While many scholars of the social sciences have long-since recognized identities such as gender and race/ethnicity as informative to understanding political perspectives, sexual identity (sexual orientation) has been largely overlooked in such explorations (Schnabel, 2018; Whitaker, 2008). Some investigations reveal that sexual identity is an important and robust predictor of political ideology and that compared to heterosexuals, lesbian, gay, and bisexual (LGB) people tend to be more liberal (e.g., Grollman, 2017; Herek, Norton, Allen, & Sims, 2010; Hertzog, 1996; Lewis, Rogers, & Sherrill, 2011; Swank, 2018a, 2018b; Swank, Fahs, & Frost, 2013). However, there is significantly less work focused on other more specific political attitudes that may help further our understandings of “sexuality gaps” (e.g., heterosexual-LGB) in political perspectives. Furthermore, sexuality gaps may also exist among mostly heterosexual (MH) and LGB people (e.g., MH-LGB) but are underexplored. In addition, although there are documented gender gaps in regard to political liberalism, it is unclear if these same gender gaps found among heterosexuals exist among MH and LGB men and women (Hertzog, 1996; Whitaker, 2008). Indeed, no studies to date have examined gender gaps among MH and LGB people they relate to political perspectives.

LGB liberal political leanings can be theorized as built from personal experiences with stigma and empathic concern for other stigmatized groups (i.e., the so-called underdog principle/hypothesis/thesis, Davis & Robinson, 1991; Schuman & Harding, 1963). LGB people may be particularly inclined to adopt liberal political attitudes because they are keenly aware of their stigmatization by various groups that are overtly hostile toward them (e.g., the religious right, family values groups, conservatives). In addition, among LGB people, those who are doubly stigmatized, including lesbian and bisexual women, may be especially likely to have liberal political attitudes.

Overall, liberal social justice perspectives may be particularly common among LGB people as a group and perhaps especially among lesbian and bisexual women due to their multiple oppressed identities and unique experiences with stigma and discrimination (Friedman & Leaper, 2010). These patterns may lead to liberal LGB political leanings and may also contribute to significant sexuality and gender gaps in political perspectives. The current study explores these relationships among college students enrolled at a university in the southern USA (N = 1940). Specifically, sexual identity (lesbian, gay, bisexual, mostly heterosexual, and heterosexual), gender (man/woman), and the intersections among sexual identity and gender are explored as they relate to politicized perspectives (liberal ideology and feminist identity) and support of politicized issues (death penalty and legal abortion). Unlike some previous studies that collapse “LGB” people into a single category (e.g., Schnabel, 2018; Swank, 2018b), by examining the unique political perspectives of lesbian, gay, bisexual, and mostly heterosexual women and men, important differences across these intersectional identities can be highlighted (Worthen, 2013). Overall, this research seeks to fill the gaps in the literature by expanding our knowledge about sexuality and gender gaps in political attitudes and contribute to new lines of inquiry that focus on MH and LGB people’s perspectives. In doing so, the current study works toward a deeper understanding of the ways college students can promote political change and advocate for social justice.

Political Perspectives and Sexualities among College Students

Colleges encourage exposure to new ideas and personal discovery (Gumprecht, 2003). In particular, universities provide students with a wide variety of experiences that can contribute to changes in cultural, intellectual, political, social, and religious values (Pascarella & Terenzini, 1991). Furthermore, moving away from home can serve as a catalyst for growth and change. Young adults who have previously lived under direct parental supervision now have the freedom to engage in new activities and free thinking which may be inclusive of more diverse political attitudes.

In addition, college can be a space for learning more about LGBQ people and exploring one’s own sexuality in ways that contrast with heteronormativity (Rupp, Taylor, Regev-Messalem, Fogarty, & England, 2014; Poynter, 2016). Especially for college students, describing oneself as MH may reflect a new set of sexual interests that have not been solidified into a sexual identity (such as bisexual). Amidst a climate of sexual exploration, college students may be more likely to identify as MH and LGB (Bogle, 2008; Rupp & Taylor, 2010; Rupp et al., 2014) and to support liberal political perspectives that align with social justice issues (e.g., women’s rights, abortion access, sexual freedoms).

Overall, college campuses are composed of students with varying political perspectives and diverse sexualities that relate to their individual cultural experiences. Universities located in conservative regions (e.g., the Bible Belt in the southern USA) are uniquely situated to explore these dynamics. The Bible Belt of the American South has its own personality, often characterized by a pervasive devotion to family, church, and state and conservative political values (Baunach, Burgess, & Courtney, 2010; Sears, 1989). However, large public universities in these areas often evoke a more liberal quality that is unlike their surrounding cultures. Students coming from small, rural southern towns near these universities are likely exposed to significantly more diverse liberal and social experiences during college and may feel inspired to embrace social justice advocacy. Indeed, colleges have often served as sites of powerful political activism (Van Dyke, 1998). Even so, despite the well-documented findings that college students are more liberal-leaning and more likely to vote than those without a college education (e.g., Hillygus, 2005), we know little about MH and LGB college student political perspectives. Thus, in order to best understand the ways MH and LGB college students can promote political change and advocate for social justice, we need to adequately explore the sexuality and gender gaps in political perspectives among college students.

Sexuality and Gender Gaps in Politicized Perspectives

Sexuality gaps in political ideology and voting patterns have been the focus of the bulk of existing research regarding LGB perspectives (e.g., Hertzog, 1996; Lewis et al., 2011; Swank, 2018a, 2018b). In addition, feminist identity as a politicized perspective has also been highlighted as significantly associated with MH and LGB attitudes toward political and social issues (e.g., Hertzog, 1996; Worthen, 2017a). Below, sexuality and gender gaps regarding liberal ideology and feminist identity are reviewed.

