Explaining Public Opinion toward Transgender People, Rights, and Candidates

Explaining Public Opinion toward Transgender People, Rights, and Candidates Abstract What explains public opinion toward transgender people, rights, and candidates? Drawing on original data from a national telephone survey of US adults, this study explains attitudes regarding (1) the personal characteristics of transgender people; (2) a variety of transgender rights; and (3) transgender candidates for public office, measured through a randomized experiment included in the survey. Results indicate majority support on most policy questions, but more tepid views of transgender people, and solid opposition to supporting a transgender candidate for office. Our analyses reflect and extend previous research on American public opinion. Respondents’ fundamental values (egalitarianism, moral traditionalism, party identity, ideology, and religiosity) and personality characteristics (need for cognitive closure) predict views of transgender people and support for their rights. A significant relationship also emerged between television use and views of transgender people, suggesting that media portrayals may play a role in shaping these perceptions. In contrast, there is no evidence that interpersonal contact with a transgender person is related to opinions. Further, many of these independent variables have little moderating effect on responses to transgender candidates, which remain negative among most subgroups. Introduction Recent years have seen increased attention to transgender people in both popular culture and political debate. While media portrayal of celebrities such as former Olympian Caitlyn Jenner and actress Laverne Cox has been largely positive, politicians have sponsored a “wave of anti-transgender legislation” (Human Rights Campaign 2016). Policies restricting transgender rights have found majority support among American legislators (as in the case of North Carolina’s “bathroom bill”) and voters (as in the passing of the Houston Equal Rights Ordinance). Likewise, the few openly transgender candidates to run for office have failed to win election beyond a handful of local races (Casey and Reynolds 2015). Although this political backlash suggests widespread antipathy toward transgender people, rights, and candidates, scholarly research to date has provided a more nuanced view of public opinion. Affect toward transgender people and support for their rights often reflect values such as egalitarianism, moral traditionalism, conservatism, and religiosity, as well as contact with LGBT people (Norton and Herek 2013; Flores 2015; Miller et al. 2017; Flores et al. 2018). Building upon this research, we assess the opinions of cisgender Americans (those whose gender identity matches their assigned sex) toward: (1) the character traits of transgender people; (2) transgender rights; and (3) a hypothetical transgender candidate, assessed via a survey experiment. Drawing from previous studies and from general theories of public opinion, we explain these opinions with a wide range of independent variables. Values—including egalitarianism, moral traditionalism, ideology, party identification, and religiosity—and psychological characteristics—specifically, a need for cognitive closure—predict support for transgender rights, though of these only moral traditionalism predicts views of transgender people. We explore the effects of media exposure, finding evidence that television viewing cultivates more favorable views of transgender people. Finally, we document opposition toward transgender candidates: respondents’ support for their party plunges when the nominee is transgender. This opposition is in many cases not moderated by the values and characteristics that otherwise predict support for transgender people and rights, suggesting a difficult path ahead for transgender candidates. Previous Studies of Opinion about Transgender People, Rights, and Candidates Several recent surveys suggest public support for transgender rights. The 2015 Attitudes towards Transgender Supportive Policies Survey reports that 65 percent of Americans believe that transgender people should “receive protections provided by anti-discrimination laws” (University of Illinois Springfield Survey Research Office 2015). Similarly, a 2016 CNN poll finds that 75 percent favor “laws that guarantee equal protection for transgender people in jobs, housing and public accommodations” (Agiesta 2016). However, such broad support does not extend to issues surrounding bathroom access: public opinion on policies requiring transgender people to use the bathroom matching the sex they were assigned at birth is much more divided (Newport 2016). Beyond such descriptive results, several recent studies have sought to understand the political and psychological predictors of these attitudes. Norton and Herek (2013) find that survey respondents who held binary conceptions of gender, higher levels of authoritarianism, less egalitarian attitudes, more conservative ideologies, and greater religiosity all rated transgender people less favorably on feeling thermometers. Similarly, Flores et al. (2018) report that moral traditionalism and partisanship—as well as mere exposure to images of transgender people—influence transphobia and discomfort with gender nonconformity. Other studies focus on predicting support for laws protecting transgender rights. Flores (2015) finds that gender and interpersonal contact with gays and lesbians (although not with transgender people) are significant predictors of support for antidiscrimination laws. Further, he shows that respondents who believed they were well informed about transgender issues were more supportive of such protections. Other research echoes these findings and shows that core values such as authoritarianism, ideology, partisan identity, and personality characteristics such as disgust sensitivity are also significant predictors (Miller et al. 2017). In contrast, only a single piece of research we know of examines attitudes toward transgender candidates. Haider-Markel et al. (2017) ask respondents to assess their likelihood of voting for a transgender candidate who shared their views on most issues. A majority of voters report they would probably or definitely vote for the candidate, with liberals, Democrats, and the least religious being most likely to support them. This self-reported support conflicts with previous work that documents strong opposition to politicians from other marginalized groups. For example, both experimental (Herrick and Thomas 1999; Golebiowska 2001) and observational (Haider-Markel 2010) studies reveal negative reactions to lesbian, gay, and bisexual (LGB) candidates. Coupled with electoral losses of transgender candidates in recent years (Casey and Reynolds 2015), one might expect greater opposition to transgender candidates than appears in respondents’ self-reports. At the same time, given previous work showing that values and partisan attachments predict support for LGB candidates much as they do for LGB rights (Haider-Markel 2010), one might also expect these predispositions to moderate opposition toward transgender candidates. In the next section, we build on these prior studies, drawing on general theories of public opinion to consider a broad range of explanations for public opinion toward transgender people, rights, and candidates. Explaining Public Opinion Toward Transgender People, Rights, and Candidates INTERPERSONAL CONTACT Numerous studies find that positive interactions with members of an outgroup reduce prejudice toward the group (Allport 1954; Pettigrew and Tropp 2006). For example, knowing an LGB person is associated with greater support for LGB rights and more positive attitudes toward LGB people (Brewer 2008; Lewis 2011; Bramlett 2012; Garner 2013). Previous studies on transgender attitudes offer mixed support for the theory. Flores (2015) finds no such effect of contact on support for nondiscrimination policies; meanwhile, Flores et al. (2018) show that mere exposure to information and images of transgender people in a survey experiment reduces transphobia and discomfort with gender nonconformity. In this study, we revisit the intergroup contact hypothesis across a wider range of dependent variables. H1: Respondents who know a transgender person will hold more positive views of transgender people, rights, and candidates. VALUES An extensive literature shows that opinions about specific groups and policies are often based on more fundamental values. These include core values such as egalitarianism (Feldman 1988) and moral traditionalism (McCann 1997), broad political orientations such as political ideology and party identification (Zaller 1992), and religiosity (Cook, Jelen, and Wilcox 1992). For example, all of these values can influence public opinion about LGB rights (Lewis and Rogers 1999; Brewer 2003; Craig et al. 2005; Olson, Cadge, and Harrison 2006; Stoutenborough, Haider-Markel, and Allen 2006; Wilcox et al. 2007) and support for LGB candidates (Haider-Markel 2010; Haider-Markel et al. 2017). Previous research has examined the role of many, but not all, of these values in shaping opinions toward transgender people. Norton and Herek (2013) find that anti-egalitarianism, conservatism, and religiosity all predicted cooler feelings toward transgender people. Building on their results, we hypothesize that: H2: More egalitarian respondents will hold more positive views of transgender people, rights, and candidates. H3: More religious respondents will hold less positive views of transgender people, rights, and candidates. H4: More conservative respondents will hold less positive views of transgender people, rights, and candidates. Expanding the scope of values under investigation, we hypothesize that: H5: More morally traditional respondents will hold less positive views of transgender people, rights, and candidates. H6: Democrats will hold more positive views of transgender people, rights, and candidates than will Republicans. NEED FOR COGNITIVE CLOSURE Personality traits can shape dispositions and orientations toward people and events (Kruglanski and Webster 1991; Webster and Kruglanski 1994; Jost et al. 2003). A need for cognitive closure (NFCC) refers to one’s tendency to impose fixed meanings on situations and to feel an aversion toward ambiguous information and experiences. Individuals with a high need for cognitive closure may also be uncomfortable with the ambiguity inherent in the concept of a gender identity that does not match the sex assigned at birth. As such, we anticipate that: H7: Respondents high in NFCC will hold less positive views of transgender people, rights, and candidates. MEDIA EXPOSURE Cultivation research (Gerbner et al. 1984) posits that the dominant themes in television content can shape viewers’ real-world perceptions. For example, media portrayals can affect views of marginalized groups including women (Dasgupta and Asgari 2004), Latinos (Mastro, Behm-Morawitz, and Ortiz 2007), and gays and lesbians (Rössler and Brosius 2001). The parasocial contact hypothesis—that exposure to outgroups via the media can have similar effects to actual contact in the real world—helps explain these results. Viewing LGB characters on screen leads to greater support for gay rights and warmer evaluations of gay people (Schiappa, Gregg, and Hewes 2006; Bond and Compton 2015; Garretson 2015). One study that exposed participants to cross-dressing comedian Eddie Izzard (Schiappa, Gregg, and Hewes 2006) reported a significant reduction in transprejudice. Given the increasingly prominent and positive portrayals of transgender people in entertainment television programming, we hypothesize that: H8: Respondents with greater levels of overall television exposure will hold more positive views of transgender people, rights, and candidates. Meanwhile, news coverage may track the politically and socially polarized debate surrounding transgender issues. Thus, we ask: RQ1: Will news media exposure among respondents predict their views of transgender people, rights, and candidates? TRANSGENDER CANDIDATES Finally, given the research cited earlier showing opposition to candidates from minority groups, we hypothesize that: H9: Respondents will express less support for transgender candidates than cisgender candidates. DEMOGRAPHIC CHARACTERISTICS Previous research suggests that various demographic characteristics are also likely to affect opinion on transgender issues (Norton and Herek 2013; Flores 2015). Sex, age, education, and race may all shape attitudes, with female, younger, better-educated, and white respondents holding more favorable opinions. As such, we control for these variables in the models that follow. Data and Methods The data for this study came from the University of Delaware Center for Political Communication’s National Agenda Poll, a random-digit-dial (RDD) telephone survey of a sample of 901 adults living in the continental United States. The survey was conducted by Princeton Survey Research Associates International (PSRAI) from November 11 to 17, 2015, using a dual sampling frame of landline (n = 451) and cell phone (n = 450) respondents provided by Survey Sampling International (SSI). Interviews were conducted in English by Princeton Data Source. The response rates for the landline and cellular samples were 7 and 5 percent (calculated using AAPOR RR3), respectively (AAPOR 2016).1 These are low in historical perspective but comparable to typical rates for surveys conducted by major polling organizations (e.g., Pew Research Center 2012). The sampling error for the full weighted sample is ±4.0 percentage points. The Online Appendix includes additional information regarding the sample, including comparison to the general US population, descriptive statistics for key variables, the weighting that is used throughout, and the sample disposition. Four respondents who identified themselves as transgender were excluded from the analysis, given that we are interested in assessing cisgender people’s views of transgender issues. VIEWS OF TRANSGENDER PEOPLE Respondents were asked how well the words “trustworthy,” “moral,” and “happy” described “most people,” with response options of “not well at all” (1), “not too well” (2), “somewhat well” (3), and “very well” (4). They were then asked how well the same words described “transgender people.” Previous research has shown that some of the most pervasive negative stereotypes held of transgender people cast them as “abnormal,” “confused,” and “abusive” (Gazzola and Morrison 2014). Views of transgender people were standardized relative to views of most people in order to correct for interpersonal incomparability in respondents’ evaluations of people in general (see King et al. 2004). Respondents’ views of “most people” were subtracted from their views of “transgender people,” to create variables that take on a positive value if the respondent viewed transgender people as more trustworthy (moral, happy) than most people, and negative