Replication Data for Issues, Groups, Or Idiots? Measuring Which Attributes are Central to Partisan Stereotypes

Abstract

When individuals picture the two parties, what do they think of? Given the dominant understanding of partisanship as a social identity, understanding the content of these mental images – individuals’ stereotypes of the two parties – is essential, as stereotypes play an important role in how identity affects attitudes and behaviors, perceptions of others, and inter-group relations. The existing literature offers three answers to this question: one that claims that people picture the two parties in terms of their constituent social groups, a second that claims that people picture the two parties in terms of policy positions, and a third that claims that people view the two parties in terms of individual traits they associate with partisans. While not mutually exclusive, these theories have different implications for the effects of partisanship and the roots of partisan animosity. This paper adjudicates between these theories by employing a new method that measures stereotype content at the collective and individual level using a conjoint experiment. An important advantage of the conjoint measure is that it allows for the direct comparison of the importance of different attributes, and different kinds of attributes, to the stereotype. Using a pre-registered 2,909-person survey, I evaluate the relative importance of issues, groups, and traits to stereotypes of partisans. I find strong evidence that issue positions and ideological labels are the central elements of partisan stereotypes. I also find that individuals who hold issue-based stereotypes are more affectively polarized than those whose stereotypes are rooted in groups or traits.