You are on page 1of 13

Sex Roles

https://doi.org/10.1007/s11199-018-0916-6

ORIGINAL ARTICLE

Gender, Masculinity Threat, and Support for Transgender Rights:


An Experimental Study
Brian F. Harrison 1 & Melissa R. Michelson 2

# Springer Science+Business Media, LLC, part of Springer Nature 2018

Abstract
We explore how gender, attitudes about traditional gender roles, and threats to masculinity and femininity affect U.S. participants’
support for transgender rights. First, we present analyses using data from the 2016 pilot survey of the American National Election
Survey (ANES) showing how men and women differ in their attitudes toward transgender people as measured by thermometer
ratings toward transgender people and questions about perceptions of whether they are victims of discrimination. Next, we
describe our randomized laboratory experiment, testing three hypotheses/predictions: (a) that men are less supportive of trans-
gender people and rights than women are, (2) that threatening a man’s masculinity increases opposition to transgender rights
whereas threatening a woman’s femininity has no effect, and (3) that this effect will be stronger among men who report that their
gender identity is very important to them. Consistent with existing scholarship, we find that women are more supportive of
transgender rights compared to men. More importantly, we also find that threatened masculinity is an even better predictor of
opposition to transgender rights than gender identity. In short, we find that attitudes toward transgender people and rights are
closely linked with the way people think and feel about their own gender identity and expectations of gender performance.

Keywords Transgender . Gender . Transgender rights . Political psychology . Gender identity . Masculinity . LGBT . Public
opinion . Bem Sex Role Inventory . BSRI . Threat

Transgender individuals and rights are increasingly part of Mack Beggs, a transgender male high school wrestler in Texas,
American culture and our national conversation as demonstrat- leapt onto the front page in early 2017 when he won the state
ed by cover-story magazine issues of National Geographic girls’ wrestling championship despite having asked state offi-
(January 2017) and Time (March 16, 2017). The ability to use cials to compete in the boys’ tournament (and again when he
public spaces like bathrooms and locker rooms that match defended his title in 2018). In July 2017, President Trump made
one’s gender identity is one of the most frequently discussed headlines when he tweeted about banning transgender people
issues surrounding transgender rights in the public sphere in the from serving in the U.S. military, reversing an Obama
United States. Other issues of transgender equality are also the Administration policy enacted in July 2015. Two federal judges
subject of public debate and policymaking, including gender have blocked the ban, noting that the administration’s justifica-
equity in sports, transgender people serving openly in the mil- tion for the ban was suspect and likely unconstitutional.
itary, and transgender candidates for public office. For example, The legal battle is ongoing. In November 2017, trans-
gender candidates made history by winning elections in
Electronic supplementary material The online version of this article several states: at least eight openly transgender people
(https://doi.org/10.1007/s11199-018-0916-6) contains supplementary won seats on city councils, school boards, and state
material, which is available to authorized users.
legislatures around the United States.
The National Center for Transgender Equality (NCTE) de-
* Melissa R. Michelson
melissa.michelson@menlo.edu
fines gender identity as a person’s Binternal knowledge of
[their] own gender—for example, your knowledge that you’re
1
Department of Political Science, Northwestern University, Scott Hall, a man, a woman, or another gender^ (National Center for
601 University Place, Evanston, IL 60208, USA Transgender Equality 2016, para 3). As they note, every-
2
Department of Political Science, Menlo College, 1000 El Camino one—not just transgender people—has a gender identity.
Real, Atherton, CA 94027, USA NCTE describes transgender people as follows:
Sex Roles

When we’re born, a doctor usually says that we’re male support of transgender people and rights, threats to gender
or female based on what our bodies look like. Most identity are even stronger predictors of support or opposition
people who were labeled male at birth turn out to actu- to transgender rights compared to gender identity alone.
ally identify as men, and most people who were labeled
female at birth grow up to be women. But some people’s
gender identity—their innate knowledge of who they Gender, Gender Ideologies, and Expectations
are—is different from what was initially expected when
they were born. Most of these people describe them- Existing research finds that U.S. attitudes toward LGBT peo-
selves as transgender. ple are often driven by feelings of discomfort and disgust
(Gadarian and Van der Vort 2017; Geller 1991; Gillig and
Scholars estimate that roughly .6% of the American popula- Murphy 2016; Haider-Markel et al. 2017; Herek 1984;
tion, approximately 1.4 million people, identifies as transgen- Miller et al. 2017). Values like authoritarianism (Norton and
der (Flores et al. 2016, p. 2). The vast majority of people are Herek 2013), beliefs in traditional gender roles (Nagoshi et al.
not transgender, also known as cisgender; their gender identity 2008), political conservatism (Norton and Herek 2013), and
is the same as the sex they were thought to be at birth. identifying as Republican (Flores 2015) are all associated with
Until recently, the fight for lesbian, gay, bisexual, and trans- more negative beliefs toward transgender people as well. Most
gender (LGBT) rights has been dominated by the issue of relevant to our research questions, surveys find that men are
same-sex marriage, often eclipsing other LGBT rights issues consistently less supportive of transgender rights than women
of greater concern to the transgender community (Harrison are. Norton and Herek (2013, p. 750) note: BMen’s greater
and Michelson 2017, chapter 7; Mananzala and Spade 2008; negativity toward transgender people is consistent with the
Nownes 2015). Increased media attention to transgender notion that they are more invested than women in adhering
equality and relevant public policy issues notwithstanding, to gender norms, presumably as a means of affirming their
relatively little is known about what drives public opinion own masculinity and heterosexuality.^
about transgender people and rights or how advocates and That gender identity is a strong and reliable predictor of
policymakers can most effectively communicate those issues. attitudes toward transgender people and rights is consistent
We do know that the U.S. public often supports transgender with research on the power of gender as a predictor of a wide
rights in the abstract while being less supportive of specific array of political attitudes (Erikson and Tedin 2015). Studies
policies that protect those rights. For example, a 2016 PRRI have also found that gender is one of the most powerful cog-
poll finds that 78% of Democrats and 61% of Republicans nitive schemas available (Brewer and Lui 1989; Starr and
support the abstract statement that laws should protect Zurbriggen 2017). People who are transgender often, by def-
LGBT people from discrimination in jobs, public accommo- inition, challenge the traditional conceptualization of gender
dations, and housing. identity as binary and immutable (Burdge 2007). We expected
The same survey, however, finds far less support for a more this challenge to affect men and women differently.
specific law allowing transgender people to use bathrooms Specifically, we hypothesized that gender differences in sup-
that correspond to their current gender identity: 64% of port for transgender people and rights would be exacerbated
Democrats, 44% of Republicans, and 51% of Independents by threats to men’s masculinity given the inherent threat to
(53% overall) report not supporting a more specific law about male gender identity often posed by transgender rights.
bathroom access (Cooper et al. 2016). Another recent survey This expectation stems from existing research that finds
finds that even those who are supportive of transgender men to be susceptible to perceived threats to masculinity, fear-
youth generally are less comfortable with allowing them ful of insufficient masculinity, and supportive of traditional
to use the restroom aligned with their gender identity or gender norms. The masculine overcompensation thesis posits
to share a room with a same-gender peer on a school that when men are concerned about the social implications of
trip (Elischberger et al. 2016). losing their masculinity (e.g., being seen as gay or otherwise
The aim of the present paper is to better understand what lacking masculine traits), they tend to overcompensate by
drives public opinion toward transgender people and the rights exhibiting extreme masculine behaviors and attitudes to create
they seek. Specifically, we investigate the impact of individual the impression that they are, indeed, masculine in socially
gender identity and attitudes toward traditional gender roles. desirable ways (Adams et al. 1996; Burke 1991; Burke and
After a brief overview of the concepts of gender identity and Stets 2009; Heise 2007). Masculine overcompensation has
gender expectations, we detail our findings from the January been cited as the cause of relatively benign behaviors like
2016 pilot survey of the American National Election Study buying sports cars at the onset of Bmid-life crises^ all the
(ANES). Next, we present results from an original laboratory way up to escalation of the Vietnam War by President
experiment at a large Midwestern research university. Overall, Johnson (Fasteau 1974; Kimmel 1996). Anti-gay hate crimes
our data show that although men and women do differ in their Bare tied closely to rigid and hierarchical ideas about
Sex Roles