values if they viewed most people as more trustworthy (moral, happy) than transgender people. Values of zero indicated no difference between views of transgender people and most people. Views were on average relatively negative: respondents viewed transgender people as less moral (M = –.07, SD = .98) and less happy (M = –.12, SD = .87), but more trustworthy (M = .08, SD = .88), than most people. The median score for each item was zero, however. The three items were averaged into an index capturing views of transgender people (M = –.06, SD = .76; α = .68). ATTITUDES TOWARD TRANSGENDER RIGHTS Using a four-point scale (strongly oppose [1]; oppose [2]; favor [3]; and strongly favor [4]), respondents were asked about four policies: (1) “laws that protect transgender students from discrimination in schools” (73.2 percent favored or strongly favored, M = 2.93, SD = .97); (2) “laws that protect transgender people from job discrimination” (72.7 percent favored or strongly favored, M = 2.94, SD = .95); (3) “allowing transgender people to serve openly in the U.S. military” (64.7 percent favored or strongly favored, M = 2.76, SD = .98); and (4) “requiring public buildings like courthouses to have gender-neutral restrooms for patrons to use” (54.0 percent favored or strongly favored, M = 2.53, SD = .98). In each case, the median response was 3 (favoring the policy). The four items were averaged into an index measuring support for transgender rights (M = 2.78, SD = .80; α = .83). SUPPORT FOR TRANSGENDER CANDIDATES The survey included an experiment that probed support for a transgender candidate. Given the extremely high likelihood of voting for a candidate of one’s own party affiliation (Green, Palmquist, and Schickler 2002), the respondent’s partisan identification as expressed earlier in the survey structured the experimental design. Respondents were asked: “Next, we’d like to ask how you would vote in a future election for public office. If the [respondent’s party] nominated a ______________ and the [opposing party] nominated a ______________, who would you be more likely to vote for?” The candidates’ assigned parties were based on the respondent’s party identity as reflected in support of or leaning toward a given party (“pure” independents and those identifying with a minor party [N = 168] were excluded). For example, a respondent who identified as a Republican would be asked the following: “Next, we’d like to ask how you would vote in a future election for public office. If the Republicans nominated a ______________ and the Democrats nominated a ______________, who would you be more likely to vote for?” Respondents were assigned to one of two conditions. In both conditions, the opposing party’s candidate was presented as cisgender (e.g., “a man”). Their own party’s candidate, however, was randomly presented either as cisgender or as transgender (e.g., “a man” versus “a transgender man”).2 INTERPERSONAL CONTACT A dichotomous variable measured whether respondents had a “close friend or family member” who is transgender (1) or not (0). Eleven percent of respondents reported such interpersonal contact. VALUES Egalitarianism was captured through agreement with two statements based on Feldman (1988): “One of the big problems in this country is that we don’t give everyone an equal chance” and “If people were treated more equally in this country we would have fewer problems.” Responses were coded from 1 (strongly disagree) to 4 (strongly agree) and averaged into a single measure (M = 2.97, SE = .97, r = .59), with higher values indicating greater egalitarianism. Moral traditionalism was measured through agreement with two statements: “We should be more tolerant of people who choose to live according to their own moral standards even if they are very different from our own” and “Modern lifestyles are contributing to the breakdown of society” (Cook, Jelen, and Wilcox 1992). Responses were coded from 1 (strongly disagree) to 4 (strongly agree) and averaged into a single measure after reverse-coding the first item (M = 2.32, SE = .82, r = .19), with higher values indicating greater traditionalism. Religiosity was measured with an item asking respondents how religious they considered themselves to be, ranging from “not at all religious” (1) to “very religious” (4). Ideology was coded from 1 (very liberal) to 5 (very conservative). Partisan identification was coded from 1 (Strong Democrat) to 6 (Strong Republican). NEED FOR COGNITIVE CLOSURE NFCC was measured with three items taken from Roets and Van Hiel (2011): “I don’t like situations that are uncertain”; “I dislike questions that could be answered in many different ways”; and “I dislike it when a person’s statement could mean many different things.” Responses ranged from 1 (strongly disagree) to 4 (strongly agree), with higher values indicating greater NFCC. The three items were combined into an index (M = 2.75, SE = .75, α = .55). FAMILIARITY WITH TERM “TRANSGENDER” Since knowledge of transgender issues predicts support for transgender rights (Flores 2015), a measure of respondents’ familiarity with the term “transgender” was included. Respondents self-identified as knowing what the term meant (coded as 2), having heard it but not being sure of its meaning (1), or never having heard it (0). Before the batteries measuring attitudes on transgender issues, interviewers clarified for all respondents that “the term transgender applies to people who live their gender in a way that does not match the sex listed on their original birth certificate or who physically change their sex.” Previous research finds that providing the definition does not alter responses to following items (Flores 2015). MEDIA EXPOSURE Respondents were asked the number of hours they spent watching television on an average day (recoded to range from 0 to 6, with higher numbers coded at the maximum). The frequency with which respondents consumed various media was measured on a scale ranging from “never” (1) to “regularly” (4). These media were “national network evening news programs such as ABC World News, CBS Evening News or NBC Nightly News”; “news magazine shows such as 60 Minutes, 20/20, or Dateline”; “the Fox News Cable Channel”; MSNBC and CNN (averaged into a single measure); “The Late Show with Stephen Colbert on CBS” and “Last Week Tonight with John Oliver” (averaged into a single measure); and “shows on the cable network E! such as Keeping Up with the Kardashians and I Am Cait.” Frequency of reading a daily newspaper and getting news online were measured on the same scale. DEMOGRAPHIC VARIABLES Age was measured in decades. Indicator variables for female, black, and Hispanic respondents reflected gender and racial/ethnic identities. Education was measured on an ordinal scale ranging from less than high school (1) to a postgraduate or professional degree (8). MODELS AND PRESENTATION OF RESULTS To estimate views of transgender people and support for transgender rights, linear regression models using survey weights are estimated. Since the dependent variable in the candidate experiment is categorical (a vote for one’s party, the other party, or a “don’t know/refused” response), the analyses for this variable employ multinomial logit models. Throughout, results are presented in conventional regression tables with coefficients and standard errors. To ease interpretation, results are discussed in terms of predicted values for linear regressions and predicted probabilities for multinomial regressions. When simulating the regression results, all numeric variables are held to their mean value in the weighted dataset and categorical variables to their modal value (female is set to one; interpersonal contact, black, and Hispanic to zero). Thus, the predicted values are for white women with no interpersonal contact who have average values of all other variables. Explaining Views of Transgender People Results from linear regression models predicting views of transgender people, using the combined scale of the three items, are shown in the first column of table 1 (estimating models for each item separately, and with an ordered logistic estimator, led to substantively consistent findings; these results are available in the Online Appendix). Table 1. OLS regression models predicting opinions toward transgender people and rights Views of transgender people (n=531) Support for transgender rights (n=531) B (s.e.) B (s.e.) Intercept –0.17 (0.36) 3.66 (0.32)*** Values  Egalitarianism 0.00 (0.04) 0.11 (0.04)**  Moral traditionalism –0.10 (0.05)* –0.15 (0.04)***  Religiosity 0.02 (0.04) –0.12 (0.03)***  Ideology –0.06 (0.05) –0.15 (0.03)***  Party ID –0.03 (0.02) –0.04 (0.02)* Need for cognitive closure 0.06 (0.05) –0.08 (0.04)* Views of transgender people 0.21 (0.04)*** Interpersonal contact 0.07 (0.09) 0.09 (0.10) Familiarity with term 0.06 (0.07) 0.02 (0.05) Media consumption  TV use 0.06 (0.02)** 0.03 (0.02)  Network news 0.07 (0.04) 0.03 (0.04)  News magazines –0.00 (0.04) –0.07 (0.04)  Fox News –0.04 (0.03) –0.05 (0.03)  CNN/MSNBC –0.03 (0.04) 0.03 (0.04)  Newspaper 0.01 (0.03) 0.03 (0.03)  Online news –0.01 (0.04) –0.01 (0.04)  E Network 0.02 (0.05) 0.07 (0.05)  Late Show/Last Week Tonight 0.03 (0.05) –0.00 (0.04) Demographics  Age (decades) –0.25 (0.25) –0.09 (0.21)  Education –0.00 (0.02) 0.02 (0.02)  Female 0.23 (0.07)** 0.33 (0.06)***  Hispanic 0.12 (0.11) –0.12 (0.09)  Black 0.25 (0.16) –0.06 (0.11) McFadden’s pseudo-R2 0.19 0.52 Views of transgender people (n=531) Support for transgender rights (n=531) B (s.e.) B (s.e.) Intercept –0.17 (0.36) 3.66 (0.32)*** Values  Egalitarianism 0.00 (0.04) 0.11 (0.04)**  Moral traditionalism –0.10 (0.05)* –0.15 (0.04)***  Religiosity 0.02 (0.04) –0.12 (0.03)***  Ideology –0.06 (0.05) –0.15 (0.03)***  Party ID –0.03 (0.02) –0.04 (0.02)* Need for cognitive closure 0.06 (0.05) –0.08 (0.04)* Views of transgender people 0.21 (0.04)*** Interpersonal contact 0.07 (0.09) 0.09 (0.10) Familiarity with term 0.06 (0.07) 0.02 (0.05) Media consumption  TV use 0.06 (0.02)** 0.03 (0.02)  Network news 0.07 (0.04) 0.03 (0.04)  News magazines –0.00 (0.04) –0.07 (0.04)  Fox News –0.04 (0.03) –0.05 (0.03)  CNN/MSNBC –0.03 (0.04) 0.03 (0.04)  Newspaper 0.01 (0.03) 0.03 (0.03)  Online news –0.01 (0.04) –0.01 (0.04)  E Network 0.02 (0.05) 0.07 (0.05)  Late Show/Last Week Tonight 0.03 (0.05) –0.00 (0.04) Demographics  Age (decades) –0.25 (0.25) –0.09 (0.21)  Education –0.00 (0.02) 0.02 (0.02)  Female 0.23 (0.07)** 0.33 (0.06)***  Hispanic 0.12 (0.11) –0.12 (0.09)  Black 0.25 (0.16) –0.06 (0.11) McFadden’s pseudo-R2 0.19 0.52 Note.—Estimates based on weighted data. Coefficients are from linear regression models with standard errors in parentheses. Omitted category for gender is male, for race is White. *p < .05; **p < .01; ***p < .001. View Large Table 1. OLS regression models predicting opinions toward transgender people and rights Views of transgender people (n=531) Support for transgender rights (n=531) B (s.e.) B (s.e.) Intercept –0.17 (0.36) 3.66 (0.32)*** Values  Egalitarianism 0.00 (0.04) 0.11 (0.04)**  Moral traditionalism –0.10 (0.05)* –0.15 (0.04)***  Religiosity 0.02 (0.04) –0.12 (0.03)***  Ideology –0.06 (0.05) –0.15 (0.03)***  Party ID –0.03 (0.02) –0.04 (0.02)* Need for cognitive closure 0.06 (0.05) –0.08 (0.04)* Views of transgender people 0.21 (0.04)*** Interpersonal contact 0.07 (0.09) 0.09 (0.10) Familiarity with term 0.06 (0.07) 0.02 (0.05) Media consumption  TV use 0.06 (0.02)** 0.03 (0.02)  Network news 0.07 (0.04) 0.03 (0.04)  News magazines –0.00 (0.04) –0.07 (0.04)  Fox News –0.04 (0.03) –0.05 (0.03)  CNN/MSNBC –0.03 (0.04) 0.03 (0.04)  Newspaper 0.01 (0.03) 0.03 (0.03)  Online news –0.01 (0.04) –0.01 (0.04)  E Network 0.02 (0.05) 0.07 (0.05)  Late Show/Last Week Tonight 0.03 (0.05) –0.00 (0.04) Demographics  Age (decades) –0.25 (0.25) –0.09 (0.21)  Education –0.00 (0.02) 0.02 (0.02)  Female 0.23 (0.07)** 0.33 (0.06)***  Hispanic 0.12 (0.11) –0.12 (0.09)  Black 0.25 (0.16) –0.06 (0.11) McFadden’s pseudo-R2 0.19 0.52 Views of transgender people (n=531) Support for transgender rights (n=531) B (s.e.) B (s.e.) Intercept –0.17 (0.36) 3.66 (0.32)*** Values  Egalitarianism 0.00 (0.04) 0.11 (0.04)**  Moral traditionalism –0.10 (0.05)* –0.15 (0.04)***  Religiosity 0.02 (0.04) –0.12 (0.03)***  Ideology –0.06 (0.05) –0.15 (0.03)***  Party ID –0.03 (0.02) –0.04 (0.02)* Need for cognitive closure 0.06 (0.05) –0.08 (0.04)* Views of transgender people 0.21 (0.04)*** Interpersonal contact 0.07 (0.09) 0.09 (0.10) Familiarity with term 0.06 (0.07) 0.02 (0.05) Media consumption  TV use 0.06 (0.02)** 0.03 (0.02)  Network news 0.07 (0.04) 0.03 (0.04)  News magazines –0.00 (0.04) –0.07 (0.04)  Fox News –0.04 (0.03) –0.05 (0.03)  CNN/MSNBC –0.03 (0.04) 0.03 (0.04)  Newspaper 0.01 (0.03) 0.03 (0.03)  Online news –0.01 (0.04) –0.01 (0.04)  E Network 0.02 (0.05) 0.07 (0.05)  Late Show/Last Week Tonight 0.03 (0.05) –0.00 (0.04) Demographics  Age (decades) –0.25 (0.25) –0.09 (0.21)  Education –0.00 (0.02) 0.02 (0.02)  Female 0.23 (0.07)** 0.33 (0.06)***  Hispanic 0.12 (0.11) –0.12 (0.09)  Black 0.25 (0.16) –0.06 (0.11) McFadden’s pseudo-R2 0.19 0.52 Note.