April 1990 believed the best way to win elections was to form coalitions with other small political parties. After some initial success in the 1990 presidential and Constitutional Assembly elections, the AD M-19 struggled to remain a viable played in Colombian politics. Some within the party, particularly Antonio Navarro Wolff, blamed the party’s declining electoral fortunes on the fact that it had to compete with limited resources against two parties that had amassed several decades of experience and had greater access to human and financial resources (Boudon 2001). Navarro also believed that the coalitions that the AD M-19 had formed with small political parties were “positive” in helping them succeed (Boudon 2001: 79). Others believed that the coalitions, instead of strengthening the AD M-19, were “stripping the party of its identity and its electoral support” (La Prensa, November 2, 1991 cited in Boudon 2001: 79). Many AD M-19 supporters, and potential supports, believed that the party “had become too friendly with MSN leader (National Salvation Movement) Alvaro Gómez, who at times occupied the far right of Colombia’s political spectrum” (Boudon 2001: 78). Finally, another division revolved around the party’s decision to accept the Health Ministry cabinet position in César Gaviria’s administration in 1990. In doing so, the AD M-19 became the first party other than the Liberal and Conservative parties to participate in
197 government. While accepting the position was never entirely popular, the situation deteriorated in 1992 when President Gaviria selected Gustavo de Roux to replace Camilo Gonzalez as the AD M-19 representative. Senator Evert Bustamente and 200 other AD M-19 supporters protested the appointment of de Roux arguing that, while he might have been approved by Navarro, the minister-select was not debated by the party (Agence France Presse July 5, 1992). The AD M-19 threatened to give up the position if the government did not increase the Health Ministry’s 1993 budget in October 1992. The AD M-19 and the government temporarily reached a budget compromise, but the AD M-19 eventually withdrew from the position shortly afterwards in protest over the government’s renewed military offensive against the country’s remaining guerrilla groups. Boudon (2001: 81) argues that any credibility that the AD M-19 might have attained by joining the Gaviria government was overshadowed by the criticism leveled against it from within the organization and possibly from outside the organization “since it created the public perception that they were part of a government that polls showed was largely unpopular.”
Discussion Both RENAMO and the AD M-19 confronted a number of difficulties in maintaining their political cohesion. While neither insurgent group suffered massive defections throughout their history, the divisions slowly ate away at the AD M-19’s support. All four insurgent groups dealt with internal challenges as insurgent groups and as political parties. During the conflict stage, only the URNG in Guatemala seems to have suffered significantly as a result. Many combatants left the URNG to take part in a revitalized civil society or because they had concluded that the URNG could not win. As a result, the URNG entered electoral politics much weaker than it would have years earlier. The M-19 also entered electoral politics much weaker than it had been years earlier, but its weakening was not the result of severe internal divisions. In the postwar period, the two larger insurgent groups, RENAMO and the FMLN, even managed to remain one of two major political parties in each country following internal divisions. The two minor political parties, the AD M-19 and the URNG, have seen their electoral fortunes decline following the divisions, particularly in the postwar period. It is not entirely surprising that the larger insurgent groups have been able to remain major political parties in spite of the divisions, while those within the minor political parties have resulted in them becoming very marginal players in each country’s political system.
198
Territory Hypothesis 5 states that insurgent groups that control territory during the war are more likely to be successful than groups that are unable to control territory. The experiences of the FMLN and the URNG support the territory hypothesis. The FMLN maintained a strong presence throughout most of El Salvador while the URNG was relatively isolated to a small number of departments at the conclusion to its civil war.
RENAMO RENAMO’s territorial presence and its success as a major political party support the territory hypothesis. During the course of the war in Mozambique, RENAMO attained a military presence in each of the country’s ten provinces, the strongest being in the central provinces of Manica and Sofala (Hall and Young 1997: 119; Weinstein 2002). Several characterizations have been put forth concerning life within RENAMO-controlled zones. Geffray (1991) argues that many communities supported RENAMO because the insurgents offered them an opportunity to live life as they had prior to FRELIMO’s rise to power. Under FRELIMO rule, communities had been forced to give up traditional values and customs.91 Following the war, a number of RENAMO personnel also characterized life within RENAMO-controlled territory in a positive light, with new recruits surprised to find clothing, shoes, housing, and even access to transportation provided to the inhabitants and insurgents living within these zones (Manning 1998: 175). Hall and Young (1997: 177) also characterize the relationship between RENAMO and local leaders in a positive light. They argue that RENAMO would “obtain the support of the legitimate owners of the land (that chief’s ancestral spirits)” before the organization would establish a new base in the area and that they developed a relationship based upon mutual self- interest. Waterhouse (1996) offers a somewhat different perspective on support within RENAMO-controlled areas. He argues that, in the central provinces, “RENAMO’s operations came to resemble those of the ancient warlords, exacting tribute from local communities in return for giving them ‘protection’” (1996: 13).
91 Geffray’s (1991) argument is summarized by Manning (1998).
199 RENAMO’s experience controlling and managing territory appears to have helped it become a successful political party. During the war, RENAMO established a presence in each of the country’s ten provinces, with the central provinces remaining its greatest source of support. RENAMO’s success in bringing the war to each of the country’s ten provinces resulted in its having “representatives at local levels in much of the country who have at least a minimal notion of the party’s political views, as well as some training in political mobilization” (Manning 1998: 188). In the postwar electoral period, FRELIMO and RENAMO continued to perform well electorally in those areas of the country in which they had shown the most support during the war. RENAMO polled strongly in the country’s central provinces and parts of the north and FRELIMO polled very strong in the country’s southern provinces and also in parts of the north (Weinstein 2002: 150-151). For example, RENAMO won seventy percent of the seats in Manica and eighty-six percent of the seats in Sofala, two of their strongholds during the war, in elections for the Assembly of the Republic in 1994.
M-19 Unlike other Colombian guerrilla groups, the M-19 focused overwhelmingly on the urban areas of the country such as Bogotá, Cali, and Baranquilla. At the departmental level, the M-19 established a strong presence in the rural areas of Cauca, Cauca Valley, and Caqueta (Eisenstadt and Garcia 1995: 271).92 While the M-19 maintained a strong presence in three departments, as well the capital district of Bogotá, Colombia is a very large country. It is comprised of thirty- two departments and the capital district. Therefore, the M-19 had a strong military presence in less than ten percent of the country’s departments and is considered to have had little nationwide military presence.
Discussion The experiences of RENAMO and the AD M-19 strongly support the territory hypothesis. RENAMO, as a major political party, occupied half of Mozambique and effectively transformed the country into RENAMO-controlled zones and FRELIMO-controlled zones. The AD M-19, a minor political party, had a military presence in only about ten percent of the
92 According to Radu & Tismaneanu (1990: 170), at one time the M-19 was operating in more than one dozen departments, but this seems to be a minority view.
200 country. Overall, the evidence from the four cases also supports the territory hypothesis. The FMLN in El Salvador and RENAMO in Mozambique both had a strong military presence in more than half of the territory in each country. The M-19 in Colombia and the URNG in Guatemala were present in less than one-fifth the national territory. The insurgent groups that controlled larger areas of territory during the war, succeeded as major political parties while those groups that had maintained a minimal territorial presence were only successful as minor political parties.
Violence Hypothesis 6 states that insurgent groups that are less violent towards civilians during the war than are government forces are more likely to be successful than those insurgent groups whose civilian-directed violence outweighs that of the government. In both El Salvador and Guatemala, government security forces committed the majority of human rights violations against civilian populations, yet the FMLN transformed itself into a major political party and the URNG transformed itself into a minor political party.
RENAMO Mozambique’s civil war was a particularly violent and devastating conflict that cost the lives of roughly one million Mozambicans. These deaths occurred directly as a result of the war, as well as indirectly through hunger and disease. Two million Mozambicans fled the country and an additional three to four million were internally displaced. All this destruction occurred in a country of roughly seventeen million people. While both sides were known to have committed violent acts against the civilian population and infrastructure, RENAMO appears to have been overwhelmingly responsible, leading to its reputation for waging a “grotesque campaign of terror against Mozambican civilians, for its status as an army of captives and for its apparent lack of political program” (Manning 1998: 161). In the early years of the war, RENAMO deliberately targeted schools, teachers, and health care professionals (Waterhouse 1996: 50; Hall and Young 1997: 128-129; Weinstein 2002: 145). Schools and clinics were burnt to the ground and the workers were often hacked to death. The surviving school children were often forced into the ranks of RENAMO contributing to its reputation as an army of child soldiers. According to Waterhouse (1996: 12), “captured
201 RENAMO documents and its own members testified that RENAMO saw civilians as legitimate targets. In the early years, its tactics included slicing off the noses, ears, and sexual organs of victims. Civilian accounts tell of RENAMO forcing people to watch their relatives murdered; of boys forced to kill parents; girls raped; children boiled alive.” RENAMO also targeted the nation’s transportation infrastructure including the railways and roads that connected it to neighboring countries. In 1988, Robert Gersony prepared a report for the U.S. State Department issued a report based upon firsthand refugee accounts of the violence. In the report, refugees recounted tales of deliberate RENAMO attacks against civilians and efforts to forcibly extract food and labor from the rural population. In summary, the “Gersony Report” found that RENAMO had been directly responsible for the deaths of at least 100,000 Mozambicans.93 Like RENAMO, FRELIMO forces resorted to forced recruitment, including that of child soldiers.94 There are also a number of examples where “FRELIMO retaliated strongly, sometimes with great brutality, against any communities suspected of supporting the guerrilla insurgents” (Waterhouse 1996: 13). As the civil war continued, RENAMO violence declined in “areas where the local Frelimo influence and presence had been eliminated and RENAMO was relatively well established” (Hall and Young 1997: 168-169). Even though RENAMO violence against the civilian population declined, it did not decline to the point where FRELIMO violence was greater. RENAMO was still generally regarded to have been much more violent against the civilian population and the country’s infrastructure than FRELIMO.
M-19 In Colombia, the armed conflict “led to extensive human rights violations on all sides, including political homicides and extrajudicial executions, forced disappearances, torture, and kidnapping. Between 1970 and 1997, over 400,000 human rights violations occurred in Colombia” (Hoskin & Murillo 1999: 48). Throughout the 1980s, the violence committed against civilian populations was relatively equally shared by a number of insurgent groups, the Colombian military, paramilitary groups, and drug cartels.
93 The Gersony Report is more formally known as the Summary of Mozambican Refugee Accounts of Principally Conflict-Related Experience in Mozambique (1988). 94 RENAMO was not the only participant resorting to forced recruitment, although FRELIMO called its recruitment efforts “conscription” and described those of RENAMO as “abduction” (Weinstein 2002: 145).
202 Individually, the M-19 was more generally known for the high profile nature of its military operations than the violence that accompanied it. In the 1970s, the M-19 portrayed itself as a type of Robin Hood insurgent group. It hijacked milk trucks, stole food, and distributed the stolen goods to the urban poor. The M-19 also kidnapped wealthy businessmen and exchanged them for ransom. As mentioned earlier, the M-19 occupied the Dominican Republic’s Embassy and held hostage over one dozen foreign ambassadors. In return for their release, the M-19 received a $1-2 million payoff and other concessions. While the M-19 was generally not considered a very violent insurgent group, one very violent incident came to define the organization and undermine much of its public support. In November 1985, the M-19 attacked the Palace of Justice and took over one hundred people hostage, including most of the country’s Supreme Court Justices. According to the M-19, the goal was to stage a public trial to highlight the government’s failure to address political and social reforms needed to end the war. During the event, the Colombian military stormed the building, killing most of the M-19 and civilians inside, including most of the country’s Supreme Court Justices. While the M-19 blames the deaths on the unwillingness of the government and Colombian military to negotiate, many blamed the entire situation on the M-19. They had broken the peace and their actions had led to the deaths of many innocent Colombians. The M- 19 would never again reach the level of public support that they had attained prior to the Palace of Justice incident.
Discussion The experiences of RENAMO and the M-19 do not support the violence hypothesis. RENAMO was considered a very ruthless insurgent organization, one that had committed the majority of human rights violations in Mozambique’s civil war, yet it has become a major political party. While every group in Colombia was responsible for attacks upon the civilian population, the M-19 did not have as violent a reputation as the other groups. In the end, the M- 19 transformed itself into a minor political party. In terms of the violence committed against the civilian population by all four, there is a great deal of variation in the histories of the insurgent groups and their electoral success. The FMLN and URNG were both responsible for fewer human rights violations against civilian populations than their respective governments, yet the FMLN went on to become a major political party and the URNG struggled as a minor political
203 party. Taken together, then, the experiences of these groups do not support the violence hypothesis.
Electoral Objective Hypothesis 7 states that former insurgent groups are more likely to succeed in parliamentary than presidential systems. Both the FMLN and URNG compete in presidential systems.