masculinity that depend on differentiating ‘real’ men from Attitudes toward Transgender People
women as well as gay and bisexual men^ (Wade 2016, para 8). and Rights
Men heavily invested in performing masculinity at extreme
levels—a phenomenon often referred to as toxic masculinity Relying on this scholarship as well as to update and replicate
(Karner 1996)—are more likely to engage in domestic vio- existing data about public attitudes toward transgender people,
lence against their female partners (Macmillan and Gartner we examine questions about transgender people and rights in
1999), to sexually harass women (Maass et al. 2003), and even the January 2016 pilot of the American National Election
to commit mass shootings (Bridges and Tober 2017; Wade Study (ANES), including 1200 respondents (570, 47.5%,
2016). Contemporary conceptualizations of toxic masculinity men; 630, 52.5%, women). The survey included two items
rely on reaffirming superiority and dominance, leading certain about transgender people. The first asked respondents how
men to abuse people and power because of what Enloe (2017, much discrimination there is against transgender people in
p. 16) calls Bthe sustainability of patriarchy.^ Gender theorist the United States. The second asked respondents to rate trans-
Judith Butler suggests that toxic masculinity also underlies gender people with a feeling thermometer, ranging from 0 to
anti-transgender violence: BTrans women have relinquished 100 (cool–warm). These questions establish a general orien-
masculinity, showing that it can be, and that is very threaten- tation toward transgender people, asking men and women to
ing to a man who wants to see his power as an intrinsic feature place them in social contexts and to assess the degree to which
of who he is^ (quoted in Tourjée 2015, para 9). As many boys their experiences rise to the level of discrimination. Given
grow into men, Bthey learn that they are entitled to feel like a established gender gaps on attitudes about transgender people,
real man, and that they have the right to annihilate anyone who we expected women to perceive higher levels of discrimina-
challenges that sense of entitlement^ (Kimmel 2012, quoted tion against transgender people and to report warmer feeling
in Wade 2016, para 9). thermometer ratings. Data support these expectations. Women
Willer et al. (2013) explore the impact of threats to both were more likely to recognize that transgender people experi-
masculinity and femininity, finding that men who are told that ence a great deal of discrimination (28.6%) compared to men
they are feminine are more supportive of war, more homopho- (20.7%), a difference of 7.9 percentage-points, and this differ-
bic, more interested in purchasing an SUV, and more support- ence is statistically significant, χ2(1) = 10.030, p = .002,
ive of dominance hierarchies and male superiority compared Cramer’s V = .09. Women were less likely to answer that
to men who are told that they are masculine. These reactions transgender people experience a little discrimination or none
are particularly strong among men with higher levels of tes- at all (women 19.2%; men 27%, a difference of 7.8 percent-
tosterone as measured by saliva tests. They do not find a age-points). This difference is also statistically significant,
parallel effect among women whose femininity is threatened, χ2(1) = 10.242, p = .001, Cramer’s V = .09. On the feeling
however; there are no measurable differences in attitudes be- thermometer question, the mean response among men was
tween women who are told that they are masculine and wom- 47.2 (SD = 31.747) degrees and the mean response among
en who are told that they are feminine. Based on these results, women was 54.7 (SD = 31.355) degrees; the difference of
we expected that men whose masculinity was threatened 7.5 degrees is both substantively large and statistically signif-
would be less supportive of transgender people and rights icant, t(1197) = 7.50, p < .001, Cohen’s d = .24.
compared to men whose masculinity was not threatened. We We further hypothesized that gender gaps would be less
did not anticipate finding differences among women regard- powerful predictors of attitudes about transgender people
less of the strength of their gender identities or whether that when controlling for attitudes about traditional gender roles.
gender identity was threatened. We explored this point using two ANES Pilot items. The first
We further hypothesize that effects should vary by strength item asked: BDo you favor, oppose, or neither favor nor op-
of gender identity. Individuals vary in the degree to which they pose requiring employers to pay women and men the same
support and find personally relevant core dimensions of amount for the same work?^ The second item asked: BDo you
gender-differentiated normative standards (that men should favor, oppose, or neither favor nor oppose requiring em-
be dominant and independent and that women should be help- ployers to offer paid leave to parents of new children?^
ful and concerned about others) (Wood et al. 1997). Responses to both items were reported on 7-point Likert
Strength of gender identity and connection to it can scales ranging from 1 (favor a great deal) to 7 (oppose a great
vary across one’s lifespan in response to an array of deal). We also include control variables for partisan identifi-
factors (Ely 1995; Martin and Ruble 2010; Sinclair cation and political ideology. Party identification is coded on a
et al. 2006; Tobin et al. 2010; White and Gardner 7-point scale ranging from 1 (strong Democrat) to 7 (strong
2009). We hypothesize that men with strong gender Republican). Partisan ideology is coded on a 7-point scale
identities will be more sensitive to manipulated threats ranging from 1 (very liberal) to 7 (very conservative).
to their masculinity compared to men with gender iden- As shown in Table 1, both of these questions mea-
tities that are less central to their identity. suring gender traditionalism are statistically significant
Sex Roles

Table 1 Predicting attitudes


toward transgender people and Predictors Transgender feeling thermometer Discrimination against transgender people
support for gender equality, 2016
ANES pilot Model 1 Model 2 Model 3 Model 4
M (SE) M (SE) M (SE) M (SE)

Gender 3.94 (2.41) 5.15* (1.75) −.05 (.07) −.10 (.07)


Equal pay −3.27* (.85) −1.75* (.60) .11* (.02) .07* (.02)
Parental leave −3.42* (.76) −1.33* (.56) .17* (.02) .13* (.02)
Party ID – −1.41* (.48) – .07* (.02)
Political ideology – −10.56* (1.06) – .16* (.04)
Constant 59.51* (4.73) 86.19* (4.13) 1.99* (.13) 1.50* (.16)

n = 1200 for Models 1 and 3; n = 1069 for Models 2 and 4. Data are weighted to be nationally representative.
Gender is coded 1 = male, 2 = female. Party ID is coded 1 = strong Democrat to 7 = strong Republican; Political
Ideology is coded 1 = very liberal to 7 = very conservative
*p < .05