—Estimates based on weighted data. Coefficients are from linear regression models with standard errors in parentheses. Omitted category for gender is male, for race is White. *p < .05; **p < .01; ***p < .001. View Large Only three independent variables were significantly related to perceptions of transgender people. First, consistent with H5, the more a respondent subscribed to traditional moral values, the less positively they viewed transgender people (β = –.10, SE = .05, p = .04). Simulating predicted values from the regression coefficients showed that respondents with the lowest levels of moral traditionalism held neutral to slightly positive views of transgender people (an expected value of .14 with 95 percent confidence intervals [–.04, .32] on the –3 to +3 scale). As hypothesized, those with the highest levels of moral traditionalism held more negative views of transgender people (–.18 [–.35, .00]). Second, consistent with cultivation theory and H8, television consumption was significantly related to perceptions of transgender people (β = .06, SE = .02, p = .002). The more TV a respondent watched, the more positive their perceptions of transgender people were: someone who did not watch any television was predicted to hold relatively negative impressions of transgender people (–.18 [–.33, –.02]), whereas someone who watched six or more hours was predicted to hold favorable perceptions (.20 [.04, .36]). Finally, in line with previous research (Norton and Herek 2013), there were gender differences: women had more positive attitudes toward transgender people than did men. Many of the independent variables expected to influence perceptions of transgender people had no statistically discernible effect, however. Egalitarianism, ideology, party identification, religiosity, NFCC, interpersonal contact, and media consumption aside from general TV use (RQ1) all appeared unrelated to perceptions of transgender people. When it comes to views of transgender people, no support emerged for H1–4 and H6–7. Explaining Attitudes Toward Transgender Rights Support for the four transgender rights items, combined into a single index, were modeled with the same independent variables used previously, as well as views of transgender people (results from models of each individual item and multinomial regression estimators are available in the Online Appendix). The results, shown in the second column of table 1, support several hypotheses. The predispositions and values that the literature identifies as strong predictors of support for minority rights largely had the expected relationships. Figure 1 shows predicted support for transgender rights across the ranges of the independent variables. These predicted values are simulated from the model in table 1, for a hypothetical respondent with mean values on all numeric variables and the modal category for categorical variables (i.e., a white woman with no interpersonal contact). Figure 1. View largeDownload slide Predicted support for transgender rights. Ninety-five percent confidence intervals. Simulated from model in table 1 using weighted data. For each simulation, all numeric variables are held at their means; interpersonal contact, black, and Hispanic are set to zero and female to one. Figure 1. View largeDownload slide Predicted support for transgender rights. Ninety-five percent confidence intervals. Simulated from model in table 1 using weighted data. For each simulation, all numeric variables are held at their means; interpersonal contact, black, and Hispanic are set to zero and female to one. For instance, as shown in the first plot of figure 1, the more egalitarian respondents were, the more likely they were to support transgender rights (as predicted by H2). Someone with the lowest support for egalitarianism was predicted to score 2.78 on the transgender rights index [2.61, 2.94], compared to someone with the greatest support for egalitarian ideals, who was predicted to score 3.09 [2.98, 3.20]. Other core values had relationships of roughly the same magnitude. Supporting H5, those who did not subscribe to morally traditional values were predicted to support transgender rights (3.18 [3.05, 3.32]) more than those who did (2.73 [2.57, 2.89]). Consistent with H3, support decreased with religiosity, from 3.20 [3.04, 3.35] for the least religious to 2.85 [2.75, 2.97] for the most religious. As previous research would lead one to expect, respondents who held egalitarian ideals were likely to support rights for transgender people, while those high in moral traditionalism and religiosity were unlikely to do so. As predicted by H4 and H6, both party identification and ideology were significant predictors of support for transgender rights, such that Republicans and conservatives were less supportive of policies protecting transgender people than Democrats and liberals, respectively. Strong Democrats were more supportive than strong Republicans (3.08 [2.94, 3.21] versus 2.85 [2.72, 2.98]). Likewise, those who identified as very liberal were more supportive than those who identified as very conservative (3.31 [3.14, 3.47] versus 2.72 [2.56, 2.86]). Psychological characteristics mattered, too. Respondents at the low end of the NFCC scale were predicted to be more supportive of transgender rights (3.12 [2.97, 3.28]) than those at the high end (2.88 [2.77, 3.01]), consistent with H7 and previous research showing that heightened NFCC results in more negative responses to nonconformity (Kruglanski and Webster 1991). Views of transgender people were a significant predictor of support for transgender rights (β = .21, SE = .04, p < .001). Respondents with the most negative views of transgender people were unlikely to support policies that benefit them (2.36 [2.13, 2.59]). Those with the most positive views, however, were highly likely to support their rights (3.62 [3.38, 3.87]). Echoing earlier research on LGB issues (Brewer 2003), those with warmer views of transgender people were more sympathetic to policies protecting them. There was no support for H8; those who watched more television were not significantly more favorable toward transgender rights. Nor was any specific form of media consumption related to opinion (RQ1). And as before, interpersonal contact had no discernible effect. Contrary to H1, knowing someone who identifies as transgender did not predict support for policies that benefit them. Instead, values such as moral traditionalism, egalitarianism, party identification, ideology, and religiosity, as well as a psychological need for closure, were significant predictors of support. Furthermore, perceptions of transgender people strongly shaped support for affording them rights. While these effects may seem small when taken individually, their combined impact is substantial. Someone high in values associated with support for transgender rights—a female respondent who was one standard deviation above the mean on egalitarianism and views of transgender people, and one standard deviation below the mean on moral traditionalism, religiosity, NFCC, ideology, and party ID—was predicted to score 3.78 [3.65, 3.92]. Her male counterpart, one standard deviation from the mean in the opposite direction on each variable, was predicted to score just 1.84 [1.71, 1.97]. Taken in combination, these values and characteristics help explain much of public opinion on transgender rights. Explaining Support for Transgender Candidates Figure 2 shows respondents’ vote choice by experimental condition. When both candidates were presented as cisgender—the norm in American politics—more than two-thirds (68 percent) said they would vote for their own party’s candidate, with just 7 percent opting for the other party and 25 percent volunteering that they did not know (a not unreasonable response given the scant information supplied). Figure 2. View largeDownload slide Reported vote choice, given experimental condition. Opposing party’s candidate described as cisgender in both conditions. Estimates based on weighted data. Pure Independents and non-partisans (N = 164) excluded from analysis. N = 359 (cisgender condition); N = 374 (transgender condition). Figure 2. View largeDownload slide Reported vote choice, given experimental condition. Opposing party’s candidate described as cisgender in both conditions. Estimates based on weighted data. Pure Independents and non-partisans (N = 164) excluded from analysis. N = 359 (cisgender condition); N = 374 (transgender condition). When presented with a transgender candidate from their own party and a cisgender candidate from the other party, respondents’ choices differed significantly. Respondents’ support for their own party dropped dramatically, from 68 percent to 37 percent, while support for the opposing party increased to nearly one-fourth (from 7 percent to 23 percent). The number of “don’t know” responses likewise increased substantially, from 25 percent to 40 percent. As predicted by H9, there was a sizeable reluctance to vote for transgender candidates, similar to the opposition toward candidates from marginalized groups documented in earlier research. When voters were presented with a transgender candidate from their own party, they were more likely to desert to the other side or to opt out of answering the question than to support their usual preference. WHAT MODERATES OPPOSITION TO TRANSGENDER CANDIDATES? These conclusions are reinforced by the multinomial regression model in the first column of table 2 (labeled “baseline model”), which shows the increased likelihood of voting for the other party (β = 1.72, SE = .31, p < .001) or giving a Don’t Know/Refused response (β = 1.06, SE = .21, p < .001) when the party’s nominee was transgender. Table 2. Multinomial logit regression models predicting vote choice in candidate experiment, with moderators Moderator used Baseline (n=733) Republican party ID (n=733) Egalitarianism (n=731) Moral traditionalism (n=729) Religiosity (n=715) B (s.e.) B (s.e.) B (s.e.) B (s.e.) B (s.e.) Vote for opposing party  Intercept –2.20 (0.26)*** –1.94 (0.32)*** –1.96 (0.80)* –2.83 (0.59)*** –1.76 (0.74)*  Transgender candidate 1.72 (0.31)*** 1.29 (0.38)*** 1.90 (0.99) 1.00 (0.78) 0.03 (0.91)  Moderator –0.88 (0.55) –0.08 (0.26) 0.27 (0.22) –0.16 (0.24)  Transgender candidate X moderator 1.38 (0.66)* –0.06 (0.32) 0.31 (0.30) 0.59 (0.30) DK/Refused  Intercept –0.99 (0.16)*** –1.24 (0.23)*** –0.53 (0.47) –1.48 (0.45)*** –0.95 (0.50)  Transgender candidate 1.06 (0.21)*** 0.94 (0.30)** 1.37 (0.66) 0.60 (0.64) –0.19 (0.70)  Moderator 0.53 (0.31) –0.16 (0.16) 0.20 (0.18) –0.03 (0.16)  Transgender candidate X moderator 0.46 (0.44) –0.10 (0.22) 0.22 (0.26) 0.45 (0.23)* Moderator used Ideology (n=701) Need for cognitive closure (n=728) Interpersonal contact (n=733) TV use (n=729) Views of transgender people (n=629) B (s.e.) B (s.e.) B (s.e.) B (s.e.) B (s.e.) Vote for opposing party  Intercept –2.39 (0.66)*** –3.06 (0.91)*** –2.23 (0.29)*** –2.53 (0.50)*** –2.20 (0.29)***  Transgender candidate 0.81 (0.82) 2.14 (1.09) 1.88 (0.34)*** 2.13 (0.60)*** 1.66 (0.34)*** Moderator used Ideology (n=701) Need for cognitive closure (n=728) Interpersonal contact (n=733) TV use (n=729) Views of transgender people (n=629) B (s.e.) B (s.e.) B (s.e.) B (s.e.) B (s.e.)  Moderator 0.07 (0.18) 0.31 (0.30) 0.23 (0.66) 0.09 (0.12) –0.10 (0.34)  Transgender candidate X moderator 0.28 (0.24) –0.15 (0.37) –1.46 (0.89) –0.12 (0.15) –0.45 (0.41) DK/Refused  Intercept –1.35 (0.48)* –1.23 (0.57)* –0.93 (0.16)*** –0.72 (0.31)* –1.12 (0.17)***  Transgender candidate 0.56 (0.63) 0.71 (0.80) 1.08 (0.22)*** 1.02 (0.42)* 0.99 (0.24)***  Moderator 0.10 (0.14) 0.09 (0.20) –0.80 (0.62) –0.10 (0.10) 0.11 (0.20)  Transgender candidate X moderator 0.56 (0.63) 0.12 (0.28) 0.21 (0.78) 0.02 (0.13) –0.42 (0.29) Moderator used Baseline (n=733) Republican party ID (n=733) Egalitarianism (n=731) Moral traditionalism (n=729) Religiosity (n=715) B (s.e.) B (s.e.) B (s.e.) B (s.e.) B (s.e.) Vote for opposing party  Intercept –2.20 (0.26)*** –1.94 (0.32)*** –1.96 (0.80)* –2.83 (0.59)*** –1.76 (0.74)*  Transgender candidate 1.72 (0.31)*** 1.29 (0.38)*** 1.90 (0.99) 1.00 (0.78) 0.03 (0.91)  Moderator –0.88 (0.55) –0.08 (0.26) 0.27 (0.22) –0.16 (0.24)  Transgender candidate X moderator 1.38 (0.66)* –0.06 (0.32) 0.31 (0.30) 0.59 (0.30) DK/Refused  Intercept –0.99 (0.16)*** –1.24 (0.23)*** –0.53 (0.47) –1.48 (0.45)*** –0.95 (0.50)  Transgender candidate 1.06 (0.21)*** 0.94 (0.30)** 1.37 (0.66) 0.60 (0.64) –0.19 (0.70)  Moderator 0.53 (0.31) –0.16 (0.16) 0.20 (0.18) –0.03 (0.16)  Transgender candidate X moderator 0.46 (0.44) –0.10 (0.22) 0.22 (0.26) 0.45 (0.23)* Moderator used Ideology (n=701) Need for cognitive closure (n=728) Interpersonal contact (n=733) TV use (n=729) Views of transgender people (n=629) B (s.e.) B (s.e.) B (s.e.) B (s.e.) B (s.e.) Vote for opposing party  Intercept –2.39 (0.66)*** –3.06 (0.91)*** –2.23 (0.29)*** –2.53 (0.50)*** –2.20 (0.29)***  Transgender candidate 0.81 (0.82) 2.14 (1.09) 1.88 (0.34)*** 2.13 (0.60)*** 1.66 (0.34)*** Moderator used Ideology (n=701) Need for cognitive closure (n=728) Interpersonal contact (n=733) TV use (n=729) Views of transgender people (n=629) B (s.e.) B (s.e.) B (s.e.) B (s.e.) B (s.e.)  Moderator 0.07 (0.18) 0.31 (0.30) 0.23 (0.66) 0.09 (0.12) –0.10 (0.34)  Transgender candidate X moderator 0.28 (0.24) –0.15 (0.37) –1.46 (0.89) –0.12 (0.15) –0.45 (0.41) DK/Refused  Intercept –1.35 (0.48)* –1.23 (0.57)* –0.93 (0.16)*** –0.72 (0.31)* –1.12 (0.17)***  Transgender candidate 0.56 (0.63) 0.71 (0.80) 1.08 (0.22)*** 1.02 (0.42)* 0.99 (0.24)***  Moderator 0.10 (0.14) 0.09 (0.20) –0.80 (0.62) –0.10 (0.10) 0.11 (0.20)  Transgender candidate X moderator 0.56 (0.63) 0.12 (0.28) 0.21 (0.78) 0.02 (0.13) –0.42 (0.29) Note.—Reference level is a vote for the respondent’s own party; coefficients show effect on response relative to voting for own party. Estimates based on weighted data. Coefficients are from multinomials logit models with standard errors in parentheses. *p < .05; **p < .01; ***p < .001. View Large Table 2. Multinomial logit regression models predicting vote choice in candidate experiment, with moderators Moderator used Baseline (n=733) Republican party ID (n=733) Egalitarianism (n=731) Moral traditionalism (n=729) Religiosity (n=715) B (s.e.) B (s.e.) B (s.e.) B (s.e.) B (s.e.) Vote for opposing party  Intercept –2.20 (0.26)*** –1.94 (0.32)*** –1.96 (0.80)* –2.83 (0.59)*** –1.76 (0.74)*  Transgender candidate 1.72 (0.31)*** 1.29 (0.38)*** 1.90 (0.99) 1.00 (0.78) 0.03 (0.91)  Moderator –0.88 (0.55) –0.08 (0.26) 0.27 (0.22) –0.16 (0.24)  Transgender candidate X moderator 1.38 (0.66)* –0.06 (0.32) 0.31 (0.30) 0.59 (0.30) DK/Refused  Intercept –0.99 (0.16)*** –1.24 (0.23)*** –0.53 (0.47) –1.48 (0.45)*** –0.95 (0.50)  Transgender candidate 1.06 (0.21)*** 0.94 (0.30)** 1.37 (0.66) 0.60 (0.64) –0.19 (0.70)  Moderator 0.53 (0.31) –0.16 (0.16) 0.20 (0.18) –0.03 (0.16)  Transgender candidate X moderator 0.46 (0.44) –0.10 (0.22) 0.22 (0.26) 0.45 (0.23)* Moderator used Ideology (n=701) Need for cognitive closure (n=728) Interpersonal contact (n=733) TV use (n=729) Views of transgender people (n=629) B (s.e.) B (s.e.) B (s.e.) B (s.e.) B (s.e.) Vote for opposing party  Intercept –2.39 (0.66)*** –3.06 (0.91)*** –2.23 (0.29)*** –2.53 (0.50)*** –2.20 (0.29)***  Transgender candidate 0.81 (0.82) 2.14 (1.09) 1.88 (0.34)*** 2.13 (0.60)*** 1.66 (0.34)*** Moderator used Ideology (n=701) Need for cognitive closure (n=728) Interpersonal contact (n=733) TV use (n=729) Views of transgender people (n=629) B (s.e.) B (s.e.) B (s.e.) B (s.e.) B (s.e.)  Moderator 0.07 (0.18) 0.31 (0.30) 0.23 (0.66) 0.09 (0.12) –0.10 (0.34)  Transgender candidate X moderator 0.28 (0.24) –0.15 (0.37) –1.46 (0.89) –0.12 (0.15) –0.45 (0.41) DK/Refused  Intercept –1.35 (0.48)* –1.23 (0.57)* –0.93 (0.16)*** –0.72 (0.31)* –1.12 (0.17)***  Transgender candidate 0.56 (0.63) 0.71 (0.80) 1.08 (0.22)*** 1.02 (0.42)* 0.99 (0.24)***  Moderator 0.10 (0.14) 0.09 (0.20) –0.80 (0.62) –0.10 (0.10) 0.11 (0.20)  Transgender candidate X moderator 0.56 (0.63) 0.12 (0.28) 0.21 (0.78) 0.02 (0.13) –0.42 (0.29) Moderator used Baseline (n=733) Republican party ID (n=733) Egalitarianism (n=731) Moral traditionalism (n=729) Religiosity (n=715) B (s.e.) B (s.e.) B (s.e.) B (s.e.) B (s.e.) Vote for opposing party  Intercept –2.20 (0.26)*** –1.94 (0.32)*** –1.96 (0.80)* –2.83 (0.59)*** –1.76 (0.74)*  Transgender candidate 1.72 (0.31)*** 1.29 (0.38)*** 1.90 (0.99) 1.00 (0.78) 0.03 (0.91)  Moderator –0.88 (0.55) –0.08 (0.26) 0.27 (0.22) –0.16 (0.24)  Transgender candidate X moderator 1.38 (0.66)* –0.06 (0.32) 0.31 (0.30) 0.59 (0.30) DK/Refused  Intercept –0.99 (0.16)*** –1.24 (0.23)*** –0.53 (0.47) –1.48 (0.45)*** –0.95 (0.50)  Transgender candidate 1.06 (0.21)*** 0.94 (0.30)** 1.37 (0.66) 0.60 (0.64) –0.19 (0.70)  Moderator 0.53 (0.31) –0.16 (0.16) 0.20 (0.18) –0.03 (0.16)  Transgender candidate X moderator 0.46 (0.44) –0.10 (0.22) 0.22 (0.26) 0.45 (0.23)* Moderator used Ideology (n=701) Need for cognitive closure (n=728) Interpersonal contact (n=733) TV use (n=729) Views of transgender people (n=629) B (s.e.) B (s.e.) B (s.e.) B (s.e.) B (s.e.) Vote for opposing party  Intercept –2.39 (0.66)*** –3.06 (0.91)*** –2.23 (0.29)*** –2.53 (0.50)*** –2.20 (0.29)***  Transgender candidate 0.81 (0.82) 2.14 (1.09) 1.88 (0.34)*** 2.13 (0.60)*** 1.66 (0.34)*** Moderator used Ideology (n=701) Need for cognitive closure (n=728) Interpersonal contact (n=733) TV use (n=729) Views of transgender people (n=629) B (s.e.) B (s.e.) B (s.e.) B (s.e.) B (s.e.)  