LGB Identities and Liberal Ideology: Sexuality Gaps

Hertzog’s (1996) detailed analysis of the voting patterns of LGB people in the 1990s indicated “There is a ‘lavender vote’ in America, as opposed merely to gay, lesbian, and bisexual voters. It is characterized mainly by strong liberalism on domestic social issues” (p. 224). Since then, studies have consistently shown that LGB people tend to identify as political liberals and support liberal perspectives about politicized issues (Grollman, 2017; Herek et al., 2010; Swank, 2018a, 2018b). In addition, scholars find that in comparison to their heterosexual counterparts, LGB people are one of the most loyal liberal Democratic Party voting blocs (Hertzog, 1996; Lewis et al., 2011; Swank, 2018a, 2018b). Studies also indicate that the robust significant differences between LGB and heterosexual political ideology, political party identification, and presidential voting patterns cannot be explained by demographic differences (e.g., race/ethnicity, age, education, income, and religion) (Grollman, 2017; Herek et al., 2010; Hertzog, 1996; Lewis et al., 2011). Overall, research confirms a sizeable sexuality gap between LGB people and heterosexuals in liberal ideology. Thus, there is strong evidence to support a relationship between being LGB and being liberal.

Among LGB people, some research indicates liberalism differences between lesbian women, gay men, and bisexuals. For example, lesbian women and gay men have been found to be more likely to identify as political liberals in comparison to bisexual men and women (Grollman, 2017; Herek et al., 2010; Lewis et al., 2011; Schaffner, MacWilliams, & Nteta, 2017). Indeed, using the American National Election Survey of 2016 (N = 2691), Swank (2018a) found that lesbian women and gay men were significantly more likely than bisexuals to have voted for the liberal 2016 US Democratic Party candidate, Hillary Clinton. Thus, a handful of studies provide support for sexuality gaps in liberal ideology among LGB people but these relationships deserve further attention.

LGB Identities and Liberal Ideology: Gender Gaps

Among heterosexuals, women tend to be more liberal about social issues and more likely to adopt liberal political ideologies than men (Eagly, Diekman, Johannesen-Schmidt, & Koenig, 2004; Lewis et al., 2011; Sidanius, Levin, Van Laar, & Sears, 2008; Whitaker, 2008; Williams & Wittig, 1997). There is also evidence to indicate a gender gap in liberal ideology among some LGB women and men. For example, examining liberalism using a US national probability sample of self-identified LGB adults (N = 662), Herek et al. (2010) found that a higher percentage of lesbian women identified as political liberals (66%) as compared to gay men (63%), bisexual women (53%), and bisexual men (45%), though differences were not statistically significant. Lewis et al.’s (2011) study with nearly 13,000 pollsters identified a similar pattern among LGB political ideologies (lesbian women, 56%; gay men, 50%; bisexual women, 33%; bisexual men, 22%). Even so, lesbian women and gay men have been found to be largely similar to one another in their liberal political leanings (Herek et al., 2010; Hertzog, 1996; Lewis et al., 2011). However, further research is needed to better flesh out these relationships.

LGB Identities and Feminism: Sexuality Gaps

Similar to the patterns reviewed above regarding liberal ideology, Swank (2018a) found that LGB people were more likely than their heterosexual counterparts to identify as feminist. In addition, Friedman and Ayres (2013) determined that non-heterosexual (sexual minority) college women (N = 107) were significantly more likely than heterosexual college women (N = 173) to participate in feminist activism. Although not specifically focused on “feminist” identities, Grollman’s (2017) analysis of the American National Election Survey (N = 4526) determined that LGB respondents were significantly more likely than heterosexual respondents to believe that sexism is a very real issue in the USA even when controlling for other demographic characteristics (i.e., gender, race and ethnicity, household income, age, marital/partner status, and region of the country). Similarly, Schnabel’s (2018) analysis of the General Social Survey (N = 5901) indicated that LGB respondents were more liberal and less sexist than heterosexuals in their perspectives about gender issues. However, Swank’s (2018b) study of social movements using the American National Election Surveys of 2012 (N = 3519) did not find a significant difference between LGB (2.3%) and heterosexual (2.0%) lifetime involvement in the women’s rights movement. In addition, among LGB people, lesbian women and gay men have been found to be more likely than bisexuals to identify as feminist (Swank, 2018a). Overall, studies provide some evidence for a heterosexual-LGB sexuality gap in feminist identity as well as sexuality gaps among LGB people (LG-B).

LGB Identities and Feminism: Gender Gaps

Heterosexual women are more likely to adopt feminist identities than heterosexual men (Hertzog, 1996; Whitaker, 2008; Williams & Wittig, 1997; Worthen, 2016). Similarly, there is support for a gender gap in feminism among some LGB women and men. For example, Worthen’s (2017a) examination of college students (N = 389) found that MH women and bisexual women were significantly more likely than MH men to identify as feminist. However, there were no significant differences in feminist identity between MH women, bisexual women, bisexual men, gay men, and lesbian women (Worthen, 2017a). Hertzog (1996) also found little support for a robust gender gap in perspectives between LGB men and women but instead, identified a “feminism gap” among LGB people demonstrated by a significant division between self-identified feminists and non-feminists that strongly relates to LGB political perspectives and voting patterns.

Yet, feminism has a complex history in the lives of LGB people that can impact the ways LGB men and women identify as feminists. The US Gay Liberation Movement of the 1970s was largely dominated by White gay cis men.Footnote 1 Scholars note that some lesbian and bisexual women retreated from the larger “gay” movement due to their experiences with sexism and this shaped their experiences with feminism (Stein, 1997; Weiss, 2004). Indeed, during the 1970s, gay men frequently adopted an assimilationist strategy (i.e., “Gay people are just like you”) while some lesbian and bisexual women opted for more radical strategies built from feminisms (Weiss, 2004). For example, emerging from radical feminism, the lesbian separatist movement created a schism between lesbian separatists and all men (including gay men) whereby lesbian rights were seen as qualitatively distinct from gay men’s rights (Hertzog, 1996; Stein, 1997; Weiss, 2004). In this way, “feminism” became fully entwined with lesbian separatism for some lesbian women. On the other hand, some lesbian women remained aligned with gay men and the Gay Liberation Movement but continued to promote liberal feminist ideals (DeBlaere et al., 2014; Szymanski, 2004). In addition, bisexual women were (and still are) sometimes rejected from lesbian-only spaces (Rust, 1995; Weiss, 2004). These historical dynamics continue to shape feminist identities among LGB men and women in differing ways. Overall, however, the gender gaps in feminist identity among MH and LGB men and women remain unclear and relatively underexplored.

