Looking the other way? Selective media exposure and the electoral punishment of corruption

From December 2019 to December 2020 Funding institution: “la Caixa” Foundation (Call for experimental research projects in the social sciences)

Team: Macarena Ares, Sofia Breitenstein, Enrique Hernández

Our project studies corruption voting by randomizing informational treatments about corruption, while also explicitly allowing some respondents to opt-in and out of these treatments. Introducing self-selection increases the realism of the experiment and allows us to address selective media exposure as an explanation for why corruption is not electorally punished.

Political corruption has multiple undesirable consequences for societies. Citizens are, therefore, expected to punish corrupt politicians at the polls. In line with this presumption, lab and survey experiments consistently show that citizens are less likely to vote for candidates that engage in corruption. However, at the same time, observational studies conducted in Spain and elsewhere conclude that corrupt politicians are frequently re-elected and just mildly punished by voters. This project aims to unravel this paradox by generating new evidence that helps us understand: why are corrupt politicians not strongly punished in elections?

Information acquisition is a prerequisite for the electoral punishment of corruption. If citizens are not exposed to information about corruption they cannot hold politicians accountable. However, pre-existing political attitudes (like one’s partisanship) affect whether and how individuals access and process political information. Studies have shown that engaged partisans self-select information that supports their preexisting beliefs. Therefore, an experimental design that randomly informs participants about corruption and disregards the fact that citizens tend to self-select information is highly unrealistic, might mask substantive heterogeneity in treatment effects, and might lead to an overestimation of the electoral consequences of information about corruption. Unlike previous studies, this project implements a novel Patient Preference Trial (PPT) experimental design that allows us to address the following research question: how does information about politicians’ engagement in corruption affect the likelihood of voting for corrupt politicians when accounting for information self-selection?