Moderator 0.07 (0.18) 0.31 (0.30) 0.23 (0.66) 0.09 (0.12) –0.10 (0.34)  Transgender candidate X moderator 0.28 (0.24) –0.15 (0.37) –1.46 (0.89) –0.12 (0.15) –0.45 (0.41) DK/Refused  Intercept –1.35 (0.48)* –1.23 (0.57)* –0.93 (0.16)*** –0.72 (0.31)* –1.12 (0.17)***  Transgender candidate 0.56 (0.63) 0.71 (0.80) 1.08 (0.22)*** 1.02 (0.42)* 0.99 (0.24)***  Moderator 0.10 (0.14) 0.09 (0.20) –0.80 (0.62) –0.10 (0.10) 0.11 (0.20)  Transgender candidate X moderator 0.56 (0.63) 0.12 (0.28) 0.21 (0.78) 0.02 (0.13) –0.42 (0.29) Note.—Reference level is a vote for the respondent’s own party; coefficients show effect on response relative to voting for own party. Estimates based on weighted data. Coefficients are from multinomials logit models with standard errors in parentheses. *p < .05; **p < .01; ***p < .001. View Large The other columns of table 2 present results from models that interacted the experimental condition with each of the measures hypothesized to moderate opposition toward a transgender candidate. Partisan differences are shown in the second column of table 2. Figure 3 shows the simulated first differences in responses when the party candidate is transgender (relative to a cisgender candidate), by party identity. Each bullet represents the change in probability of giving each response when the candidate is transgender, with brackets representing 95 percent confidence intervals around the estimates. Figure 3. View largeDownload slide First differences in vote choice given transgender nominee, by party ID. Bullets represent first differences in probability of each response given shift from cisgender to transgender nominee, with 95 percent confidence intervals. Simulated from model in second column of table 2. Figure 3. View largeDownload slide First differences in vote choice given transgender nominee, by party ID. Bullets represent first differences in probability of each response given shift from cisgender to transgender nominee, with 95 percent confidence intervals. Simulated from model in second column of table 2. Negative responses to transgender candidates were common among both Democrats and Republicans. Voters were less likely to support their party when the nominee is transgender, regardless of party (for Democrats, there was a predicted change in probability of support of –.26 [–.37, –.14]; for Republicans, –.38 [–.51, –.26]). While the estimated decline in support among Republicans was larger than that for Democrats, the uncertainty around the estimates precludes us from concluding that the two groups responded differently. Figure 4 presents predicted first differences in the probability of supporting a transgender nominee from one’s own party (relative to a cisgender nominee), across the full range of values of other moderating variables. Figure 4. View largeDownload slide First differences in probability of voting for own party given transgender nominee, by moderators. Ninety-five percent confidence intervals. Simulated from models in table 2. Figure 4. View largeDownload slide First differences in probability of voting for own party given transgender nominee, by moderators. Ninety-five percent confidence intervals. Simulated from models in table 2. Respondents had negative responses to a transgender nominee regardless of their position on most of these measures. Contrary to the earlier hypotheses, those who subscribed to egalitarian beliefs and those who did not; those who held traditional moral values and those who did not; liberals and conservatives; those with a high need for closure and those with a low need for closure; those who watched lots of TV and those who watched none—all were less likely to support a transgender candidate than a cisgender one. In some cases, values do appear to temper opposition to transgender candidates. Respondents who were not at all religious were not significantly less likely than the very religious to support a transgender nominee (a first difference in the probability of support of –.09 [–.29, .11]). The very religious, however, were substantially less likely to do so (–.43 [–.55, –.30]). The difference in responses from the least and most religious was statistically significant at the p < .05 level (.34 [.07, .61]), indicating that religiosity did increase opposition to transgender candidates, consistent with H3. The estimated first difference was also indistinguishable from zero for two other subsets of respondents: those who knew a transgender person (–.22 [–.47, .03]) and those with the most positive views of transgender people (.03 [–.35, .41]) were not significantly less likely to support a transgender nominee than a cisgender one. In neither case, however, are the differences between respondents at each end of the scale statistically significant, and so we do not overemphasize this finding. There was thus some suggestive evidence that the least religious, those who knew a transgender person, and those with positive views of transgender people did not discriminate against transgender candidates. However, the other variables failed to moderate opposition to transgender candidates, as hypothesized. Conclusions The visibility of transgender people in the media and political debate has increased dramatically in recent years. That visibility has not been accompanied by uniformly positive attitudes toward transgender people, rights, and candidates. Our data reveal that a non-trivial proportion of the public has negative views of transgender people, perceiving them as less trustworthy, happy, and moral than cisgender people. While there is majority support for transgender workplace protections, military service, and student protections, the results show a more divided public on the issue of gender-neutral bathrooms, as we might expect given lower support for “body-centric” policies (Miller et al. 2017). And finally, the experiment reveals stronger opposition to transgender candidates than that suggested by self-reports of willingness to support a candidate (Haider-Markel et al. 2017). Beyond these descriptive findings, the results offered support for—and illuminate the limitations of—several broader theories of public opinion. Broad values helped explain opinions about specific groups and issues. In particular, moral traditionalism predicted negative views of transgender people and opposition toward transgender rights. These results echo earlier studies showing moral traditionalism’s relationship with attitudes toward LGB rights (Brewer 2008). The other values studied (egalitarianism, religiosity, ideology, and party) had similarly expected relationships with support for transgender rights (but not with views of transgender people). As hypothesized, more egalitarian, more liberal, less religious, and more Democratic respondents expressed greater support for transgender rights. Likewise, there was support for theories linking need for cognitive closure to political attitudes. Respondents who had a high need for closure were less supportive of transgender rights, indicating the political consequences of seemingly nonpolitical psychological characteristics. In contrast, there was no evidence supporting interpersonal contact theory. Consistent with Flores (2015), respondents who knew a transgender person were not significantly more likely to view transgender people positively nor be more supportive of their rights. Only 11 percent of respondents reported knowing a transgender person, however, making it difficult to discern any true impact of contact. Experimental studies in which exposure to transgender people is manipulated (and hence more prevalent than in probability sampling-based survey research) may offer a greater opportunity to discern contact’s actual effects (see Flores et al. 2018). There was support for parasocial contact theory. Consistent with the notion that positive media portrayals can cultivate more positive attitudes toward marginalized groups, there was a positive association between TV consumption (but not other forms of media exposure) and views of transgender people. Absent a larger-scale content analysis of media portrayals of transgender people and measures of more specific media consumption, we are hesitant to say more here. Still, there is indication that respondents with greater exposure to media portrayals of transgender people viewed them more positively. Indeed, views of transgender people as a group are central to understanding attitudes toward transgender rights. The more positive a person’s views of transgender people, the more they tended to support protecting their rights. Akin to research on support for LGB rights (Brewer 2003), those with more positive views of transgender people were more supportive of policies benefiting them. This latter finding suggests one potential trajectory for public opinion on these issues. Brewer (2003, 2008) shows that increasingly positive perceptions of gay people resulted in an upswing in support for gay rights during the 1990s. This change reflected more positive media portrayals of LGB people (Garretson 2015). Given the media’s increasing (and mostly sympathetic) attention to transgender people, we might expect a similar pattern of change to occur for transgender issues. The finding that greater TV viewing cultivated more positive perceptions of transgender people suggests that such a mechanism may come into play. Throughout, these effects are arguably more limited than expected given the centrality of such variables to public opinion on other issues. Given the relative newness of the issue for many citizens, the relationship between their predispositions and transgender rights may take time to crystallize. However, the emergence of significant relationships in these early data suggests that public opinion on transgender rights already exhibits predictable structure. At the same time, it is possible that social desirability bias may inflate the apparent support for transgender people and rights reported here. The strong opposition to transgender candidates even among those predisposed to support them (i.e., Democrats, egalitarians, and those low in moral traditionalism) suggests potential conflict between public and private opinions on these issues. Although values and psychological traits helped explain attitudes toward transgender people and rights, they were much less useful in explaining attitudes toward transgender candidates. The solid opposition toward transgender politicians uncovered here was in many cases not lessened by variables we would expect to have a moderating effect. There was some suggestion that the least religious, those who knew a transgender person, and those with positive views of transgender people were less negatively predisposed toward a transgender candidate. Perhaps more striking, though, is the fact that opposition did not vary systematically by egalitarianism, moral traditionalism, ideology, NFCC, or media consumption. This suggests deep opposition even among groups that we would expect to be positively predisposed, and the significant obstacles that transgender candidates face. Of course, there are limitations to these conclusions: the experiment excluded those who do not identify with either major party, who may respond differently to transgender candidates than partisans do. Further, voters may react differently to experimental stimuli that include visual images or more detailed candidate information. Certainly, when actual transgender candidates run for office, they come with rich biographical and policy profiles that will shape public opinion as well, making this a particularly fruitful area for future research. It is difficult to speculate about how the public’s attitudes on this newly salient issue might affect future policy directions. On the one hand, the data showed sizeable support for several transgender rights issues, suggesting a more tolerant public than might be expected given recent restrictive laws. On the other hand, the same variables that have polarized political debate in general predicted attitudes on transgender rights too, suggesting that greater elite attention to the issue might lead to more divided responses among the public (Zaller 1992). Given the links between descriptive and substantive representation found elsewhere in the literature (e.g., Haider-Markel 2010), perhaps most striking was the widespread opposition to transgender candidates. These results suggest that any near-term change in policy will likely be the result of sympathetic allies rather than the election of transgender politicians themselves. Supplementary Data Supplementary data are freely available at Public Opinion Quarterly online. The authors thank the editors and reviewers for useful comments and suggestions, and Benjamin E. Bagozzi for extremely helpful guidance regarding the analysis of the candidate experiment. The data for this study were collected by the University of Delaware’s Center for Political Communication, with funding from the William P. Frank Foundation. All errors remain the authors’ own. Appendix: Question wording Need for cognitive closure, egalitarianism, and moral traditionalism. Now, please tell me how strongly you agree or disagree with each of the following statements. The first one is: [INSERT ITEM; RANDOMIZE ITEMS]—do you strongly agree, somewhat agree, somewhat disagree, or strongly disagree with this statement? I don’t like situations that are uncertain. I dislike questions that could be answered in many different ways. I dislike it when a person’s statement could mean many different things. One of the big problems in this country is that we don’t give everyone an equal chance. If people were treated more equally in this country we would have fewer problems. We should be more tolerant of people who choose to live according to their own moral standards even if they are very different from our own. Modern lifestyles are contributing to the breakdown of society. MEDIA CONSUMPTION On an average day, about how many hours do you watch television, including watching media on a computer, tablet, or phone? How often do you watch [INSERT; RANDOMIZE]: regularly, sometimes, hardly ever, or never? National network evening news programs such as ABC World News, CBS Evening News or NBC Nightly News News magazine shows such as 60 Minutes, 20/20, or Dateline The Fox News Cable Channel MSNBC CNN The Late Show with Stephen Colbert on CBS Last Week Tonight with John Oliver Shows on the cable network E! such as Keeping Up with the Kardashians and I Am Cait How often do you read a