RENAMO Mozambique is a semi-presidential system. The president is directly elected by popular vote to serve a five-year term. The prime minister and the cabinet are selected by the president. Since the advent of competitive elections, the same political party (FRELIMO) has occupied the presidency and position of prime minister. The president also “appoints provincial governors, who in turn control appointments to every other administrative post, right down to the district level” (Weinstein 2002: 152). This gives the president tremendous power. For example, while RENAMO won a majority of the vote in five of Mozambique’s ten provinces in 1994, the FRELIMO president selected FRELIMO governors to rule in those provinces.
M-19 Colombia is a presidential system where the president is elected by popular vote to serve a four-year term without the possibility of reelection.95 During the National Front years, the presidency alternated between the Liberal and Conservative parties. In 1974, the requirement that the presidency must alternate between candidates from the Liberal and Conservative parties expired.
Discussion The FMLN, URNG, and AD M-19 compete in strong presidential systems. RENAMO competes in a semi-presidential system. Although Mozambique is semi-presidential, both positions have historically been controlled by FRELIMO. These systems are generally considered less favorable to the emergence and success of new political parties than
95 In Colombia, President Alvaro Uribe is currently challenging the constitutionality of this law.
204 parliamentary systems. Given the varied performance of the four new political parties, the electoral objective does not appear to have a strong impact on determining the relative success of these new political parties. Former insurgent groups have experienced both minor and major party success in these presidential-dominant systems.
Presidential Runoff Hypothesis 8 states that former insurgent groups are more likely to succeed in systems with majority presidential runoffs than simple plurality. Both the FMLN in El Salvador and the URNG in Guatemala compete under the more favorable majority runoff systems, though this factor does not seem to have played a role in explaining their varied electoral performances.
RENAMO Mozambican electoral rules require that the president attain greater than fifty percent of the national vote in order to win a first round victory. If no candidate surpasses fifty percent, a second round is scheduled involving the two leading candidates. In both 1994 and 1999, the electoral contest was decided in the first round. FRELIMO secured first round victories with fifty-three and fifty-two percent of the vote to RENAMO’s thirty-four and forty-eight percent. Although RENAMO challenged the results of the second election, the requirement of a runoff election appears neither to have helped nor hindered RENAMO’s performance.
M-19 Following the expiration in 1974 of the National Front’s requirement that the executive rotate between the Liberal and Conservative parties until the approval of a new constitution in 1991, the Colombian president was elected by a simple plurality. The 1991 constitution amended the electoral law to require that the president be elected with a majority of the vote. If no candidate surpasses fifty percent of the vote in the first round, a second round is held involving the two leading candidates. In 1990, César Gaviria won the presidency with 48% of the vote. The AD M-19 finished a strong third with more than 12% of the vote. Unfortunately for the AD M-19, these elections were determined by a simple plurality leaving them with no impact on a second round. Following the constitutional change in 1990, Colombia has needed a second round election to determine the presidency in 1994.
205
Discussion All four insurgent groups have competed in presidential systems that employ a majority runoff formula for electing the president. This electoral rule is generally considered more favorable to the emergence and success of new political parties. Therefore, none of the insurgent groups should have been inhibited by these rules. Their varied performances as political parties suggests that having a majority runoff formula instead of simple plurality is not a major factor in explaining the difference between the major (FMLN and RENAMO) and minor (URNG and AD M-19) parties.
Legislative Rules Hypothesis 9 states that former insurgent groups are more likely to succeed under proportional representation electoral rules than under single member district voting. Both the FMLN and the URNG compete under the more favorable proportional representation rules for electing members of the legislature. In both countries’ municipal elections conducted under single member district voting, the parties performed significantly worse than they did in elections for the legislature using proportional representation. The experiences of the FMLN and URNG in proportional representation elections and single-member district elections support the legislative rules hypothesis.
RENAMO The Assembly of the Republic is unicameral and comprises 250 seats distributed between Mozambique’s ten provinces and the capital district.96 The 250 seats are distributed throughout the country’s ten provinces and the capital district based upon the total number of people registered to vote in each. Legislators are elected by popular vote for 5-year terms through proportional representation. The 1990 Constitution, promulgated by FRELIMO, called for the
96 The Mozambican constitution requires that two seats are distributed according to the voting preference of Mozambicans living abroad (one for those living in Africa and one for those living in other areas of the world), but the electoral commission was unprepared for the logistics of carrying out overseas voting in 1994 and 1999. Therefore, all 250 seats were distributed based upon voting conducted within the country’s ten provinces and capital district.
206 election of parliamentarians through single-member-district (SMD) voting, but this was changed to proportional representation through the peace negotiations.97
M-19 The Colombian legislature is bicameral and includes the Senate (upper house) and the Chamber of Representatives (lower house). Since 1991, the Chamber of Representatives has allowed for up to 166 seats. 161 seats are elected from the country’s thirty-two departments and the capital district.98 Up to five additional seats are distributed for representatives of ethnic groups, political minorities, and Colombians residing abroad (Jones 1995). Following the constitutional reforms, 100 members are elected to the Senate through a single nationwide district. Two additional seats are set aside for the country’s indigenous population (Jones 1995).99
Discussion Each former insurgent group has competed under relatively favorable electoral rules in elections to the legislature. Colombia, El Salvador, and Guatemala employed proportional representation prior to the incorporation of the insurgents as new political parties. In Mozambique, the electoral law established by FRELIMO in 1990 was changed from single member district voting to proportional representation as part of the negotiation process. Therefore, the electoral rules favored each of these new political parties.100
District Magnitude Hypothesis 10 states that former insurgent groups are more likely to succeed when the district magnitude is high and when there is a nationwide district. In the analyses on the FMLN and URNG, there was mixed support for the district magnitude hypothesis. The FMLN competed under a relatively more favorable average departmental district magnitude when compared to the URNG in each group’s first election, but not in each group’s second election. I
97 Voter tallies are translated into parliamentary seats using the D’Hondt method after excluding blank and invalid ballots, as well as those that belonged to parties that did not clear the five percent threshold. 98 Each department is guaranteed two seats. Departments receive an additional seat for every 250,000 inhabitants. 99 In both chambers, Colombia uses the Hare quota method to translate votes into seats. This method tends to favor small political parties. 100 In terms of the electoral formulas used to translate votes into seats, Mozambique and Guatemala use the D’Hondt formula while Colombia and El Salvador use the largest remainder-Hare quota formula.
207 also showed that the FMLN and URNG would have fared better had the seats set aside for the national district been distributed among the geographical districts.
RENAMO The average district magnitude for the Assembly of the Republic in 1994 and 1999 was a highly favorable 22.7. 250 seats for the Mozambican legislature are distributed between the country’s ten provinces and capital district.
M-19 Since passage of the 1991 Constitution, the district magnitude for the Chamber of Representatives has been 4.88 based upon 161 seats. Since 1991, the Colombian Senate has been elected from a single nationwide district with a district magnitude of 100.101
Discussion In terms of district magnitude, the experiences of RENAMO and the AD M-19 are consistent with the hypothesis. RENAMO (22.7), a major party, clearly has competed under a more favorable district magnitude compared to the AD M-19 (4.88), a minor party. 102 The FMLN (4.57) competed under a more favorable average district magnitude compared to the URNG in the 1999 election (3.96), but not the 2004 election (5.52).103 The FMLN (20) and URNG (22 and 31) were somewhat helped by the addition of a nationwide district, but, as discussed earlier, both parties would have gained more seats had the seats been redistributed at the department level rather than through an additional national district. When one looks at the experiences of all four political parties, there is no clear support for the hypothesis. While RENAMO clearly benefited from the existence of high district magnitudes, the AD M-19, the FMLN, and the URNG have all competed in systems with relatively equal district magnitudes. These results suggest that district magnitude is not a key explanation for the parties’ divergent performance.
101 The district magnitudes exclude the seats elected from Colombia’s minority populations. 102 The numbers in parentheses are average district magnitude. 103 The district magnitudes for El Salvador and Guatemala are based upon department level seats, while that for Colombia is based upon provincial seats for the Chamber of Representatives.
208 Electoral Threshold Hypothesis 11 states that former insurgent groups are more likely to succeed in systems with low electoral thresholds. To date, the existence of an electoral threshold has not affected the FMLN or the URNG. Both political parties have surpassed the percentage of the votes needed to remain a legal party in each system, though the URNG is close to losing its legal standing if it continues its downward trajectory.
RENAMO Mozambique uses a favorable proportional representation system in comparison to a single-member district voting system. The favorability of PR, though, is offset by a very high electoral threshold. Political parties must attain five percent of the national vote even if the party is only running candidates on a provincial list. The threshold was agreed upon by RENAMO and FRELIMO in order to make it extremely difficult for third parties to challenge them. Small political parties are discouraged from fielding their own candidates under these rules. In the 1994 and 1999 elections, only three parties cleared this five percent threshold – FRELIMO, RENAMO, and the UD. As mentioned previously, though, the UD vote is generally considered a fluke occurrence. No other political party has ever cleared three percent of the national vote. According to the Mozambique Peace Process Bulletin (Issue 22, April 1999), a study by Paul Fauvet of the Mozambique Information Agency found that the high electoral threshold only affected voting in one province, Nampula. After rerunning the 1994 electoral results without the five percent electoral threshold, Fauvet found that three small political parties would have gained additional seats, FRELIMO would have lost one, and RENAMO would have lost two seats. The electoral threshold has hindered other new political parties, but not RENAMO. RENAMO was sufficiently large to actually benefit from the high electoral threshold.
M-19 Colombia also uses proportional representation, though unlike Mozambique, it has a low threshold to entry. According to the 1991 Constitution, Colombian parties receive legal recognition if they win one seat in either chamber of congress or if one of its candidates receives at least 50,000 votes (Taylor 1995 cited in Boudon 2000). The low threshold to entry increases
209 the probability that a third party will develop into an effective electoral alternative. In the December 1990 Constituent Assembly election, representatives were chosen from a single nationwide district. The electoral rules for the Constituent Assembly provided few barriers that would prevent the emergence of new political parties and was instrumental in convincing the M- 19 to participate in the democratic political process (Shugart 1992).
Discussion The experiences of RENAMO and the AD M-19 do not support the electoral threshold hypothesis. In terms of an electoral threshold, new political parties in Mozambique confront much greater obstacles to success than their counterparts in Colombia. In order to win seats on a provincial list, RENAMO had to surpass five percent of the vote in every province. While this has not been a problem for RENAMO or FRELIMO, almost every other political party has struggled because of this requirement. In Colombia, the AD M-19 had to overcome minimal electoral requirements. The experiences of all four insurgent groups do not support the electoral threshold hypothesis either. While RENAMO has faced the greatest obstacles to success, it has become a major political party. In fact, in Mozambique, the existence of a high electoral threshold appears to have helped prevent other opposition parties from challenging RENAMO. The FMLN needed to surpass three percent of the national vote while the URNG needed to surpass four percent or win a seat at the department level. Finally, the AD M-19 simply needed to win a seat. RENAMO, a major political party, competed under the most difficult electoral threshold and the two minor political parties, the URNG and the AD M-19, competed under the most favorable electoral thresholds.
Abstention Hypothesis 12 states that former insurgent groups are more likely to succeed in systems with high levels of voter abstention. In El Salvador and Guatemala, both the FMLN and URNG entered electoral politics as voter turnout rates were in decline. Therefore, both political parties were able to appeal to large groups of potential voters who lacked confidence in the existing political parties. The experiences of the FMLN and URNG support the abstention hypothesis.
210 RENAMO Mozambique had no prior history of democratic elections prior to the 1994 elections. Therefore, the history of voter turnout and abstention had no effect on the performance of the new political party, RENAMO.
M-19 For most of Colombia’s history, the Liberal and Conservative parties have dominated party politics in the country. Since the formation of the National Front in 1958, the Liberal Party and Conservative Party have alternated political power in Colombia. The National Front provided that the two parties would alternate control of the presidency, share all administrative, judicial, and legislative posts equally, and prevent new political parties from upsetting the balance of power.104 Voter turnout declined during the years of the National Front and appears to have leveled off at just below fifty percent in the 1970s and 1980s (Archer 1995: 193). The AD M-19 was entering the electoral process at a point in time when voters were beginning to tune out the traditional political parties.
Discussion In many ways, both RENAMO and the AD M-19 entered electoral politics at opportune moments in each country’s history. RENAMO participated in the country’s first election, before any other opposition party was able to establish itself as a viable alternative to FRELIMO. In Colombia, while the National Front regime began to break down in the 1970s, the Liberal and Conservative parties still tended to dominate in the 1980s. At the end of the decade, it did appear as if the country was ready to accept a new electoral alternative. The AD M-19 began to compete when voter loyalties to the two main political parties were declining and voter abstention was on the rise. The AD M-19, FMLN, and URNG all began to compete as political parties after voters had had the opportunity to evaluate existing electoral alternatives over a number of elections. In all three countries, voter turnout rates were below fifty percent and declining prior to the incorporation of the insurgents, suggesting that the public was losing confidence in the existing political parties. RENAMO, on the other hand, competed in Mozambique’s first ever election
104 Alternation of the executive ended in 1974 and the sharing of all bureaucratic posts in 1978.
211 where it was not clear that voters had any party loyalties. All four political parties competed under relatively favorable conditions as it related to voter abstention rates in each country. Therefore, voter abstention provides little leverage over explaining their varied electoral performances.