predictors of attitudes toward transgender people. As and femininity) should also be significant predictors of
shown in Model 1, respondents who oppose equal pay attitudes about transgender people. As we noted, the
for men and women or who oppose paid parental leave concept of transgender identity likely threatens those
have cooler feelings toward transgender people and per- for whom the concept of a traditional, immutable gender
ceive less discrimination. This finding persists in Model is essential to their own individual identity. Identity loss
2 when controlling for partisan identification and polit- theory, originating from economic decision-making, sug-
ical ideology. As shown in Model 3, respondents who gests that because many people view gender as having
have cooler feelings toward transgender people or who only two discrete social categories, violating these ex-
oppose paid parental leave perceive less discrimination pectations Bevokes anxiety and discomfort in oneself
against transgender people. Again, this finding persists and in others. Gender identity, then, changes the ‘pay-
when controlling for party identification and ideology. offs’ from different actions^ (Akerlof and Kranton 2000,
These attitudes are more consistent and stronger predic- pp. 716–717). In other words, when a conventional
tors than gender identity alone, which is not statistically Bmale or female^ conceptualization of gender expression
significant in three of the four models. This suggests is threatened, attitudes and behaviors toward that which
that different conceptualizations and expectations of threatens the identity (i.e. transgender people and rights)
gender identity and gender roles (as indicated by re- should change as well.
sponses to the two gender equity items) may be a stron- Existing research shows that identity threat translates to
ger predictor of attitudes toward transgender individuals more negative attitudes toward gay and lesbian people. As
and support for transgender rights. It may also be the we noted, Willer et al. (2013) found increased homophobia
case that attitudes toward gender roles and transgender among men whose masculinity was threatened. In part, their
people are all driven by a common underlying factor of finding stems from the same threat to gender traditionalism
conservative ideology. We control for political ideology that LGBT identity and rights represent. One area of current
in the analysis of our subsequent experimental data. debate that is likely to trigger a visceral threat to masculinity is
These results are consistent with recent scholarship that bathroom access for transgender people. As Cavanagh (2010,
finds that perceptions of and expectations about masculinity p. 4) notes: BNowhere are the signifiers of gender more pain-
and femininity, known predictors of support for traditional fully acute and subject to surveillance than in sex-segregated
gender roles, are stronger predictors of political attitudes than washrooms. The perceived loss of a binary gender axis, the
is self-reported gender identity. In particular, McDermott grid upon which normative heterosexuality depends for co-
(2016) argues that traditional understandings of gender- gency and intelligibility, incites anxiety about gender
based differences in political attitudes and behaviors are actu- incoherence.^ The policy issue of transgender bathroom
ally based on gender expectations and traits of masculinity and access is a threat to the underlying ontological structure
femininity. We found support for this idea: support for gender of the strict binary gender expression (Roughgarden
roles as measured by the 2016 ANES Pilot study was more 2013). For some people, the idea that they might
predictive of attitudes toward transgender people and rights encounter a transgender person while using a public
than reported gender identity. bathroom or locker room is perceived as a real (and intensely
Building on these initial findings, we hypothesized personal) potential threat (Schlit and Westbrook 2015;
that threats to traditional gender norms (i.e. masculinity Westbrook and Schilt 2014).
Sex Roles

Priming Gender Identity Method

To test the effect of individual-level identity on attitudes Participants


toward transgender people and rights, we rely on the
prominent theory of priming (Druckman 2004; Iyengar Participants were 182 undergraduate students at a large U.S.
and Kinder 1987; Krosnick and Brannon 1993; McLeish Midwestern university (Mage = 19.31, SD = .15, range = 17–
and Oxoby 2008). Previous research has demonstrated 39). After providing informed consent, they were asked a set
that priming can effectively manipulate gender self- of demographic questions including their gender identity, year
concepts among both men (McCall and Dasgupta of birth, political ideology [ranging from 1 (very progressive)
2007) and women (Haines and Kray 2005; Rudman to 5 (very conservative)], their racial and ethnic identity, and
and Phelan 2010). Social identity theory (Tajfel 1981; their gender. The sample was roughly equal in terms of gender
Tajfel and Turner 1986) suggests that people have both identity: n = 89, 48.9%, men; n = 93, 51.1%, women. About
personal identities as unique individuals and also social half of respondents reported they were White, non-Hispanic
identities as members of groups. (n = 90, 49.45%), with the remainder identifying as Black
Although social identity is often thought of in cate- (n = 12, 6.59%), Hispanic or Latinx (n = 10, 5.49%), Asian
gorical terms, significant evidence demonstrates that any American (n = 32, 17.58%), other (n = 4, 2.20%), or more
given social identity can be a more important source of than one category (n = 34, 18.68%). When asked to describe
identity for some members of a group than for others. their political views generally speaking, most reported their
For example, although all women may identify as mem- ideology as progressive (n = 119, 64.38%) whereas 46
bers of the social category Bwomen,^ there is consider- (25.27%) respondents reported their ideology as moderate,
able variation in the degree to which each individual and the remaining 17 (9.34%) respondents reported their ide-
woman considers that membership as central to or as ology as conservative. Most respondents reported their marital
an important part of her self-identity (Burn et al. status as single, never married (n = 176, 96.7%), and none of
2000; Schmader 2002). As a result, identity importance the respondents had any children. Most respondents
or salience is a key measure of identity priming effects (n = 148, 81.32%) reported their annual income as be-
as well. Klar (2013) notes that priming can increase an low $20,000, and most respondents (n = 154, 84.62%)
identity’s salience and, subsequently, related concern for reported that none of their Bimmediate family members,
identity-based interests, particularly when a strongly- relatives, neighbors, co-workers, or close friends^ was
held identity is threatened. Priming gender identity and transgender or gender non-conforming.
threats to gender norms can test the power of gender
construction in shaping attitudes and behavior, particu- Procedure and Materials
larly toward issues like transgender rights that draw so
heavily on individual concepts of gender. Men and women were randomly separated into six groups,
with 29–31 individuals randomly assigned to one of six con-
ditions. Replicating Willer et al. (2013), two-thirds of our
participants were assigned to take the Bem Sex Role
The Present Study Inventory (BSRI) and one third were assigned to complete
the Big 5 personality test (Goldberg 1992; Tupes and
Next, we turn to a survey experiment that experimentally Christal 1961). Of those assigned to the BSRI, half (one-
manipulates threats to participants’ masculinity or femi- third of all participants) were randomly assigned to the gender
ninity to test their effects on attitudes and behaviors to- threat condition and half (one-third of all participants) were
ward transgender people and key rights for that commu- assigned to the non-threat condition. Participants assigned to
nity. We test three hypotheses/predictions: (a) men will be one of the BSRI conditions were asked to answer the 60 ques-
less supportive of transgender rights compared to women tions that compose the original BSRI which captures how
(Hypothesis 1); (b) threatening a male respondent’s mas- strongly a respondent endorses qualities considered socially
culinity will increase opposition to transgender rights desirable in women and in men. In other words, it measures
(Hypothesis 2a); we predict that threatening a female re- how well a respondent endorses traditional gender roles.
spondent’s femininity will neither increase nor decrease Although participants were administered the real BSRI test
opposition to transgender rights (Prediction 2b); and (c) in our experiment, the scores they were shown on a scale of 0–
a threat to masculinity will increase opposition to trans- 50 were false. (The marked scales we used to share this
gender rights more significantly among men who self- false feedback with participants is available in an online
report that gender is essential to their identity compared supplement.) Replicating the design by Willer et al.
to men who do not (Hypothesis 3). (2013), upon completing the BSRI respondents received
Sex Roles

