Measuring affective polarization: Matching concepts to measures

Abstract Affective polarization is the extent to which partisans like their own party relative to their dislike of opposing parties. High levels of affective polarization in the United States and many other countries have prompted widespread scholarly interest and practical concern. Yet, despite nearly 15 years of scholarship, significant uncertainty remains about how best to measure affective polarization. We offer a comprehensive review focused on the measurement of affective polarization, underscoring how seemingly innocuous choices can yield substantially different results. These include the target of evaluation (e.g., engaged ideologues, ordinary voters), the dimensions of evaluation (e.g., liking/disliking, moralization), the measurement context (e.g., whether respondents are interviewed by an interviewer or complete surveys independently) and the political context (e.g., the power relationship between parties). We conclude that the next generation of inquiry will need to more carefully align measures of affective polarization with specific research questions.