People who identify as lesbian, gay, bisexual, and transgender (LGBT+) often encounter prejudice, stigma, and abuse both in the United States and around the world (Velasco Reference Velasco2023). Although public acceptance of gay and lesbian individuals has been increasing over time in the United States, transgender people face unique challenges (Campbell, Hinton, and Anderson Reference Campbell, Hinton and Anderson2019; Lewis et al. Reference Lewis, Flores, Haider-Markel, Miller, Tadlock and Taylor2017, Reference Lewis, Flores, Haider-Markel, Miller and Taylor2022; Nownes Reference Nownes2019). Fortunately, studies have shown that prejudices can be reduced through positive experiences of contact and conversations with transgender people (Broockman and Kalla Reference Broockman and Kalla2016; Kalla and Broockman Reference Kalla and Broockman2020). Mechanisms related to perspective taking, framing, and identity priming can lessen public transphobia in this process (Flores et al. Reference Flores, Haider-Markel, Lewis, Miller, Tadlock and Taylor2018, Reference Flores, Haider-Markel, Lewis, Miller, Tadlock and Taylor2020; Harrison and Michelson Reference Harrison, Michelson, Brown and Gershon2020). In this research note, we inquire whether framing transgender rights within the broader LGBT+ rights struggle translates into greater public support for transgender rights specifically.
Appeals to a common superordinate identity have been shown to reduce intergroup prejudice and boost tolerance in various contexts (Hewstone and Greenland Reference Hewstone and Greenland2000; Wenzel, Mummendey, and Waldzus Reference Wenzel, Mummendey and Waldzus2008). Less clear, however, is whether such appeals can work to reduce prejudice when the superordinate identity is uncommon. Our motivating example involves cisgender, heteronormative majorities whose affinities for the broader social categorization of LGBT+ identity are potentially greater than for the transgender subgroup (Campbell, Hinton, and Anderson Reference Campbell, Hinton and Anderson2019; Lewis et al. Reference Lewis, Flores, Haider-Markel, Miller, Tadlock and Taylor2017, Reference Lewis, Flores, Haider-Markel, Miller and Taylor2022). We ask whether appeals to the broader LGBT+ rights struggle reduce prejudices against transgender rights in specific arenas of political contestation.
We focus on the recently salient and contentious issue of transgender participation in sports in the United States. Using a randomized survey experiment with an online representative sample of 1,366 Americans from May 19–25, 2021, we find that framing transgender rights in the context of LGBT+ rights significantly increases support for transgender sports participation based on self-identified gender. Our treatment effects are particularly noteworthy in boosting support among conservatives and partisan Republicans, who have often been at the forefront of opposition to transgender athletes. Our results underscore the importance of support from the broader LGBT+ community to reducing social stigma against transgender minorities.
We find that framing transgender rights in the context of LGBT+ rights significantly increases support for transgender sports participation based on self-identified gender.
THEORY AND HYPOTHESES
People who identify as lesbian, gay, bisexual, and transgender, among other subgroups, are often represented as members of the broader superordinate identity category of LGBT+. Some argue that a superordinate category is efficient for the discussion of many shared experiences of prejudice and discrimination faced by gender and sexual minorities within a broader cisgender, heteronormative population and that it facilitates sexual minority/majority comparisons (Harcourt Reference Harcourt2013). Many studies also show potentially important individual and collective benefits of a broad and encompassing LGBT+ identity. In a recent meta-analysis, Hinton et al. (Reference Hinton, de la Piedad Garcia, Kaufmann, Koc and Anderson2022) found that people who identify more closely with the broader LGBT+ community are less likely to conceal their sexual identity, exhibit less uncertainty over their identity, feel less internalized stigma, and display greater identity pride and cohesion with people across other LGBT+ subgroup categories. LGBT+ identification also has been associated with greater recognition and sensitivity to shared experiences of discrimination (Armstrong Reference Armstrong2002). Although all LGBT+ subgroups share many common experiences of discrimination and prejudice, Lewis et al. (Reference Lewis, Flores, Haider-Markel, Miller, Tadlock and Taylor2017) underscore the gap in public support for gay and lesbian versus transgender people, who face greater stigma. This raises questions about whether appeals to a broader and more socially accepted superordinate LGBT+ rights framing could enhance public support for transgender rights.
