Recent studies suggest that online abuse directed at politicians can have negative effects on their public engagement and continued participation in politics. This article considers the broader consequences of such online abuse by testing whether exposure to online abuse of politicians also decreases the prospective political participation of ordinary citizens. In a preregistered survey experiment with 2,000 participants from Denmark, we find that exposing citizens to cases of online abuse of politicians does not have any statistically significant, or substantively meaningful, negative effects on citizens’ prospective political participation. This result holds across multiple measures of political participation and when distinguishing citizens by their gender and level of conflict avoidance. If anything, exploratory analyses indicate that online abuse of politicians may in some cases mobilize citizens who have been bystanders to such abuse.

Many worry that uncivil discourse can undermine democratic processes. Yet, what exactly does it mean for discourse to be uncivil? I argue that there is systematic variation in perceptions of incivility based on the identity of those targeted by uncivil speech. Specifically, I show—via a conjoint survey experiment—that White Americans are less likely to view statements directed at Black Americans as uncivil but more likely to perceive incivility when the target is a woman or a co-partisan. These results suggest an identity-laden aspect of incivility such that it is acceptable to treat Black Americans with less civility but less acceptable to do so for women and co-partisans. The results have implications for how we assess discourse and how that discourse affects the public.