Bipolarization Hypothesis 13 states that former insurgent groups are more likely to succeed in a presidential system when bipolarization has not yet set in. The URNG competed in a system with a much weaker level of bipolarization compared to the FMLN. Given that the FMLN went on to become a major political party and the URNG a minor political party, there is no support for the bipolarization hypothesis.
RENAMO Mozambique had no prior history of democratic elections prior to the 1994 elections. Mozambique was a one-party state under FRELIMO since the country’s independence in 1975. Therefore, the level of bipolarization had no effect on the performance of the new political party, RENAMO. On the other hand, RENAMO benefited from competing in the country’s first multiparty election. While FRELIMO was clearly a likely major political party, no other political party or organization has established themselves during the war as the country’s main political opposition party.
M-19 In Colombia, the bipolar nature of the political system was frequently cited as a reason for the emergence and support for the country’s insurgent groups. As mentioned in the previous section, the National Front divided political power in the country between the Liberal Party and the Conservative Party. At the ballot box, these two parties captured over ninety percent of all congressional seats from 1978 to 1990 and over 85 percent of the presidential vote between 1974 and 1990 (Archer 1995: 191-194). During the second half of the National Front years, Colombians’ attitudes toward the two major parties were beginning to decline. Voter identification with the Liberal and Conservative parties declined to 70 percent by 1970 and to 60 percent by the late 1980s (Archer 1995: 178).
212 In the late 1980s, Colombian attitudes toward political parties were highly negative.105 By the time the M-19 had chosen to try its hand at electoral politics, Colombia’s “traditional parties were in disarray” and “their elite leadership increasingly marginalized” (Archer 1995: 164). Though bipolarization was still strong, the M-19 entered the political system just when it appeared voters were ready to support a third candidate.
Discussion In terms of bipolarization, RENAMO appears to have benefited much more so than the M-19. In Mozambique, RENAMO entered at a relatively favorable moment, before any other party could establish itself as a political alternative to FRELIMO. In Colombia, the AD M-19 was at a disadvantage because two major political parties dominated the Colombian political system for several decades prior to their incorporation. Given the experience of the four insurgent groups as political parties, bipolarization does not appear to be a very good predictor of new party success. In Mozambique, RENAMO had to compete in a system dominated by a single political party. FRELIMO, though, had never been challenged electorally. The other major party, the FMLN in El Salvador, had to compete where a relatively bipolar system comprised of the PDC and ARENA was breaking down. In many ways, this was similar to the experience of the AD M-19 in Colombia where bipolarization was weakening just at the point of its insertion into party politics. Finally, in Guatemala, the URNG did not have to confront a strong bipolar system. Most of the political parties that the URNG competed against in 1999 did not even exist ten years earlier. Given the level of bipolarization and performance of all four political parties, it does not appear as if bipolarization is a good explanation. In the strong bipolar systems of Colombia and El Salvador, the AD M-19 has become a minor political party and the FMLN has become a major political party. In the two systems without a strong bipolarization, Mozambique and Guatemala, RENAMO transformed itself into a major political party and the URNG transformed itself into a minor political party.
105 In terms of evaluating Colombian institutions, political parties finished eighth behind the media, Roman Catholic Church, armed forces, justice system, economic peak organizations, government and congress (Archer 1995: 195).
213 Economic Crisis Hypothesis 14 states that former insurgent groups are more likely to succeed when there is an economic crisis [that voters do not blame on the group]. The electoral experiences of the FMLN and URNG do not support the economic crisis hypothesis. In Guatemala, the URNG entered following several years of anemic economic growth, yet it could only become a minor political party. The FMLN entered electoral politics following a period of strong economic growth that accompanied the end of the war. The FMLN, however, went on to become a major political party in El Salvador.
RENAMO Upon independence, the socialist FRELIMO set out to modernize the health and education infrastructure of Mozambique. Initially, health care was “the jewel in FRELIMO’s socialist crown” (Waterhouse 1996: 54). In an effort to undercut FRELIMO’s advances, RENAMO targeted those successes of the government, making health clinics and schools attractive targets. According to government sources, “by 1982 RENAMO had destroyed 840 schools, twelve health clinics, twenty-four maternity clinics, 174 health posts, two centre for the handicapped and 900 shops, while kidnapping fifty-two foreign technicians and killing twelve” (Hall and Young 1997: 129). By the end of the war, RENAMO had destroyed or forced the government to abandon over 3,000 schools (Waterhouse 1996: 50). In addition to health clinics and schools, bridges, roads, and railway lines were all targets. At the end of its civil war, Mozambique had suffered years of war, economic mismanagement, and natural disasters. FRELIMO and RENAMO began negotiating in Rome when Mozambican “economy was in shambles, still suffering from nationalization and collectivism and unable to attract foreign investment” (Weinstein 2002: 149). Mozambique “was among the poorest countries in the world, with an average annual income per head of only $88.00, and two thirds of its population living in ‘absolute poverty’, according to the World Bank. It was the most indebted country in the world (relative to income), and the most aid- dependent” (Waterhouse 1996: 14). Mozambique had been poor at the start of the conflict and remained poor throughout the war years with production levels declining to half of what they were upon independence (Waterhouse 1996: 42). In addition, an early 1990s drought contributed to the deterioration of the economy. From the peace signing in Rome until
214 RENAMO’s first electoral competition, the Mozambican economy continued to suffer. Mozambique suffered negative growth rates of real gross domestic product per capita in 1992 (- 0.07), 1993 (-5.92), and 1994 (-1.04).106
M-19 While the civil war in Colombia dragged on through the 1980s, its economy was one of the best in the region. Colombia was the only Latin American country that did not experience a negative rate of growth during the 1980s (Bushnell 1993: 268). A report by the Inter-American Development Bank in 1988 confirmed that Colombia had the lowest rate of inflation in Latin America, one of the five highest per capita GDP growth rates, and the lowest foreign debt per capita for the previous five years (Kline 1996: 178). The Colombian economy continued to grow right up to the M-19’s incorporation with growth domestic product per capita increases of 1.72 (1988), 2.02 (1989), and 2.37 (1990).107
Discussion The experiences of RENAMO and the AD M-19 are consistent with the economic crisis hypothesis. Mozambique suffered negative rates of growth in the two years prior to its performance in the 1994 elections. RENAMO went on to become a major political party in Mozambique finishing second in presidential and legislative elections. In Colombia, the growth rate of real GDP per capita was positive in the two years prior to the AD M-19’s political competition. The AD M-19 then went on to become a minor political party in Colombia. The results of all four political parties and their relation to the economic crisis hypothesis are much more mixed. Mozambique had the weakest growth of all the countries while El Salvador had the strongest growth. Yet, RENAMO and the FMLN both performed well enough in the first postwar elections to become major political parties. Colombia experienced moderately strong growth and Guatemala experienced moderately weak growth in the two years prior to the first postwar election in each country. Yet, both the AD M-19 and the URNG both
106 Alan Heston, Robert Summers and Bettina Aten, Penn World Table Version 6.1, Center for International Comparisons at the University of Pennsylvania (CICUP), October 2002. 107 Alan Heston, Robert Summers and Bettina Aten, Penn World Table Version 6.1, Center for International Comparisons at the University of Pennsylvania (CICUP), October 2002.
215 performed as minor political parties. Therefore, there is no consistent relationship between major political party performance and poor economic conditions.
Social Cleavages Hypothesis 15 states that identity-based insurgent groups are more likely to be successful if they represent a significant percentage of the population. The conflict in Guatemala provided limited support for the identity-based hypothesis. URNG membership and popular support came primarily from the country’s Mayan Indians who comprised over 40% of the country’s population. However, the URNG captured no more than 12% in any single election on its way to becoming a minor political party. It is possible that the URNG would have failed as a political party if the indigenous population in the country did not comprise such a large percentage of the overall population, but it could not guarantee major party success. The conflict in El Salvador had no identity-based component.
RENAMO Upon independence in 1975, RENAMO inherited a country “deeply divided by indigenous tradition, colonial policies, and regional differences in its development. Within its boundaries, over 100 linguistic dialects were spoken by matrilineal clans in the north and patrilineal peoples in the south. Religious faiths ranged from Animist to Christian to Muslim, with combinations of all three. Many of its inhabitants had never even heard of a Mozambican nation” (Waterhouse 1996: 25). While the conflict exacerbated regional differences in the country, the conflict did not break down along regional lines (Manning 1998; Weinstein 2002). Manning (1998: 168) concludes that “there are certainly important rivalries between ethnic groups in the same region (as for example between Sena and Ndau in Sofala, or Macua and Erati in Nampula), but the most important source of division within Frelimo, and between Frelimo and its opposition, is that between the southern region of Mozambique, on the one hand, and the center and north, on the other.” As seen in the election results 1994, voting did break down along regional lines. FRELIMO performed well in the southern provinces, winning a wide majority of seats in the most southern regions of the country – Maputo City (17 out of 18), Maputo province (12 out of 13), Gaza (15 out of 16), and Inhambane (13 out of 18). RENAMO performed well in its
216 traditionally strong central provinces – Manica (9 out of 13) and Sofala (18 out of 21). There was a clear provincial pattern to voting for FRELIMO and RENAMO in 1994, but “no indication surfaced of ethnically motivated voting” (Turner et al. 1998: 162).
M-19 In elections for the 1991 Constituent Assembly, two indigenous candidates were selected. The representation of indigenous peoples was “an enormous surprise given their lack of resources or campaign experience, the miniscule size of the indigenous population, and its low level of voter registration” (Van Cott 2001: 47).108 Following the demobilization of the indigenous Quintín Lame insurgent group, it was provided with a seat at the assembly, but was unable to vote. In the Constituent Assembly, the AD M-19 worked closely with the indigenous representatives and even supported some of the territorial autonomy ambitions (Van Cott 2001: 48). While the AD M-19 worked closely with the indigenous representatives, the AD M-19 were not associated with indigenous groups or demands during its conflict or, for the most part, in the postwar period as the indigenous had their own groups through which they channeled demands if they so desired.
Discussion Each conflict in this study has broken down more along geographic cleavages than along ethnic, racial, or linguistic cleavages. None of the cases provide a good test of the social cleavage hypothesis. In El Salvador, there is no ethnic, social, or linguistic cleavage. In Colombia, the ethnic population is quite small and did not have a connection with the AD M-19. In Guatemala and Mozambique, there were large ethnic communities with which both were associated, but neither were indigenous-based insurgent groups.
Opposition Political Parties Hypothesis 16 states that insurgent groups are more likely to be successful if parties with similar programs do not already exist. In Guatemala and El Salvador, both the URNG and FMLN competed shortly after similar-thinking leftist groups had competed in party politics. Given that the FMLN became a major political party and the URNG became a minor political
108 According to the 1993 Colombian national census, indigenous people comprised 3.7% of the population.
217 party, the existence of parties with similar programs is not a good explanation for the varied performance of the two political parties.