daily newspaper: regularly, sometimes, hardly ever, or never? Do you use the Internet or email at least occasionally? [IF “YES”] How often do you get news online: regularly, sometimes, hardly ever, or never? Familiarity with term “transgender.” Next, how familiar are you with the term transgender: Do you know what this term means, have you heard of it but are not sure what it means, or have you never heard of the term transgender before? Support for transgender rights Please tell me how strongly you favor or oppose each of the following. The first one is: [INSERT ITEM; RANDOMIZE ITEMS]—do you strongly favor, favor, oppose, or strongly oppose this? Laws that protect transgender people from job discrimination Allowing transgender people to serve openly in the U.S. military Requiring public buildings like courthouses to have gender-neutral restrooms for patrons to use Laws that protect transgender students from discrimination in schools Interpersonal contact Now, thinking about the people you know, do you have a close friend or family member who is transgender? Views of transgender people Next, please tell me how well each of the following describes most people. The first one is: [INSERT ITEM; RANDOMIZE ITEMS]—does this describe most people very well, somewhat well, not too well, or not well at all? Trustworthy Moral Happy Next, please tell me how well each of the following describes transgender people. The first one is: [INSERT ITEM; RANDOMIZE ITEMS IN SAME ORDER AS IN PREVIOUS QUESTION]—does this describe transgender people very well, somewhat well, not too well, or not well at all? Trustworthy Moral Happy Candidate experiment The precise wording of the candidate question depended on the respondent’s own party identification. Respondents identifying as Republicans or leaning toward the Republican Party were asked one of the following eight conditions; Democratic respondents received identical questions but the party affiliations of the candidates were reversed. The order of the candidates was rotated in each condition. Next, we’d like to ask how you would vote in a future election for public office. If the Republicans nominated a man and the Democrats nominated a man, who would you be more likely to vote for? the Republicans nominated a woman and the Democrats nominated a woman, who would you be more likely to vote for? the Republicans nominated a man and the Democrats nominated a woman, who would you be more likely to vote for? the Republicans nominated a woman and the Democrats nominated a man, who would you be more likely to vote for? the Republicans nominated a transgender man and the Democrats nominated a man, who would you be more likely to vote for? the Republicans nominated a transgender woman and the Democrats nominated a man, who would you be more likely to vote for? the Republicans nominated a transgender man and the Democrats nominated a woman, who would you be more likely to vote for? the Republicans nominated a transgender woman and the Democrats nominated a woman, who would you be more likely to vote for? Religiosity How religious do you consider yourself to be—very religious, somewhat religious, not too religious, or not at all religious? Ideology In general, would you describe your political views as very conservative, somewhat conservative, moderate, somewhat liberal, or very liberal? Party identity In politics today, do you consider yourself a Republican, Democrat, or Independent? [IF “DEMOCRAT”/“REPUBLICAN”] Would you call yourself a strong [Democrat/Republican] or not a very strong [Democrat/Republican]? [IF “INDEPENDENT”] As of today, do you lean more to the Republican Party or more to the Democratic Party? Age What is your age? Education What is the highest level of school you have completed or the highest degree you have received? Race Are you, yourself, of Hispanic or Latino origin, such as Mexican, Puerto Rican, Cuban, or some other Spanish background? What is your race? Are you white, black, Asian, or some other race? Gender Would you describe yourself as transgender, or not? Footnotes 1. See Online Appendix Table A5. 2. The gender of both candidates was randomized within conditions. Gender was not associated with vote choice, and so we collapse the conditions here. The Online Appendix includes details of all conditions and presents conjoint analysis showing no effect of the candidates’ gender. References Agiesta , Jennifer . 2016 . “ Poll: 6-in-10 Oppose Bills Like the North Carolina Transgender Bathroom Law .” CNN . Available at http://www.cnn.com/2016/05/09/politics/poll-transgender-bathroom- law-north-carolina/. Allport , Gordon W . 1954 . The Nature of Prejudice . Reading, MA : Addison-Wesley . American Association for Public Opinion Research . 2016 . Standard Definitions: Final Dispositions of Case Codes and Outcome Rates for Surveys , 9th ed . AAPOR . Bond , Bradley J. , and Benjamin L. Compton . 2015 . “ Gay On-Screen: The Relationship Between Exposure to Gay Characters on Television and Heterosexual Audiences’ Endorsement of Gay Equality .” Journal of Broadcasting and Electronic Media 59 : 717 – 32 . Google Scholar CrossRef Search ADS Bramlett , Brittany H . 2012 . “ The Cross-Pressures of Religion and Contact with Gays and Lesbians, and Their Impact on Same-Sex Marriage Opinion .” Politics & Policy 40 : 13 – 42 . Google Scholar CrossRef Search ADS Brewer , Paul R . 2003 . “ The Shifting Foundations of Public Opinion About Gay Rights .” Journal of Politics 65 : 1208 – 20 . Google Scholar CrossRef Search ADS Brewer , Paul R . 2008 . Value War: Public Opinion and the Politics of Gay Rights . Lanham, MD : Rowman and Littlefield . Casey , Logan S. , and Andrew Reynolds . 2015 . Standing Out: Transgender and Gender Variant Candidates and Elected Officials Around the World . LGBTQ Representation and Rights Initiative at the University of North Carolina at Chapel Hill . Cook , Elizabeth A. , Ted G. Jelen , and Clyde C. Wilcox . 1992 . Between Two Absolutes: Public Opinion and the Politics of Abortion . Boulder, CO : Westview . Craig , Stephen C. , Michael D. Martinez , James G. Kane , and Jason Gainous . 2005 . “ Core Values, Value Conflict, and Citizens’ Ambivalence about Gay Rights .” Political Research Quarterly 58 : 5 – 17 . Dasgupta , Nilanjana , and Shaki Asgari . 2004 . “ Seeing Is Believing: Exposure to Counterstereotypic Women Leaders and Its Effect on the Malleability of Automatic Gender Stereotyping .” Journal of Experimental Social Psychology 40 : 642 – 58 . Google Scholar CrossRef Search ADS Feldman , Stanley . 1988 . “ Structure and Consistency in Public Opinion: The Role of Core Beliefs and Values .” American Journal of Political Science 32 : 416 – 40 . Google Scholar CrossRef Search ADS Flores , Andrew R . 2015 . “ Attitudes Toward Transgender Rights: Perceived Knowledge and Secondary Interpersonal Contact .” Politics, Groups, and Identities 3 : 398 – 416 . Google Scholar CrossRef Search ADS Flores , Andrew R. , Donald P. Haider-Markel , Daniel C. Lewis , Patrick R. Miller , Barry L. Tadlock , and Jami K. Taylor . 2018 . “ Challenged Expectations: Mere Exposure Effects on Attitudes About Transgender People and Rights .” Political Psychology 39:197–216. doi: 10.1111/pops.12402 Garner , Andrew . 2013 . “ Ambivalence, the Intergroup Contact Hypothesis, and Attitudes about Gay Rights .” Politics & Policy 41 : 241 – 66 . Google Scholar CrossRef Search ADS Garretson , Jeremiah J . 2015 . “ Exposure to the Lives of Lesbians and Gays and the Origin of Young People’s Greater Support for Gay Rights .” International Journal of Public Opinion Research 27 : 277 – 88 . Google Scholar CrossRef Search ADS Gazzola , Stephanie B. , and Melanie A. Morrison . 2014 . “ Cultural and Personally Endorsed Stereotypes of Transgender Men and Transgender Women: Notable Correspondence or Disjunction ?” International Journal of Transgenderism 15 : 76 – 99 . Google Scholar CrossRef Search ADS Gerbner , George , Larry Gross , Michael Morgan , and Nancy Signorielli . 1984 . “ Political Correlates of Television Viewing .” Public Opinion Quarterly 48 : 283 – 300 . Google Scholar CrossRef Search ADS Golebiowska , Ewa A . 2001 . “ Group Stereotypes and Political Evaluation .” American Politics Research 29 : 535 – 65 . Google Scholar CrossRef Search ADS Green , Donald P. , Bradley Palmquist , and Eric Schickler . 2002 . Partisan Hearts and Minds: Political Parties and the Social Identity of Voters . New Haven, CT : Yale University Press . Haider-Markel , Donald P . 2010 . Out and Running: Gay and Lesbian Candidates, Elections, and Policy Representation . Washington, DC : Georgetown University Press . Haider-Markel , Donald P. , Patrick R. Miller , Andrew R. Flores , Daniel C. Lewis , Barry L. Tadlock , and Jami K. Taylor . 2017 . “ Bringing ‘T’ to the Table: Understanding Individual Support of Transgender Candidates for Public Office .” Politics, Groups, and Identities 5 : 399 – 417 . Google Scholar CrossRef Search ADS Herrick , Rebekah , and Sue Thomas . 1999 . “ The Effects of Sexual Orientation on Citizen Perceptions of Candidate Viability .” In Gays and Lesbians in the Democratic Process , edited by Ellen D. B. Riggle and Barry L. Tadlock , pp. 170 – 91 . New York : Columbia University Press . Human Rights Campaign . 2016 . “Anti-Transgender Legislation Spread Nationwide, Bills Targeting Transgender Children Surge.” Available at http://hrc-assets.s3-website-us-east-1.amazonaws.com//files/assets/resources/HRC-Anti-Trans-Issue-Brief-FINAL-REV2.pdf. Jost , John T. , Jack Glaser , Arie W. Kruglanski , and Frank J. Sulloway . 2003 . “ Political Conservatism as Motivated Social Cognition .” Psychological Bulletin 129 : 339 – 75 . Google Scholar CrossRef Search ADS PubMed King , Gary , Christopher J. L. Murray , Joshua A. Salomon , and Ajay Tandon . 2004 . “ Enhancing the Validity and Cross-Cultural Comparability of Measurement in Survey Research .” American Political Science Review 98 : 191 – 207 . Google Scholar CrossRef Search ADS Kruglanski , Arie W. , and Donna M. Webster . 1991 . “ Group Members’ Reactions to Opinion Deviates and Conformists at Varying Degrees of Proximity to Decision Deadline and of Environmental Noise .” Journal of Personality and Social Psychology 61 : 212 – 25 . Google Scholar CrossRef Search ADS PubMed Lewis , Gregory B . 2011 . “ The Friends and Family Plan: Contact with Gays and Support for Gay Rights .” Policy Studies Journal 39 : 217 – 38 . Google Scholar CrossRef Search ADS Lewis , Gregory B. , and Marc A. Rogers . 1999 . “ Does the Public Support Equal Employment Rights for Gays and Lesbians ?” In Gays and Lesbians in the Democratic Process , edited by Ellen D. B. Riggle and Barry L. Tadlock , pp. 118 – 45 . New York : Columbia University Press . Mastro , Dana , Elizabeth Behm-Morawitz , and Michelle Ortiz . 2007 . “ The Cultivation of Social Perceptions of Latinos: A Mental Models Approach .” Media Psychology 9 : 347 – 65 . Google Scholar CrossRef Search ADS McCann , James A . 1997 . “ Electoral Choices and Core Value Change: The 1992 Presidential Campaign .” American Journal of Political Science 4 : 564 – 83 . Google Scholar CrossRef Search ADS Miller , Patrick R. , Andrew R. Flores , Donald P. Haider-Markel , Daniel C. Lewis , Barry L. Tadlock , and Jami K. Taylor . 2017 . “ Transgender Politics as Body Politics: Effects of Disgust Sensitivity and Authoritarianism on Transgender Rights Attitudes .” Politics, Groups, and Identities 5 : 4 – 24 . Google Scholar CrossRef Search ADS Newport , Frank . 2016 . “ Disentangling Attitudes Toward Transgender Bathroom Use .” Gallup , May 20. Available at http://www.gallup.com/opinion/polling-matters/191774/disentangling-attitudes-toward-transgender-bathroom.aspx. Norton , Aaron T. , and Gregory M. Herek . 2013 . “ Heterosexuals’ Attitudes Toward Transgender People: Findings from a National Probability Sample of U.S. Adults .” Sex Roles 68 : 738 – 53 . Google Scholar CrossRef Search ADS Olson , Laura R. , Wendy Cadge , and James T. Harrison . 2006 . “ Religion and Public Opinion About Same-Sex Marriage .” Social Science Quarterly 87 : 340 – 60 . Google Scholar CrossRef Search ADS Pettigrew , Thomas F. , and Linda R. Tropp . 2006 . “ A Meta-Analytic Test of Intergroup Contact Theory .” Journal of Personality and Social Psychology 90 : 751 – 83 . Google Scholar CrossRef Search ADS PubMed Pew Research Center . 2012 . “ Assessing the Representativeness of Public Opinion Surveys .” Pew Research Center , May 15. Available at http://www.people-press.org/2012/05/15/assessing-the-representativeness-of-public-opinion-surveys/. Roets , Arne , and Alain Van Hiel . 2011 . “ Item Selection and Validation of a Brief, 15-Item Version of the Need for Closure Scale .” Personality and Individual Differences 50 : 90 –9 4 . Google Scholar CrossRef Search ADS Rössler , Patrick , and Hans-Bernd Brosius . 2001 . “ Do Talk Shows Cultivate Adolescents’ Views of the World? A Prolonged‐Exposure Experiment .” Journal of Communication 51 : 143 – 63 . Google Scholar CrossRef Search ADS Schiappa , Edward , Peter B. Gregg , and Dean E. Hewes . 2006 . “ Can One TV Show Make a Difference? Will & Grace and the Parasocial Contact Hypothesis .” Journal of Homosexuality 51 : 15 – 37 . Google Scholar CrossRef Search ADS PubMed Stoutenborough , James W. , Donald P. Haider-Markel , and Mahalley D. Allen . 2006 . “ Reassessing the Impact of Supreme Court Decisions on Public Opinion: Gay Civil Rights Cases .” Political Research Quarterly 59 : 419 – 33 . Google Scholar CrossRef Search ADS University of Illinois Springfield Survey Research Office . 2015 . Examining Attitudes towards the “T” in LGBT: Public Support for Transgender Rights and Supportive Policies . Available at https://transgenderinclusivepolicysurvey.wordpress.com. Webster , Donna M. , and Arie W. Kruglanski . 1994 . “ Individual Differences in Need for Cognitive Closure .” Journal of Personality and Social Psychology 67 : 1049 – 62 . Google Scholar CrossRef Search ADS PubMed Wilcox , Clyde , Paul R. Brewer , Shauna Shames , and Celinda Lake . 2007 . “ If I Bend This Far I Will Break? Public Opinion about Same-Sex Marriage .” In The Politics of Same-Sex Marriage , edited by Craig A. Rimmerman and Clyde Wilcox , pp. 215 – 242 . Chicago : University of Chicago Press . Zaller , John R . 1992 . The Nature and Origins of Mass Opinion . Cambridge : Cambridge University Press . Google Scholar CrossRef Search ADS © The Author(s) 2018. Published by Oxford University Press on behalf of the American Association for Public Opinion Research. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Public Opinion Quarterly Oxford University Press

Explaining Public Opinion toward Transgender People, Rights, and Candidates

Public Opinion Quarterly , Volume Advance Article (2) – May 15, 2018

Loading next page...
 
/lp/ou_press/explaining-public-opinion-toward-transgender-people-rights-and-pitEQY1OKM
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the American Association for Public Opinion Research. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
ISSN
0033-362X
eISSN
1537-5331
D.O.I.