Sexuality and Gender Gaps in Attitudes toward Politicized Issues

Sexuality gaps in attitudes toward specific politicized issues have not been frequently examined. While there are numerous politicized issues that could be explored, those that are evenly divided in the general population (close to 50/50) tend to be most contentious. Public opinion polls indicate that the USA is split in regard to the death penalty (55% favor) and legal abortion (48% pro-choice) (Gallup, 2017, 2018). Below, sexuality and gender gaps in regard to support of the death penalty and legal abortion are discussed.

LGB Identities and Death Penalty Attitudes: Sexuality Gaps

Two existing studies indicate that gay and lesbian individuals are less likely than heterosexuals to support the death penalty for murderers (Hertzog, 1996; Worthen, Sharp, & Rodgers, 2012). Hertzog’s (1996) summarization of findings from a 1990 CBS exit poll show that heterosexuals support the death penalty for those convicted for first-degree murder at nearly double the rate of gay men and lesbian women, indicating a sexuality gap in capital punishment perspectives. Worthen et al.’s (2012) analysis of General Social Survey data (N = 2649) also demonstrates a sexuality gap in death penalty support; however, gender plays a significant role in these differences. In particular, heterosexual men were significantly more likely than heterosexual women, gay men, and lesbian women to support capital punishment (Worthen et al., 2012). In contrast, among LGB people, neither being gay/lesbian nor bisexual were found to be significantly correlated with death penalty support in Worthen, Rodgers, and Sharp’s (2014) study among college students. No other research could be located that has examined LGB death penalty support, thus the current research provides a much-needed continued exploration into this area of inquiry.

LGB Identities and Death Penalty Attitudes: Gender Gaps

Heterosexual men are more likely to support capital punishment than heterosexual women are (Cochran & Sanders, 2009; Hertzog, 1996; Unnever & Cullen, 2006; Worthen et al., 2012; Whitaker, 2008). However, there is mixed evidence regarding a gender gap in death penalty support among LGB women and men. In the only study to date that has investigated death penalty attitudes and the intersections between lesbian/gay identities and gender identity, Worthen et al. (2012) found no significant differences in mean levels of death penalty support between gay men and lesbian women. However, in logistic regression models with heterosexual men as the reference category, being a gay man was found to exert a significant negative effect on death penalty support even when controlling for background variables including years of schooling, total family income, age, southern US region, rural county, and race/ethnicity. In contrast, being a lesbian woman was not significantly related to capital punishment in this exploratory study (Worthen et al., 2012). Such findings suggest that if there is a gender gap in gay/lesbian death penalty support, it may operate in the direction opposite from the gender gap in heterosexual death penalty support. However, because there is only one existing study on this subject, much more research is needed to draw any conclusions about trends and patterns in LGB death penalty attitudes.

LGB Identities and Abortion Attitudes: Sexuality Gaps

Studies show that LGB people are more likely to support legalized abortion and less likely to support restrictions on abortion when compared to their heterosexual counterparts even when controlling for demographics including gender, race and ethnicity, household income, age, marital/partner status, and region of the country (Grollman, 2017; Hertzog, 1996; Schnabel, 2018). However, Swank’s (2018b) study of social movements did not find a significant difference between LGB (1.8%) and heterosexual (2.9%) lifetime participation in the Right to Life Movement that opposes legalized abortion. No studies could be located that examined sexuality gaps in abortion attitudes among MH or LGB people. Even so, there is some evidence to indicate a significant heterosexual-LGB sexuality gap in abortion attitudes and to support a relationship between being LGB and supporting legalized abortion and access to abortion.

LGB Identities and Abortion Attitudes: Gender Gaps

Among heterosexuals, there is mixed support for differences between men and women in their attitudes toward abortion rights (Bolzendahl & Brooks, 2005; Patel & Johns, 2009; Whitaker, 2008). In addition, there is little evidence indicating a gender gap in abortion attitudes among LGB women and men. In the only study to date that has partially examined this, Hertzog (1996) found that gay men support abortion rights as strongly as lesbian women; thus, there was no “gender gap” among lesbian women and gay men in their abortion attitudes. Even so, considering the multiple layers of oppression that lesbian and bisexual women endure, they may be more inclined than gay and bisexual men to support the rights of women, including those related to legal abortion. No existing research has specifically examined MH or bisexual men’s and women’s attitudes toward abortion, thus, the current study is the first to do so.

Theoretical Framework: Social Justice Perspectives among LGB People

Although there are various theoretical explanations for LGB peoples’ strong liberalism (e.g., intergenerational transmission of liberal values and embeddedness in liberal LGB communities), the strongest support for these patterns focuses on arguments that emphasize how the stigmatized status of belonging to a devalued group shapes both distrust of the majority (and accompanying hierarchal systems that uphold the status of the majority) and empathy with the minority (Swank, 2018a, 2018b; see also Hertzog, 1996; Lewis et al., 2011; Schanbel 2018; Worthen et al., 2012). This empathic concern can include one’s own minority group as well as other minority groups who experience stigma. Identifying and empathizing with the minority aligns with the so-called underdog principle/hypothesis/thesis (Davis & Robinson, 1991; Schuman & Harding, 1963). Put simply, LGB people’s experiences with being a stigmatized “underdog” group lead them to empathize with other underdog groups. Indeed, studies show that LGB people are inclined to support other oppressed and disadvantaged groups, such as African-Americans and women, as well as issues that directly impact marginalized groups such as immigration, affirmative action, criminal justice reform, and feminism (Grollman, 2017; Schnabel, 2018; Swank, 2018a; Worthen et al., 2012). Swank (2018a) notes that this “sense of shared oppressions across social identities can, in turn, lead to embracing policies that try to counteract social injustices” (p. 25). Indeed, Swank (2018b) finds that LGB people are significantly more likely than heterosexual people to be involved in social justice movements including peace, environmentalism, Occupy Wall Street, and LGB movements. Thus, LGB liberal political attitudes can be theorized as embedded in an overarching social justice perspective built from personal experiences with stigma and empathic concern for other stigmatized groups.