A superordinate identity is generally understood as a higher-order social category that encompasses both in-groups and out-groups. Theoretical mechanisms include in-group projection, where people who positively associate their in-group values and characteristics with the superordinate category will project those positive associations onto other outgroups when defined in terms of the same superordinate identity (Wentzel, Mummendey, and Waldzus Reference Wenzel, Mummendey and Waldzus2008). Key scope conditions, however, include the extent to which in-group and out-group members recognize themselves and one another as prototypical members of both the higher-order identity and each subgroup (i.e., dual identification). In an LGBT+ context, however, we assume that cisgender heterosexual people, allyship notwitstanding, do not identify with either LGBT+ superordinate or transgender subgroup categories. Both are out-groups from the standpoint of sexual and gender orientation.
Our research departs then from the conventional approach to superordinate identity by posing a more novel question: Can subgroup prejudice and discrimination be reduced by framing the subgroup’s struggle for rights and recognition in terms of a more socially acceptable superordinate out-group rights struggle, as visualized in figure 1? Here, the subgroup and superordinate identities are both out-groups. Although many cisgender heterosexual majorities display LGBT+ affinities, support, and solidarity, Lewis et al. (Reference Lewis, Flores, Haider-Markel, Miller, Tadlock and Taylor2017) show how these LGBT+ affinities are greater than those for the subgroup transgender community. This motivates us to inquire whether framing a transgender rights debate in terms of the broader LGBT+ group rights struggle could be effective in bolstering support for transgender rights.
Conceptualizing Superordinate Framing of Out-Group Rights
Under what conditions would such framing strategies work? We argue that LGBT+ group rights framing effects are likely moderated by the extent to which people accept the broader superordinate LGBT+ group in its struggle for rights and recognition. In the United States, such acceptance may be conditional to partisanship, ideology, gender, religiosity, urban–rural and regional effects, and other salient correlates of LGBT+ support (Becker and Jones Reference Becker and Jones2021; Thompson Reference Thompson2023). Or, it may be possible that superordinate LGBT+ rights framings can increase transgender support across such moderating conditions, albeit from different baselines. We test the following hypothesis.
Hypothesis 1. (Superordinate Rights Framing). Public support for transgender subgroup rights increases when first framed in the context of LGBT+ group rights.
RATIONALE FOR CASE SELECTION: TRANSGENDER PARTICIPATION IN SPORTS
Transgender participation in sports provides a compelling and salient arena of contestation over transgender inclusivity rights within which to evaluate our hypothesis. Both professional and recreational sports provide an intriguing and potentially difficult case for the expansion of transgender rights because they encompass a historically highly gender-segregated set of institutions (Ferez Reference Ferez2012; Flores et al. Reference Flores, Haider-Markel, Lewis, Miller, Tadlock and Taylor2020). There is consequently a strong association of binary, heteronormative gender divisions in sports institutions and culture that are exclusionary to transgender and transgender intersectional people (Anderson and Travers Reference Anderson and Travers2017; Cunningham and Pickett Reference Cunningham and Pickett2018; McClearen Reference McClearen2023).
At present in the United States, efforts to address questions of transgender participation in sports have run the gamut from formal and informal methods of inclusion, as well as exclusion, that vary across different sports, localities, states, and regions of the country (Anderson and Travers Reference Anderson and Travers2017; Flores et al. Reference Flores, Haider-Markel, Lewis, Miller, Tadlock and Taylor2020). Although issues of transgender participation in sports may ultimately be addressed in courts of law, we focus on the court of public opinion, which we argue is vitally important for the acceptance of any legal decisions handed down in terms of compliance and accommodation, as has often been the case in other landmark judicial proceedings on identity group rights (Bell Reference Bell2004).
Existing research finds an American public divided over whether transgender sports participation should be based on self-identified gender (which we refer to as a more inclusive rights framework) or gender assigned at birth (which we treat as a more exclusive or restrictive perspective). Flores et al. (Reference Flores, Haider-Markel, Lewis, Miller, Tadlock and Taylor2020) report that cisgender men are more opposed to transgender participation in sports based on self-identifying gender criteria than are cisgender women, and people who identify as Evangelical score highly on attitudes of authoritarianism, moral binary and gender traditionalism, binary gender conformity, levels of religious attendance, greater sports fandom, and are more opposed to inclusive transgender participation than other groups. In contrast, people who identify as partisan Democrats and ideological liberals are more favorable to inclusive transgender sports participation. Race is not a significant factor in predicting attitudes toward transgender sports, and education is only a positive predictor of support for inclusive transgender participation among cisgender women.