a false BSRI score which was either threatening or non- 3. Suppose your state had a ballot initiative in an upcoming
threatening to their masculinity or femininity. Men ran- election where you could vote on a law that would require
domly assigned to the threatening condition were told that public bathrooms and locker rooms to be divided in the
their BSRI score was in the feminine range (BYour Score: traditional way, male or female. This law would require
32^), whereas those randomly assigned to a non- transgender people to use the bathroom of the sex they
threatening condition were told that their BSRI score fell were assigned at birth, not the one with which they cur-
in the masculine range (BYour Score: 11^). Similarly, rently identify. How would you vote on such a measure?
women in the threatening condition were told that their Note that a Byes^ vote would mean requiring bathrooms
BSRI score was masculine (BYour Score: 18^), whereas divided by male and female; a Bno^ vote would mean
those in the non-threatening condition were told that their allowing transgender people to choose which bathroom
BSRI score was feminine (BYour Score: 39^). they want to use. (1) Definitely vote yes, (2) probably vote
Assignment to the Big 5 treatment acts as a placebo: par- yes, (3) undecided, (4) probably vote no, (5) definitely
ticipants in these groups also answered a series of survey items vote no.
about themselves and the same dependent variables about
transgender rights. Participants in the Big 5 treatment groups The first two items had binary responses and re-
were asked a traditional set of 18 items meant to measure their sponses on the ballot item were collapsed into two re-
score on the Big 5 personality items, items which do not ref- sponses: 1 (definitely or probably vote yes), 0
erence gender or gender expectations. Thus, individuals (undecided and probably/definitely no).
assigned to these conditions compose the baseline for com- After the questions about sexism and gender roles, respon-
parison for the other treatment groups. After completing these dents were asked about their party identification and their
items, all participants in the Big 5 treatment groups received personal identity. Partisanship was coded as a traditional 7-
the same false report, the specifics of which are reported in the point Likert scale from 1 (strong Democrat) to 7 (strong
online supplement. Republican). The personal identity question asked: BWe’d like
you to think about your own personal identity; that is, the
Post-Manipulation Measures: Support ways you define yourself as a person. How important are each
for Transgender Rights of these characteristics [gender, race and ethnicity, age, marital
status, employment status, religion, or sexual orientation] to
All participants were then asked a set of questions about trans- your own personal identity?^ Responses were recorded on a
gender rights, sexism and gender roles, partisanship, and their scale from 1 (essential to your identity) to 4 (not important at
personal identity characteristics. The dependent variables of all). Because very few respondents (n = 20, 11.1% [men: n =
interest for the present study include three items: (a) the first 16, 17.98%; women: n = 4, 4.30%]) chose answers 3 or 4 to
asks respondents about transgender access to public bath- characterize their gender identity, we dichotomized this vari-
rooms and locker rooms that match their gender identity able into just two categories, collapsing answers 2–4. The
(Bathrooms), (b) the second is about whether the government survey ended with a debriefing item that noted that their test
should protect transgender people against job discrimination score had been false and reminded them of contact informa-
(ENDA), and (c) the third asks how respondents would vote tion for the Institutional Review Board.
on a hypothetical ballot initiative on transgender access to
public bathrooms and locker rooms that match their
gender identity (Ballot). The following presents these
materials verbatim: Results

1. As you may know, transgender people are those who Preliminary Analyses
identify as a gender different from the one from when they
were born. Some think that bathrooms should be orga- We conducted a randomization check of ideology and real
nized by sex; others think that people should be able to BSRI scores for men and women. Men had the same average
choose which bathroom or locker room fits their gender BSRI score in both the threatening (M = 60.13, SD = 15.27)
identity. What about you? Do you think people who are and non-threatening (M = 60.83, SD = 20.31) conditions,
transgender should: (1) be allowed to use the bathrooms t(1) = .15, p = .882. The same is true for women in the threat-
and locker rooms of their preferred gender or (2) have to ening (M = 52.92, SD = 14.62) and non-threatening (M =
use the bathrooms and locker rooms of the gender they 48.68, SD = 19.66) conditions, t(1) = −.82, p = .415. In addi-
were born? tion, men, F(4, 84) = .50, p = .735, and women, F(4, 88) =
2. Do you favor or oppose laws to protect transgender peo- 1.63, p = .173, did not vary by political ideology across all
ple against job discrimination? three conditions (threatening, non-threatening, and control).
Sex Roles

Hypothesis Testing: Simple Comparisons repored that they would vote to protect transgender ac-
cess rights compared to 24 (83%) in the non-threat con-
First, we examined differences in attitudes about transgender dition and 24 (80%) in the Big5 condition, χ2(2) = 4.79,
rights between men and women. Overall, 88 (94.6%) women p = .091, V = .232.
in the sample supported transgender access to public bath- In contrast, women assigned to the threat condition were
rooms and locker rooms compared to 75 (84.3%) men, generally no more or less supportive of transgender rights
χ 2 (1) = 5.22, p = .022, Cramer’s V = .17. Similarly, 82 compared to women assigned to the non-threat and baseline
(88.2%) women said they would definitely or probably vote conditions, with one exception. Among women assigned to
to protect transgender access on a hypothetical ballot initiative the threat condition, 26 (84%) in the threat treatment condition
compared to 66 (74.2%) men, χ 2 (1) = 5.88, p = .015, said that they would vote to support transgender access rights
Cramer’s V = .18. However, no significant difference was on a hypothetical ballot measure compared to 31 (100%)
found in support for laws to protect transgender people against women in the non-threat condition and 25 (81%) women in
job discrimination, with both men (78, 95.1%) and women the Big5 condition; this difference is statistically significant,
(87, 98.9%) reporting strong levels of support, χ2(1) = 2.08, χ2(2) = 6.39, p = .04, V = .262. Thus, our Prediction 2b was
p = .149. In sum, although we found that a majority of both only partially supported.
women and men had favorable attitudes, we found that our No statistically significant responses were found between
first hypothesis about differences between women and men on women assigned to the threat, non-threat, and baseline condi-
transgender rights was consistently supported on items about tions for the other two dependent variables, although women
transgender access to public bathrooms and locker rooms but in the threat condition were consistently less supportive than
not on an item about laws to protect transgender people women in the non-threat condition. These findings suggest
against job discrimination. that women who are assigned to a condition that threatens
Next, we used Chi-square tests to examine our second hy- their femininity think differently about transgender rights
pothesis regarding gender identity threat for men and women when asked about their attitudes compared to when they are
and its impact on support for transgender rights (see Table 2). asked about their own future hypothetical behavior.
Men assigned to the threat condition (i.e., told that they were Further research is needed to confirm whether this is a
in the feminine range on the false BSRI test) were consistently robust difference.
less supportive of transgender rights compared to those Before moving on to test Hypothesis 3, we used Chi-square
assigned to the non-threat condition (i.e., reportedly in the tests to examine whether men and women who self-report that
masculine range) or to the baseline Big 5 condition. In support gender is essential to their identity are less supportive of trans-
of Hypothesis 2a, only 21 (70%) men assigned to the threat gender rights (see Table 3). Overall, 67 (72.0%) female re-
condition said that they supported transgender access rights to spondents said that gender was essential to their identity as
public bathrooms and locker rooms compared to 26 (90%) did 43 (48.3%) male respondents. Among men who reported
assigned to the non-threat condition and 28 (93%) in the that gender is essential to their identity, 32 (74%) supported
Big5 condition, χ2(2) = 7.10, p = .029, V = .283. There were transgender access rights to public bathrooms and locker
also statistically significant differences on the hypothetical rooms that match their gender identity compared to 43
ballot measure about access to public bathrooms and locker (94%) men with non-essential gender identities, χ2(1) =
rooms: only 18 (60%) men assigned to the threat condition 6.09, p = .014, Cramer’s V = .26. Similarly, 26 (60.5%) men

Table 2 Support for transgender rights by gender and experimental condition

Men Women

Non-threatening Threatening Big5 Non-threatening Threatening Big5


control control

Q1: Bathrooms (% yes access) 89.7% 70%a,b 93.3% 96.8% 90.3% 96.8%
(26/29) (21/30) (28/30) (30/31) (28/31) (30/31)
Q2: ENDA (% yes protect) 96.2% 92.9% 96.4% 100% 96.4% 100%
(25/26) (26/28) (27/28) (30) (27/28) (30)
Q3: Ballot Bathrooms (% yes protect access) 82.2% 60%a,b 80% 100% 83.9%a 80.7%
(24/29) (18/30) (24/30) (31) (26/31) (25/31)