While anyone can adopt a liberal social justice perspective, LGB people may be particularly inclined to do so. This is perhaps because LGB people endure stigma and oppression due to both their minority status as well as the stereotype that they are immoral and threaten traditional values (Rubin, 1984; Schanbel 2018). In addition, identifying as LGB often involves a “coming out” process built from reflexive engagement with self, family, community, and culture (D'Augelli, 1994). As a result, LGB people may be likely to embrace liberal social justice perspectives because they are keenly aware of how they are stigmatized by various groups that are overtly hostile toward them (e.g., the religious right, family values groups, conservatives). Furthermore, among LGB people, those who occupy doubly marginalized identities, including lesbian and bisexual women, may be especially likely to have liberal social justice perspectives. Indeed, multiple stigmatized statuses enhance oppression experiences and shape awareness of injustices (Collins, 2000; Crenshaw, 1991; Greene, 1996; Harnois, 2015, 2017). In contrast, those who may be exploring their sexualities but have not yet identified as LGB (such as those who identify as “mostly heterosexual”) may be less likely than LGB-identified people to adopt liberal social justice perspectives.

Current Study

Overall, social justice perspectives may be particularly common among LGB people as a group and perhaps especially among lesbian and bisexual women due to their multiple oppressed identities. These patterns may lead to liberal LGB political leanings and may also contribute to significant sexuality and gender gaps in political perspectives. The current study explores these relationships through the following hypotheses whereby “liberal” is measured in terms of liberal ideology, feminist identity, and attitudes toward the death penalty and legal abortion:

Hypothesis 1a

There are sexuality gaps in heterosexuals’ political perspectives as compared to all others (heterosexual-LGB, heterosexual-MH, heterosexual-B, and heterosexual-LG) whereby heterosexuals are significantly less liberal than all others.

Hypothesis 1b

There are sexuality gaps in MH people’s political perspectives as compared to LGB people (MH-LGB, MH-B, and MH-LG) whereby MH people are significantly less liberal than LGB people.

Hypothesis 1c

There are sexuality gaps in LGB people’s political perspectives whereby lesbian/gay people are significantly more liberal than bisexual people.

Hypothesis 2a

There are gender gaps among MH peoples’ political perspectives whereby MH women are significantly more liberal than MH men.

Hypothesis 2b

There are gender gaps among LGB peoples’ political perspectives (bisexual women-bisexual men and lesbian women-gay men) whereby LB women are significantly more liberal than GB men.

Methods

Data and Sample Characteristics

The dataset was derived from student responses to an online survey written by the author and distributed by mass email to all students enrolled in a large southern Bible Belt university (~ 24,000). Respondents were offered a recruitment incentive (the chance to win an iPad). The full sample (N = 1940) was 58% women, 80% heterosexual, 11% mostly heterosexual, 5% lesbian/gay, and 4% bisexual. In regard to race/ethnicity, the full sample was 76% Caucasian/White, 5% African-American/Black, 7% Asian/Pacific Islander, 5% Native American/Alaskan Native, 7% other race, and 7% Hispanic/Latinx (of any race). The average age was 24.6 years and 12% were freshmen. The MH and LGB subsample (N = 389) was 67% women, 57% mostly heterosexual, 24% lesbian/gay, and 19% bisexual. For race/ethnicity, the MH and LGB subsample was 76% Caucasian/White, 5% African-American/Black, 7% Asian/Pacific Islander, 5% Native American/Alaskan Native, 7% other race, and 7% Hispanic/Latinx (of any race). The average age was 23.72 and 11% were freshmen.

Measures

Dependent Variables

For liberal ideology, respondents were asked to identify as (1) extremely conservative, (2) conservative, (3) moderate, (4) liberal, or (5) extremely liberal. For feminist identity, respondents were asked, “Do you think of yourself as a feminist?” with four response options: (1) No, I do not consider myself to be a feminist and I disagree with feminism; (2) No, I do not consider myself to be a feminist; (3) Yes, I consider myself to be a feminist; and (4) Yes, I consider myself to be a strong feminist. Death penalty support and was estimated by asking respondents, “Are you in favor of the death penalty for persons convicted of murder?” Legal abortion support was estimated by asking respondents, “Do you think abortion should be legal?” Both questions had four response options: (1) Never under any circumstances, (2) Only under certain extreme circumstances, (3) Sometimes under certain circumstances, and (4) Always under any circumstances. In addition, a Liberal Scale was constructed by summing all of the dependent variables together (range = 4–17), with death penalty support reverse coded to indicate more liberal-leaning perspectives (Tables 1 and 2).

Table 1 Mean values of dependent variables with ANOVA results identifying sexuality gaps in political perspectives (N = 1940)
Full size table
Table 2 Mean values of dependent variables with ANOVA results identifying sexuality and gender gaps in political perspectives (N = 1940)
Full size table

Independent Variables

Students were asked to provide their gender (man or woman) and their sexual identity (exclusively heterosexual, mostly heterosexual, bisexual, mostly homosexual and exclusively homosexual). All those identifying as something other than “exclusively heterosexual” were included in the MH and LGB subsample (N = 389). Those identifying as mostly homosexual (N = 31) and exclusively homosexual (N = 64) were collapsed into the two categories of gay or “lesbian” in accordance with gender identities of “man” and “woman,” respectively. Mostly heterosexual was kept as a distinct category for reasons reviewed above.

Control Variables

Because race and ethnicity have been found to be related to political ideology and feminist identity as well as death penalty and abortion attitudes (Collins, 2000; Peffley & Hurwitz, 2007; Wilcox, 1990), race (Caucasian/White, Black/African-American, Asian/Pacific Islander, Native American/Alaskan Native, or “Other”) and Hispanic/Latinx ethnicity (regardless of race) are included as controls with Caucasian/White as the reference category. Previous research also indicates that younger people and those early in their student career think about politicized issues differently than older people and those later in their tenure at a university (Pascarella & Terenzini, 1991; Sidanius et al., 2008; Worthen, 2017b). Thus, student classification (collapsed to “freshman” as compared to all others) and age are also included as controls.

Method of Analysis

In the first set of analyses, the mean values of the dependent variables were compared by sexual identity and gender using ANOVAs and post hoc Tukey-Kramer tests. LGB people were combined into one group for ANOVA test 1 and then separated into two groups, lesbian/gay and bisexual people, for ANOVA test 2. In the second set of analyses, OLS regressions were used to explore the effects of sexual identity and gender on liberal ideology, feminist identity, death penalty support, legal abortion support, and the liberal scale. The same set of models was estimated with exclusive heterosexuals (N = 1551) as the reference category in Table 3 and with mostly heterosexuals (N = 222) as the reference category in Table 4. In Model 1, the baseline effects of being LGB were estimated along with controls. Model 2 considers the sexual identity categories of mostly heterosexual, bisexual, and lesbian/gay. Model 3 adds woman and the interaction effects of gender and sexual identity.