Working from the assumption that public views surrounding transgender sports participation are likely moderated by these attributes, we inquire whether framing the debate around broader LGBT+ rights could increase inclusive transgender support among any of those demographics. However, we are especially interested in framing effects along partisan and gender lines. Could partisan Republicans and ideological conservatives be more amenable to inclusive transgender rights if situated within a broader LGBT+ framework? We reason that an LGBT+ superordinate identity, although still an out-group identity, could dampen Republican and conservative opposition to inclusive transgender subgroup rights in the context of sports.
RESEARCH DESIGN
Our research design uses a randomized sequencing of survey items related to superordinate LGBT+ rights and transgender participation in sports. The treatment group receives priming on LGBT+ rights first, followed by transgender rights in sports. The control group has the reverse sequencing. Hence, our treatment both primes subjects to think about LGBT+ rights and frames subsequent transgender rights in a broader LGBT+ context. To assess support for LGBT+ rights, we focus on areas of legal contestation at the national, state, and local levels, in part because the issue of transgender participation in sports has involved legal contestation at multiple levels of the US federal system. There are no overarching federal guidelines to direct state and local community actions, which often vary widely (Anderson and Travers Reference Anderson and Travers2017; Flores et al. Reference Flores, Haider-Markel, Lewis, Miller, Tadlock and Taylor2020). Our LGBT+ rights items prime subjects to think about LGBT+ rights, broadly construed, which should then provide a frame for interpreting subsequent questions about transgender subgroup rights to participation in sports. Our legal rights frame is as follows:
$ {\displaystyle \begin{array}{c}\left( LGBT+ Legal\ Rights\ Treatment\right)\; How\; do\; you\; think\ LGBT+\hskip3.12em rights\ should\ be\ addressed\ legally?\end{array}} $
$ {\displaystyle \begin{array}{c}( Response\ options\ range\ from\hskip0.5em 4= strongly\ agree\hbox{--} 1=\hskip0.22em strongly\hskip3.12em disagree\ for\ each\ item,\hskip0.5em which\ is\ presented\hskip3.12em in\ randomized\ order)\end{array}} $
-
• Congress needs to enact a uniform set of national legal protections.
-
• The Supreme Court needs to make clear what rights are/are not protected under the Constitution.
-
• Let each state decide what’s best for them.
-
• Let each local community decide what’s best for them.
We then measure public support for transgender participation in sports using the following item, which provides background context to increase public issue salience. In both the treatment and outcome variables, items are presented in random order, and the LGBT+ treatment block is also randomized before or after the item on transgender sports participation.
(Transgender Participation in Sports) As of 2020, there are no federal laws regulating transgender athletes in sports, and state laws vary widely for whether athletes can compete in sports based on their self-identified gender or their assigned gender at birth. Which statement best captures your views on the issue of transgender athletes in sports?
-
0. Transgender athletes must compete in sports according to their gender assigned at birth.
-
1. Transgender athletes should be allowed to compete in sports according to their self-identified gender.
To test our main hypothesis, we follow Gomila (Reference Gomila2021) in using ordinary least squares (OLS) regression to estimate treatment effects and provide logistic regression analysis in our online appendix for comparison. We use the following model:
$ \hskip3.12em {Y}_i\left( Support\ for\ Inclusive\ Transgender\ Rights\right)\hskip7.12em ={\beta}_0+{\beta}_1{\left( LGBT\hskip-2pt + Rights\ Treatment\right)}_i+{\beta}_n{X}_i+{e}_i $
where our dependent variable, Support for Inclusive Transgender Rights, is coded 1 if respondents support allowing transgender athletes to compete in sports according to their self-identified gender and 0 if not. We test our main hypothesis with our LBGT+ Rights Treatment, which is coded 1 if respondents receive priming on LGBT+ rights before transgender rights and 0 in the reverse sequence. We also include responses to specific LGBT+ rights preferences, including those who favor more national or state and local legal frameworks for addressing LGBT+ rights. Finally, we control for a wide range of demographics, including gender, age, education, income, race, region, urban–rural location, partisanship, ideology, and potential moderated treatment effects using interaction terms.
