Biases in Corruption Suspicion: Exploring the Role of Affective Polarization

The corruption literature, perhaps especially in the area of gender and corruption, has investigated the role of biases, such as stereotypes, that express descriptive and injunctive norms relevant to how people assess the corrupt acts of others. But it has not so far been integrated with the literature on affective polarization, a newer literature that focuses specifically on biases that emerge among political camps, considering their tendencies to distrust (as well as to dislike) each other. This paper begins by discussing the opportunity of forging stronger links across these two literatures. It then reports the results of a conjoint experiment, fielded in late 2024, as part of a representative survey of the Brazilian population (N=3,000). The experiment aims to measure the extent to which different individual attributes randomised among conjoint profiles - including politics, but also gender, race, religion, and profession - predict respondents’ tendencies to suspect another person of giving or receiving a bribe. Results will be assessed on-average, as well as by respondent subgroup, considering both ingroup dynamics anticipated by the affective polarization literature, and more generalised biases across the population (for example, as much of the gender literature posits) with which the corruption literature has more commonly engaged. Individual-level tendencies for bias will also be analyzed in terms of how respondents score on several classic measures of affective polarization, including social distance measures, emotion ratings (e.g. feelings of disgust), and “feeling thermometers” to measure both ingroup-outgroup differences in liking as well as meta-polarization (that is, mis-estimation of what the other group thinks of your group).