10.1093/poq/nfy009
Publisher site
See Article on Publisher Site

Abstract

Abstract What explains public opinion toward transgender people, rights, and candidates? Drawing on original data from a national telephone survey of US adults, this study explains attitudes regarding (1) the personal characteristics of transgender people; (2) a variety of transgender rights; and (3) transgender candidates for public office, measured through a randomized experiment included in the survey. Results indicate majority support on most policy questions, but more tepid views of transgender people, and solid opposition to supporting a transgender candidate for office. Our analyses reflect and extend previous research on American public opinion. Respondents’ fundamental values (egalitarianism, moral traditionalism, party identity, ideology, and religiosity) and personality characteristics (need for cognitive closure) predict views of transgender people and support for their rights. A significant relationship also emerged between television use and views of transgender people, suggesting that media portrayals may play a role in shaping these perceptions. In contrast, there is no evidence that interpersonal contact with a transgender person is related to opinions. Further, many of these independent variables have little moderating effect on responses to transgender candidates, which remain negative among most subgroups. Introduction Recent years have seen increased attention to transgender people in both popular culture and political debate. While media portrayal of celebrities such as former Olympian Caitlyn Jenner and actress Laverne Cox has been largely positive, politicians have sponsored a “wave of anti-transgender legislation” (Human Rights Campaign 2016). Policies restricting transgender rights have found majority support among American legislators (as in the case of North Carolina’s “bathroom bill”) and voters (as in the passing of the Houston Equal Rights Ordinance). Likewise, the few openly transgender candidates to run for office have failed to win election beyond a handful of local races (Casey and Reynolds 2015). Although this political backlash suggests widespread antipathy toward transgender people, rights, and candidates, scholarly research to date has provided a more nuanced view of public opinion. Affect toward transgender people and support for their rights often reflect values such as egalitarianism, moral traditionalism, conservatism, and religiosity, as well as contact with LGBT people (Norton and Herek 2013; Flores 2015; Miller et al. 2017; Flores et al. 2018). Building upon this research, we assess the opinions of cisgender Americans (those whose gender identity matches their assigned sex) toward: (1) the character traits of transgender people; (2) transgender rights; and (3) a hypothetical transgender candidate, assessed via a survey experiment. Drawing from previous studies and from general theories of public opinion, we explain these opinions with a wide range of independent variables. Values—including egalitarianism, moral traditionalism, ideology, party identification, and religiosity—and psychological characteristics—specifically, a need for cognitive closure—predict support for transgender rights, though of these only moral traditionalism predicts views of transgender people. We explore the effects of media exposure, finding evidence that television viewing cultivates more favorable views of transgender people. Finally, we document opposition toward transgender candidates: respondents’ support for their party plunges when the nominee is transgender. This opposition is in many cases not moderated by the values and characteristics that otherwise predict support for transgender people and rights, suggesting a difficult path ahead for transgender candidates. Previous Studies of Opinion about Transgender People, Rights, and Candidates Several recent surveys suggest public support for transgender rights. The 2015 Attitudes towards Transgender Supportive Policies Survey reports that 65 percent of Americans believe that transgender people should “receive protections provided by anti-discrimination laws” (University of Illinois Springfield Survey Research Office 2015). Similarly, a 2016 CNN poll finds that 75 percent favor “laws that guarantee equal protection for transgender people in jobs, housing and public accommodations” (Agiesta 2016). However, such broad support does not extend to issues surrounding bathroom access: public opinion on policies requiring transgender people to use the bathroom matching the sex they were assigned at birth is much more divided (Newport 2016). Beyond such descriptive results, several recent studies have sought to understand the political and psychological predictors of these attitudes. Norton and Herek (2013) find that survey respondents who held binary conceptions of gender, higher levels of authoritarianism, less egalitarian attitudes, more conservative ideologies, and greater religiosity all rated transgender people less favorably on feeling thermometers. Similarly, Flores et al. (2018) report that moral traditionalism and partisanship—as well as mere exposure to images of transgender people—influence transphobia and discomfort with gender nonconformity. Other studies focus on predicting support for laws protecting transgender rights. Flores (2015) finds that gender and interpersonal contact with gays and lesbians (although not with transgender people) are significant predictors of support for antidiscrimination laws. Further, he shows that respondents who believed they were well informed about transgender issues were more supportive of such protections. Other research echoes these findings and shows that core values such as authoritarianism, ideology, partisan identity, and personality characteristics such as disgust sensitivity are also significant predictors (Miller et al. 2017). In contrast, only a single piece of research we know of examines attitudes toward transgender candidates. Haider-Markel et al. (2017) ask respondents to assess their likelihood of voting for a transgender candidate who shared their views on most issues. A majority of voters report they would probably or definitely vote for the candidate, with liberals, Democrats, and the least religious being most likely to support them. This self-reported support conflicts with previous work that documents strong opposition to politicians from other marginalized groups. For example, both experimental (Herrick and Thomas 1999; Golebiowska 2001) and observational (Haider-Markel 2010) studies reveal negative reactions to lesbian, gay, and bisexual (LGB) candidates. Coupled with electoral losses of transgender candidates in recent years (Casey and Reynolds 2015), one might expect greater opposition to transgender candidates than appears in respondents’ self-reports. At the same time, given previous work showing that values and partisan attachments predict support for LGB candidates much as they do for LGB rights (Haider-Markel 2010), one might also expect these predispositions to moderate opposition toward transgender candidates. In the next section, we build on these prior studies, drawing on general theories of public opinion to consider a broad range of explanations for public opinion toward transgender people, rights, and candidates. Explaining Public Opinion Toward Transgender People, Rights, and Candidates INTERPERSONAL CONTACT Numerous studies find that positive interactions with members of an outgroup reduce prejudice toward the group (Allport 1954; Pettigrew and Tropp 2006). For example, knowing an LGB person is associated with greater support for LGB rights and more positive attitudes toward LGB people (Brewer 2008; Lewis 2011; Bramlett 2012; Garner 2013). Previous studies on transgender attitudes offer mixed support for the theory. Flores (2015) finds no such effect of contact on support for nondiscrimination policies; meanwhile, Flores et al. (2018) show that mere exposure to information and images of transgender people in a survey experiment reduces transphobia and discomfort with gender nonconformity. In this study, we revisit the intergroup contact hypothesis across a wider range of dependent variables. H1: Respondents who know a transgender person will hold more positive views of transgender people, rights, and candidates. VALUES An extensive literature shows that opinions about specific groups and policies are often based on more fundamental values. These include core values such as egalitarianism (Feldman 1988) and moral traditionalism (McCann 1997), broad political orientations such as political ideology and party identification (Zaller 1992), and religiosity (Cook, Jelen, and Wilcox 1992). For example, all of these values can influence public opinion about LGB rights (Lewis and Rogers 1999; Brewer 2003; Craig et al. 2005; Olson, Cadge, and Harrison 2006; Stoutenborough, Haider-Markel, and Allen 2006; Wilcox et al. 2007) and support for LGB candidates (Haider-Markel 2010; Haider-Markel et al. 2017). Previous research has examined the role of many, but not all, of these values in shaping opinions toward transgender people. Norton and Herek (2013) find that anti-egalitarianism, conservatism, and religiosity all predicted cooler feelings toward transgender people. Building on their results, we hypothesize that: H2: More egalitarian respondents will hold more positive views of transgender people, rights, and candidates. H3: More religious respondents will hold less positive views of transgender people, rights, and candidates. H4: More conservative respondents will hold less positive views of transgender people, rights, and candidates. Expanding the scope of values under investigation, we hypothesize that: H5: More morally traditional respondents will hold less positive views of transgender people, rights, and candidates. H6: Democrats will hold more positive views of transgender people, rights, and candidates than will Republicans. NEED FOR COGNITIVE CLOSURE Personality traits can shape dispositions and orientations toward people and events (Kruglanski and Webster 1991; Webster and Kruglanski 1994; Jost et al. 2003). A need for cognitive closure (NFCC) refers to one’s tendency to impose fixed meanings on situations and to feel an aversion toward ambiguous information and experiences. Individuals with a high need for cognitive closure may also be uncomfortable with the ambiguity inherent in the concept of a gender identity that does not match the sex assigned at birth. As such, we anticipate that: H7: Respondents high in NFCC will hold less positive views of transgender people, rights, and candidates. MEDIA EXPOSURE Cultivation research (Gerbner et al. 1984) posits that the dominant themes in television content can shape viewers’ real-world perceptions. For example, media portrayals can affect views of marginalized groups including women (Dasgupta and Asgari 2004), Latinos (Mastro, Behm-Morawitz, and Ortiz 2007), and gays and lesbians (Rössler and Brosius 2001). The parasocial contact hypothesis—that exposure to outgroups via the media can have similar effects to actual contact in the real world—helps explain these results. Viewing LGB characters on screen leads to greater support for gay rights and warmer evaluations of gay people (Schiappa, Gregg, and Hewes 2006; Bond and Compton 2015; Garretson 2015). One study that exposed participants to cross-dressing comedian Eddie Izzard (Schiappa, Gregg, and Hewes 2006) reported a significant reduction in transprejudice. Given the increasingly prominent and positive portrayals of transgender people in entertainment television programming, we hypothesize that: H8: Respondents with greater levels of overall television exposure will hold more positive views of transgender people, rights, and candidates. Meanwhile, news coverage may track the politically and socially polarized debate surrounding transgender issues. Thus, we ask: RQ1: Will news media exposure among respondents predict their views of transgender people, rights, and candidates? TRANSGENDER CANDIDATES Finally, given the research cited earlier showing opposition to candidates from minority groups, we hypothesize that: H9: Respondents will express less support for transgender candidates than cisgender candidates. DEMOGRAPHIC CHARACTERISTICS Previous research suggests that various demographic characteristics are also likely to affect opinion on transgender issues (Norton and Herek 2013; Flores 2015). Sex, age, education, and race may all shape attitudes, with female, younger, better-educated, and white respondents holding more favorable opinions. As such, we control for these variables in the models that follow. Data and Methods The data for this study came from the University of Delaware Center for Political Communication’s National Agenda Poll, a random-digit-dial (RDD) telephone survey of a sample of 901 adults living in the continental United States. The survey was conducted by Princeton Survey Research Associates International (PSRAI) from November 11 to 17, 2015, using a dual sampling frame of landline (n = 451) and cell phone (n = 450) respondents provided by Survey Sampling International (SSI). Interviews were conducted in English by Princeton Data Source. The response rates for the landline and cellular samples were 7 and 5 percent (calculated using AAPOR RR3), respectively (AAPOR 2016).1 These are low in historical perspective but comparable to typical rates for surveys conducted by major polling organizations (e.g., Pew Research Center 2012). The sampling error for the full weighted sample is ±4.0 percentage points. The Online Appendix includes additional information regarding the sample, including comparison to the general US population, descriptive statistics for key variables, the weighting that is used throughout, and the sample disposition. Four respondents who identified themselves as transgender were excluded from the analysis, given that we are interested in assessing cisgender people’s views of transgender issues. VIEWS OF TRANSGENDER PEOPLE Respondents were asked how well the words “trustworthy,” “moral,” and “happy” described “most people,” with response options of “not well at all” (1), “not too well” (2), “somewhat well” (3), and “very well” (4). They were then asked how well the same words described “transgender people.” Previous research has shown that some of the most pervasive negative stereotypes held of transgender people cast them as “abnormal,” “confused,” and “abusive” (Gazzola and Morrison 2014). Views of transgender people were standardized relative to views of most people in order to correct for interpersonal incomparability in respondents’ evaluations of people in general (see King et al. 2004). Respondents’ views of “most people” were subtracted from their views of “transgender people,” to create variables that take on a positive value if the respondent viewed transgender people as more trustworthy (moral, happy) than most people, and negative values if they viewed most people as more trustworthy (moral, happy) than transgender people. Values of zero indicated no difference between views of transgender people and most people. Views were on average relatively negative: respondents viewed transgender people as less moral (M = –.07, SD = .98) and less happy (M = –.12, SD = .87), but more