Table 3 OLS regression results predicting the effects of sexual identity and gender on political perspectives with interaction effects (N = 1940)
Full size table
Table 4 OLS regression results predicting the effects of gender and sexual identity on political perspectives with interaction effects among LGB and mostly heterosexual respondents (N = 389)
Full size table

Results

Mean Comparisons

Sexuality Gaps

In Table 1, ANOVA and post hoc Tukey-Kramer test results reveal significant sexual identity differences in means whereby heterosexuals are less liberal than MH, LGB, lesbian/gay, and bisexual people in terms of liberal ideology, feminist identity, and support for legal abortion. In contrast, heterosexuals report significantly higher levels of support for the death penalty. This illustrates a significant sexuality gap between heterosexuals and all others for all of the current study’s measures of political perspectives, fully supporting Hypothesis 1a. Interestingly, there are only two significant differences among MH and LGB people. Mostly heterosexuals are significantly more likely than LGB people to support the death penalty (partially supporting Hypothesis 1b) and lesbian/gay people are significantly less likely than bisexuals to support legal abortion (contradicting Hypothesis 1c). Thus, for the most part, there is not a robust sexuality gap in political perspectives among MH and LGB people and Hypotheses 1b and 1c lack strong support.

In Fig. 1, box and whisker plots reveal visual sexuality gaps with heterosexuals as the least liberal as indicated by the lowest median (11) and lowest minimum value (4) on the liberal scale. Mostly heterosexuals follow with a slightly higher median (13) and minimum value (5). Looking at LGB people as a group, the median is the same (13) but the minimum value is higher (6). Separating LGB people into two groups, lesbian/gay people indicate the same median (13) and the same minimum value (6) on the liberal scale as LGB people while bisexuals indicate the same median (13) but highest minimum values (9). ANOVA and post hoc Tukey-Kramer test results reveal significant sexual identity differences in means with heterosexuals identifying significantly lower on the liberal scale than all others, fully supporting Hypothesis 1a. In addition, among MH and LGB people, mostly heterosexuals are significantly less liberal than bisexuals, only partially supporting Hypothesis 1b. In contrast, there are no significant differences between lesbians/gays and bisexuals, thus, Hypothesis 1c is not supported.

Fig. 1
figure 1

Box and Whisker plot of liberal scale with ANOVA results identifying sexuality gaps. ANOVA results (test 1: groups 1–3, F = 151.40, df(2, 1937), p < .001; test 2: groups 1, 2, 4, and 5, F = 102.35, df(3, 1936), p < .001) and post hoc Tukey-Kramer test results by group number (see Table 1 for group numbers): group 1 ≠ (groups 2–5); group 2 ≠ group 5

Full size image

Sexuality and Gender Gaps

In Table 2, there were no significant differences in liberal ideology among MH and LGB people, contradicting Hypotheses 2a and 2b. For feminist identity, MH men were significantly less likely to indicate a feminist identity in comparison to MH women and bisexual women supporting Hypothesis 2a. Gay men were less likely to support the death penalty than bisexual men. In addition, MH men were less likely to support legal abortion than bisexual women. Across Table 2, there is evidence of sexuality and gender gaps in liberalism but mixed support regarding Hypothesis 2a and no support for Hypothesis 2b. See Footnote 2 for a summary of significant differences regarding MH and LGB comparisons to heterosexual men and women.

In Fig. 2, box and whisker plots reveal visual sexuality and gender gaps with heterosexual men as the least liberal as indicated by the lowest median (10) and lowest minimum value (4) on the liberal scale. Heterosexual women follow with a slightly higher median (11) and minimum value (5) as well as a different interquartile ranges (3 and 4, respectively). MH men report the same median (11) but a slightly higher minimum value (7) while MH women have a higher median (13) but lower minimum value (5) and smaller interquartile range (3 and 2, respectively). Lesbian women and gay men are similar on the liberal scale with a median of (13) but minimum values are lower among gay men (6) than among lesbian women (7). Bisexual men indicate a lower median (12) but higher minimum value (9) on the liberal scale. Bisexual women indicate the highest median (14) and minimum value (10) of all groups. ANOVA and post hoc Tukey-Kramer test results reveal significant sexual and gender identity differences in means with heterosexual men and women identifying significantly lower on the liberal scale than all others (except bisexual men). In addition, among MH and LGB respondents, MH men and women are significantly less liberal than bisexual women and bisexual women are more liberal than lesbian women. Overall, in exploring differences in means, Hypotheses 2a and 2b are not well-supported.

Fig. 2
figure 2

Box and whisker plot of liberal scale with ANOVA results identifying sexuality and gender gaps. ANOVA results, F = 46.53, df(7, 1932), p < .001; post hoc Tukey-Kramer test results by group number (see Table 2 for group numbers): group 1 ≠ (groups 2, 4, and 7); group 2 ≠ (groups 7 and 8); group 4 ≠ (groups 7 and 8); group 5 ≠ (groups 7 and 8); group 6 ≠ (groups 7 and 8); group 7 ≠ group 8

Full size image

OLS Regression Results

Sexuality Gaps

Table 3 estimates the effects of sexual identity and gender as they relate to political perspectives with exclusively heterosexual as the reference category. Across all model 1s, LGB is positively and significantly related to liberal ideology, feminist identity, support of legal abortion, and the liberal scale. In addition, LGB is negatively and significantly related to death penalty support. Thus, results for model 1 identify heterosexual-LGB gaps in liberalism. Similarly, across all model 2s, all three measures of sexual identity considered separately follow the same patterns, demonstrating heterosexual-mostly heterosexual, heterosexual-bisexual, and heterosexual-lesbian/gay gaps in liberalism whereby heterosexuals are less liberal than all other groups. Overall, Hypothesis 1a is fully supported.