RENAMO The first election in Mozambican history occurred in 1994, an election in which RENAMO participated. Therefore, RENAMO did not have to compete against other opposition parties that had gained political experience or popular support prior to the incorporation of RENAMO. The nonexistence of opposition political parties prior to the transition of RENAMO appears to have contributed to its successful performance in the postwar. RENAMO was the only political party with national recognition and national offices. Given that for a political party to win an assembly seat, it needed to win not only a certain percentage of the vote in a single province, but five percent of the national vote the smaller opposition parties were at a distinct advantage. As mentioned earlier, a Trust Fund was set up specifically to aid RENAMO’s transition to political party. ONUMOZ also provided technical and financial assistance to all other opposition political parties (Turner et al. 1998: 155). The Trust Fund for Assistance to Registered Parties in Mozambique set aside over $3 million to assist seventeen opposition parties that had not participated in the country’s civil war (Waterhouse 1996: 23; Turner et al. 1998: 157). It does not appear as if the effort to strengthen opposition political parties was remotely successful. Some international donors claim that many political parties that had received international funds were headed by “charlatans” who were uninterested in the long-term development of a political project (Waterhouse 1996: 23). It is possible that the trust fund undermined a third party alternative. Given that equal payments of the fund were given to each party “regardless of size or credibility,” parties did not see it in their self-interest to work together for the 1994 election (Turner et al. 1998: 174). As a result, only the UD surpassed the five percent electoral threshold needed to occupy seats in the assembly. In 1999, many small political parties entered into an electoral alliance with RENAMO to form the RENAMO Electoral Union. In return for supporting RENAMO’s presidential candidate, the leaders of sixteen small political parties were offered relatively safe positions on
218 RENAMO’s parliamentary lists almost guaranteeing themselves seats in the assembly (MPPB October 8, 1999).109
M-19 In 1986, the Colombian Communist Party and the Revolutionary Armed Forces of Colombia (FARC) established the leftist Unión Patriótica (UP), or Patriotic Union, to compete in the electoral process and present itself as a viable alternative to the existing two-party system. In the 1986 presidential elections, the UP presidential candidate, Jaime Pardo Leal, finished third with less than five percent of the vote. In the legislative elections, the UP garnered 4.4% of the vote (Shugart 1992: 134). This was the best performance for a leftist party in Colombian politics. During the campaign, the UP was a target of death squads during what is known as Colombia’s “dirty war” period. At least 1,500 members of the Patriotic Union were killed, including Pardo in October 1987. As the M-19 prepared to leave political violence behind, it made contacts with the UP to explore an alliance between the two political parties. A coalition of the left, led by Bernardo Jaramillo and Carlos Pizarro, might have proven a formidable challenge to the status quo, but within one year the leaders of both political parties would be dead (Dudley 2004: 156).110 Jaramillo’s successor in the UP, Diego Montana Cuellar, continued Jaramillo’s movement away from the Colombian Communist Party and towards the M-19. Following Montana’s break from the UP, many from the UP’s aperturista faction followed Montana into the new AD M-19.111 Some speculate that the AD M-19 might have benefited in the 1990 presidential and constituent assembly elections from protest votes against the assassinations of Jaramillo and Pizarro (Garcia 1990, May 22 1990, IPS-Inter Press Service; Dudley 2004: 165). The UP continued to participate in electoral politics for another decade, winning one Senate seat and two Chamber of Representative seats in 1991. Prior to the 1994 presidential vote, there was a new challenge for the AD M-19 from the ruling Liberal Party. The Liberal Party candidate, Ernesto Samper Pizano, “presented himself as a social democrat and… entrenched himself on the center-left, essentially co-opting many of the
109 In 2004, RENAMO and FRELIMO won all seats with 92% of the vote combined. Eighteen other political parties shared the remaining 8%; none of the attained more than 2% individually. 110 Prior to the 1990 presidential elections, three presidential candidates, including both Jaramillo of the UP and Pizarro of the M-19, were assassinated. 111 The aperturista faction wanted to moderate the UP and separate itself from the Colombian Communist Party. The UP would continue to compete in the Colombian political system until 2002.
219 planks that Navarro and the AD M-19 had proposed in 1993. He promised to revive the peace process, soften the apertura (opening), increase government spending on public education, develop a national health system, and create jobs” (Boudon 2001: 84). In the first round April elections, the AD M-19 received less than four percent of the vote. Because of the similarity in platforms, several former M-19 guerrillas and union members formed a new coalition, Alianza por Colombia, and came out in support of Samper in the second round, which he eventually won (Latin America Weekly Report May 12, 1994).
Discussion The experiences of RENAMO and the AD M-19 support the opposition party hypothesis. In Mozambique, RENAMO participated in the first election in the country’s history in 1994. No opposition political parties existed in Mozambique prior to the peace process and, therefore, RENAMO did not have to compete against other anti-FRELIMO parties that had built up an electoral history in the country. In Colombia, the AD M-19 was not the only leftist party to compete. Another leftist party associated with the country’s insurgent groups, the Patriotic Union, competed in the 1986 presidential and legislative elections. Some claim that the FARC’s decision to create the UP and compete in electoral politics had “stolen the stage” from the M-19 (Dudley 2004: 55). The M-19’s popularity would never again reach the heights that it had prior to the siege on the Palace of Justice and the reincorporation into society of the UP. In 1990, though, the AD M-19 and a breakaway faction from the UP attained more than 12% of the vote while capitalizing on the outrage over the assassinations of the leaders of both parties. In terms of the four groups in this study, there is limited support for the opposition parties’ hypothesis. While other leftist parties existed in El Salvador, Guatemala, and Colombia, none had to deal with a strong opposition party that shared a similar platform and had developed a loyal electoral following. These environments should have been favorable to the FMLN, the URNG, and the AD M-19. RENAMO, on the other hand, clearly benefited from the absence of any other anti-FRELIMO parties, as they were the only other political party with a national presence and name recognition.
220 DISCUSSION
In the previous two chapters, I found that organizational factors tended to be the most important explanations for the major party success of the FMLN in El Salvador and the minor party success of the URNG in Guatemala, as both competed under relatively favorable environmental conditions. In this chapter, I extended the discussion of former insurgent group transitions to political parties in Mozambique and Colombia. In the analysis of the RENAMO and the M-19, organizational characteristics were again more significant in explaining the divergent performance of each insurgent group as a political party. Following the war in Mozambique, RENAMO went on to become a major political party, finishing second in elections for the presidency and the legislature in 1994 and 1999. RENAMO counted on a much greater number of armed combatants, had dedicated resources into constructing political structures during the war, maintained relative unity, and established RENAMO-controlled zones throughout roughly half the country. For RENAMO, its inability to connect with mass-based movements and its overwhelming responsibility for violent committed against the civilian population do not appear to have seriously impeded its performance. In terms of the environmental conditions effecting RENAMO, most were favorable to RENAMO’s success. While it competed in an unfavorable presidential system relative to a new party that competes in a parliamentary system, most other factors were favorable to RENAMO. Rules for electing the presidency were done by majority runoff rather than simple plurality. Elections for the legislature were conducted under proportional representation. District magnitude was high. While the electoral threshold of five percent nationwide is a difficult obstacle to new political parties, it has actually helped RENAMO by preventing other opposition parties from making inroads in the political system. RENAMO also competed under favorable economic conditions as Mozambique’s economy was suffering from years of FRELIMO mismanagement and the difficulties of neoliberal economic reforms. Though there exist a number of ethnic groups in Mozambique, the war and postwar political parties was formed more along geographic lines. Finally, because RENAMO participated in the first election in the country’s history, strong bipolarization and the existence of other opposition parties were not problems.
221 In Colombia, the M-19 (or AD M-19 as it became as a political party), after showing great promise following the 1990 presidential and Constituent Assembly elections and the 1991 legislative elections, struggled as a minor party in the 1994 presidential and legislative elections. The AD M-19 counted on few armed combatants, was unable to build a serious political base, had unorganized ties with the country’s mass-based movements, and a limited geographic presence. The AD M-19 remained relatively unified, although there were some divisions in the postwar period. Instead, the AD M-19 seemed to be hurt more by the deaths of key leaders both during the war and in the postwar phase. While the AD M-19 was not a very violent organization overall, a great deal of its reputation emanated from its siege on the Palace of Justice where dozens of people died, including many of the nation’s Supreme Court justices. In terms of the environment in which the AD M-19 competed, it was not a determining factor in the performance of the new political party. In 1990, the AD M-19 competed in a presidential system with elections via simple plurality. Later that year, the AD M-19 won a significant share of seats in elections for a constituent assembly and played an influential role in rewriting the country’s constitution. Under the new system, the rules for electing the president were changed to a more favorable majority runoff. Members to the bicameral legislature were elected using proportional representation. The low average district magnitude for the Chamber of Representatives was unfavorable while that for the Senate was highly favorable. Colombia uses a low electoral threshold that facilitates new party emergence and success. Historically, the Liberal and Conservative parties dominated the Colombian political system, but their stranglehold on politics appeared to be breaking down in the 1980s, as voters appeared willing to support new political alternatives. A strong economy worked against the AD M-19. Strong identity-based cleavages were not salient factors in the AD M-19’s performance. Finally, there was no leftist political party with a strong presence in the party system that inhibited the AD M- 19. The AD M-19 benefited from some defections from the UP, but might also have been hurt by the cooptation of some of its platform by the Liberal Party prior to the 1994 elections.
222
Map 8.1 Mozambique, Perry-Castañeda Library Map Collection, 1995.
223
Map 8.2 Colombia, Perry Castañeda Library Map Collection, 2001.
224 TABLE 8.1 Postwar Presidential Elections in Mozambique (Votes and % of Total Votes)
Political Party 1994 1999
FRELIMO 2,633,740 (53.30%) 2,338,333 (52.29%)
RENAMO 1,666,695 (33.73%) 2,133,655 (47.71%)a
PADEMO 141,905 (2.87%)
UNAMO 120,708 (2.44%)
MONAMO-PMSD 115,442 (2.34%)
Unknown 58,848 (1.19%)
PIMO 51,070 (1.03%)
FUMO-PCD 37,767 (0.76%)
Independent 34,588 (0.70%)
SOL 32,036 (0.65%)
Independent 24,238 (0.49%)
PPPM 24,208 (0.49%)
Valid Votes 4,941,515 4,471,988
a RENAMO competed as the lead partner in the RENAMO-UE coalition in 1999. Source: http://africanelections.tripod.com/mz.html.
225 TABLE 8.2 Postwar Legislative Elections in Mozambique (Votes and % of Total Votes)
Political Party 1994 Political Party 1999
FRELIMO 2,115,793 (44.33%) FRELIMO 2,005,713 (48.54%)
RENAMO 1,803,506 (37.78%) RENAMO-UE 1,603,811 (38.81%)
UD 245,793 (5.15%) UD 61,122 (1.48%)
PT 26,961 (0.56%) PT 111,139 (2.69%)
SOL 79,622 (1.67%) SOL 83,440 (2.02%)
PIMO 58,590 (1.235) PIMO 29,456 (0.71%)
AP 93,031 (1.95%) PALMO 101,970 (2.47%)
FUMO-PCD 66,527 (1.39%) UMO 64,117 (1.55%)
PCN 60,635 (1.27%) PANDELIMO 33,247 (0.80%)
PACODE 52,446 (1.10%) PANAOC 24,527 (0.59%)
PPPM 50,793 (1.06%) PPLM 11,628 (0.28%)
PRD 48,030 (1.01%) PASOMO 2,153 (0.05%)
PADEMO 36,689 (0.77%)
UNAMO 34,809 (0.73%)
Valid Votes 4,773,225 Valid Votes 4,132,323
Source: http://africanelections.tripod.com/mz.html.
226 TABLE 8.3 Postwar Legislative Elections in Mozambique (Number of and % of Seats)
Political Party 1994 1999
FRELIMO 129 (51.6%) 133 (53.2%)
RENAMO 112 (44.8%) 117 (46.8%)
UD 9 (3.6%) -
Source: http://africanelections.tripod.com/mz.html.
227 TABLE 8.4 Postwar Presidential Elections in Colombia (Votes and % of Total Votes)
Political Party 1990 1994 (1st Round) 1994 (2nd Round)
PL 2,891,808 (47.81%) PL 2,623,210 (45.30%)
AD M-19 754,740 (12.48%) AD M-19 219,241 (3.79%)
PSC 735,374 (12.15%) PSC 2,604,771 (44.98%)
MUP 37,442 (0.61%) MUP 64,131 (1.11%)
MOE 7,429 (0.12%) MOE 22,923 (0.40%)
MSN 1,433,913 (23.71%) MCCN 55,190 (0.95%)
PNC 33,645 (0.55%) CN 34,437 (0.59%)
PN 9,468 (0.15%) MADN 29,246 (0.50%)
PST 9,173 (0.15%) JEGA 17,397 (0.30%)
PDC 8,148 (0.13%) CCC 11,704 (0.20%)
MAC 2,411 (0.03%) Protestemos 9,059 (0.16%)
ME88 996 (0.01%) FM 9,055 (0.16%)
MCI 6,020 (0.10%)
MNP 5,711 (0.10%)
SL 4,368 (0.08%)
OPPN 3,797 (0.07%)
CGT 3,319 (0.06%)
CNG 2,637 (0.05%)
Source: Political Database of the Americas. (1999) Colombia: Presidential Elections 1990. [Internet]. Georgetown University and the Organization of American States. 31 October 2001.
Source: Political Database of the Americas (1999) Colombia: Presidential Election 1994 First Round. [Internet]. Georgetown University and the Organization of American States. 24 April 2003.
228 TABLE 8.5 Postwar Legislative Elections for Colombian Senate (Number of and % of Seats)
Political Party 1990 Political Party 1991 Political Party 1994
PL 66 (57.89%) PL 56 (56%) PL 56 (56%)
PSC 38 (33.3%) PSC 9 (9%) PC 20 (20%)
MNC 1 (1%) UP 1 (1%) MNC 1 (1%)
UP-FN 1 (1%) MUM 1 (1%) MSN 2 (2%)
Coalition PC 1 (1%) MNC 1 (1%) MNCa 1 (1%) and other MNP 1 (1%) CCC 1 (1%) Coalition PL 1 (1%) and other PNC 1 (1%) MLC 1 (1%)
UP Alliance – 1 (1%) AD M-19 9 (9%) UC 1 (1%) PPG MSN 5 (5%) MNP 1 (1%) Coalition PL- 1 (1%) PC UC 1 (1%) CI 1 (1%)
Coalition PL 1 (1%) MUC 1 (1%) LIDER 1 (1%) and other LIDER 1 (1%) MUM 1 (1%) Coalition PL 1 (1%) and other MAIC 1 (1%) ANAPO 1 (1%)
Coalition PL 2 (1.75%) FP 1 (1%) PCC 1 (1%) and other CI 1 (1%) NFD/MURCO 1 (1%)
NFD 8 (8%) MCI 1 (1%)
LC 1 (1%) NFD 4 (4%)
RC 1 (1%) PC-FP 2 (2%)
E, T and C.S. 1 (1%)
BDR-MOIR 1 (1%)
AR 1 (1%)
Total Seats 114 100 100
Source: Stephen Taylor, personal correspondence.
