We analyze the winner-loser gap in affective polarization. Using data from 37 countries over years 1996 – 2016 from the Comparative Study of Electoral Systems (CSES), we find that winners’ affective polarization is significantly greater than losers’, and the difference is due to winners’ consistently higher in-party favoritism. These findings are robust when controlling for partisanship and ideological distance to winning party. Although much of the literature focuses on the impact of out-party dislike on affective polarization, our results align with research in social identity theory indicating that intergroup discrimination is driven primarily by in-group favoritism rather than out-group dislike. Given that winners are more likely than losers to support their favored party’s violation of democratic norms, our work suggests that in-party favoritism is an important but overlooked contributor to problematic implications of affective polarization.