Gender Gaps

Model 3 includes gender (woman) and interaction effects between sexuality and gender. The same patterns found in models 1 and 2 continue for mostly heterosexual and lesbian/gay; however, bisexual is no longer significant in the model estimating death penalty support. In addition, woman is positively related to liberal ideology, feminist identity, and the liberal scale but negatively related to support of the death penalty. The interaction effect representing MH women is positively related to support of legal abortion indicating an MH woman-MH man gap in these perspectives whereby MH women are more liberal than MH men, supporting Hypothesis 2a. The interaction effect representing bisexual women is positively related to liberal ideology and negatively related to death penalty support indicating a bisexual woman-bisexual man gap whereby bisexual women are more liberal than bisexual men, supporting Hypothesis 2b. In contrast, the interaction effect representing lesbian women operates in the opposite direction and is positively related to death penalty support and negatively related to liberal ideology, feminist identity, and the liberal scale indicating a lesbian woman-gay man gap whereby gay men are more liberal than lesbian women, contradicting Hypothesis 2b.

Controls and Goodness of Fit

Among the controls, African-American/Black is negatively related to death penalty while Asian/Pacific Islander is positively related to liberal ideology. Native American/Alaskan Native and other race are positively related to liberal ideology, feminist identity, and the liberal scale while Hispanic/Latinx is negatively related to feminist identity. Age is positive and significant in nearly all models while freshman is negative and significant in models estimating support of legal abortion and the liberal scale. The R2 values improve from model 1 (ranging from .02 to .06) to model 3 (ranging from .09 to .16).

OLS Regression Results

Sexuality Gaps

Table 4 estimates the effects of sexual identity and gender as they relate to political perspectives with mostly heterosexual as the reference category. Across all model 1s, LGB is only significantly related to death penalty support (negative) and the liberal scale (positive). Thus, there is only partial support in model 1 for MH-LGB gaps in liberalism (Hypothesis 1b). Similarly, across all model 2s, when measures of sexual identity are considered separately (bisexual and lesbian/gay) there are mixed results. Bisexual is positively related to feminist identity, support of legal abortion, and the liberal scale. Lesbian/gay is negatively related to death penalty support. Thus, there are some MH-B and MH-LG gaps in liberalism whereby mostly heterosexuals are less liberal than bisexuals and gays/lesbians, partially supporting Hypothesis 1b.

Gender Gaps

As seen previously, model 3 includes gender (woman) and interaction effects between sexuality and gender. Bisexual remains positive and significant only in the model estimating legal abortion support. Lesbian/gay remains negatively related to death penalty support but is also now positively related to liberal ideology, feminist identity, and the liberal scale. In addition, woman is positively and significantly related to feminist identity, support of legal abortion, and the liberal scale. The interaction effect representing bisexual women is positively related to liberal ideology and negatively related to death penalty support indicating a bisexual woman-bisexual man gap in these perspectives whereby bisexual women are more liberal than bisexual men, supporting Hypothesis 2b. In contrast, the interaction effect representing lesbian women operates in the opposite direction and is negatively related to liberal ideology, feminist identity, support of legal US abortion, and the liberal scale indicating a lesbian woman-gay man gap whereby gay men are more liberal than lesbian women, contradicting Hypothesis 2b.

Controls and Goodness of Fit

Among the controls, African-American/Black is negatively related to liberal ideology, legal abortion support, and the liberal scale. Freshman is negative and significant in all models except liberal ideology. No other controls are significant. The R2 values improve from model 1 (ranging from .02 to .06) to model 3 (ranging from .03 to .09), but are lower than those found in Table 3.

Discussion

The current study examined political perspectives with special attention to sexuality and gender gaps. Consistent with previous work (e.g., Grollman, 2017; Herek et al., 2010; Hertzog, 1996; Lewis et al., 2011; Swank, 2018a, 2018b) and Hypothesis 1a, there was a robust sexuality gap in liberalism between heterosexuals and all others. Indeed, similar to Hertzog’s (1996) work that identified a “lavender vote,” the current study’s findings also uncovered a distinct “lavender liberalism” among LGB college students. In addition, through including mostly heterosexuals, results demonstrated that exclusive heterosexuals are significantly less liberal than all others. This is especially informative because research indicates that MH-identified individuals are a growing and visible group on college campuses with estimates ranging between 7.7 and 22.6% of college women and 2.8 and 8.3% of college men (Savin-Williams & Vrangalova, 2013; see also Thompson & Morgan 2008). Thus, the current study’s findings both bolster existing work through identifying similar sexuality gaps among heterosexual-LGB college students (as found in more general population samples) and extend beyond previous studies through including MH people as a distinct group that also differs from exclusive heterosexuals. Overall, these sexuality gap findings support the stereotype that “all the gays are liberal.”

When other sexuality gaps were explored, findings were less consistent. For example, although it was hypothesized that MH people would be less liberal than LGB people (Hypothesis 1b), results were mixed. There were few significant differences in mean comparisons and the main effects of being MH and being LGB were similar in Table 3, model 2s, where the reference category was exclusive heterosexuals. However, in Table 4 where the reference category was mostly heterosexuals, some noteworthy differences emerged. First, in model 1s, compared to being MH, being LGB was significantly related to two measures of liberalism (death penalty support-negative and the liberal scale-positive). Second, in model 2s, being bisexual was significantly positively related to three measures of liberalism (feminist identity, legal abortion support, and the liberal scale) and being lesbian/gay was significantly negatively related to death penalty support. Thus, when compared to being MH, being LGB may be more strongly related to being liberal. In particular, MH college students may believe their romantic/sexual interests in the same sex are temporary or experimental, especially in the “hook up” and “casual sex” culture prominent on college campuses (Bogle, 2008; Rupp & Taylor, 2010). As a result, MH college students may not be (or feel) a part of the LGB community and this likely shapes their liberal social justice perspectives in ways that differ from LGB college students. By investigating MH people (instead of only LGB people) and comparing MH people to LGB people (instead of only comparing to exclusive heterosexuals), the current study allows for a more nuanced understanding of these relationships and demonstrates some significant MH-LGB sexuality gaps in liberalism.