SAMPLING AND DATA COLLECTION
Our preregistered, IRB-approved survey experiment was administered by Dynata as part of a stratified random sample of online participants using key demographics of gender, age, education, race and ethnicity, urban–rural, and regional strata (see the online appendix for a more detailed discussion of the sampling methodology). A total of 1,366 respondents completed our survey between May 19 to May 25, 2021. Women represented 52% of the sample with very few nonbinary respondents (4 of 1,366). More people in our sample identified as Democrats (44%) than Republicans (32%) or Independents (20%), and our sample was predominantly White (78%), with 12% identifying as African American and 12% as non-White Latino. Balance tests, reported in the online appendix, indicate that respondents were well-balanced across treatment and control groups. We provide a full range of summary statistics on demographic controls in the online appendix.
RESULTS
We begin with an overview of support for transgender inclusion in sports in our control group. Of 680 responses, 30.4% stated that transgender athletes should be able to participate in sports based on their self-identified gender, whereas 69.6% preferred that transgender athletes compete in sports based on their assigned gender at birth. In the treatment group (N = 683), we assess whether priming on superordinate LGBT+ rights affects support for inclusive transgender rights in sports. In figure 2, we show that priming on LGBT+ rights led to a 9.1 percentage-point increase in overall support for transgender participation in sports based on self-identified gender (unpaired t-test = 3.53, p < 0.0002, Cohen’s d = 0.19). We consider this an encouraging finding with meaningful implications for boosting public support for transgender inclusion in sports. Critically, the level of support for overall LGBT+ rights remains the same across the control and treatment groups, indicating that discussion of transgender rights has no reverse negative effect on support for LBGT+ rights (see the online appendix).
Priming on LGBT+ rights led to a 9.1 percentage-point increase in overall support for transgender participation in sports based on self-identified gender.
Average Treatment Effects of LGBT+ Rights Framing
Average treatment effects with 95% confidence intervals.
We find that treatment effects are robust to extended demographic controls in table 1. Using OLS regression, model 1 shows a 9.1 percentage-point increase in support for transgender rights in the treatment group relative to control. Model 2 indicates that treatment effects are robust to the inclusion of LGBT+ priming items. Specifically, respondents who agree more with national remedies for addressing LGBT+ rights (Congress or the Supreme Court) are more favorable toward transgender inclusion in sports, whereas those who favor more state and local control over the issue of transgender sports are more opposed to inclusive transgender participation. Both national and state/local variables are generated using interim correlation indices (see online appendix for details). Finally, model 3 includes interaction effects for LGBT+ acceptance and extended demographic controls for many variables used by Flores et al. (Reference Flores, Haider-Markel, Lewis, Miller, Tadlock and Taylor2020). First, we find that our treatment effect is moderated by pretreatment LGBT+ acceptance vs. homophobia. The baseline treatment effect (20.1% increase relative to control) is strong and positive for people who report feeling very safe around LGBT+ people (49%) but declines by an average rate of 7.7% for people who report feeling only somewhat safe (34%), somewhat unsafe (11%), or very unsafe (6%) around LGBT+ people. Consistent with Flores et al. (Reference Flores, Haider-Markel, Lewis, Miller, Tadlock and Taylor2020), we find that Democrats are more supportive of inclusive transgender rights than Republicans or Independents, ideological liberals and moderates are more supportive of inclusive rights than conservatives, cisgender women and a small subsample of nonbinary respondents are more supportive of transgender inclusion than cisgender men, Evangelical Christians are more opposed to transgender inclusion, and race is largely not a salient factor (with the noted exception of a small subsample who identify as mixed race and are more supportive of inclusive rights). Regional and urban–rural effects are also not significant. Finally, variance inflation factor tests indicate that multicollinearity is not a cause for concern in model 3.
LGBT+ Rights Framing and Transgender Inclusion in Sports (OLS Regression)
Notes: Robust standard errors are in parentheses. *** p<0.01, ** p<0.05, * p<0.1.
