Perceived Social (Dis)Similarity and Affective Polarization

While the sources of affective polarization have been studied extensively in political science, less is known about how citizens’ perceptions of social groups shape affective polarization and which group perceptions are most consequential. This paper investigates the social foundations of affective polarization using original survey data and novel measures. Building on research into social sorting, we extend existing work in two directions. First, we focus on subjective perceptions of social sorting, recognizing that many citizens are misinformed about the social composition of electorates. Second, we distinguish between two dimensions of sorting: social similarity (similarity with voters of one's party) and social dissimilarity (from other parties’ voters). We test whether affective polarization increases when individuals perceive their party’s voters as socially similar and opposing voters as socially dissimilar. We test our expectations using original survey data collected in Belgium and Canada, which we use to develop novel measures of subjective social sorting. We find a strong association between social similarity and social dissimilarity, on the one hand, and individual affective polarization, on the other. Our paper contributes to research on the link between social groups and parties, as well as to a deeper understanding of the social roots of affective polarization.