As noted above, a sexuality gap in political perspectives between lesbian/gay and bisexual people has been found in previous literature whereby lesbian/gay people are more liberal than bisexual people (Grollman, 2017; Herek et al., 2010; Lewis et al., 2011; Schaffner et al., 2017; Swank, 2018a). Yet, the findings of the current study do not support this sexuality gap and are largely inconsistent with Hypothesis 1c. In fact, the results indicate an opposite pattern whereby bisexual people are more liberal than lesbian/gay people. For example, in mean comparisons found in Table 1, bisexuals were significantly more likely to support legal abortion than lesbian/gay people. In addition, although the main effects of being bisexual and being lesbian/gay were similar in Table 3, model 2s, where the reference category was exclusive heterosexuals, in Table 4 where the reference category was mostly heterosexuals, different patterns emerged. In particular (as previously noted), in model 2s, compared to being MH, being bisexual was positively related to three measures of liberalism examined in the current study while being lesbian/gay was only related to one measure. Thus, there may be a more robust relationship between being bisexual and being liberal than there is between being lesbian/gay and being liberal. Bisexual college students may be a particularly liberal-leaning group because they may feel motivated to connect to university groups and programs that support social justice in ways that differ from lesbian/gay college students. In particular, bisexual individuals often lack a concrete “bisexual community” and can sometimes feel displaced or erased from the larger “LGBTQ community” (Weiss, 2004). Thus, bisexual college students may look to other types of groups that support oppressed people, such as social justice groups. These experiences may have a unique impact on liberalism among bisexual college students.

Looking at the gender gaps among MH peoples’ political perspectives (Hypothesis 2a), there is some supporting evidence. For example, comparing means, MH women are significantly more feminist than MH men in Table 2. In addition, in Table 3, model 3s, there is one significant finding related to the MH gender gap in liberalism: compared to MH men, MH women are more likely to support legal abortion. Thus, there is a liberalism gender gap among MH people across some measures. MH women may be more liberal than MH men because they may be more likely to embrace social justice issues. Indeed, research indicates that MH women acknowledge their MH feelings at younger ages than MH men and are overall more likely to identify as MH than men are (Savin-Williams & Vrangalova, 2013). Coming to terms with not being “straight” at earlier ages coupled with enduring marginalization as “not straight” women likely impacts MH women’s liberalism and social justice perspectives in ways that differ from MH men. Such findings support the continued exploration of the MH gender gap in political attitudes.

Some evidence supporting gender gaps among bisexuals’ political perspectives (Hypothesis 2b) were also found in the current study. Although there were no significant differences in means, in Table 3, model 3s, where the reference category was exclusive heterosexuals, compared to being a bisexual man, being a bisexual woman was significantly related to two measures of liberalism (liberal ideology-positive and death penalty support-negative). In addition, in Table 4, model 3s, where the reference category was mostly heterosexuals, compared to being a bisexual man, being a bisexual woman was significantly related to the same two measures of liberalism (liberal ideology-positive and death penalty support-negative). Thus, there is a gender gap among bisexuals across multiple measures of liberalism. Bisexual women may be particularly liberal because they endure distinct overlapping experiences with stigma that differ from the stigma experiences of bisexual men. For example, bisexual women are hypersexualized in media, including pornography, and are often the object of the male gaze (Rupp & Taylor, 2010; Worthen, 2013). These oppressive experiences can motivate bisexual women to push back against these social injustices specifically, but can also stimulate more general liberal leanings toward politicized issues. Thus, bisexual women’s experiences with marginalization can impact their liberalism in unique ways that differ from bisexual men.

It is noteworthy that bisexual women are the most liberal group examined in the current study. This is most visible in Fig. 2, where bisexual women were significantly higher on the liberal scale than MH men, heterosexual men, and heterosexual women. As historically contentious within both the LGBTQ community generally and in lesbian spaces specifically (Rust, 1995; Weiss, 2004), bisexual women may be uniquely situated to be reflexive about both their gender and sexual identities in ways that relate to their political, social, and cultural cognizance. Indeed, some personal accounts of bisexual college women’s experiences indicate a very real socio-political awareness (Howard & Stevens, 2000) and may differ from lesbian women’s experiences with activism, feminism, and liberalism in important ways (Friedman & Leaper, 2010). Thus, there may be a specific bisexual woman consciousness that stimulates liberal social justice perspectives. This may be particularly common on college campuses where activism is visible (Van Dyke, 1998). Indeed, bisexual women’s negative perceptions of campus climate coupled with their experiences of feeling unsafe on college campuses and in residence halls and their high likelihood of enduring nonconsensual sexual contact in college (Cantor et al., 2015; Dugan & Yurman, 2011; Evans & Broido, 2002), likely impact bisexual women’s political awareness in unique ways. Together, these experiences may motivate bisexual women to become actively involved in advocacy including initiatives that support campus climate changes as well more general liberal social justice issues.

The current study’s examination of gender gaps among lesbian women’s and gay men’s political perspectives whereby lesbian women were theorized to be significantly more liberal than gay men (Hypothesis 2b) revealed perhaps the most surprising results. In fact, findings were in the opposite direction from hypothesized. For example, in Table 3, model 3s, where the reference category was exclusive heterosexuals, compared to being a gay man, being a lesbian woman was significantly negatively related to three measures of liberalism (liberal ideology, feminist identity, and the liberal scale) and positively related to death penalty support. Similarly, in Table 4, model 3s, where the reference category was mostly heterosexuals, compared to being a gay man, being a lesbian woman was negatively related to four measures of liberalism (liberal ideology, feminist identity, legal abortion, and the liberal scale). Thus, results from the current study indicate a gender gap among lesbian women’s and gay men’s political perspectives but it is gay men who are more liberal than lesbian women. This does not support the current study’s framework that lesbian women’s doubly stigmatized status contributes to increased liberalism and contrasts to previous work that has provided some evidence to indicate that lesbian women might be more liberal than (or mostly similar to) gay men (Herek et al., 2010; Hertzog, 1996; Lewis et al., 2011). Instead, these findings support Worthen et al.’s (2012) exploratory study focused on gay and lesbian people’s capital punishment attitudes whereby gay men were found to be significantly more empathic than lesbian women. There are two likely reasons for this. First, because heterosexuality is often conceptualized as a core component of masculinity, gay men (who are perceived of as violating masculinity norms) can sometimes be marginalized in different ways than lesbian women (Kimmel, 2009; McCreary, 1994; Worthen, 2013). Indeed, generally, college students indicate less favorable attitudes toward gay men than lesbian women (e.g., Herek, 1988; Worthen, 2012, 2013). This type of marginalization can motivate gay men toward liberal social justice activism and empathic concern for others. Second, the current study’s findings may indicate that lesbian college women are a distinct group concerned with politicized issues not examined here in ways that uniquely shape their liberal perspectives. Overall, however, the findings from the current study support a gender gap in liberalism whereby gay men are more liberal than lesbian women.

