From Individuals to Partisans: The Role of Social Categorization in Political Polarization

Abstract

Political polarization is increasingly viewed not as a process of diverging political opinions but of rising levels of partisan prejudice. This so-called affective polarization is argued to influence how members of different political groups relate with one another, with potentially severe consequences for social cohesion and democratic functioning. While many researchers have therefore set out to find ways of reducing partisan prejudice, they have so far neglected a crucial prerequisite for most of the presumed negative consequences of partisan prejudice: social categorization. Prejudice toward the political outgroup can only affect how we relate to outgroup members when they are identified as such. Given that political categories are not tied to any invariant perceptual features, the present dissertation asks how individuals in their day-to-day lives come to perceive each other as partisans and to what extent this tendency is malleable. The first dissertation article showed across a series of five experiments that individuals inferred a wide range of political categories from behavioral information (e.g., expressions of political opinions). Moreover, these inferences were shown to occur spontaneously and to be robust in that they persisted even when alternative situational reasons for the behaviors were provided. The second article is premised on the idea that these robust categorizations are grounded in the behavioral cues being perceived as highly diagnostic (i.e., uniquely tied to the inferred category), and showed across three experiments that perceptions of diagnosticity can be reduced through counterstereotypical exemplars highlighting the occurrence of the cue outside of the inferred category. The third article builds on findings suggesting that partisan prejudice strongly depends on entitativity, that is, the degree to which category members are perceived as a coherent unit. It showed that the use of group labels—a known driver of entitativity—in social media posts predicted expressions of partisan prejudice in the respective comments. The dissertation concludes with a discussion of how the constructs of the three articles relate to each other and why they should be considered a driver and potential leverage point of polarization, rather than a mere epiphenomenon of it.