trustworthy (M = .08, SD = .88), than most people. The median score for each item was zero, however. The three items were averaged into an index capturing views of transgender people (M = –.06, SD = .76; α = .68). ATTITUDES TOWARD TRANSGENDER RIGHTS Using a four-point scale (strongly oppose [1]; oppose [2]; favor [3]; and strongly favor [4]), respondents were asked about four policies: (1) “laws that protect transgender students from discrimination in schools” (73.2 percent favored or strongly favored, M = 2.93, SD = .97); (2) “laws that protect transgender people from job discrimination” (72.7 percent favored or strongly favored, M = 2.94, SD = .95); (3) “allowing transgender people to serve openly in the U.S. military” (64.7 percent favored or strongly favored, M = 2.76, SD = .98); and (4) “requiring public buildings like courthouses to have gender-neutral restrooms for patrons to use” (54.0 percent favored or strongly favored, M = 2.53, SD = .98). In each case, the median response was 3 (favoring the policy). The four items were averaged into an index measuring support for transgender rights (M = 2.78, SD = .80; α = .83). SUPPORT FOR TRANSGENDER CANDIDATES The survey included an experiment that probed support for a transgender candidate. Given the extremely high likelihood of voting for a candidate of one’s own party affiliation (Green, Palmquist, and Schickler 2002), the respondent’s partisan identification as expressed earlier in the survey structured the experimental design. Respondents were asked: “Next, we’d like to ask how you would vote in a future election for public office. If the [respondent’s party] nominated a ______________ and the [opposing party] nominated a ______________, who would you be more likely to vote for?” The candidates’ assigned parties were based on the respondent’s party identity as reflected in support of or leaning toward a given party (“pure” independents and those identifying with a minor party [N = 168] were excluded). For example, a respondent who identified as a Republican would be asked the following: “Next, we’d like to ask how you would vote in a future election for public office. If the Republicans nominated a ______________ and the Democrats nominated a ______________, who would you be more likely to vote for?” Respondents were assigned to one of two conditions. In both conditions, the opposing party’s candidate was presented as cisgender (e.g., “a man”). Their own party’s candidate, however, was randomly presented either as cisgender or as transgender (e.g., “a man” versus “a transgender man”).2 INTERPERSONAL CONTACT A dichotomous variable measured whether respondents had a “close friend or family member” who is transgender (1) or not (0). Eleven percent of respondents reported such interpersonal contact. VALUES Egalitarianism was captured through agreement with two statements based on Feldman (1988): “One of the big problems in this country is that we don’t give everyone an equal chance” and “If people were treated more equally in this country we would have fewer problems.” Responses were coded from 1 (strongly disagree) to 4 (strongly agree) and averaged into a single measure (M = 2.97, SE = .97, r = .59), with higher values indicating greater egalitarianism. Moral traditionalism was measured through agreement with two statements: “We should be more tolerant of people who choose to live according to their own moral standards even if they are very different from our own” and “Modern lifestyles are contributing to the breakdown of society” (Cook, Jelen, and Wilcox 1992). Responses were coded from 1 (strongly disagree) to 4 (strongly agree) and averaged into a single measure after reverse-coding the first item (M = 2.32, SE = .82, r = .19), with higher values indicating greater traditionalism. Religiosity was measured with an item asking respondents how religious they considered themselves to be, ranging from “not at all religious” (1) to “very religious” (4). Ideology was coded from 1 (very liberal) to 5 (very conservative). Partisan identification was coded from 1 (Strong Democrat) to 6 (Strong Republican). NEED FOR COGNITIVE CLOSURE NFCC was measured with three items taken from Roets and Van Hiel (2011): “I don’t like situations that are uncertain”; “I dislike questions that could be answered in many different ways”; and “I dislike it when a person’s statement could mean many different things.” Responses ranged from 1 (strongly disagree) to 4 (strongly agree), with higher values indicating greater NFCC. The three items were combined into an index (M = 2.75, SE = .75, α = .55). FAMILIARITY WITH TERM “TRANSGENDER” Since knowledge of transgender issues predicts support for transgender rights (Flores 2015), a measure of respondents’ familiarity with the term “transgender” was included. Respondents self-identified as knowing what the term meant (coded as 2), having heard it but not being sure of its meaning (1), or never having heard it (0). Before the batteries measuring attitudes on transgender issues, interviewers clarified for all respondents that “the term transgender applies to people who live their gender in a way that does not match the sex listed on their original birth certificate or who physically change their sex.” Previous research finds that providing the definition does not alter responses to following items (Flores 2015). MEDIA EXPOSURE Respondents were asked the number of hours they spent watching television on an average day (recoded to range from 0 to 6, with higher numbers coded at the maximum). The frequency with which respondents consumed various media was measured on a scale ranging from “never” (1) to “regularly” (4). These media were “national network evening news programs such as ABC World News, CBS Evening News or NBC Nightly News”; “news magazine shows such as 60 Minutes, 20/20, or Dateline”; “the Fox News Cable Channel”; MSNBC and CNN (averaged into a single measure); “The Late Show with Stephen Colbert on CBS” and “Last Week Tonight with John Oliver” (averaged into a single measure); and “shows on the cable network E! such as Keeping Up with the Kardashians and I Am Cait.” Frequency of reading a daily newspaper and getting news online were measured on the same scale. DEMOGRAPHIC VARIABLES Age was measured in decades. Indicator variables for female, black, and Hispanic respondents reflected gender and racial/ethnic identities. Education was measured on an ordinal scale ranging from less than high school (1) to a postgraduate or professional degree (8). MODELS AND PRESENTATION OF RESULTS To estimate views of transgender people and support for transgender rights, linear regression models using survey weights are estimated. Since the dependent variable in the candidate experiment is categorical (a vote for one’s party, the other party, or a “don’t know/refused” response), the analyses for this variable employ multinomial logit models. Throughout, results are presented in conventional regression tables with coefficients and standard errors. To ease interpretation, results are discussed in terms of predicted values for linear regressions and predicted probabilities for multinomial regressions. When simulating the regression results, all numeric variables are held to their mean value in the weighted dataset and categorical variables to their modal value (female is set to one; interpersonal contact, black, and Hispanic to zero). Thus, the predicted values are for white women with no interpersonal contact who have average values of all other variables. Explaining Views of Transgender People Results from linear regression models predicting views of transgender people, using the combined scale of the three items, are shown in the first column of table 1 (estimating models for each item separately, and with an ordered logistic estimator, led to substantively consistent findings; these results are available in the Online Appendix). Table 1. OLS regression models predicting opinions toward transgender people and rights Views of transgender people (n=531) Support for transgender rights (n=531) B (s.e.) B (s.e.) Intercept –0.17 (0.36) 3.66 (0.32)*** Values  Egalitarianism 0.00 (0.04) 0.11 (0.04)**  Moral traditionalism –0.10 (0.05)* –0.15 (0.04)***  Religiosity 0.02 (0.04) –0.12 (0.03)***  Ideology –0.06 (0.05) –0.15 (0.03)***  Party ID –0.03 (0.02) –0.04 (0.02)* Need for cognitive closure 0.06 (0.05) –0.08 (0.04)* Views of transgender people 0.21 (0.04)*** Interpersonal contact 0.07 (0.09) 0.09 (0.10) Familiarity with term 0.06 (0.07) 0.02 (0.05) Media consumption  TV use 0.06 (0.02)** 0.03 (0.02)  Network news 0.07 (0.04) 0.03 (0.04)  News magazines –0.00 (0.04) –0.07 (0.04)  Fox News –0.04 (0.03) –0.05 (0.03)  CNN/MSNBC –0.03 (0.04) 0.03 (0.04)  Newspaper 0.01 (0.03) 0.03 (0.03)  Online news –0.01 (0.04) –0.01 (0.04)  E Network 0.02 (0.05) 0.07 (0.05)  Late Show/Last Week Tonight 0.03 (0.05) –0.00 (0.04) Demographics  Age (decades) –0.25 (0.25) –0.09 (0.21)  Education –0.00 (0.02) 0.02 (0.02)  Female 0.23 (0.07)** 0.33 (0.06)***  Hispanic 0.12 (0.11) –0.12 (0.09)  Black 0.25 (0.16) –0.06 (0.11) McFadden’s pseudo-R2 0.19 0.52 Views of transgender people (n=531) Support for transgender rights (n=531) B (s.e.) B (s.e.) Intercept –0.17 (0.36) 3.66 (0.32)*** Values  Egalitarianism 0.00 (0.04) 0.11 (0.04)**  Moral traditionalism –0.10 (0.05)* –0.15 (0.04)***  Religiosity 0.02 (0.04) –0.12 (0.03)***  Ideology –0.06 (0.05) –0.15 (0.03)***  Party ID –0.03 (0.02) –0.04 (0.02)* Need for cognitive closure 0.06 (0.05) –0.08 (0.04)* Views of transgender people 0.21 (0.04)*** Interpersonal contact 0.07 (0.09) 0.09 (0.10) Familiarity with term 0.06 (0.07) 0.02 (0.05) Media consumption  TV use 0.06 (0.02)** 0.03 (0.02)  Network news 0.07 (0.04) 0.03 (0.04)  News magazines –0.00 (0.04) –0.07 (0.04)  Fox News –0.04 (0.03) –0.05 (0.03)  CNN/MSNBC –0.03 (0.04) 0.03 (0.04)  Newspaper 0.01 (0.03) 0.03 (0.03)  Online news –0.01 (0.04) –0.01 (0.04)  E Network 0.02 (0.05) 0.07 (0.05)  Late Show/Last Week Tonight 0.03 (0.05) –0.00 (0.04) Demographics  Age (decades) –0.25 (0.25) –0.09 (0.21)  Education –0.00 (0.02) 0.02 (0.02)  Female 0.23 (0.07)** 0.33 (0.06)***  Hispanic 0.12 (0.11) –0.12 (0.09)  Black 0.25 (0.16) –0.06 (0.11) McFadden’s pseudo-R2 0.19 0.52 Note.—Estimates based on weighted data. Coefficients are from linear regression models with standard errors in parentheses. Omitted category for gender is male, for race is White. *p < .05; **p < .01; ***p < .001. View Large Table 1. OLS regression models predicting opinions toward transgender people and rights Views of transgender people (n=531) Support for transgender rights (n=531) B (s.e.) B (s.e.) Intercept –0.17 (0.36) 3.66 (0.32)*** Values  Egalitarianism 0.00 (0.04) 0.11 (0.04)**  Moral traditionalism –0.10 (0.05)* –0.15 (0.04)***  Religiosity 0.02 (0.04) –0.12 (0.03)***  Ideology –0.06 (0.05) –0.15 (0.03)***  Party ID –0.03 (0.02) –0.04 (0.02)* Need for cognitive closure 0.06 (0.05) –0.08 (0.04)* Views of transgender people 0.21 (0.04)*** Interpersonal contact 0.07 (0.09) 0.09 (0.10) Familiarity with term 0.06 (0.07) 0.02 (0.05) Media consumption  TV use 0.06 (0.02)** 0.03 (0.02)  Network news 0.07 (0.04) 0.03 (0.04)  News magazines –0.00 (0.04) –0.07 (0.04)  Fox News –0.04 (0.03) –0.05 (0.03)  CNN/MSNBC –0.03 (0.04) 0.03 (0.04)  Newspaper 0.01 (0.03) 0.03 (0.03)  Online news –0.01 (0.04) –0.01 (0.04)  E Network 0.02 (0.05) 0.07 (0.05)  Late Show/Last Week Tonight 0.03 (0.05) –0.00 (0.04) Demographics  Age (decades) –0.25 (0.25) –0.09 (0.21)  Education –0.00 (0.02) 0.02 (0.02)  Female 0.23 (0.07)** 0.33 (0.06)***  Hispanic 0.12 (0.11) –0.12 (0.09)  Black 0.25 (0.16) –0.06 (0.11) McFadden’s pseudo-R2 0.19 0.52 Views of transgender people (n=531) Support for transgender rights (n=531) B (s.e.) B (s.e.) Intercept –0.17 (0.36) 3.66 (0.32)*** Values  Egalitarianism 0.00 (0.04) 0.11 (0.04)**  Moral traditionalism –0.10 (0.05)* –0.15 (0.04)***  Religiosity 0.02 (0.04) –0.12 (0.03)***  Ideology –0.06 (0.05) –0.15 (0.03)***  Party ID –0.03 (0.02) –0.04 (0.02)* Need for cognitive closure 0.06 (0.05) –0.08 (0.04)* Views of transgender people 0.21 (0.04)*** Interpersonal contact 0.07 (0.09) 0.09 (0.10) Familiarity with term 0.06 (0.07) 0.02 (0.05) Media consumption  TV use 0.06 (0.02)** 0.03 (0.02)  Network news 0.07 (0.04) 0.03 (0.04)  News magazines –0.00 (0.04) –0.07 (0.04)  Fox News –0.04 (0.03) –0.05 (0.03)  CNN/MSNBC –0.03 (0.04) 0.03 (0.04)  Newspaper 0.01 (0.03) 0.03 (0.03)  Online news –0.01 (0.04) –0.01 (0.04)  E Network 0.02 (0.05) 0.07 (0.05)  Late Show/Last Week Tonight 0.03 (0.05) –0.00 (0.04) Demographics  Age (decades) –0.25 (0.25) –0.09 (0.21)  Education –0.00 (0.02) 0.02 (0.02)  Female 0.23 (0.07)** 0.33 (0.06)***  Hispanic 0.12 (0.11) –0.12 (0.09)  Black 0.25 (0.16) –0.06 (0.11) McFadden’s pseudo-R2 0.19 0.52 Note.—Estimates based on weighted data. Coefficients are from linear regression models with standard errors in parentheses. Omitted category for gender is male, for race is White. *p < .05; **p < .01; ***p < .001. View Large Only three independent variables were significantly related to perceptions of transgender people. First, consistent with H5, the more a respondent subscribed to traditional moral values, the less positively they viewed transgender people (β = –.10, SE = .05, p = .04). Simulating predicted values from the regression coefficients showed that respondents with the lowest levels of moral traditionalism held neutral to slightly positive views of transgender people (an expected value of .14 with 95 percent confidence intervals [–.04, .32] on the –3 to +3 scale). As hypothesized, those with the highest levels of moral traditionalism held more negative views of transgender people (–.18 [–.35, .00]). Second, consistent with cultivation theory and H8, television consumption was significantly related to perceptions of transgender people (β = .06, SE = .02, p = .002). The more TV a respondent watched, the more positive their perceptions of transgender people were: someone who did not watch any television was predicted to hold relatively negative impressions of transgender people (–.18 [–.33, –.02]), whereas someone who watched six or more hours was predicted to hold favorable perceptions (.20 [.04, .36]). Finally, in line with previous research (Norton and Herek 2013), there were gender differences: women had more positive attitudes toward transgender people than did men. Many of the independent variables expected to influence perceptions of transgender people had no statistically discernible effect, however. Egalitarianism, ideology, party identification, religiosity, NFCC, interpersonal contact, and media consumption aside from general TV use (RQ1) all appeared unrelated to perceptions of transgender people. When it comes to views of transgender people, no support emerged for H1–4 and H6–7. Explaining Attitudes Toward Transgender Rights Support for the four transgender rights items, combined into a single index, were modeled with the same independent variables used previously, as well as views of transgender people (results from models of each individual item and multinomial regression estimators are available in the Online Appendix). The results, shown in the second column of table 1, support several hypotheses. The predispositions and values that the literature identifies as strong predictors of support for minority