Numbers in parentheses are the number responding Byes^/cell size


a
indicates a statistically significant difference (p < .05) between the threatening vs. non-threatening conditions within each gender
b
indicates a statistically significant difference (p < .05) between the threatening vs. Big5 control conditions within each gender
Sex Roles

Table 3 Support for transgender


rights, by gender and strength of Men Women
gender identity
Not Essential Essential Not Essential Essential

Q1: Bathrooms (% yes access) 93.5 74.4a 96.2 94


(43/46) (32/43) (25/26) (63/67)
Q2: ENDA (% yes protect) 97.6 92.5 100 98.4
(41/42) (37/40) (25/25) (62/63)
Q3: Ballot Bathrooms (% yes protect access) 87.0 60.5a 92.3 86.6
(40/46) (26/43) (24/26) (58/67)
a
indicates a statistically significant difference (p < .05) across subgroups (essential vs. nonessential) within a
gender

who reported that their gender is essential to their identity 3.10, 95% CI [1.17, 8.23]. Conservative respondents were
would vote to protect transgender access rights on a hypothet- slightly less likely to say that they would vote for access rights
ical ballot initiative compared to 40 (87%) men with non- (ß = −1.59, SE = .30), Wald(1) = 27.56, p < .001, odds ratio =
essential gender identities, χ2(1) = 8.14, p = .004, Cramer’s .20, 95% CI [.11, .37].
V = .30. Among women, there were no statistically significant We next conducted a set of binomial logistic regressions on
differences in support for transgender rights across levels of male respondents only, testing the effects of assignment to the
gender identity strength. gender threat or cued gender conditions on support for trans-
gender rights. These models also include race, conservative
Hypothesis Testing: Logistic Regressions ideology, age, and whether the respondent personally knew
anyone who was transgender. Models including an interaction
We next conducted a set of binomial logistic regressions to term testing the effect of a threatened gender identity among
further explore our hypotheses using more complex models respondents who report that their gender is essential to their
that included a wider range of measures. First, we tested the identity, testing Hypothesis 3, failed to converge so that vari-
effects of gender, race, conservative ideology, age, and wheth- able was dropped from the models. The first logistic model,
er the respondent personally knew anyone who was transgen- predicting support for transgender access rights among male
der on their likelihood of supporting transgender rights. The respondents, was significant, χ 2 (7) = 20.19, p = .01.
logistic model predicting support for transgender access rights Conservative respondents were slightly less likely to say they
is significant, χ2(5) = 33.69, p < .001. Men were slightly less support transgender access rights (ß = −1.13, SE = .43),
likely to say they support access rights (ß = −1.45, SE = .64), Wald(1) = 6.81, p = .01, odds ratio = .32, 95% CI [.14, .75].
Wald(1) = 5.11, p = .02, odds ratio = .24, 95% CI [.07, .83]. The model predicting support among male respondents for
Additionally, Anglo (non-Hispanic White) respondents were laws to protect transgender people against job discrimination
nearly four times more likely to support transgender access was not statistically significant, χ2(7) = 8.25, p = .31. The
rights (ß = 1.33, SE = .62), Wald(1) = 4.54, p = .03, odds ra- model predicting votes among male respondents on a hypo-
tio = 3.77, 95% CI [1.11, 12.78]. More conservative partici- thetical ballot measure in support of transgender access rights
pants were slightly less likely to support transgender access was significant, χ2(7) = 30.43, p < .001. Conservative respon-
rights (ß = −1.49, SE = .35), Wald(1) = 18.06, p < .001, odds dents were slightly less likely to say they would vote to sup-
ratio .23, 95% CI [.69, 1.91]. port transgender access rights (ß = −1.54, SE = .40),
The logistic model predicting support for laws to protect Wald(1) = 14.67, p < .001, odds ratio = .21, 95% CI [.10, .47].
transgender people against job discrimination was statistically Multivariate logistic regression coefficient estimates can-
significant, χ2(5) = 13.38, p = .02, but none of the individual not be calculated for the subsample of women respondents
coefficient estimates are statistically significant at the p < .05 because there is so little variation in their responses to the
level. The logistic model predicting votes on a hypothetical dependent variables. (For example, only one female respon-
ballot measure in support of transgender access rights was also dent said that she opposed laws to protect transgender people
significant, χ2(5) = 50.69, p < .001. Here, men were slightly against job discrimination.) As a result, in many cases, these
less likely to say they would vote for transgender access rights multivariate models do not directly test Hypothesis 2a,
(ß = 1.39, SE = .51), Wald(1) = 7.56, p = .01, odds ratio = .25, Prediction 2b, or Hypothesis 3. However, there were several
95% CI [.09, .67]. White participants were more than three interesting findings in the results. Men were overall less sup-
times more likely to say that they would vote for access rights portive of expanding rights for transgender people compared
(ß = 1.13, SE = .50), Wald(1) = 5.15, p = .02, odds ratio = to women. Conservative political ideology was also a fairly
Sex Roles

consistent predictor of opposition to transgender rights. to those assigned to the BSRI non-threatening condition; this
Finally, Anglo (non-Hispanic White) respondents were con- difference is large and statistically significant for the item
sistently more supportive compared to Respondents of Color. asking about a hypothetical ballot measure. In contrast, there
were no differences in responses to the item asking about
support for a law protecting transgender individuals against
Discussion job discrimination. In part, this reflects the very high levels of
support across all six conditions.
We hypothesized that men would be less supportive of trans- Our final hypothesis was that this effect would be stronger
gender rights compared to women and that threatening the among men with stronger gender identities. Although this
masculinity of men would increase their opposition to trans- hypothesis was not confirmed, we did find that men who
gender rights, particularly if they self-reported that their gen- reported that their gender identity is essential to them were
der identity was essential to their overall identity. Although far more likely to oppose transgender rights compared to
men were overall less supportive than women were, in many men who reported that their gender identity is not essential
cases, threats to masculinity led men to be even less support- to them and to women. Finally, the indicator variable of
ive of transgender rights to access public bathrooms and lock- whether the respondent knew a transgender person was not a
er rooms that match their gender identity. Men who self- statistically significant predictor of attitudes about transgender
reported that gender is an essential part of their identity were rights. This finding, although tangential to our study,
less supportive of those rights but this measure did not interact contributes to a growing literature on the effect of
with threats to masculinity. interpersonal contact on attitudes about transgender people.
Our experimental study investigated the role of gender, For example, Flores (2015) found no effect whereas King
strength of gender identity, and threats to masculinity and et al. (2009) and Tadlock et al. (2017) found a positive effect
femininity on support for transgender rights. Overall, we of contact.
found that men are consistently less supportive of transgender
rights compared to women. Men assigned to a condition Limitations and Future Research Directions
where they were told that they were feminine were less likely
to support transgender rights compared to men assigned to The present study is limited due to several factors of our ex-
conditions that either reported that they were masculine or that perimental design. First, our experimental sample is U.S. un-
did not prime gender. Men who self-reported that gender is dergraduate students, which potentially limits the external va-
essential to their identity were also less supportive of trans- lidity of our results, particularly given that young people are
gender rights compared to men who did not. There were no more supportive of transgender rights compared to older peo-
statistically significant differences among women who report- ple. Our sample also predominantly identified as progressive
ed that gender is essential to their identity compared to other in terms of political ideology in proportions not representative
women, but women who were assigned to a threat condition of the general public. In addition, our experiment did not in-
that reported that they are masculine were less likely to say clude a manipulation check for threats to masculinity. Future
that they would vote to protect the rights of transgender people research may want to include explorations of how respondents
to access public bathrooms and locker rooms that match their are reacting to assignment to conditions meant to threaten their
gender identity. In sum, we found that importance of gender gender identity.
identity to one’s self-concept is an important motivator of Another limitation of our paper is that we do not have
attitudes and behavior toward transgender people and rights. different question wordings and experimental conditions to
In theory, it may be possible to motivate individuals to revisit test how people think and feel about transgender men com-
and revise those attitudes if perceptions of gender norms and pared to transgender women. Our dependent variables fo-
expectations can be mitigated. cused, in part, on transgender access rights to public bath-
Men were less likely to support bathroom and locker room rooms and locker rooms, a public debate that generally focus-
rights for transgender people and were more likely to say that es on the rights of transgender women. Thus, different results
they would vote for a ballot measure restricting those rights. may have been generated by survey questions more focused
Women, as expected, were generally more supportive of those on transgender rights issues that bring to mind transgender
rights. Men assigned to the masculinity-threatening condition men. Transgender women do not represent all transgender
were more likely than other respondents were to oppose ac- people writ large, although they do remain the most stigma-
cess rights and more likely to say they would vote to restrict tized group of people within the transgender community.
those rights. These differences are statistically significant, Many messages label these individuals as Btroubled men^
supporting Hypothesis 2a. Women assigned to the rather than as transgender women; they reject the validity of
femininity-threatening condition also tended to be less sup- the identity group by refusing to use the word Btransgender.^
portive of bathroom and locker room access rights compared Additional work should identify and test differences in
Sex Roles

