Professor Weitz-Shapiro's research seeks to understand variation in political responsiveness and accountability in middle-income democracies. She has three major recently completed or ongoing projects that explore questions of political accountability in Latin America through the lens of clientelism, corruption, and state audit institutions.
A. Curbing clientelism: politics, poverty, and social policy in Argentina
This project explores the phenomenon of clientelism—the individualized, contingent exchange of goods or services for political support. In recent years, much attention has been paid to cases like Brazil and Mexico where the top-down reform of government social policy has been credited with greatly reducing clientelism. But in a context where the national government makes no effort to curb this practice, might some local-level politicians nonetheless willingly abandon the use of clientelism in the distribution of goods and services?
To answer this question, Weitz-Shapiro argues for the importance of what she calls the “costs of clientelism”: although clientelism is an effective tool for getting votes from the poor, the practice simultaneously alienates non-poor voters, who see it as a signal of low quality governance and gain no direct benefits from clientelism. Thus, in a context of intense political competition, clientelism is valuable to politicians whose constituents are mostly poor, but its electoral costs grow as the size of the middle class increases. Weitz-Shapiro argues that while political competition alone is unlikely to create incentives for politicians to curb clientelism, the combination of high levels of political competition and a large middle class makes possible a subnational path away from clientelism and toward improved governance.
In “What wins votes: Why some politicians opt out of clientelism,” (AJPS 2012), Professor Weitz-Shapiro supports this theory through an examination of social policy implementation in a sample of Argentine municipalities. She further refines and tests this theory in Curbing clientelism in Argentina: politics, poverty, and social policy (Cambridge University Press, 2014). The book employs additional empirical tests to provide evidence for the costs of clientelism among non-poor voters and to examine how partisanship mediates the abandonment of clientelism.
B. Vertical accountability and control of corruption: the role of the public
While clientelism provides direct benefits to some citizens, corruption imposes costs on the vast majority of citizens while offering little or no benefits. Citizens in democracies at all levels of development report widespread dislike of corruption. Given this, what might explain the persistence of corruption among elected officials? Professor Weitz-Shapiro's second major project explores the phenomenon of political corruption, or the abuse of public office for private gain. While other researchers have emphasized institutional explanations for corruption, in this project, carried out jointly with Matthew S. Winters (University of Illinois at Urbana-Champaign), the authors focus on citizens themselves. In a democracy, if corrupt politicians are elected, it is because they have received the support of voters. Given popular distaste for corruption, why would this be the case, under what conditions are citizens most likely to actually punish corrupt politicians, and what are the broader implications of political corruption for the political system?
In a series of papers that draw on original experiments embedded in two nationally-representative surveys in Brazil (conducted in 2010 and 2013), Weitz-Shapiro and Winters argue and provide evidence that the extent to which citizens punish elected officials for corruption depends on two key factors: (1) the nature of information available about corruption, and (2) citizens’ abilities to process that information. They argue that citizens are most likely to act on specific, credible allegations of corruption, and that politically sophisticated citizens are the most sensitive to variation in information quality. In doing so, this project offers the most comprehensive theory and analysis to date of when and why citizens will act on their anti-corruption preferences. Through this project, they also contribute to the literature in political psychology by providing some of the very first evidence that the credibility of political information sources, isolated from citizen affinity for those sources, affects whether citizens incorporate new information. An extension in Argentina is currently underway: in that part of the project, Weitz-Shapiro and Winters will examine how citizen responses to corruption information change when incentives created by a source’s credibility and an individual's own political allegiances are in conflict.
C. Horizontal accountability and control of corruption: the role of government institutions
Guillermo O’Donnell famously distinguished between vertical accountability—the ability of citizens to hold the state to account through elections—and horizontal accountability, or the ability of the state to supervise itself. In this project, Professor Weitz-Shapiro emphasizes the connections between the two. Reliable information about government performance allows citizens to punish politicians for wrongdoing, and independent government oversight bodies are well positioned to produce and disseminate the information that voters need, yet in many cases they fail to perform this function. What explains why these oversight institutions sometimes perform the functions with which they are charged and at other times fail to do so?
In the first part of this project, Weitz-Shapiro focuses on the process by which the officials heading these agencies are appointed. She argues that variation in the intensity of political competition, party structure, and legislative career incentives shape the types of officials who are appointed to top positions on these courts, as well as their subsequent incentives. She will test these expectations using an original dataset on the identities of top-level officials in Argentina and Brazils’ subnational audit courts. To date, Weitz-Shapiro has completed a new, individual-level dataset on the 189 councilors who lead Brazil’s 27 state audit courts and she is currently collaborating with Transparencia Brasil to expand that dataset to include all councilors who have served on the Brazilian state audit courts at any time since the promulgation of Brazil’s democratic constitution in 1988. Initial results from the analysis of data on current councilors suggest, among other findings, that higher quality officials are more likely to be appointed during election years, and that constraints on executive branch discretion may create opportunities for legislative branch overreach. A working paper that presents these results (joint with Miriam Hinthorn and Camila Moraes) is available here: http://ssrn.com/abstract=2623788.
A second part of this project, currently in the planning stages, will focus on the incentives and performance of street-level bureaucrats within government oversight agencies. In collaboration with Ana de la O (Yale) and Lucas Gonzalez (UNSAM, Argentina), this part of the project will employ an experimental approach and be carried out in collaboration with Argentine provincial auditing agencies.