Attitude formation based on political views: An expectancy-value model of interpersonal likability in politics

How affective polarization in Western democracies results from the interplay of ideological differences and group identification remains unclear. To better understand these processes and address the call for falsifiable theories, we propose a formalized expectancy-value model of interpersonal likability in politics (EVIL-P). It argues that people (dis)like targets with (dis)similar political views (i.e., value), especially if these views are rare (i.e., expectancy). By reanalyzing data from the American National Election Studies (N = 38,868), the British Election Study (N = 24,522), and nine studies on ratings of societal groups that vary in political views (N = 6,587), we find support for EVIL-P. Expectancy and value interacted to predict people’s attitudes. Moreover, targets with moderate political views were liked most on average, and targets with more extreme political views had more extreme fans and haters. We discuss how EVIL-P contributes to research on affective polarization and helps testing related phenomena.