Corruption and Partisan Feelings: A Valence Perspective on Explaining Affective Polarization

Affective polarization - broadly defined as the mutual antipathy between different political camps in society - has become an increasingly troublesome phenomenon in contemporary politics, as it can threaten social cohesion and adherence to democratic norms. The bulk of the research on the drivers of affective polarization has focused either on policy disagreement or the strength and overlap of political and social identities. In this paper, we argue that corruption – which is broadly regarded as a valence issue in politics - is a hitherto overlooked factor that helps us gain a better understanding of the structural foundations of affective polarization. Corruption is a phenomenon that has been associated with various adverse outcomes, such as lower institutional trust and weaker economic performance. We contend that while corruption tends to decrease the overall confidence in political institutions (including parties), partisan motivated reasoning leads voters to ascribe most of the blame to their out-parties, whereas in-party (i.e. the party the person feels closest to) feelings are more immune to it. Consequently, higher levels of corruption should increase the gap between in- and out-party feelings, signifying greater affective polarization. As such, our theory combines the rational/political and the tribalist/identitarian approaches on the foundations of affective polarization. We test this hypothesis first at the system level, using the classic party like-dislike scale from the Comparative Study of Electoral Systems (CSES) to calculate affective polarization for ca 200 country-year cases. Controlling for the other variables that have been associated with affective polarization, we identify the effect of corruption on both the average in- and out-party ratings at the country level. We then move to the individual level, employing original survey data collected recently in four European countries (Estonia, Hungary, Italy, France). Survey items capturing feelings towards parties, ideological distance perceptions, partisanship, and a unique question that asked respondents to evaluate party-level corruption - i.e. how corrupt do they think are the individual parties in their respective country – allows us provide a more fine-grained analysis of the relationship between corruption and affective polarization. As a third and final step, we have designed a survey experiment to test whether being exposed to information about corruption regarding both in- and out-parties is linked to more negative affective evaluations for the out-party compared to the in-party. The experiment will be fielded in Italy in early 2025, using actual news stories about politicians from different parties being involved in corrupt behaviour. The experiment should help us tease out the causal mechanism behind the correlational relationship between corruption and affective polarization. Our findings will improve the existing knowledge on the adverse outcomes of corruption from a political behaviour/party politics perspective. Moreover, our study will contribute to the burgeoning literature on the determinants of affective polarization by improving our understanding of how the selective distribution of socio-economic resources shapes partisan feelings.