Abstract
Affective polarization measures account for partisans' feelings towards their own party versus its opponent(s), but not for how likely partisans are to encounter co-partisans versus out-partisans. However, the intensity of out-party dislike and the probability with which this comes into play both determine the social impact of cross-party hostility. We develop an affective fractionalization measure that accounts for both factors, and apply it to longitudinal survey data from 20 Western publics. From this perspective, countries with fewer dominant parties may be more harmonious because partisans have lower probabilities of interacting with political opponents. At the party level, partisans of smaller, more radical parties are particularly troubled because they strongly dislike out-partisans and have few co-partisans. Affective fractionalization has increased in most Western publics over time, primarily because of growing party-system fragmentation.