Explaining Heterogeneity in Women’s Support for Female Candidates

• In Event: Primaries and Candidates

Abstract

Although theories of descriptive representation often posit that citizens are drawn to individuals from groups with which they identify and despite the fact that women remain largely politically underrepresented, most previous studies find no overwhelming gender gap in favor of female candidates: women do not consistently support same-gender candidates. Why is gender identification such a weak predictor of candidate evaluation and voting behavior? This paper proposes that gender has been found to play only a minor role in elections due to the neglect of two strong but countervailing identities at the subgroup level – feminism and non-feminism – whose impact is distinct and goes above and beyond gender itself. Among women, gender identification has different implications as a result of competing interpretations about what ‘identifying as a woman’ means. The distinction between group-level and subgroup-level identities is crucial because several scholars have shown that identifying as a woman is less politically consequential than is identifying as a feminist. However, measuring the consequences and implications of feminist/non-feminist self-identification is complicated by the fact that many women with pro-feminist beliefs reject the feminist “label.” In other words, feminist ideology and feminist identity are correlated but distinct. The pervasiveness of individuals’ disdain for certain politically salient labels has been shown in regard to political ideology (Ellis & Stimson 2012) and partisanship (Klar & Krupnikov 2016), but not in relation to gender. What we know is that a large majority of the U.S. electorate support gender equality, but only a minority self-identify as feminists. Thus, three distinct gender subgroups need to be operationalized: 1) “feminists,” who both accept the feminist label and hold a feminist ideology, 2) “non-labelers” or “egalitarians,” who eschew the feminist label yet endorse feminist principles, and 3) “non-feminists,” who reject both feminist self-labeling and feminist beliefs. In this paper, I argue that, rather than failing to have a positive attachment to their gender group, many women who endorse feminist beliefs still vote against a female candidate because that candidate is explicitly associated with feminism. It is the aversive reaction to the label, rather than the actual substance of their politics, that drives many women away from candidates of their gender group. Relying on evidence from three studies – the 2016 American National Election Study, a survey-based experiment conducted on Amazon’s Mechanical Turk, and a survey conducted on YouGov, I show that feminist/non-feminist labeling and ideology are key to explaining and predicting what kinds of candidates are more likely to win voter support. Moreover, I analyze how views about feminism as well as beliefs about gender equality and women’s role relate to individuals’ willingness to embrace the feminist label. Finally, I discuss how feminist/non-feminist self-identification impacts political participation and attitudes about public policies.

Author

• Marzia Oceno, University of Michigan, Ann Arbor