How “Gay” Can You Go? The Effect of Gender Typicality on Electability

Recently, scholars have argued that political candidates’ sexual orientation has become a non-issue in elections in Western democracies. However, in evaluating the claim that sexual orientation is a non-issue for voters, a key question must be asked: What kind of gay candidates face little to no discrimination at the ballot box? So far, most researchers have treated all gay candidates alike, failing to consider that a key feature distinguishing them—gender typicality—might be critical to their electoral prospects. The electoral bias against gay candidates might be a function of either bias based on sexual orientation or bias based on gender typicality—or their interaction. In this paper, I explore both to pry apart their influence. Relying on the literatures on gender categorization and cognitive fluency, I argue that gender atypicality has a negative effect on a gay candidate’s probability of being elected. Furthermore, building on the literatures on stereotypes and cognition, I contend that straight gender atypical political candidates—also seen as counterstereotypical of their gender group—will also pay a price for deviating from gender norms. Especially likely to discriminate against gender atypical or gay candidates, I argue, are voters who need order and certainty (NOC). I expect that both outright anti-gay hostility and NOC will moderate the penalty levied against gay candidates. However, I argue that these two mechanisms operate independently: while anti-gay hostility penalizes gayness irrespective of gender typicality, need for order and certainty penalizes both gayness and gender atypicality or, more generally, counterstereotypicality. Given that ideology and partisanship correlate with both NOC and anti-gay hostility, an implication of my theory is that there will be systematic partisan/ideological differences in how voters react to gay and gender atypical candidates. In testing my theory, I use a pre-registered survey experiment where I present respondents with hypothetical co-partisan primary candidates for U.S. Congress. To manipulate a candidate’s gender typicality, I rely on manipulating their physical appearance, specifically their face and voice. Compared to facial stimuli, vocal stimuli have been severely underutilized in political science research. This is despite the fact that, like faces, voices too impart ample cues about the speaker that have been shown to impact judgement and voting—people rely on voices to infer, among other things, a speaker’s leadership abilities, emotional state, health, race, age, immigration status, and their gender (typicality). Overall, in this paper I make three main contributions. First, I break new ground by bringing gender typicality into the debate about what attributes shape the electability of political candidates. Second, I advance scholarly understanding of some of the psychological mechanisms undergirding identity-based discrimination in voting. Finally, I showcase the promise of voice stimuli, long underutilized in political science, for studying social identities, discrimination, and political psychology and behavior in general.