attitudes among transgender men, transgender women, and help advocates and policymakers to frame their proposals to
gender non-conforming individuals as well. mitigate perceived threats to masculinity.
Another avenue of possible future research is to investigate Another practice implication from our findings stems from
how real-world transgender rights issues other than bathroom the stronger levels of support for transgender bathroom access
access may trigger threats to masculinity (or femininity). For rights when framed as transgender people Bshould be
example, perceptions of threats to concepts like traditional allowed^ to exercise that right compared to when it was posed
family values, traditional American society and culture, con- as a hypothetical ballot measure. These differences suggest
servative political values and ideologies, and children are like- that policy changes that are made administratively, without
ly triggered in different ways by different transgender rights the need for action on the part of the general public, may be
issues. These threats should be addressed in a more robust more palatable. In other words, future efforts to recognize and
fashion in future work. Further, there are likely emotional protect the rights of transgender people may be more success-
reactions other than threat that are at play. Additional work ful if they are advanced directly by policymakers rather than
should investigate which emotions are being triggered by dis- via public referenda.
cussions of transgender people and explore ways to ei-
ther inoculate against such emotional reactions or to
disarm those emotions to encourage rational and calm Conclusion
discussion. Additional research might also find different
ways of discussing transgender rights issues in ways In many ways, contemporary anti-transgender rhetoric paral-
that evoke positive emotions. lels historical attacks against lesbian and gay people: The va-
Respondents’ gender identity, threats to gender identity, lidity of those identity groups was called into question, with
and conceptualizations of gender expression are just some of many people describing lesbian and gay people as pedophiles,
the predictors of attitudes toward transgender people. As we sexual deviants, or mentally ill. Similarly, today’s opponents
noted, attitudes about transgender people and rights have other to transgender equality often call into question the legitimacy
predictors as well, including feelings of discomfort and dis- of transgender identity, as in the fight against transgender
gust, authoritarian values, political conservativism, and rights in Houston, Texas that referred to transgender women
Republican partisanship. Future experimental work might ex- as Btroubled men^ (Moyer 2015, para 11).
plore the links between those attitudes and how they are inter- There has been some progress toward increasing accep-
related as well as possible mechanisms for breaking those tance of transgender people, in part due to portrayals in the
links and crafting persuasive messages that open individuals media (Gillig et al. 2017). This extension of equality to more
to the possibility of attitudinal shifts. Finally, our dependent and more groups, progressing from equality for lesbians and
variables refer to access to public bathrooms and locker rooms gay men and then expanding to include transgender people is
as is consistent with actual text from ballot initiatives around reminiscent of Charles Taylor’s social imaginary—the idea
the United States. It is possible, however, that respondents that in the modern moral order, the presumption of equal-
have differing attitudes toward bathrooms versus locker ity is Bapplied in more and more contexts, ending with the
rooms so that these questions constitute double-barreled ques- multiple equal treatment or nondiscrimination provisions,
tions. Future research might tease out any differences in public which are an integral part of most entrenched charters^
opinion in terms of access to these spaces. (Taylor 2004, p. 160).
Despite these advances, however, not enough attention is
paid to important issues facing transgender people in the
Practice Implications United States today. For example, transgender people are vic-
tims of harassment and violent assault, including murder, at an
Our paper identifies several implications for practitioners and alarmingly high rate. Rates of violence are especially high for
policymakers as U.S. culture becomes more attuned to the fact transgender women and particularly against Black transgender
that gender is not, in fact, a binary construction. As awareness women. According to the National Coalition of Anti-Violence
of transgender people increases, so too will discussions of Programs (McBride 2017), at least 28 transgender people
whether and how public outreach and policies might need to were murdered in 2017. The Human Rights Campaign reports
be revised to better include them as full members of society. that 102 transgender people have been killed since January
Moving beyond the current headlines about gender-neutral 2013, 88 of whom were transgender women and nearly all
bathrooms may mean revisiting other elements of sex and of whom were Black or Hispanic (Crary 2017). Nearly half
gender in our culture, including our need for sex-marked iden- (46%) of respondents to the 2015 U.S. Transgender Survey
tity documents or for sex segregation in competitive sports conducted by the National Center for Transgender Equality
(Davis 2017). Attention to how these policies might spark reported being verbally harassed in the previous year because
backlash among those with strong masculine identities might they were transgender and nearly 9% reported being
Sex Roles