Next, following Flores et al. (Reference Flores, Haider-Markel, Lewis, Miller, Tadlock and Taylor2020), we consider other potential interaction effects between our treatment group and various demographic moderators, focusing on gender and partisanship. Figure 3 visualizes interactions between our treatment/control groups, men and women, and partisanship. Democratic women are the most supportive of transgender inclusion in our study (47.6% in the control group), but our LGBT+ framing treatment group boosts their support to 61% (unpaired t-test = 2.34, p < 0.01, Cohen’s d = 0.28). In contrast, treatments had no significant effect on Democratic men or Independents of either cisgender. Among Republican men and women, LGBT+ framing increases transgender support among Republican women from 13.5% to 24.2% (unpaired t-test = 2.16, p < 0.02, Cohen’s d = 0.28) and among Republican men from 16.4% to 26.3% (unpaired t-test = 1.71, p < 0.04, Cohen’s d = 0.25). Further conditioning shows that increased support among Republicans is primarily among non-Evangelicals and those who feel very safe around LGBT+ people. However, given widespread opposition among Republicans to transgender inclusion, we find these treatment effects noteworthy and encouraging.
LGBT+ Framing Effects by Gender and Party ID
Average treatment effects with 95% confidence intervals.
DISCUSSION AND CONCLUSION
Our research note illustrates how situating transgender rights in the broader context of LGBT+ rights has positive effects on public support for inclusive transgender participation in sports, even among many staunch opponents of such rights. Republican support for inclusive transgender participation in sports almost doubled when framed around broader LGBT+ rights contestation. As such, our findings have meaningful policy implications for bolstering public support and acceptance of transgender rights by framing debates in terms of the broader LGBT+ rights struggle. Nevertheless, there remain important barriers to transgender rights. Even after our treatment, a majority of respondents remain opposed to transgender sports participation based on self-identified gender, and the issue is likely to remain contentious across local, state, and national political domains for the foreseeable future.
In summary, more research should address whether support for inclusive transgender participation is conditional on gender identity (such as transgender women participating in women’s sports compared to transgender men in men’s sports). Future studies should also clarify the public’s understanding of the relationship between transgender people and the broader LGBT+ community, as well as the mechanisms underlying LGBT+ framing effects. For example, the superordinate category of LGBT+ includes lesbian, gay, and bisexual sexual orientations, whereas the transgender subgroup is a gender identity. Does this imply that people are more comfortable with rights for people marginalized by their sexual orientation, rather than by their gender identification? Does the public understand the differences between these concepts? Finally, is LGBT+ framing the only way to encourage people to be more supportive of transgender rights? Transgender people sometimes feel marginalized under the bigger “LGBT+ umbrella.” Although the LGBT+ framing may be helpful, transgender advocacy needs to be informed by those with trans-specific competence.
Republican support for inclusive transgender participation in sports almost doubled when framed around broader LGBT+ rights contestation.
The struggle to create a more inclusive society for transgender people continues. On January 9, 2025, a federal judge in Kentucky struck down Title IX protections for transgender students and student-athletes issued earlier in the year by the Department of Education under the Biden administration. In L.W. vs Skrmetti, the Supreme Court also recently heard oral arguments in a lawsuit against the state of Tennessee for its ban on gender-affirming healthcare for transgender youth. Several states have sued to block those rules from going into effect, and candidates weighed in on the issue of transgender rights in the 2024 election cycle to attract voters. While the outcome of the 2024 election speaks to ongoing public debate over transgender identity and rights, our simple experiment shows how appeals to a superordinate LGBT+ identity can shift public opinion in productive ways in the struggle for greater transgender recognition and inclusion.
In L.W. vs Skrmetti, the Supreme Court also recently heard oral arguments in a lawsuit against the state of Tennessee for its ban on gender-affirming healthcare for transgender youth.
SUPPLEMENTARY MATERIAL
To view supplementary material for this article, please visit https://doi.org/10.1017/S1049096524001045.
ACKNOWLEDGMENTS
This collaborative project was based on our collective efforts to pool resources to study topics of shared interest in our department. We thank Phillip Ayoub, as well as the editors and anonymous reviewers at PS, for their many helpful comments. We also thank High Point University for research-development funding for this project and Frank Markowitz at Dynata for data collection. Any errors or omissions are our own.
DATA AVAILABILITY STATEMENT
Research documentation and data that support the findings of this study are openly available at the PS: Political Science & Politics Harvard Dataverse at https://doi.org/10.7910/DVN/LDRXW2.
CONFLICTS OF INTEREST
The authors declare that there are no ethical issues or conflicts of interest in this research.