The exploration of gender (“woman”) in the regression models also deserves some discussion. In Table 3, model 3s, where the reference category was exclusive heterosexuals, being a woman was positively related to liberal ideology, feminist identity, and the liberal scale, negatively related to death penalty support, and not significantly related to abortion attitudes. These results are in line with previous studies (e.g., Cochran & Sanders, 2009; Hertzog, 1996; Lewis et al., 2011; Whitaker, 2008; Williams & Wittig, 1997). In Table 4, model 3s, where the reference category was mostly heterosexuals, being a woman was also positively related to feminist identity as well as the liberal scale. However, being a woman was no longer significantly related to death penalty support in Table 4 and was positively related to legal abortion support. Thus, in explorations of gender among only MH and LGB people, different gender gaps are revealed. Such findings suggest that the commonly expected and perhaps tacitly taken-for-granted gender gaps in political attitudes (i.e., death penalty support) are not universal but rather have important relationships to sexual identity that deserve further attention (see also, Worthen et al., 2012). In addition, while previous work does not provide strong evidence to indicate gender gaps in abortion attitudes (Bolzendahl & Brooks, 2005; Patel & Johns, 2009; Whitaker, 2008), the current study’s findings demonstrate that abortion attitudes may be shaped by the unique experiences of MH and LGB women and their alignment with women’s rights. Thus, explorations of political perspectives should be careful not to focus their analyses solely on comparisons to exclusive heterosexuals because such investigations may mask important findings related to sexuality and gender.

Finally, because many have speculated that sexuality gaps in political perspectives may stem from demographic differences among heterosexual and LGB people, it is noteworthy that the current study demonstrated robust sexuality and gender gaps even when including race, ethnicity, age, and student classification as controls among a college-educated sample. This is consistent with other non-college-student-focused work that highlights a unique relationship between being LGB and being liberal with an assortment of sociodemographic controls, including education (e.g., Grollman, 2017; Herek et al., 2010; Lewis et al., 2011; Swank, 2018a, 2018b; Schnabel, 2018). In addition, because the current study focused on students in the southern Bible Belt, the findings demonstrate particularly informative patterns that can help further our understandings of the potential ways MH and LGB college students engage with social justice advocacy in a culture that is typically characterized by a pervasive devotion to family, church, and state and conservative political values (Baunach et al., 2010; Sears, 1989).

Overall, the current study uncovered evidence of lavender liberalism consistent with other studies focused on LGB liberal political attitudes (Grollman, 2017; Herek et al., 2010; Hertzog, 1996; Lewis et al., 2011; Swank, 2018a, 2018b; Swank et al., 2013) and further expanded beyond existing research by incorporating MH people into these investigations. In doing so, the current study is the first to examine MH political perspectives and contributes to previous work that has highlighted the importance of exploring MH people as a distinct group (Savin-Williams & Vrangalova, 2013; Thompson & Morgan 2008). In addition, the current study moves beyond clustering LGB people into a single group as seen in some previous work (e.g., Schnabel, 2018; Swank, 2018b) and is the first to investigate gender gaps in political perspectives among both MH and LGB people. In doing so, some surprising findings emerged. For example, against predicted patterns, bisexuals were more liberal than gays/lesbians and gay men were more liberal than lesbian women. There was also evidence of a particular bisexual woman consciousness that may be uniquely shaped by overlapping layers of political, social, and cultural cognizance. Furthermore, this study is unique because it explored college students in southern Bible Belt. In sum, through examining liberalism sexuality and gender gaps of multiple kinds, the current study offers significant contributions to our understandings of the ways MH and LGB college students may embrace social justice perspectives and promote political changes.

Limitations and Future Research

The use of a non-generalizable relatively small, predominately heterosexual sample at one university located in a southern US state is a limitation of the current study. In particular, the LGB population was quite small (some cells are less than the optimal size for adequate power; see VanVoorhis & Morgan, 2007) so findings here should be considered provisional. Specifically, the significant differences found between lesbian women and gay men contrast with previous work (e.g., Herek et al., 2010; Hertzog, 1996; Lewis et al., 2011) and are in contrast to the theoretical framework and hypotheses set forth here, thus, these results may be unique to the current study’s population. Even so, there were other significant sexuality gaps found in the current study even among these small groups that deserve further attention. For example, there was a sexuality gap found in mean comparisons whereby bisexual men were significantly more likely than gay men to support the death penalty (see Table 2). Additional research with larger LGB samples could interrogate this finding further. There were also significant results in the current study related to race and ethnicity that could inform other work that focuses on intersectionalities (e.g., race, ethnicity, gender, class) and LGB perspectives. Thus, future studies would benefit from more diverse cross-national investigations of both college and non-college populations. In addition, this study utilized limited measures of political perspectives. Furthermore, the measurements of gender and sexual identity were limited and there were no measures of socioeconomic status available. Thus, future work might incorporate more response options to better capture these identities. It would also be beneficial to examine how embeddedness in LGB communities, families, relationships, friendships, and marriages may increase exposure and adherence to liberal perspectives (see DeBlaere et al., 2014; Grollman, 2017; Lewis et al., 2011; Szymanski, 2004). Finally, it would be especially valuable to examine how additional perspectives known to be associated with politicized attitudes among heterosexuals (e.g., religiosity, patriarchal gender norms, authoritarianism; see Cochran & Sanders, 2009; Inglehart & Welzel, 2005; McDaniel & Ellison, 2008; Unnever, Cullen, & Fisher, 2005; Whitaker, 2008; Worthen, 2017b) relate to liberalism among MH and LGB people. Such future work can not only allow us to better understand MH and LGB people’s liberal social justice perspectives but also can contribute to deeper understanding of the ways college students engage with political advocacy.