229 TABLE 8.6 Postwar Legislative Elections for Colombian Chamber of Representatives (Number of and % of Seats)
Political Party 1990 Political Party 1991 Political Party 1994
PL 119 (59.8%) PL 87 (54%) PL 88 (54.66%)
PSC 62 (31.1%) PSC 25 (15.53%) PC 40 (24.84%)
MNC 3 (1.5%) UP 3 (1.86%) MUM 1 (0.62%)
UP 1 (0.5%) MUM 1 (0.62%) MNC 6 (3.73%)
AD M-19 1 (0.5%) MNC 6 (3.73%) MNP 1 (0.62%)
Coalition 1 1 (0.5%) AD M-19 13 (8.07%) MNC (2) 1 (0.62%)
Coalition 2 1 (0.5%) MSN 11 (6.83%) AD M-19 1 (0.62%)
Others 1 1 (0.5%) MUC 1 (0.62%) MSN 1 (0.62%)
Others 2 1 (0.5%) MNC 1 (0.62%) MFP 4 (2.48%)
Coalition 3 1 (0.5%) MUC (2) 1 (0.62%) MNC (3) 2 (1.24%)
Coalition 4 1 (0.5%) LIDER 1 (0.62%) CI 1 (0.62%)
Coalition 5 1 (0.5%) MAIC 1 (0.62%) MIR 1 (0.62%)
Coalition 6 1 (0.5%) FP 1 (0.62%) MURCO 1 (0.62%)
Coalition 7 1 (0.5%) Coalition AS 1 (0.62%) PL-UP 1 (0.62%)
Coalition 8 1 (0.5%) MRD 1 (0.62%) MCAU 1 (0.62%)
Others 4 1 (0.5%) IR 1 (0.62%) VP 2 (1.24%)
Coalition 9 1 (0.5%) MCI 2 (1.24%) CPC 1 (0.62%)
Alianza 1 (0.5%) MURCO 1 (0.62%) CDM 1 (0.62%) UP/MO CH 3 (1.86%) Ganas 1 (0.62%) NF 1 (0.62%) Ind 1 (0.62%) ARAC 1 (0.62%) CCDM 1 (0.62%) MPP 1 (0.62%) MH 1 (0.62%)
Total Seats 199 161 161
Source: Stephen Taylor, personal correspondence.
230 TABLE 8.7 Postwar Legislative Elections for Colombian Constituent Assembly (Number of and % of Seats)
Political Party 1990
PL 22 (31.43%)
PSC 7 (10%)
AD M-19 19 (27.14%)
MSN 11 (15.71%)
UC 2 (2.86%)
CP 2 (2.86%)
ASI 1 (1.43%)
ME 1 (1.43%)
ONIC 1 (1.43%)
NC 1 (1.43%)
PNP 1 (1.43%)
UC 2 (2.86%)
Total Seats 70
Source: Stephen Taylor, personal correspondence.
231 TABLE 8.8 Global Insurgent Group Electoral Performance
Country Opposition Group Election 1 Election 1 Election 2 Election 2 ENP Performance ENP Performance
Colombia AD M-19 3.03a Major Party 2.75b Minor
Colombia AD M-19 3.06c Major Party 2.79d Failed
Colombia AD M-19 4.08e Major Party - -
Mozambique RENAMO 2.20 Major Party 1.99 Major Party
Guatemala URNG 2.35 Minor Party 4.56 Minor Party
El Salvador FMLN 3.1 Major Party 4.1 Major Party
a These results are for the 1991 Chamber of Representatives election. b These results are for the 1994 Chamber of Representatives election. c These results are for the 1991 Senate elections. d These results are for the 1994 Senate elections. e These are the 1990 elections for the Constituent Assembly.
232 CHAPTER 9 CONCLUSION
Overview of the Cases and Findings While a number of insurgent groups have disarmed, demobilized, and transformed themselves into political parties, we know very little about what has transpired afterwards. In this dissertation, I asked why some former insurgent groups have been more successful than others as political parties. In many ways these insurgent groups confront similar challenges to those faced by all new political parties. They must devise a political platform, develop a nationwide organization, select candidates for office, and fundraise. On the other hand, these insurgent groups are not “new” groups. Many have long and controversial histories of having acted outside the law, employed terrorism, and relied upon illegal fundraising activities. I argued that certain insurgent groups are more likely to succeed as political parties as a result of a variety of organizational factors (size, previous experience with political parties, contacts with mass- based organizations, internal unity, control of territory, and minimal violence committed against civilians) and environmental conditions (electoral rules, other political parties, voters, economic crises, social cleavages). To test these hypotheses, I combined qualitative and quantitative analyses of two Central American insurgent groups currently undergoing such transitions: the Farabundo Martí National Liberation Front (FMLN) in El Salvador and the Guatemalan National Revolutionary Unit (URNG) in Guatemala. First, I conducted a structured, focused comparison of the FMLN and URNG and their transitions to political parties. For this part of my dissertation, I relied heavily upon data I collected from fieldwork in El Salvador and Guatemala. Second, I conducted a number of statistical analyses utilizing electoral results, political violence and socio-economic data to determine whether the explanations also hold for within nation analyses. Finally, I pursued the generalizability of the findings through a limited comparison of similar insurgent transitions in Mozambique and Colombia.
FMLN and URNG Through an analysis of the experiences of the FMLN and the URNG, my findings suggest that the FMLN drew upon certain experiences both prior to and during the war to
233 succeed as a political party. The URNG, on the other hand, was unable to count on these experiences, and therefore was less successful. The FMLN relied upon a much larger insurgent force, prior experience with electoral politics and an important political wing, and a stronger relationship with mass-based organizations to succeed as a major political party. The FMLN also developed a nationwide organization during the war that was relatively easy to transfer to a national political party organization. Finally, the FMLN was an attractive political alternative because it did not have a reputation for committing violence against civilians. The URNG’s minor party success was based upon a number of factors. The URNG had relatively few combatants. It had no prior experience with electoral politics and a weak political wing. The URNG maintained a tenuous relationship with the country’s mass-based organizations. The URNG suffered from defections during the war and throughout the postwar. The URNG was also geographically isolated at the end of the war, which made it difficult to compete nationally in the first postwar elections. Finally, while the URNG did not target civilians on a scale comparable to the military and state security forces, it drew a lot of criticism because it could not protect those it placed in danger. These characteristics of the URNG and its war-time performance seriously limited its electoral performance and explain its performance as a minor political party. Environmental conditions played a marginal role in the electoral performance of each group. Both political parties competed under relatively similar institutional conditions. Both the FMLN and URNG competed in presidential systems that employed majority runoff elections. In terms of elections for the legislature, both parties competed under favorable proportional representation electoral rules instead of single member district voting. Both political parties would have benefited to a certain extent by the eradication of a national district and distribution of those seats at the department level instead of the current two list system. Neither the FMLN nor the URNG was adversely affected by the existence of an electoral threshold. Both El Salvador and Guatemala had historically high levels of voter abstention that provided a pool of potential voters for each group. The URNG appears to have been at a slight advantage compared to the FMLN when it came to bipolarization. Bipolarization was not an issue in Guatemala, while in El Salvador the two largest political parties had attained over 80% of the national vote in the three presidential elections prior to the incorporation of the FMLN. Economic conditions were also more favorable to the URNG, as Guatemala had very low rates of economic growth in
234 the years immediately prior to its insertion into the political party system. El Salvador’s economy had grown very well in the early 1990s, particularly when it is compared to the anemic growth of the 1980s. Neither of the insurgent groups provided a good test of the social cleavages hypothesis. The URNG was comprised primarily of the country’s indigenous peoples, but, for the most part, the URNG was not an identity-based organization and had difficulty maintaining relations with the country’s organized and unorganized indigenous sectors. Less than 1% of the Salvadoran population is indigenous and there was no identity component to the FMLN. Finally, both political parties had to compete against leftist opposition political parties that had already participated in elections prior to their transition. The performance of these other opposition groups does not appear to have tremendously hindered either organization. Instead, the FMLN and URNG captured some of the support that had previously gone to the CD and FDNG. Overall, environmental conditions appear to have played a marginal role in explaining the difference between the two political parties. Environmental conditions were generally favorable to both political parties. If anything, environmental conditions were more favorable to the URNG. Instead, it was the FMLN’s organizational characteristics as an insurgent group that explains its superior performance relative to the URNG. It is also likely that the FMLN, because of how well it was prepared organizationally for the challenges of electoral politics, would have succeeded under a more challenging electoral environment. The URNG, on the other hand, might not have performed nearly as well had the environment in which it competed been less favorable.
Within Nation Analysis In terms of helping to explain the electoral performance of the FMLN in the 1994 Salvadoran elections, there is limited support for both organizational factors and environmental conditions. Previously, I argued that the FMLN’s success in capturing and controlling territory throughout El Salvador better prepared it for the challenges of electoral competition, relative to the experience of the URNG in Guatemala. Given this finding, I also expected that the FMLN would have performed better as a political party in the areas it controlled during the war than in the areas that it did not. The analysis of the relationship between the territory that the FMLN controlled during the war and its performance in the 1994 postwar election is mixed. ANOVA results support the hypothesized relationship, but when controlling for other factors using OLS,
235 FMLN-control is no longer statistically significant. FMLN-controlled zones support the FMLN no better than zones not controlled by the FMLN. The second main relationship tested in this chapter relates to the distribution of political violence throughout the country. I hypothesized that given the disproportionately violent response of the Salvadoran security forces against the civilian population, we should have seen FMLN support much greater in these ex-conflict zones. When analyzing municipal level data, conflict zones did tend to vote for the FMLN more so than non-conflict zones, even when controlling for other factors using OLS regression. In terms of the environmental variables, only opposition strength appears to be helpful in explaining the performance of the FMLN as a new political party. In those municipalities where the left performed well in 1991, prior to the FMLN’s participation in elections, the FMLN also did well in 1994. The other controls – voter abstention and bipolarization – did not have any systematic impact on FMLN performance.
Comparability How do the performances of other insurgent transitions to political parties compare to the FMLN and URNG? In Chapter 8, I extended the discussion of former insurgent group transitions to political parties to Mozambique and Colombia. Following the war in Mozambique, RENAMO became a major political party. It finished second in the 1994 and 1999 presidential and legislative elections. In Colombia, the M-19 showed great promise in the 1990 presidential and Constituent Assembly elections and the 1991 legislative elections. However, it was unable to capitalize on these strong performances in the 1994 election, and struggled as a minor party thereafter. What explains the relative performances of RENAMO and the M-19? RENAMO counted on a greater number of combatants, had dedicated resources into constructing political structures during the war, maintained relative unity, and established RENAMO-controlled zones throughout roughly half the country. For RENAMO, its inability to connect with mass-based movements and its overwhelming responsibility for violence committed against the civilian population do not appear to have seriously impeded its performance. The AD M-19 counted on few armed combatants, had difficulties maintaining a relationship with the country’s mass-based movements, and had a very limited geographic presence. While the AD M-19 remained relatively unified, there were several divisions in the postwar period that the small organization
236 struggled to overcome. Unlike RENAMO, the AD M-19 was not considered a very violent organization, which might have helped this small group attain minor success as a political party. In terms of the environmental conditions affecting RENAMO, most were favorable to its success. The electoral rules for electing the president were conducted by majority runoff rather than simple plurality. Elections for the legislature were conducted under proportional representation with high district magnitudes. A high electoral threshold of 5% nationwide actually helped RENAMO. It prevented other opposition parties from making inroads in the political system and challenging RENAMO as the country’s primary opposition party. RENAMO also competed under favorable economic conditions as the country was one of the poorest in the world. Mozambique was suffering from years of war, FRELIMO economic mismanagement, and disruptive neoliberal economic reforms. Though Mozambique is comprised of a number of ethnic groups, the war and postwar political parties divided the country more along geographic lines. Finally, because RENAMO participated in the first election in the country’s history, strong bipolarization and the existence of other opposition parties were not problems and were more favorable than anything else. The environment in which the AD M-19 competed does not appear to have been a determining factor in the performance of the new political party. In 1990, the AD M-19 finished third in the presidential elections, won a significant share of seats in elections for a constituent assembly and played an influential role in rewriting the country’s constitution. The AD M-19 competed for a legislature using proportional representation. A low average district magnitude for the lower chamber was unfavorable while a high district magnitude for the upper chamber was highly favorable. Colombia uses a low electoral threshold that facilitates new party emergence and success. Bipolarization was historically strong in Colombia, but the two major parties’ stranglehold on politics appeared to be breaking down in the 1980s. A strong economy worked against the AD M-19. Strong identity-based cleavages were not salient factors in the AD M-19’s performance. Finally, there was no leftist political party with a strong presence in the party system that inhibited the AD M-19, although we did see some efforts by the Liberal Party to co-opt some of the M-19’s leftist platform. There are several interesting results from the analyses in this dissertation of four insurgent groups that have made the transition to political parties. First, given very different organizational backgrounds, civil war characteristics, and electoral environments, all four groups
237 experienced minor party or major party success for at least two elections following their initial electoral competition. While the two largest insurgent groups, the FMLN and RENAMO, were very successful, a very small insurgent group, the M-19 with only 2.5-3.75 combatants per 100,000 Colombians, still attained major party success in the 1990 Constituent Assembly elections and minor party success thereafter. This should provide hope to other small insurgent groups in the world today contemplating disarming and competing in democratic elections – even though they may not be large in number, there is hope as a political party. Another interesting finding relates to the role of violence in civil wars. RENAMO was the most violent of the insurgent groups in this study when it came to violence perpetrated against the civilian population, yet it still managed to become a major political party in Mozambique. Finally, those insurgent groups that put the greatest effort into establishing and developing political wings (FMLN and RENAMO) were not only the most successful insurgent groups, but the most successful political parties. This leads one to believe that a political wing is not only important to attaining one’s wartime goals, but to the group’s future as a political party. Overall, environmental conditions do not appear to explain the difference between the two major political parties and the two minor political parties. FMLN and RENAMO would probably have done well under any electoral arrangements. The environmental conditions are more critical to the performances of smaller insurgent groups like the URNG and the AD M-19. Had the environments in which these groups competed been a little less favorable to new parties, especially Guatemala, their ability to persist as minor parties might have been tested much sooner.
