The Masculine/Feminine Double Bind: A Survey Experiment of Gendered Elections

• In Event: Candidates' Personal Characteristics and Their Performance

Abstract

While most studies show no aggregate penalty against women in elections, recent scholarship on intraparty races has demonstrated that candidates—especially women—do face a “double-bind” with regard to their self-presentation; voters prefer women who fit traditional profiles (i.e. married/parent/caregiver) while preferring candidates who fulfill different criteria. However, research has relied on 1) high level competition, which masks selection effects and perceived candidate competency and 2) text-heavy survey experiments lacking common audio/visual signals of gendered presentation. Because gender biases are not always or primarily explicit, experimental evaluations of candidates are more likely to reflect real world evaluations with the additional of these audio/visual cues than text-only cues. Therefore, existing findings of differences in candidate evaluations may be magnified (and/or extend to a wider range of voters) when treatments are more realistic and therefore have the chance to activate implicit biases.

Building on the experimental vignettes in previous studies, I conduct a survey experiment using audio recordings of campaign positions with manipulation to create a more masculine or more feminine timbre. In short, I am able to better test the validity of current research on voter biases against female and/or feminine candidates. Using a nationally diverse sample of verified voters, I find that across both parties, women face a penalty for feminine self-presentation, though evidence of the “double-bind” regarding likeability and competence is weaker than in previous work. I also find that among Republicans, this more feminine self-presentation puts women candidates at a significant disadvantage in intraparty elections.

Author

• Sierra Davis Thomander, Stanford University