Precarious or Policed Sexualities? How Race and Gender Affect the Categorization of Sexual Behaviors

Download data and study materials from OSF

Principal investigators:

Stefan Vogler

NORC at the University of Chicago



Christopher Petsko

Duke University



Sample size: 3010

Field period: 05/22/2019-10/25/2019

Do the inferences we make about a person’s sexuality depend on their race and gender? We conducted an experiment using a nationally representative sample of U.S. adults (N = 3,010) to examine how respondents perceive the sexual orientation of Black men, white men, Black women, and white women who engaged in a single same-sex sexual encounter. Replicating past work, we found that men’s perceived heterosexuality decreased more than did women’s. However, we found no evidence that a person’s race on its own influenced respondents’ perceptions of that person’s sexuality. Instead, we found evidence that—at least for some judgments—a person’s race and gender influenced respondents’ perceptions of their sexuality in tandem. Significant gender differences in judgments of heterosexuality emerged in perceptions of Black (but not white) people, whereas significant gender differences in perceptions of bisexuality emerged in perceptions of white (but not Black) people. Gender differences in judgments of homosexuality emerged in perceptions of both Black and white people. How precarious a person’s sexual orientation seems is best understood from a framework that builds on the insights of research about status and intersectionality, and recognizes that sexuality is perceived based on intersections of gender, race, and other social axes.

H1: Men’s perceived heterosexuality will decrease more than women’s when they engage in a single same-sex sexual encounter.

H2: Black men’s perceived heterosexuality will decrease more than white men’s for engaging in a single same-sex sexual encounter.

H3: Black women’s perceived heterosexuality will decrease more than white women’s for engaging in a single same-sex sexual encounter.

Experimental Manipulations
Using vignettes in which individuals engage in either same- or different-sex sexual behavior, we manipulated the individuals' genders and races (using racialized names as proxies), as well as whether they engaged in a same- or different-sex sexual interaction.
Our dependent variables were ratings of participants' certainty that an individual in a vignette was either heterosexual, bisexual, or homosexual.
Summary of Results
We anticipated that men’s heterosexuality would be regarded as more precarious than women’s heterosexuality. In addition, we hypothesized that this may be especially true for Black targets relative to White targets. When we subjected ratings of targets’ heterosexuality to a 2 (target race) × 2 (target gender) × 2 (target behavior) ANOVA, we found a sizable main effect of target behavior, F (1, 2897) = 878.71, p < .001, suggesting that engaging in just one same-sex sexual encounter caused targets to seem substantially less heterosexual to participants (M = 54.14, SE = 0.91) than they would have seemed otherwise (M = 81.16, SE = 0.65), M diff = –27.02, 95% CI[–28.81, –25.23], β = –0.96. Moreover, and in support of the view that men’s heterosexuality is more precarious than women’s, this effect was moderated by targets’ gender groups (target gender × target behavior interaction: F(1, 2897) = 17.01, p < .001). The nature of this interaction was such that engaging in a single same-sex sexual behavior (vs. not) decreased the perceived heterosexuality of men (M diff = –30.78, 95% CI[–33.32, –28.24], β = –1.10) significantly more than it decreased the perceived heterosexuality of women (M diff = –23.26, 95% CI[–25.78, –20.75], β = –0.83). Interestingly, participants did not exhibit a tendency to regard Black targets’ heterosexualities as more or less precarious than White targets’ heterosexualities (target gender × target behavior interaction: F(1, 2897) = 0.01, p = .90). Nor did participants’ perceptions of men’s vs. women’s heterosexualities depend on these targets’ racial groups (three-way interaction: F(1, 2897) = 0.23, p = .63). Thus, men’s heterosexuality was indeed more precarious than women’s, but there was no evidence that this effect depended upon whether these men and women were Black vs. White.