Doctoral Dissertation Research in Political Science: Corruption and Electoral Accountability in Brazil

Trustees Of Boston University, Boston

Investigators

Abstract

This award satisfies Division B, Title V, Sec. 543 of the Consolidated and Further Continuing Appropriations Act of 2013 (P.L. 113-6, enacted on March 26, 2013). The project addresses corruption in emerging democracies, an issue critical to political stability in the world. It develops and tests a theory to understand what circumstances raise or lower the level of electoral accountability when candidates are suspected of corruption, focusing on the case of Brazil. Recent massive demonstrations in major cities against misuse of public funds illustrate the importance of improved knowledge about the conditions that mitigate corruption. Brazil is an emerging economic power of growing international influence and its political performance and stability are of great interest to the economic interests and national security of the United States. The intellectual merit of this research lies in its capacity to use different methodological approaches to test a theory on the circumstances in which voters are more likely to punish candidates accused of corruption. Two theoretical arguments are at the foundation of the project. First, electoral accountability increases when the distance between candidate ideologies is low. By contrast, when there is greater ideological polarization there should be lower levels of punishment to candidates accused of corruption, as voters will be more likely to choose a candidate based on their issue positions. Second, electoral accountability will be lower when incumbent candidates accused of corruption can claim credit for delivering material benefits. These hypotheses are tested with a multi-method design incorporating election data, online survey experiments, and interviews with candidates for municipal elections in Brazil. The project takes advantage of newly available data on candidates' public records of corruption to perform a large-N observational analysis. The online survey experiments with Brazilian eligible voters over 18 years old will use a 2 x 2 factorial design to assess treatment by treatment interactions; one treatment being existence (absence) of corruption allegations involving the candidate and the other treatment being high (low) degree of party polarization in the first experiment, and positive (negative) candidate records of social provision and public works while in office in the second experiment. The tested recruitment platform develops representative samples of voters in understudied countries at minimal cost. Interviews with local elites are used for the purpose of understanding causal linkages between the interventions and outcomes of theoretical interest. The broader impacts are related to the study's potential contributions to knowledge about democratization and development. The project contributes to an emerging research agenda on the circumstances in which voters hold their representatives accountable by focusing on the case of a new democracy in the developing world. Hence, the study will be of interest not only for scholars but also for the broader international community, including policy makers, foreign aid donors, non-governmental organizations and international organizations.