Contributions The dissertation is an important advancement in the field of comparative politics for several reasons. First, it is one of the first works to focus on the development of the FMLN and URNG as political parties, both individually and in a comparative framework. Those who follow the FMLN in El Salvador tend to lament its failure to win the presidency while overlooking its success at almost every other level. On the other hand, those that follow the URNG tend to blame its electoral failure on its weakened condition at the end of the war while overlooking the fact that it attained roughly 10% in its first electoral contest, albeit as part of an alliance. While the FMLN has not won the presidency in El Salvador, it has successfully transformed itself into a
238 major political party. The URNG was also successful in its first electoral contest, though not nearly as successful as the FMLN. After thirty-six years in the jungle and little knowledge about the demands of constructing a viable political party, the URNG quickly became a minor political party, finishing third, in the 1999 presidential and legislative elections. Second, several criticisms have been leveled against the FMLN and URNG such as their inability to come to grips with the fall of the Soviet Union, the “discrediting” of Marxism, the holding on to power by “dinosaurs” from the country’s revolutionary past and their inability to practice internal democracy. If these two groups (and others) are criticized for the same shortcomings, what explains why one has been so successful while the other has not? I point to a number of important explanations that I believe explain their varied performance. Given that there are several other insurgent groups that have made similar transitions to political parties, this dissertation provides a valuable framework for more comparative studies of Central American groups as well as groups from other regions of the world. Third, the measure of individual party success will be useful to those who study political parties, particularly new political parties. The measure of new party success that I developed in Chapter 3 and employed in the analyses of the FMLN, URNG, RENAMO, and M-19 provide a more reliable measure of success that can better carry across political contexts. In the more established party systems, any new political party that captures a seat or a percentage of the national vote is likely to be considered successful; in less established party systems, this is unlikely to be the case. While there are likely to be many new political parties that attain a few seats or percentage of the vote in any election, we are also likely to witness new political parties winning 20% or greater of the vote. This measure will help us distinguish between new political parties that succeed as minor political parties from those that succeed as major political parties.
Future Extensions In the future, I plan to continue to explore the experiences of the FMLN and URNG as they continue to compete as political parties in El Salvador and Guatemala. This dissertation explores the electoral performance of these former insurgent groups, but it does not consider the substantive impact their performance has had on policy. While the FMLN in El Salvador has performed very well as a political party and the URNG has not, both countries are equally ruled by right-wing political parties in line with the Washington Consensus. It remains to be seen,
239 then, what significant impact the performances of the FMLN and URNG have had on policy. Has the FMLN been able to translate its electoral success into the implementation of the conditions of the peace accords, or policy more generally? Has the URNG, as a minor political party, been able to influence the policy process in Guatemala and the implementation of the conditions of the peace accords? I also intend to explore the relationships of these groups with other actors in their respective political systems in greater detail. In El Salvador, three new political parties have splintered from the FMLN in the post war (Democratic Party, the Renovating Movement, and the United Democratic Front). The first two failed as political parties, while the United Democratic Front is scheduled to participate in the March 2006 elections. In Guatemala, former insurgents from the URNG have competed in a number of political parties. The relationships between the insurgent groups and the political wings in each country deserve further analysis. In El Salvador, the FMLN’s success came at the electoral expense of the FDR/CD. In Guatemala, the URNG’s success came at the electoral expense of the FDNG. In this dissertation, I had originally hypothesized that the performance of these like-minded opposition parties would have negative consequences for the insurgent groups in their transitions to political parties. Instead, the transitions of the FMLN and URNG undermined the electoral viability of the more moderate leftist parties, the FDR/CD and the FDNG. I believe this is a very interesting avenue of research. First, is the destruction of like-minded opposition political parties caused by the transition of insurgent groups to political parties a more general occurrence than simply those in El Salvador and Guatemala? Second, why did the political wing fail rather than the former insurgent group? In Chapter 7, I explored how well the hypotheses used to explain success across insurgent groups helps to explain the performance of a single insurgent group in a single country. I explored the performance of the FMLN in the 1994 “elections of the century”. I found that areas most severely affected by the violence tended to support the FMLN more than the areas less affected. However, I did not run any comparable analyses of the URNG performance in Guatemala’s 1999 elections. It is quite possible that the impact of violence will be different for the URNG. In Chapter 5, I discussed how some Guatemalans blamed the guerrillas for the violence that the state committed against their communities. If this is indeed the case, support
240 for the URNG in the community where government massacres occurred might be lower than in those areas where government and guerrilla violence was not as intense. In addition to the questions directly related to the FMLN and URNG, there are several extensions of the dissertation to other insurgent transitions that I intend to pursue. The Central America region can be a tremendous investigatory lab for one interested in civil war resolution. Many students previously turned to the region in order to better understand the processes of revolutionary violence. Today, the region’s revolutionaries have mostly transformed themselves into political parties – the FSLN in Nicaragua, the PUD in Honduras, the FMLN and URNG – though they have done so in very different ways (Allison 2006). The FSLN was victorious in overthrowing the Somoza regime in 1979 and then winning the country’s presidential elections in 1984. Ex-Contras in Nicaragua have also tried the electoral route to power in the 1990s. Finally, the PUD in Honduras is comprised of several former guerrilla groups that merged to form a political party after they had been defeated by the country’s military and accepted a general amnesty. It is important that we continue to explore the similarities and differences in each group’s transformation into political party and identify how they have been able to overcome the challenges of new party politics. Finally, I also intend to expand the study of former insurgent groups globally to include South American, European, African and Asian cases that have negotiated settlements with their respective governments and entered the political process as political parties. I would also like to perhaps investigate the transition of all former insurgent groups to political parties, not simply those that did so after negotiated settlements. Some groups received amnesty or simply gave up fighting before participating in elections. Their insertion into the electoral process was not the result of negotiations between the government and the opposition group. I believe this dissertation and related projects will be useful in helping to understand the past, present and future of insurgent organizations. I believe the data and the papers derived from the project will be of use to scholars interested in political violence, civil war resolution, and the emergence and success of new political parties.
241 APPENDIX A INTERVIEWS
El Salvador
Name: Óscar Martínez Peñate Position: Political Scientist Organization: Universidad Francisco Gavidia Location: San Salvador, El Salvador Date: April 30, 2004 Duration: 5:30 P.M. – 6:30 P.M.
Name: Sonia Baires Position: Member of the RN and FMLN; currently teaches at the UCA Organization: UCA Location: San Salvador, El Salvador Date: April 30, 2004 Duration: 2:00 P.M. – 3:30 P.M.
Name: Ana Guadeloupe Martínez Position: ERP Organization: FMLN/Partido Democrata/ Location: San Salvador, El Salvador Date: May 1, 2004 Duration: 10:00 A.M. – 12:00 P.M.
Name: Héctor Cordova Position: Deputy in Legislative Assembly Organization: FMLN Location: San Salvador, El Salvador Date: May 3, 2004 Duration: 10:30 A.M. – 11:30 A.M.
Name: Facundo Guardado Position: FPL; Former FMLN commander; 1999 presidential candidate; co-founder of Renovating Movement Organization: Unknown Location: San Salvador, El Salvador Date: May 6, 2004 Duration: 7:30 A.M. – 9:30 A.M.
Name: Héctor Silva Position: CDU Deputy; CDU presidential candidate; former mayor of San Salvador for the FMLN-CD (1997-2003)
242 Organization: CDU Location: San Salvador, El Salvador Date: May 2004 Duration: 90 minutes
Name: Alvaro Artiga-Gonzalez Position: Political Science Organization: University of Central America Location: San Salvador, El Salvador Date: May 4, 2004 Duration: 1 hour
Name: Roberto Cañas Position: Member of RN Organization: Currently teaches at UES Location: San Salvador, El Salvador Date: May 5, 2004 Duration: 1 hr.
Name: David Holiday Position: Director, Organization: Citizens’ Participation and Governance Project, Creative Associates International, Inc. Location: San Salvador, El Salvador Date: March 24, 2004 Duration: 90 minutes
Guatemala
Name: Alberto Ramírez Recinos (aka Tino) Position: Director of Encuentro de Expresiones Organization: Former member of ORPA and URNG Location: Quetzaltenango, Guatemala Date: April 13, 2004 Duration: 2 hours
Name: Juan Francisco García Position: Currently Political Analyst; former member of FAR Organization: INCEP Location: Guatemala City, Guatemala Date: April 14, 2004 Duration: 1 hour
Name: Tania Irasema Palencia Prado Position: National Coordinator – Guatemala Program
243 Location: World Lutheran Federation Date: April 19, 2004 Duration: One hour
Name: Rodrigo Asturias (aka Gaspar Ilom) Position: Former commander of ORPA and URNG; 2003 presidential candidate Organization: URNG Location: Guatemala City, Guatemala (Party Headquarters) Date: April 22, 2004 Duration: 10:40 A.M. – 11:45 A.M.
244 REFERENCES
Agence France Presse (1992). “M-19 Protests Nomination of One of its Members as Health Minister.” July 5.
Alden, Chris (1995). “The UN and the Resolution of Conflict in Mozambique.” The Journal of Modern African Studies 33, 1: 103-128.
Alvarez, Francisco A. (1988). “Transition before the Transition: The Case of El Salvador.” Latin American Perspectives 15, 1 (Winter): 78-92.
Atlas, Pierre M. and Roy Licklider (1999). “Conflict among Former Allies after Civil War Settlement: Sudan, Zimbabwe, Chad, and Lebanon.” Journal of Peace Research, 36 (1): 35-54.
Balboa, Juan (2003). “En Guatemala el sistema politico está colapsado, asegura Nineth Montenegro.” La Jornada, México D.F., November 17.
Baloyra-Herp, Enrique A. (1983). “Reactionary Despotism in Central America.” Journal of Latin American Studies 15: 295-319.
Baloyra-Herp, Enrique A. (1995). “Elections, Civil War, and Transition in El Salvador, 1982- 1994: A Preliminary Evaluation.” In Elections and Democracy in Central America Revisited, edited by Mitchell A. Seligson and John A. Booth. NC: University of North Carolina Press, pp. 45-65.
Berrington, Hugh (1985). “New Parties in Britain: Why Some Live and Most Die.” International Political Science Review 6(4): 441-461.
Blondel, Jean (1968). “Party Systems and Patterns of Government in Western Democracies.” Canadian Journal of Political Science 1(2): 180-203.
Bosch, Brian J. (1999). The Salvadoran Officer Corps and the Final Offensive of 1981. NC: McFarland & Company, Inc. Publishers.
Boudon, Lawrence (2001). “Colombia’s M-19 Democratic Alliance: A Case Study in New-Party Self Destruction.” Latin American Perspectives 28: 73-92.