physically attacked in the previous year because of their gen- Burdge, B. J. (2007). Bending gender, ending gender: Theoretical foun-
dations for social work practice with the transgender community.
der identity (James et al. 2016).
Social Work, 52, 243–250. https://doi.org/10.1093/sw/52.3.243.
We need to have a better understanding of how the mass Burke, P. J. (1991). Identity processes and social stress. American
public thinks and feels about transgender people before we Sociological Review, 56, 836–849 www.jstor.org/stable/2096259.
will be able to enact the kind of change that the transgender Burke, P. J., & Stets, J. E. (2009). Identity theory. New York: Oxford
community deserves. Our paper serves as our first step by University Press.
Burn, S. M., Aboud, R., & Moyles, C. (2000). The relationship between
investigating how expectations of gender performance and gender social identity and support for feminism. Sex Roles, 42,
one’s own gender identity (and threats to it) explain support 1081–1089. https://doi.org/10.1023/A:1007044802798.
for or opposition to transgender rights. As Lazarsfeld et al. Cavanagh, S. L. (2010). Queering bathrooms: Gender, sexuality, and the
(1948, p. 27) note: Ba person thinks, politically, as he is social- hygienic imagination. Toronto: University of Toronto Press.
Cooper, B., Cox, D., Lienesch, R., & Jones, R. P. (2016). Majority of
ly. Social characteristics determine political preferences.^
Americans oppose laws requiring transgender individuals to use
Thus, gender identity—and particularly how a person per- bathrooms corresponding to sex at birth. Retrieved from https://
ceives gender norms and their own individual gender con- www.prri.org/research/poll-lgbt-transgender-bathroom-bill-
cept—will be an important predictor of how an individual presidential-election/.
reacts to issues that call into question norms about that identity Crary, D. (2017). Killings of transgender people hit a record high in 2017,
advocacy groups say. Time Magazine. Retrieved from http://time.
(e.g., transgender rights issues that threaten the gender binary com/5029561/transgender-murders-2017/.
and thus male and female identities). A better understanding Davis, H. F. (2017). Beyond trans: Does gender matter? New York: New
of the precise interplay between gender identity, threats to York University Press.
masculinity or femininity, and attitudes about transgender Druckman, J. N. (2004). Political preference formation: Competition,
deliberation, and the (ir)relevance of framing effects. American
(and non-gender conforming) people will allow for messaging Political Science Review, 98, 671–686. https://doi.org/10.1017/
and framing of issues that maximize public support. S0003055404041413.
Although the ability to exercise one’s rights should not be Elischberger, H. B., Glazier, J. J., Hill, E. D., & Verduzco-Baker, L.
up for a public vote, the reality of U.S. politics is that the rights (2016). ‘Boys don’t cry’—or do they? Adult attitudes toward and
beliefs about transgender youth. Sex Roles, 75, 197–214. https://doi.
of marginalized communities are often influenced by public
org/10.1007/s11199-016-0609-y.
opinion. Increasing acceptance of and support for transgender Ely, R. J. (1995). The power in demography: Women’s social construc-
people will also reduce the very real threat of acts of violence tions of gender identity at work. Academy of Management Journal,
and homicide that afflict this community. Additional research 38, 589–634. http://www.jstor.org/stable/256740.
is needed to identify the root causes of attitudes toward trans- Enloe, C. (2017). Exposing and challenging the persistence of patriarchy.
Berkeley, CA: University of California Press.
gender people and how those attitudes can be shifted. Erikson, R. S., & Tedin, K. L. (2015). American public opinion: Its
origins, content and impact (9th ed.). New York: Routledge.
Compliance with Ethical Standards Fasteau, M. F. (1974). The male machine. New York: McGraw-Hill.
Flores, A. (2015). Attitudes toward transgender rights: Perceived knowl-
Experimental data for this paper was obtained via the undergraduate edge and secondary interpersonal contact. Politics, Groups, and
research lab at Northwestern University. All participants provided in- Identities, 3, 398–416. https://doi.org/10.1080/21565503.2015.
formed consent prior to participating in the experiment. Prior to data 1050414.
collection, the study was approved by Menlo College’s Institutional Flores, A., Herman, J. L., Gates, G. J., & Brown, T. N. T. (2016). How
Review Board. many adults identify as transgender in the United States? Retrieved
from https://williamsinstitute.law.ucla.edu/wp-content/uploads/
How-Many-Adults-Identify-as-Transgender-in-the-United-States.
pdf.
Gadarian, S. K., & Van der Vort, E. (2017). The gag reflex: Disgust
References rhetoric and gay rights in American politics. Political Behavior,
23, 1–23. https://doi.org/10.1007/s11109-017-9412-x.
Geller, W. W. (1991). Attitudes toward gays and lesbians: A longitudinal
Adams, H. E., Wright Jr., L. W., & Lohr, B. (1996). Is homophobia
study. Eric Documents, ED 340970. Retrieved from https://archive.
associated with homosexual arousal? Journal of Abnormal
org/stream/ERIC_ED340970/ERIC_ED340970_djvu.txt.
Psychology, 105, 440–445. https://doi.org/10.1037/0021-843X.
Gillig, T. K., & Murphy, S. T. (2016). Fostering support for LGBTQ
105.3.440.
youth?: The effects of a gay adolescent media portrayal on young
Akerlof, G. A., & Kranton, R. E. (2000). Economics and identity. The
viewers. International Journal of Communication, 10, 3828–3850.
Quarterly Journal of Economics, 3, 715–753. https://doi.org/10.
Retrieved from http://ijoc.org/index.php/ijoc/article/view/5496.
1162/003355300554881.
Gillig, T. K., Rosenthal, E. L., Murphy, S. T., & Langrall Folb, K. L.
Brewer, M. B., & Lui, L. N. (1989). The primacy of age and sex in the (2017). More than a media moment: The influence of televised
structure of person categories. Social Cognition, 7, 262–274. https:// storylines on viewers’ attitudes toward transgender people and pol-
doi.org/10.1521/soco.1989.7.3.262. icies. Sex Roles. Advance online publication. https://doi.org/10.
Bridges, T., & Tober, T. L. (2017). Masculinity & violence, and the 1007/s11199-017-0816-1.
violence of masculinity. The Huffington Post. Retrieved from Goldberg, L. R. (1992). The structure of phenotypic personality traits.
https://www.huffingtonpost.com/entry/masculinity-violence-and- American Psychologist, 48, 26–34. https://doi.org/10.1037/0003-
the-violence-of-masculinity_us_5a0d2d59e4b023a796fed40c. 066X.48.1.26.
Sex Roles

