New Left Party Success and Failure in Latin America
Why do strong parties take root in some contexts, but not in others? Political parties are the basic building blocks of representative democracy. They reduce information costs for voters, enhance executive accountability, and contribute to democratic governability by facilitating legislative organization and aggregating the interests of powerful societal groups. Yet parties remain weak in much of the world, with negative consequences for democracy. Where strong parties have failed to take root or collapsed and not been rebuilt, democracy has fallen into crisis (e.g., Peru in the 1990s, Venezuela in the 2000s). Where successful party-building has occurred, democracies have become consolidated (e.g., Ghana, Brazil, Taiwan).
Despite the importance of strong parties, our understanding of the conditions for successful party-building remains limited. The dominant theories of party and party system development arise from studies of successful party systems, especially those in North America and Western Europe during the 19th and early 20th centuries. Classic works such as Duverger (1954), Lipset and Rokkan (1967), Shefter (1994), and Aldrich (1995) ask how electoral institutions, social cleavages, and access to patronage shape emerging parties, leaving aside the more fundamental question of whether durable parties will emerge at all.
The focus on successful cases has handicapped more recent analyses of party development in new democracies. In third-wave Latin America, numerous left-wing parties and coalitions formed and became serious contenders for national power. While some of them survived and became institutionalized, most disappeared after one or a few elections. Scholars of the Latin American new left have focused almost exclusively on the survivors, not asking what distinguished Latin America's new left successes from their more numerous failed counterparts. To date, no systematic comparison of new left survivors and failures exists.
My dissertation addresses this problem. I present a theory of party-building based on a systematic comparison of Latin America's new left successes and failures. The dissertation argues that most new parties do not survive because they do not build strong (i.e., territorially implanted) organizations and have sources of cohesion. As a result, they do not withstand early crises. Paradoxically, organizationally strong, cohesive parties are more likely to form under conditions of adversity. Specifically, office-seekers with low access to state resources and mass media have an opportunity and incentive to do the difficult work of organization-building. Shared struggle in contexts of polarization or conflict strengthens internal group identity and disincentivizes elite defection, generating cohesion. New party-builders are more likely to experience this cluster of adverse conditions under authoritarian rule.
I demonstrate the theory through a nested analysis of Latin America's new left parties. The dissertation's empirical chapters situate four case studies in a broader, medium-n comparison. According to the dissertation's operationalization, fifteen new left parties became serious contenders for national power during the third wave in Latin America. Eleven of the fifteen were founded under democracy, and four under authoritarian rule. Of the eleven founded under democracy, seven failed. Of the four founded under authoritarian rule, three survived. Using evidence from fourteen months of interviews and archival research in Brazil, Mexico, Peru, and Argentina, I examine four of these new left cases in depth: two successes, Brazil's Workers' Party (PT) and Mexico's Party of the Democratic Revolution (PRD), and two failures, Argentina's Front for a Country in Solidarity (FREPASO) and Peru's United Left (IU). I supplement the four in-depth case studies with extensive secondary research on the full range of new left survivors and failures, focusing on four shadow cases: Colombia's M-19, El Salvador's FMLN, Uruguay's Broad Front (FA), and Venezuela's Radical Cause (LCR).