Boudon, Henry Lawrence (1997). New Party Persistence and Failure: A Comparative Analysis of Colombia’s M-19 Democratic Alliance and Venezuela’s Radical Cause. Unpublished Ph.D. Dissertation, University of Miami.
Boudon, Lawrence (2000). “Party System Deinstitutionalization: The 1997-98 Colombian Elections in Historical Perspective.” Journal of Interamerican Studies and World Affairs, Fall 2000,
245 Bracamonte, José Angel Moroni and David E. Spencer (1995). Strategy and Tactics of Salvadoran FMLN Guerrillas: Last Battle of the Cold War, Blueprint for Future Conflicts. CT: Praeger.
Brockett, Charles D. (2005). Political Movements and Violence in Central America. NY: Cambridge University Press.
Brooke, James (1990). “Colombian Rebels’ Arms Go, But Where’s Bolivar Sword?” New York Times, April 14, 1990 Section 1, Page 3, Column 3.
Brooke, James (1990). “Colombia Rebels Shun Arms and Win Votes.” New York Times, May 31, 1990, Section A, Page 15, Column 1.
Bushnell, David (1993). The Making of Modern Colombia. CA: University of California Press.
Byrne, Hugh (1996). El Salvador’s Civil War: A Study of Revolution. CO: Lynne Rienner Publishers.
Carrigan, Ana (1993). The Palace of Justice: A Colombian Tragedy. NY: Four Walls Eight Windows.
CEH (1999). Informe de la Comisión para el Esclarecimiento Histórico. Guatemala: UNOPS.
Censo de Población y IV de Vivienda, El Salvador 1992 (1992). El Salvador: Dirección General de Estadística y Censos de El Salvador
Coleman, Joseph (1990). “Colombian Rebels Turn Over Arms.” United Press International, March 8, 1990.
“Colombian Government Promotes Normalization Program with M-19.” The Xinhua General Overseas News Service, June 5, 1990.
Cox, Gary (1997). Making Votes Count: Strategic Coordination in the World’s Electoral Systems. NY: Cambridge University Press.
Downs, Anthony (1957). An Economic Theory of Democracy. NY: Harper & Row Publishers.
Duch, Raymond M. (2001). "A Developmental Model of Heterogeneous Economic Voting in New Democracies." American Political Science Review 98(4):895–910
Duch, Raymond M. and Harvey D. Palmer (2002). “Strategic Voting in Post-Communist Democracy?” British Journal of Political Science 32(1): 63-91.
Dudley, Steven (2004). Walking Ghosts: Murder and Guerrilla Politics in Colombia. NY: Routledge.
246 Dunkerley, James (1985). Central America: The Collapse of the Military System. In The Political Dilemmas of Military Regimes, eds. Christopher Clapham and George Philip. London: Croom Helm. 171-200.
Dunkerley, James (1988). Power in the Isthmus: A Political History of Modern Central America.. NY: Verso.
Duverger, Maurice (1954). Political Parties: Their Organization and Activity in the Modern State. NY: John Wiley & Sons, Inc.
Eisenstadt, XXX and XXX Garcia (1995).
Ellner, Steve (1988). Venezuela’s Movimiento al Socialismo: From Guerrilla Defeat to Innovative Politics. NC: Duke University Press.
Epstein, Leon D. (1967). Political Parties in Western Democracies. NY: Praeger.
Fauriol, Georges (1985). Latin American Insurgencies. Washington, D.C.: Georgetown University Center for Strategic and International Studies.
Fearon, James D. and David Laitin (2003). “Ethnicity, Insurgency, and Civil War”. American Political Science Review 97 (1): 75-90.
FMLN Official Website http://www.fmln.org.sv/.
Garber, Larry (1998). “Introduction.” In Postconflict Elections, Democratization & International Assistance, ed. Krishna Kumar. CO: Lynne Rienner Publishers, Inc.
Garcia, Maria Isabel (1990). “Ex-Guerrilla Presidential Candidate Expects Wide Support.” IPS- Inter Press Service May 22, 1990.
Goméz, Luz Inés (1996). “Colombia’s Movimiento 19 de Abril: the Ambiguities of Latin America’s Second Wave Revolutionaries.” 1996 Berkeley McNair Journal, Volume 4.
Goodwin, Jeff (2001). No Other Way Out: States and Revolutionary Movements, 1945-1991. NY: Cambridge University Press.
Gross, Liza (1995). Handbook of Leftist Guerrilla Groups in Latin America and the Caribbean. CO: Westview Press.
Hall, Margaret and Tom Young (1997). Confronting Leviathan: Mozambique Since Independence. OH: Ohio University Press.
Halloran, Richard (1987). Latin Guerrillas Joining Forces, U.S. Officers Say. New York Times March 3, p. A8.
247 Hammond, John L. (1998). Fighting to Learn: Popular Education and Guerrilla War in El Salvador. NJ: Rutgers University Press.
Harmel, Robert (1985). “On the Study of New Parties.” International Political Science Review 6(4): 403-418.
Harmel, Robert and John D. Robertson (1985). “On the Formation and Success of New Parties.” International Political Science Review 6(4): 501-23.
Hauss, Charles and David Rayside (1978). “The Development of New Parties in Western Democracies Since 1945.” In Political Parties: Development and Decay, edited by Louis Maisel and Joseph Cooper. CA: Sage Publications, 31-57.
Herman, Edward S. and Frank Brodhead (1984). Demonstration Elections: US-Staged Elections in the Dominican Republic, Vietnam, and El Salvador. MA: South End.
Heston, Alan, Robert Summers and Bettina Aten, 2002. Penn World Table Version 6.1, Center for International Comparisons at the University of Pennsylvania (CICUP), October.
Holiday, David (2000). “Guatemala’s Precarious Peace.” Current History (Feb): 78-84.
Hoskin, Gary & Gabriel Murillo (1999). “Can Colombia Cope?” Journal of Democracy 10, 1 (1999), pp. 36-50.
Hug, Simon (2001). Altering Party Systems: Strategic Behavior and the Emergence of New Political Parties in Western Democracies. MI: The University of Michigan Press.
Huntington, Samuel P. (1968). Political Order in Changing Societies. CT: Yale University Press. Jackman, Robert and Karin Volpert (1996). “Conditions Favouring Parties of the Extreme Right in Western Europe.” British Journal of Political Science 26: 501-526.
Janda, Kenneth, 1980. Political Parties: a Cross-National Survey. New York: The Free Press: a Division of Macmillan Publishing Co., Inc.
Jaramillo, Velia. 2003. “Guatemala: Los ‘Excommandantes’ se dividen en la política.” Proceso. México November 4, 2003.
Jonas, Susanne (1991). The Battle for Guatemala: Rebels, Death Squads, and U.S. Power. CO: Westview Press.
Jonas, Susanne (1995). “Electoral Problems and the Democratic Project in Guatemala.” In Elections and Democracy in Central America Revisited, edited by Mitchell A. Seligson and John A. Booth. NC: The University of North Carolina Press, pp. 25-44.
Jones, Mark P. (1994). “Presidential Election Laws and Multipartism in Latin America.” Political Research Quarterly 47: 41-57.
248
Kalyvas, Stathis N. (1999). “Wanton and Senseless? The Logic of Massacres in Algeria.” Rationality and Society 11: 3, 243-285.
Kalyvas, Stathis N. (2001). “’New’ and ‘Old’ Civil Wars: A Valid Distinction?” World Politics 54: 1, 99-118.
Kaminski, Marek M. (2002). “Do Parties Benefit from Electoral Manipulation? Electoral Laws and Heresthetics in Poland, 1989-93.” Journal of Theoretical Politics 14(3) 325-358.
Key, V.O., Jr. (1958). Politics, Parties, and Pressure Groups. NY: Thomas Y. Crowell Company.
Kim, Jae-On and Mahn-Geum Ohm (1992). “A Theory of Minor-Party Persistence: Election Rules, Social Cleavages, and the Number of Parties.” Social Forces 70 (3): 575-599.
Kitschelt, Herbert (1988). “Left-Libertarian Parties: Explaining Innovation in Competitive Party Systems.” World Politics 40(2): 194-234.
Kitschelt, Herbert (1989). The Logics of Party Formation: Ecological Politics in Belgium and West Germany. Ithaca and London: Cornell University Press.
Kline, Harvey F. (1996). “Colombia: The Attempt to Replace Violence with Democracy.” In Latin American Politics and Development, edited by Howard J. Wiarda and Harvey F. Kline. CO: Westview Press, pp. 173-199.
Kopecký, Petr (1995). “Developing Party Organizations in East-Central Europe.” Party Politics 1(4): 515-534.
Kumar, Krishna (1998). Postconflict Elections, Democratization & International Assistance. CO: Lynne Rienner Publishers, Inc.
Laakso, Markku and Rein Taagepera (1978). “Effective Number of Parties: A Measure with Application to West Europe.” Comparative Political Studies 12: 3-27.
LaFeber, Walter (1993). Inevitable Revolutions: The United States in Central America. Second Edition. NY: Norton.
Landau, Saul (1993). The Guerrilla Wars of Central America: Nicaragua, El Salvador, Guatemala. NY: St. Martin’s Press.
Latin America Weekly Report (1991). “Left in Disarray Over Recent Losses; Dissident Factions Threaten Splits in ADM-19 and UP.” December 5, p. 3.
Latin America Weekly Report (1994). “Samper’s New Allies.” May 12, p. 204.
249 Latin America Weekly Report (1991). “What Does Navarro Stand For; Interview with the Rising Star Who Led the M-19 Guerrilla.” March 28, p. 8.
Lawson, Kay (1980). Political Parties & Linkage: a Comparative Perspective, edited by Kay Lawson. NY: The Vail-Ballou Press.
Lawson, Kay and Peter H. Merkl (1988). When Parties Fail: Emerging Alternative Organizations. NJ: Princeton University Press.
Le Bot, Yvon (1995). La guerra en tierras mayas. México: Fondo de Cultura Económica.
Lewis, Paul (2000). Political Parties in Post-Communist Eastern Europe. NY: Routledge.
Lewis-Beck, Michael and Mary Stegmaeir (2000). “Economic Determinants of Electoral Outcomes.” Annual Review of Political Science 3: 183-219.
Licklider, Roy (1995). “The Consequences of Negotiated Settlements in Civil Wars, 1945- 1993.” American Political Science Review 89: 681-90.
Lijphart, Arend (1975). “The Comparable-Cases Strategy in Comparative Research.” Comparative Political Studies 8 (2): 158-177.
Lijphart, Arend (1984). Democracies: Patterns of Majoritarian and Consensus Government in Twenty-One Countries. CT: Yale University Press.
Lijphart, Arend (1994). Electoral Systems and Party Systems: A Study of Twenty-Seven Democracies, 1945-1990. Oxford: Oxford University Press.
Lijphart, Arend (1999). Patterns of Democracy: Government Forms and Performance in Thirty- Six Countries. CT: Yale University Press.
Linz, Juan J. (1990). “The Perils of Presidentialism.” Journal of Democracy 1: 51-69.
Lipset, Seymour Martin (1963). Political Man: The Social Basis of Politics. NY: Doubleday.
Lipset, Seymour Martin and Stein (1967). “Cleavage Structures, Party Systems, and Voter Alignments: An Introduction.” In Seymour M. Lipset and Stein Rokkan, eds., Party Systems and Voter Alignments: Cross National Perspectives, NY: Free Press, 1-64.
Luciak, Ilja A. (2001). After the Revolution: Gender and Democracy in El Salvador, Nicaragua, and Guatemala. MD: The Johns Hopkins University Press.
Lungo Uclés, Mario (1996). El Salvador in the Eighties: Counterinsurgency and Revolution. PA: Temple University Press.
250 Mainwaring, Scott and Timothy R. Scully (1995). Building Democratic Institutions: Party Systems in Latin America. CA: Stanford University Press.
Mainwaring, Scott and Timothy R. Scully (2003). Christian Democracy in Latin America: Electoral Competition and Regime Conflicts, edited by Scott Mainwaring and Timothy R. Scully. CA: Stanford University Press.
Mair, Peter (1997). Party System Change: Approaches and Interpretations. NY: Oxford University Press Inc.
Manning, Carrie (2004). “Armed Opposition Groups into Political Parties: Comparing Bosnia, Kosovo, and Mozambique.” Studies in Comparative International Development 39(1): 54-76.
Manning, Carrie (1998). “Constructing Opposition in Mozambique: Renamo as Political Party.” Journal of Southern African Studies 24(1): 161-189.
McCleary, Rachel M. (1997). “Guatemala’s Postwar Prospects.” Journal of Democracy 8 (2): 129-143.
McClintock, Cynthia (1998). Revolutionary Movements in Latin America: El Salvador’s FMLN and Peru’s Shining Path. Washington, D.C.: United States Institute of Peace Press.
McDonald, Ronald H. and J. Mark Ruhl (1989). Party Politics and Elections in Latin America. CO: Westview Press, Inc.
