Democratizing urban Brazil : voters, reformers, and the pursuit of political accountability

This study tackles a question of longstanding and central interest to political scientists: why have democratic transitions in some developing societies provoked profound transformations in mass and elite political behaviors, while the adoption of competitive elections elsewhere has done little to enhance democratic accountability? Most recent scholarship focuses almost exclusively on formal institutional arrangements in trying to answer this question. In contrast, I contend that the deepening of democratic practices in developing countries depends as much on these societies’ distributions of socioeconomic and social capital resources as it does on the design of their macropolitical institutions. In settings where most individuals lack the resources to knowledgeably evaluate candidates for political office on the basis of policy proposals or incumbent voting records, voters largely cast their ballots for politicians who provide private goods and benefits. Drawing from rational choice and collective action theories, my central line of line of reasoning is that social resource contexts decisively influence the extent to which individuals obtain the basic political information needed to hold politicians accountable in elections. Voter sophistication, in turn, shapes formal and informal political behaviors, and in doing so influences the campaign and governance ix strategies of elected public officials. The pursuit of political accountability in developing settings thus hinges critically on the resources and political sophistication of voters. I test my argument by examining various elite and mass political behaviors in three mid-sized state capital cities that collectively represent a cross-section of Brazil’s urban population. By investigating multiple subnational cases from a single country, I am able to hold fully constant institutional factors so that noninstitutional variables that shape subnational democratization can be systematically examined. Three distinct modes of inquiry—in-depth case studies, medium-n statistical analyses of city-regional data, and large-n statistical modeling that incorporates individual-level data—all confirm my main theoretical expectation: despite sharing identical party, electoral, and governance systems, dramatically different patterns of subnational governance have emerged across and within Brazil’s socially-diverse urban centers since the military left power in the 1980s.