Haider-Markel, D., Miller, P., Flores, A., Lewis, D. C., Tadlock, B., & McCall, C., & Dasgupta, N. (2007). The malleability of men’s gender
Taylor, J. (2017). Bringing BT^ to the table: Understanding individ- self-concept. Self and Identity, 6, 173–188. https://doi.org/10.1080/
ual support of transgender candidates for public office. Politics, 15298860601115328.
Groups, and Identities, 5, 1–19. https://doi.org/10.1080/21565503. McDermott, M. L. (2016). Masculinity, femininity, and American politi-
2016.1272472. cal behavior. New York: Oxford University Press.
Haines, E. L., & Kray, L. J. (2005). Self-power associations: The posses- McLeish, K. N., & Oxoby, R. J. (2008). Social interactions and the sa-
sion of power impacts women’s self-concepts. European Journal of lience of social identity. Journal of Economic Psychology, 32, 172–
Social Psychology, 35, 643–662. https://doi.org/10.1002/ejsp.252. 178. https://doi.org/10.1016/j.joep.2010.11.003.
Harrison, B. F., & Michelson, M. R. (2017). Listen, we need to talk: How Miller, P. R., Flores, A. R., Haider-Markel, D. P., Lewis, D. C., Tadlock,
to change attitudes about LGBT rights. New York: Oxford B. L., & Taylor, J. K. (2017). Transgender politics as body politics:
University Press. Effects of disgust sensitivity and authoritarianism on transgender
Heise, D. R. (2007). Expressive order: Confirming sentiments in social rights attitudes. Politics, Groups, and Identities, 5, 4–24. https://
actions. New York: Springer. doi.org/10.1080/21565503.2016.1260482.
Herek, G. M. (1984). Attitudes toward lesbians and gay men: A factory- Moyer, J. W. (2015). Why Houston’s gay rights ordinance failed: Fear of
analytic study. Journal of Homosexuality, 10, 39–51. https://doi.org/ men in women’s bathrooms. Retrieved from https://www.
10.1300/J082v10n01_03. washingtonpost.com/news/morning-mix/wp/2015/11/03/why-
Iyengar, S., & Kinder, D. R. (1987). News that matters: Television and houstons-gay-rights-ordinance-failed-bathrooms/.
American opinion. Chicago: University of Chicago Press. Nagoshi, J. L., Adams, K. A., Terrell, H. K., Hill, E. D., Brzuzy, S., &
James, S. E., Herman, J. L., Rankin, S., Keisling, M., Mottet, L., & Anafi, Nagoshi, C. T. (2008). Gender differences in correlates of homopho-
M. (2016). The report of the 2015 U.S. transgender survey. bia and transphobia. Sex Roles, 59, 521–531. https://doi.org/10.
Washington, DC: National Center for Transgender Equality 1007/s11199-008-9458-7.
Retrieved from http://www.transequality.org/sites/default/files/ National Center for Transgender Equality. (2016). Frequently asked ques-
docs/usts/USTS%20Full%20Report%20-%20FINAL%201.6.17. tions about transgender people. Retrieved from http://www.
pdf. transequality.org/issues/resources/frequently-asked-questions-
Karner, T. (1996). Fathers, sons, and Vietnam: Masculinity and betrayal about-transgender-people.
in the life narratives of Vietnam veterans with post traumatic stress Norton, A. T., & Herek, G. M. (2013). Heterosexuals’ attitudes toward
disorder. American Studies, 37, 63–94. Retrieved from http://www. transgender people: Findings from a national probability sample of
jstor.org/stable/40642783. U.S. adults. Sex Roles, 68, 738–753. https://doi.org/10.1007/
Kimmel, M. (1996). Manhood in America: A cultural history. New York: s11199-011-0110-6.
The Free Press. Nownes, A. J. (2015). Interest groups and transgender politics:
Kimmel, M. (2012). Masculinity, mental illness, and guns: A lethal Opportunities and challenges. In J. K. Taylor & D. P. Haider-
equation? Retrieved from http://www.cnn.com/2012/12/19/living/ Markel (Eds.), Transgender rights and politics: Groups, issue fram-
men-guns-violence/. ing, & policy adoption (pp. 83–107). Ann Arbor: University of
King, M. E., Winter, S., & Webster, B. (2009). Contact reduces Michigan Press.
transprejudice: A study on attitudes towards transgenderism and Roughgarden, J. (2013). Evolution's rainbow: Diversity, gender, and sex-
transgender civil rights in Hong Kong. International Journal of uality in nature and people. Berkeley, CA: University of California
S e x u a l H e a l t h , 2 1 , 1 7 – 3 4 . h t t p s : / / d o i . o rg / 1 0 . 1 0 8 0 / Press.
19317610802434609. Rudman, L. A., & Phelan, J. E. (2010). The effect of priming gender roles
Klar, S. (2013). The influence of competing identity primes on political on women’s implicit gender beliefs and career aspirations. Social
preferences. Journal of Politics, 75, 1108–1124. https://doi.org/10. Psychology, 41, 192–202. https://doi.org/10.1027/1864-9335/
1017/S0022381613000698. a000027.
Krosnick, J. A., & Brannon, L. A. (1993). The impact of the Gulf war on Schlit, K., & Westbrook, L. (2015). Bathroom battlegrounds and penis
the ingredients of presidential evaluations: Multidimensional effects panics. Contexts, 14, 26–31. https://doi.org/10.1177/
of political involvement. American Political Science Review, 87, 1536504215596943.
963–975. https://doi.org/10.2307/2938828. Schmader, T. (2002). Gender identification moderates stereotype threat
Lazarsfeld, P. F., Berelson, B., & Hazel Gaudet, H. (1948). The people’s effects on women’s math performance. Journal of Experimental
choice: How the voter makes up his mind in a presidential campaign Social Psychology, 38, 194–201. https://doi.org/10.1006/jesp.2001.
(2nd ed.). New York: Columbia University Press. 1500.
Maass, A., Cadinu, M., Guarnieri, G., & Grasselli, A. (2003). Sexual Sinclair, S., Hardin, C. D., & Lowery, B. S. (2006). Self-stereotyping in
harassment under social identity threat: The computer harassment the context of multiple social identities. Journal of Personality and
paradigm. Journal of Personality and Social Psychology, 85, 853– Social Psychology, 90, 529–542. https://doi.org/10.1037/0022-
870. https://doi.org/10.1037/0022-3514.85.5.853. 3514.90.4.529.
Macmillan, R., & Gartner, R. (1999). When she brings home the bacon: Starr, C. R., & Zurbriggen, E. L. (2017). Sandra Bem’s gender schema
Labor-force participation and the risk of spousal violence against theory after 34 years: A review of its reach and impact. Sex Roles,
women. Journal of Marriage and the Family, 61, 947–958. https:// 76, 566–578. https://doi.org/10.1007/s11199-016-0591-4.
doi.org/10.2307/354015. Tadlock, B. L., Flores, A. R., Haider-Markel, D. P., Lewis, D. C., Miller,
Mananzala, R., & Spade, D. (2008). The nonprofit industrial complex and P. R., & Taylor, J. K. (2017). Testing contact theory and attitudes on
trans resistance. Sexuality Research & Social Policy, 5, 53–71. transgender rights. Public Opinion Quarterly. https://doi.org/10.
https://doi.org/10.1525/srsp.2008.5.1.53. 1093/poq/nfx021.
Martin, C. L., & Ruble, D. N. (2010). Patterns of gender development. Tajfel, H. (1981). Human groups and social categories. New York:
Annual Review of Psychology, 61, 353–381. https://doi.org/10.1146/ Cambridge University Press.
annurev.psych.093008.100511. Tajfel, H., & Turner, J. C. (1986). The social identity theory of intergroup
McBride, S. (2017). HRC & Trans people of color coalition release relations. In S. Worchel & W. G. Austin (Eds.), Psychology of inter-
report on violence against the transgender community. Retrieved group relations (pp. 7–24). Chicago, IL: Nelson-Hall.
from https://www.hrc.org/blog/hrc-trans-people-of-color-coalition- Taylor, C. (2004). Modern social imaginaries. Durham, NC: Duke
release-report-on-violence-against-the. University Press.
Sex Roles

Tobin, D. D., Menon, M., Menon, M., Spatta, B. C., Hodges, E., & Perry, Westbrook, L., & Schilt, K. (2014). Doing gender, determining gender:
D. G. (2010). The intrapsychics of gender: A model of self-sociali- Transgender people, gender panics, and the maintenance of the sex/
zation. Psychological Review, 117, 601–622. https://doi.org/10. gender/sexuality system. Gender & Society, 28, 32–57. https://doi.
1037/a0018936. org/10.1177/0891243213503203.
Tourjée, D. (2015). Why do men kill trans women? Gender theorist Judith White, J. B., & Gardner, W. L. (2009). Think women, think warm:
Butler explains. Broadly. Retrieved from https://broadly.vice.com/ Stereotype content activation in women with a salient gender iden-
en_us/article/z4jd7y/why-do-men-kill-trans-women-gender- tity, using a modified Stroop task. Sex Roles, 60, 247–260. https://
theorist-judith-butler-explains. doi.org/10.1007/s11199-008-9526-z.
Tupes, E. C., & Christal, R. E. (1961). Recurrent personality factors Willer, R., Rogalin, C., Conlon, B., & Wojnowicz, M. T. (2013).
based on trait ratings. No. ASD-TR-61-97. Personal Research Overdoing gender: A test of the masculine overcompensation thesis.
Lab, Lackland AFB, Texas. Retrieved from http://www.dtic.mil/ American Journal of Sociology, 118, 980–1022. https://doi.org/10.
get-tr-doc/pdf?AD=AD0267778. 1086/668417.
Wade, L. (2016). The hypermasculine violence of Omar Mateen and Wood, W., Christensen, P. N., Hebl, M. R., & Rothgerber, H. (1997).
Brock Turner. New Republic. Retrieved from https://newrepublic. Conformity to sex-typed norms, affect, and the self-concept.
com/article/134270/hypermasculine-violence-omar-mateen-brock- Journal of Personality and Social Psychology, 73, 523–535.
turner. https://doi.org/10.1037/0022-3514.73.3.523.

You might also like