Michels, Robert (1962). Political Parties: A Sociological Study of the Oligarchical Tendencies of Modern Democracies. NY: The Free Press.
Middlebrook, Kevin J. (2000). Conservative Parties, the Right, and Democracy in Latin America, Edited by Kevin J. Middlebrook. MD: The Johns Hopkins University Press.
Ministerio de Economía, Dirección General de Estadística y Censos. Encuesta de Hogares de Propositos Multiples, 1998.
Montgomery, Tommie Sue (1982). Revolution in El Salvador: Origins and Evolution. CO: Westview Press.
Montgomery, Tommie Sue (1993).”Armed Struggle and Popular Resistance in El Salvador: The Struggle for Peace.” In Barry Carr and Steve Ellner (eds.), The Latin American Left: From the Fall of Allende to Perestroika. CO: Westview Press.
Montgomery, Tommie Sue (1995). Revolution in El Salvador: From Civil Strife to Civil Peace. Second Edition, CO: Westview Press.
Müller-Rommel, Ferdinand (1985). “The Greens in Western Europe: Similar but Different.” International Political Science Review 6(4): 483-499.
251
Offe, Claus (1992). “Capitalism by Democratic Design? Democratic Theory Facing the Triple Transition in East-Central Europe.” In Győrgy Lengyl, Claus Offe, and Jochen Tholen (eds.), Economic Institutions, Actors and Attitudes: East Central Europe in Transition. Budapest and Bremen, 11-22.
Paige, Jeffrey M. (1997). Coffee and Power: Revolution and the Rise of Democracy in Central America. MA: Harvard University Press.
Pearce, Jenny (1984). Promised Land: Peasant Rebellion in Chalatenango, El Salvador. London: Latin American Bureau.
Perera, Victor (1993). Unfinished Conquest: The Guatemalan Tragedy. CA: University of California Press.
Peters, B. Guy. 1998. Comparative Politics Theory and Methods. NY: New York University Press.
Poguntke, Thomas (1987). “New Politics and Party Systems.” West European Politics 10(1): 76- 88.
Powell, G. Bingham, Jr. (1982). Contemporary Democracies: Participation, Stability, and Violence. MA: Harvard University Press.
Radu, Michael and Vladimir Tismaneanu (1990). Latin American Revolutionaries: Groups, Goals, and Methods. NY: Pergamon-Brassey’s International Defense Publishers, Inc.
Rae, Douglas W. (1967). The political Consequences of Electoral Laws. CT: Yale University Press.
Reding, Andrew (1997). Democracy and Human Rights in Guatemala. World Policy Paper.
Reich, Gary (2004). “The Evolution of New Party Systems: Are Early Elections Exceptional?” Electoral Studies 23(2): 235-250.
Ribera, Ricardo (1996). Los Partidos Políticos en El Salvador entre 1979 y 1992: Evolución y Cambios. San Salvador: FLACSO Programa El Salvador.
Rochon, Thomas R. (1985). “Mobilizers and Challengers: Toward a Theory of New Party Success.” International Political Science Review 6(4): 419-439.
Rose, Richard and Thomas T. Mackie (1988). “Do Parties Persist or Fail? The Big Trade-off Facing Organizations.” In When Parties Fail: Emerging Alternative Organizations, edited by Kay Lawson and Peter H. Merkl. NJ: Princeton University Press, Pp. 533-558.
252 Ryan, Jeffrey J. (1994). “The Impact of Democratization on Revolutionary Movements.” Comparative Politics 27 (October): 22-44.
Sartori, Giovani (1976). Parties and Party Systems: A Framework for Analysis. Cambridge: Cambridge University Press.
Seligson, Mitchell A. and John A. Booth (1995). Elections and Democracy in Central America Revisited, edited by, Mitchell A. Seligson and John A. Booth. NC: UNC Press.
Seligson, Mitchell A. and Vincent McElhinny (1996). “Low-Intensity Warfare, High-Intensity Death: The Demographic Impact of the Wars in El Salvador and Nicaragua.” Canadian Journal of Latin American and Caribbean Studies, 21 (42): 211-241.
Sharpless, Richard E. “Colombia.” pp. 187-205 in Charles D. Ameringer (ed.), Political Parties of the Americas, 1980s to 1990s Canada, Latin America, and the West Indies. CT: Greenwood Press.
Shugart, Matthew Soberg (1992). “Guerrillas and Elections: An Institutionalist Perspective on the Costs of Conflict and Cooperation.” International Studies Quarterly 36: 121-52.
Shugart, Matthew Soberg and John M. Carey (1992). Presidents and Assemblies: Constitutional Design and Electoral Dynamics. NY: Cambridge University Press.
Shvetsova, Olga (2003). “Endogenous Selection of Institutions and Their Exogenous Effects.” Constitutional Political Economy 14(3): 191-212.
Sichar, Gonzalo (1999). Historia de los Partidos Políticos Guatemaltecos: Distintas siglas de (casi) una misma ideología. 2a Edición. Guatemala: Editorial Nojib’sa.
Sichar, Gonzalo (n.d.). “Izquierda en Guatemala: Del levantamiento a la pérdida del horizonte.” http://www.inisoc.org/sichar4.htm (September 25, 2004).
Sieder, Rachel (1998). Guatemala After the Peace Accords. UK: Institute of Latin American Studies, University of London.
Spence, Jack (2004). War and Peace in Central America: Comparing Transitions Toward Democracy and Social Equity in Guatemala, El Salvador, and Nicaragua. MA: Hemisphere Initiatives.
Spence, Jack, David Dye, and George Vickers (1994). El Salvador’s Elections of the Century: Results, Recommendations, Analysis. MA: Hemisphere Initiatives.
Spence, Jack and George Vickers (1994). Toward a Level Playing Field?: A Report on the Post- War Salvadoran Electoral Process. MA: Hemisphere Initiatives.
253 Spence, Jack, David R. Dye, Paul Worby, Carmen Rosa de Leon-Escribano, George Vickers and Mike Lanchin (1998). Promise and Reality: Implementation of the Guatemalan Peace Accords. MA: Hemisphere Initiatives.
Stanley, William and David Holiday (2002). “Broad Participation, Diffuse Responsibility: Peace Implementation in Guatemala.” In Ending Civil Wars: The Implementation of Peace Agreements, edited by Stephen John Stedman, Donald Rothchild, and Elizabeth M. Cousens. CO: Lynne Rienner Publishers. Pp. 421-462.
Stedman, Stephen John (1993). “The End of the Zimbabwean Civil War.” In Stopping the Killing: How Civil Wars End, edited by Roy Licklider. NY: New York University Press. Pp. 125-163.
Stedman, Stephen John (1997). “Spoiler Problems in Peace Processes.” International Security 22 (2): 5-53.
Taagepera, Rein and Matthew Shugart (1989). Seats and Votes: The Effects and Determinants of Electoral Systems. CT: Yale University Press.
Tarrow, Sidney (1994). Power in Movement: Social Movements, Collective Action and Politics. NY: Cambridge University Press.
Tavits, Margit (2004). “Party Systems in Making: The Emergence and Success of New Parties in Eastern Europe.” Paper presented at the Annual Meeting of the Midwest Political Science Association, Chicago, IL.
Tavits, Margit and Taavi Annus (2004). “Learning to Make Votes Count: The Role of Democratic Experience.” Paper Presented at the Annual Meeting of the Southern Political Science Association, New Orleans, January 8-10.
Taylor, Ruth and John Marshall (1996). “A Rocky Road to the Final Peace.” Report on Guatemala (Winter): 2-5, 14.
Tilly, Charles (1978). From Mobilization to Revolution. MA: Addison-Wesley Publishing Co.
Tilly, Charles. 2002. “Violence, Terror, and Politics as Usual.” Boston Review, Summer 2002.
Torres-Rivas, Edelberto, Secundino González M., Horacio Boneo, Fabrice Lehoucq, and David L. Wall (2001). Construyendo la democracia electoral en Guatemala. Guatemala: FLACSO.
Tucker, Joshua A. (2002). “The First Decade of Post-Communist Elections and Voting: What Have We Studied, and How Have We Studied It?” Annual Review of Political Science 5: 271- 304.
Turner, A.W. (1993). “Postauthoritarian Elections: Testing Expectations About First Elections.” Comparative Political Studies 26(3): 330-349.
254
Turner, J. Michael, Sue Nelson, and Kimberly Mahling-Clark. 1998. “Mozambique’s Vote for Democratic Governance.” In Postconflict Elections, Democratization & International Assistance, edited by Krishna Kumar. CO: Lynne Rienner. Pp.153-175.
URNG Official Website. www.urng.org.gt
U.S. Library of Congress
Van Cott, Donna Lee (2001). “Explaining Ethnic Autonomy Regimes in Latin America.” Studies in Comparative International Development 35, 4: 30-58.
Van Cott, Donna Lee (2003). “Institutional Change and Ethnic Parties in South America.” Latin American Politics and Society 45, 2: 1-39.
Vinegrad, Anne (1998). “From Guerrillas to Politicians: The Transition of the Guatemalan Revolutionary Movement in Historical and Comparative Perspective.” In Guatemala After the Peace Accords, edited by Rachel Sieder. UK: University of London; Institute of Latin American Studies, Pp. 207-27.
Waterhouse, Rachel (1996). Mozambique: Rising from the Ashes. UK: Oxfam.
Weinberg, Leonard (1991) "Turning to Terror: The Conditions under Which Political Parties Turn to Terrorist Activities" Comparative Politics 23 (4): 423-437..
Weinstein, Jeremy M. 2002. “Mozambique: A Fading U.N. Success Story.” Journal of Democracy 13(1): 141-156.
Wickham-Crowley, Timothy P. (1989). “Winners, Losers, and Also-Rans: Toward a Comparative Sociology of Latin American Guerrilla Movements.” Susan Eckstein, ed., Power and Protest: Latin American Social Movements. CA: University of California Press, pp. 132-181.
Wickham-Crowley, Timothy P. (1992). Guerrillas and Revolution in Latin America: A Comparative Study of Insurgents and Regimes Since 1956. NJ: Princeton Univ. Press.
Wittman, D (1973). “Parties as Utility Maximizers.” American Political Science Review 67: 490- 498.
Wittman, D (1983). “Candidate Motivation: A Synthesis of Alternative Theories.” American Political Science Review 77: 142-157.
Wittman, D (1990). “Spatial Strategies when Candidates have Polity Preferences.” In J. Elenow and M. Hinich, eds., Advances in the Spatial Theory of Voting. Cambridge: Cambridge University Press.
255 Wood, Elisabeth Jean (2000). Forging Democracy from Below: Insurgent Transitions in South Africa and El Salvador. NY: Cambridge University Press.
Wood, Elisabeth Jean (2003). Insurgent Collective Action and Civil War in El Salvador. NY: Cambridge University Press.
Zamora, Rubén (2002). “A Diez Años de Chapultepec: Un Balance del Proceso Salvadoreño.” In A cinco años de la firma de la paz en Guatemala: un balance crítico. Contributors Rubén Zamora, Carlos F. Chamorro, Francisco Leal B., Luis Pásara, Edelberto Torres-Rivas, Juan Alberto Fuentes K., and Raquel Zelaya. Guatemala: FLACSO, pp. 17-46.
Zamora, Rubén (2003). La izquierda partidaria Salvadoreña: entre la identidad y el poder. El Salvador: FLACSO Programa El Salvador.
256 BIOGRAPHICAL SKETCH
Michael E. Allison was born in Queens, New York on April 9, 1974. He graduated from Regis High School in New York City, New York in 1992 and from Fairfield University in Fairfield, Connecticut in 1996. In 1995, he studied abroad at the Universidad del Salvador in Buenos Aires, Argentina. He spent 1997 in El Salvador on a Fulbright Fellowship where he researched the country’s political party system, attended classes at the Universidad Centroamericana “José Simeón Cañas” (UCA) and the Universidad de El Salvador (UES), and taught English as a second language at the Centro de Intercambio y Solidaridad (CIS). In 1999, Michael enrolled in the doctoral program in political science at Florida State University. At FSU, his research and teaching largely focused upon the comparative politics and international relations of Latin America. In particular, he studied civil war and civil war resolution in Latin America and the rest of the developing world. Michael also worked extensively as a research assistant on Paul Hensel’s Issue Correlates of War (ICOW) project. He earned his Masters of Science degree in 2001. He was awarded a Clogg Scholarship from the Political Methodology section of the American Political Science Association to attend the Inter-university Consortium for Political and Social Research (ICPSR) summer program in quantitative research methods in 2001. Michael has conducted fieldwork in El Salvador and Guatemala and has traveled extensively throughout Central and South America. He co-authored an article with Damarys Canache and has a solo article forthcoming in Latin American Politics and Society. Michael and his wife, Marie, were married in Dedham, Massachusetts in August 2004. Michael begins teaching at the University of Scranton in Scranton, Pennsylvania in fall 2006.
257