Affective Polarization and Punishing Democratic Violators: Experimental Evidence from Turkey

Politics in many countries are characterized by high levels of affective polarization, that is, citizens’ increasing levels of antipathy, distrust, and negative sentiments across partisan lines. Theoretically, scholars consider rising affective polarization as a potential driver of democratic backsliding as it politicizes liberal democratic norms and heightens partisan cue taking. These, in turn, might lead citizens to have lower tolerance for outparty groups and prioritize partisanship over democratic norms. Therefore, citizens become less likely to push back would-be autocrats from their own in-group and sometimes even actively support their undemocratic behavior. Despite these theoretical conjectures about how affective polarization could erode democracy, we lack empirical evidence to support them. In fact, some recent studies focusing on the case of the U.S. failed to register a causal association between affective polarization and democratic attitudes. This paper contributes to the affective polarization and democratic backsliding literature by presenting data from an original, nationally representative survey experiment fielded in Turkey. Turkey is an appropriate context to address this question, as the country has recently experienced both high levels of affective polarization and significant democratic backsliding. In addition, it is a multi-party system that better mirrors other political contexts around the world than the two-party system of the U.S. In our survey we manipulate respondents’ affective polarization and observe whether these manipulations lead to a change in respondents' propensity to punish political candidates who violate democratic norms. In particular, we first deploy interventions to reduce respondents’ affective polarization and then present them with a candidate choice conjoint experiment where some candidates make proposals that violate democratic principles such as electoral fairness, civil liberties, and separation of powers. In this way, we are able to ascertain whether respondents in depolarization treatment groups are more likely to punish democratic violators than respondents in the control group. Through this novel design, we find that the lowered affective polarization levels indeed lead to higher propensity to punish democratic violators. In particular, respondents in successful depolarization treatment groups give lower support to candidates who violate democratic norms, such as ignoring court decisions, restricting protest rights of outparty supporters, and manipulating electoral districts to gain more seats. Respondents are also less likely to vote for such candidates. Overall, our research suggests that affective polarization might play a causal role in democratic backsliding by making partisans less likely to punish politicians who violate democratic norms.