rights largely had the expected relationships. Figure 1 shows predicted support for transgender rights across the ranges of the independent variables. These predicted values are simulated from the model in table 1, for a hypothetical respondent with mean values on all numeric variables and the modal category for categorical variables (i.e., a white woman with no interpersonal contact). Figure 1. View largeDownload slide Predicted support for transgender rights. Ninety-five percent confidence intervals. Simulated from model in table 1 using weighted data. For each simulation, all numeric variables are held at their means; interpersonal contact, black, and Hispanic are set to zero and female to one. Figure 1. View largeDownload slide Predicted support for transgender rights. Ninety-five percent confidence intervals. Simulated from model in table 1 using weighted data. For each simulation, all numeric variables are held at their means; interpersonal contact, black, and Hispanic are set to zero and female to one. For instance, as shown in the first plot of figure 1, the more egalitarian respondents were, the more likely they were to support transgender rights (as predicted by H2). Someone with the lowest support for egalitarianism was predicted to score 2.78 on the transgender rights index [2.61, 2.94], compared to someone with the greatest support for egalitarian ideals, who was predicted to score 3.09 [2.98, 3.20]. Other core values had relationships of roughly the same magnitude. Supporting H5, those who did not subscribe to morally traditional values were predicted to support transgender rights (3.18 [3.05, 3.32]) more than those who did (2.73 [2.57, 2.89]). Consistent with H3, support decreased with religiosity, from 3.20 [3.04, 3.35] for the least religious to 2.85 [2.75, 2.97] for the most religious. As previous research would lead one to expect, respondents who held egalitarian ideals were likely to support rights for transgender people, while those high in moral traditionalism and religiosity were unlikely to do so. As predicted by H4 and H6, both party identification and ideology were significant predictors of support for transgender rights, such that Republicans and conservatives were less supportive of policies protecting transgender people than Democrats and liberals, respectively. Strong Democrats were more supportive than strong Republicans (3.08 [2.94, 3.21] versus 2.85 [2.72, 2.98]). Likewise, those who identified as very liberal were more supportive than those who identified as very conservative (3.31 [3.14, 3.47] versus 2.72 [2.56, 2.86]). Psychological characteristics mattered, too. Respondents at the low end of the NFCC scale were predicted to be more supportive of transgender rights (3.12 [2.97, 3.28]) than those at the high end (2.88 [2.77, 3.01]), consistent with H7 and previous research showing that heightened NFCC results in more negative responses to nonconformity (Kruglanski and Webster 1991). Views of transgender people were a significant predictor of support for transgender rights (β = .21, SE = .04, p < .001). Respondents with the most negative views of transgender people were unlikely to support policies that benefit them (2.36 [2.13, 2.59]). Those with the most positive views, however, were highly likely to support their rights (3.62 [3.38, 3.87]). Echoing earlier research on LGB issues (Brewer 2003), those with warmer views of transgender people were more sympathetic to policies protecting them. There was no support for H8; those who watched more television were not significantly more favorable toward transgender rights. Nor was any specific form of media consumption related to opinion (RQ1). And as before, interpersonal contact had no discernible effect. Contrary to H1, knowing someone who identifies as transgender did not predict support for policies that benefit them. Instead, values such as moral traditionalism, egalitarianism, party identification, ideology, and religiosity, as well as a psychological need for closure, were significant predictors of support. Furthermore, perceptions of transgender people strongly shaped support for affording them rights. While these effects may seem small when taken individually, their combined impact is substantial. Someone high in values associated with support for transgender rights—a female respondent who was one standard deviation above the mean on egalitarianism and views of transgender people, and one standard deviation below the mean on moral traditionalism, religiosity, NFCC, ideology, and party ID—was predicted to score 3.78 [3.65, 3.92]. Her male counterpart, one standard deviation from the mean in the opposite direction on each variable, was predicted to score just 1.84 [1.71, 1.97]. Taken in combination, these values and characteristics help explain much of public opinion on transgender rights. Explaining Support for Transgender Candidates Figure 2 shows respondents’ vote choice by experimental condition. When both candidates were presented as cisgender—the norm in American politics—more than two-thirds (68 percent) said they would vote for their own party’s candidate, with just 7 percent opting for the other party and 25 percent volunteering that they did not know (a not unreasonable response given the scant information supplied). Figure 2. View largeDownload slide Reported vote choice, given experimental condition. Opposing party’s candidate described as cisgender in both conditions. Estimates based on weighted data. Pure Independents and non-partisans (N = 164) excluded from analysis. N = 359 (cisgender condition); N = 374 (transgender condition). Figure 2. View largeDownload slide Reported vote choice, given experimental condition. Opposing party’s candidate described as cisgender in both conditions. Estimates based on weighted data. Pure Independents and non-partisans (N = 164) excluded from analysis. N = 359 (cisgender condition); N = 374 (transgender condition). When presented with a transgender candidate from their own party and a cisgender candidate from the other party, respondents’ choices differed significantly. Respondents’ support for their own party dropped dramatically, from 68 percent to 37 percent, while support for the opposing party increased to nearly one-fourth (from 7 percent to 23 percent). The number of “don’t know” responses likewise increased substantially, from 25 percent to 40 percent. As predicted by H9, there was a sizeable reluctance to vote for transgender candidates, similar to the opposition toward candidates from marginalized groups documented in earlier research. When voters were presented with a transgender candidate from their own party, they were more likely to desert to the other side or to opt out of answering the question than to support their usual preference. WHAT MODERATES OPPOSITION TO TRANSGENDER CANDIDATES? These conclusions are reinforced by the multinomial regression model in the first column of table 2 (labeled “baseline model”), which shows the increased likelihood of voting for the other party (β = 1.72, SE = .31, p < .001) or giving a Don’t Know/Refused response (β = 1.06, SE = .21, p < .001) when the party’s nominee was transgender. Table 2. Multinomial logit regression models predicting vote choice in candidate experiment, with moderators Moderator used Baseline (n=733) Republican party ID (n=733) Egalitarianism (n=731) Moral traditionalism (n=729) Religiosity (n=715) B (s.e.) B (s.e.) B (s.e.) B (s.e.) B (s.e.) Vote for opposing party  Intercept –2.20 (0.26)*** –1.94 (0.32)*** –1.96 (0.80)* –2.83 (0.59)*** –1.76 (0.74)*  Transgender candidate 1.72 (0.31)*** 1.29 (0.38)*** 1.90 (0.99) 1.00 (0.78) 0.03 (0.91)  Moderator –0.88 (0.55) –0.08 (0.26) 0.27 (0.22) –0.16 (0.24)  Transgender candidate X moderator 1.38 (0.66)* –0.06 (0.32) 0.31 (0.30) 0.59 (0.30) DK/Refused  Intercept –0.99 (0.16)*** –1.24 (0.23)*** –0.53 (0.47) –1.48 (0.45)*** –0.95 (0.50)  Transgender candidate 1.06 (0.21)*** 0.94 (0.30)** 1.37 (0.66) 0.60 (0.64) –0.19 (0.70)  Moderator 0.53 (0.31) –0.16 (0.16) 0.20 (0.18) –0.03 (0.16)  Transgender candidate X moderator 0.46 (0.44) –0.10 (0.22) 0.22 (0.26) 0.45 (0.23)* Moderator used Ideology (n=701) Need for cognitive closure (n=728) Interpersonal contact (n=733) TV use (n=729) Views of transgender people (n=629) B (s.e.) B (s.e.) B (s.e.) B (s.e.) B (s.e.) Vote for opposing party  Intercept –2.39 (0.66)*** –3.06 (0.91)*** –2.23 (0.29)*** –2.53 (0.50)*** –2.20 (0.29)***  Transgender candidate 0.81 (0.82) 2.14 (1.09) 1.88 (0.34)*** 2.13 (0.60)*** 1.66 (0.34)*** Moderator used Ideology (n=701) Need for cognitive closure (n=728) Interpersonal contact (n=733) TV use (n=729) Views of transgender people (n=629) B (s.e.) B (s.e.) B (s.e.) B (s.e.) B (s.e.)  Moderator 0.07 (0.18) 0.31 (0.30) 0.23 (0.66) 0.09 (0.12) –0.10 (0.34)  Transgender candidate X moderator 0.28 (0.24) –0.15 (0.37) –1.46 (0.89) –0.12 (0.15) –0.45 (0.41) DK/Refused  Intercept –1.35 (0.48)* –1.23 (0.57)* –0.93 (0.16)*** –0.72 (0.31)* –1.12 (0.17)***  Transgender candidate 0.56 (0.63) 0.71 (0.80) 1.08 (0.22)*** 1.02 (0.42)* 0.99 (0.24)***  Moderator 0.10 (0.14) 0.09 (0.20) –0.80 (0.62) –0.10 (0.10) 0.11 (0.20)  Transgender candidate X moderator 0.56 (0.63) 0.12 (0.28) 0.21 (0.78) 0.02 (0.13) –0.42 (0.29) Moderator used Baseline (n=733) Republican party ID (n=733) Egalitarianism (n=731) Moral traditionalism (n=729) Religiosity (n=715) B (s.e.) B (s.e.) B (s.e.) B (s.e.) B (s.e.) Vote for opposing party  Intercept –2.20 (0.26)*** –1.94 (0.32)*** –1.96 (0.80)* –2.83 (0.59)*** –1.76 (0.74)*  Transgender candidate 1.72 (0.31)*** 1.29 (0.38)*** 1.90 (0.99) 1.00 (0.78) 0.03 (0.91)  Moderator –0.88 (0.55) –0.08 (0.26) 0.27 (0.22) –0.16 (0.24)  Transgender candidate X moderator 1.38 (0.66)* –0.06 (0.32) 0.31 (0.30) 0.59 (0.30) DK/Refused  Intercept –0.99 (0.16)*** –1.24 (0.23)*** –0.53 (0.47) –1.48 (0.45)*** –0.95 (0.50)  Transgender candidate 1.06 (0.21)*** 0.94 (0.30)** 1.37 (0.66) 0.60 (0.64) –0.19 (0.70)  Moderator 0.53 (0.31) –0.16 (0.16) 0.20 (0.18) –0.03 (0.16)  Transgender candidate X moderator 0.46 (0.44) –0.10 (0.22) 0.22 (0.26) 0.45 (0.23)* Moderator used Ideology (n=701) Need for cognitive closure (n=728) Interpersonal contact (n=733) TV use (n=729) Views of transgender people (n=629) B (s.e.) B (s.e.) B (s.e.) B (s.e.) B (s.e.) Vote for opposing party  Intercept –2.39 (0.66)*** –3.06 (0.91)*** –2.23 (0.29)*** –2.53 (0.50)*** –2.20 (0.29)***  Transgender candidate 0.81 (0.82) 2.14 (1.09) 1.88 (0.34)*** 2.13 (0.60)*** 1.66 (0.34)*** Moderator used Ideology (n=701) Need for cognitive closure (n=728) Interpersonal contact (n=733) TV use (n=729) Views of transgender people (n=629) B (s.e.) B (s.e.) B (s.e.) B (s.e.) B (s.e.)  Moderator 0.07 (0.18) 0.31 (0.30) 0.23 (0.66) 0.09 (0.12) –0.10 (0.34)  Transgender candidate X moderator 0.28 (0.24) –0.15 (0.37) –1.46 (0.89) –0.12 (0.15) –0.45 (0.41) DK/Refused  Intercept –1.35 (0.48)* –1.23 (0.57)* –0.93 (0.16)*** –0.72 (0.31)* –1.12 (0.17)***  Transgender candidate 0.56 (0.63) 0.71 (0.80) 1.08 (0.22)*** 1.02 (0.42)* 0.99 (0.24)***  Moderator 0.10 (0.14) 0.09 (0.20) –0.80 (0.62) –0.10 (0.10) 0.11 (0.20)  Transgender candidate X moderator 0.56 (0.63) 0.12 (0.28) 0.21 (0.78) 0.02 (0.13) –0.42 (0.29) Note.—Reference level is a vote for the respondent’s own party; coefficients show effect on response relative to voting for own party. Estimates based on weighted data. Coefficients are from multinomials logit models with standard errors in parentheses. *p < .05; **p < .01; ***p < .001. View Large Table 2. Multinomial logit regression models predicting vote choice in candidate experiment, with moderators Moderator used Baseline (n=733) Republican party ID (n=733) Egalitarianism (n=731) Moral traditionalism (n=729) Religiosity (n=715) B (s.e.) B (s.e.) B (s.e.) B (s.e.) B (s.e.) Vote for opposing party  Intercept –2.20 (0.26)*** –1.94 (0.32)*** –1.96 (0.80)* –2.83 (0.59)*** –1.76 (0.74)*  Transgender candidate 1.72 (0.31)*** 1.29 (0.38)*** 1.90 (0.99) 1.00 (0.78) 0.03 (0.91)  Moderator –0.88 (0.55) –0.08 (0.26) 0.27 (0.22) –0.16 (0.24)  Transgender candidate X moderator 1.38 (0.66)* –0.06 (0.32) 0.31 (0.30) 0.59 (0.30) DK/Refused  Intercept –0.99 (0.16)*** –1.24 (0.23)*** –0.53 (0.47) –1.48 (0.45)*** –0.95 (0.50)  Transgender candidate 1.06 (0.21)*** 0.94 (0.30)** 1.37 (0.66) 0.60 (0.64) –0.19 (0.70)  Moderator 0.53 (0.31) –0.16 (0.16) 0.20 (0.18) –0.03 (0.16)  Transgender candidate X moderator 0.46 (0.44) –0.10 (0.22) 0.22 (0.26) 0.45 (0.23)* Moderator used Ideology (n=701) Need for cognitive closure (n=728) Interpersonal contact (n=733) TV use (n=729) Views of transgender people (n=629) B (s.e.) B (s.e.) B (s.e.) B (s.e.) B (s.e.) Vote for opposing party  Intercept –2.39 (0.66)*** –3.06 (0.91)*** –2.23 (0.29)*** –2.53 (0.50)*** –2.20 (0.29)***  Transgender candidate 0.81 (0.82) 2.14 (1.09) 1.88 (0.34)*** 2.13 (0.60)*** 1.66 (0.34)*** Moderator used Ideology (n=701) Need for cognitive closure (n=728) Interpersonal contact (n=733) TV use (n=729) Views of transgender people (n=629) B (s.e.) B (s.e.) B (s.e.) B (s.e.) B (s.e.)  Moderator 0.07 (0.18) 0.31 (0.30) 0.23 (0.66) 0.09 (0.12) –0.10 (0.34)  Transgender candidate X moderator 0.28 (0.24) –0.15 (0.37) –1.46 (0.89) –0.12 (0.15) –0.45 (0.41) DK/Refused  Intercept –1.35 (0.48)* –1.23 (0.57)* –0.93 (0.16)*** –0.72 (0.31)* –1.12 (0.17)***  Transgender candidate 0.56 (0.63) 0.71 (0.80) 1.08 (0.22)*** 1.02 (0.42)* 0.99 (0.24)***  Moderator 0.10 (0.14) 0.09 (0.20) –0.80 (0.62) –0.10 (0.10) 0.11 (0.20)  Transgender candidate X moderator 0.56 (0.63) 0.12 (0.28) 0.21 (0.78) 0.02 (0.13) –0.42 (0.29) Moderator used Baseline (n=733) Republican party ID (n=733) Egalitarianism (n=731) Moral traditionalism (n=729) Religiosity (n=715) B (s.e.) B (s.e.) B (s.e.) B (s.e.) B (s.e.) Vote for opposing party  Intercept –2.20 (0.26)*** –1.94 (0.32)*** –1.96 (0.80)* –2.83 (0.59)*** –1.76 (0.74)*  Transgender candidate 1.72 (0.31)*** 1.29 (0.38)*** 1.90 (0.99) 1.00 (0.78) 0.03 (0.91)  Moderator –0.88 (0.55) –0.08 (0.26) 0.27 (0.22) –0.16 (0.24)  Transgender candidate X moderator 1.38 (0.66)* –0.06 (0.32) 0.31 (0.30) 0.59 (0.30) DK/Refused  Intercept –0.99 (0.16)*** –1.24 (0.23)*** –0.53 (0.47) –1.48 (0.45)*** –0.95 (0.50)  Transgender candidate 1.06 (0.21)*** 0.94 (0.30)** 1.37 (0.66) 0.60 (0.64) –0.19 (0.70)  Moderator 0.53 (0.31) –0.16 (0.16) 0.20 (0.18) –0.03 (0.16)  Transgender candidate X moderator 0.46 (0.44) –0.10 (0.22) 0.22 (0.26) 0.45 (0.23)* Moderator used Ideology (n=701) Need for cognitive closure (n=728) Interpersonal contact (n=733) TV use (n=729) Views of transgender people (n=629) B (s.e.) B (s.e.) B (s.e.) B (s.e.) B (s.e.) Vote for opposing party  Intercept –2.39 (0.66)*** –3.06 (0.91)*** –2.23 (0.29)*** –2.53 (0.50)*** –2.20 (0.29)***  Transgender candidate 0.81 (0.82) 2.14 (1.09) 1.88 (0.34)*** 2.13 (0.60)*** 1.66 (0.34)*** Moderator used Ideology (n=701) Need for cognitive closure (n=728) Interpersonal contact (n=733) TV use (n=729) Views of transgender people (n=629) B (s.e.) B (s.e.) B (s.e.) B (s.e.) B (s.e.)  Moderator 0.07 (0.18) 0.31 (0.30) 0.23 (0.66) 0.09 (0.12) –0.10 (0.34)  Transgender candidate X moderator 0.28 (0.24) –0.15 (0.37) –1.46 (0.89) –0.12 (0.15) –0.45 (0.41) DK/Refused  Intercept –1.35 (0.48)* –1.23 (0.57)* –0.93 (0.16)*** –0.72 (0.31)* –1.12 (0.17)***  Transgender candidate 0.56 (0.63) 0.71 (0.80) 1.08 (0.22)*** 1.02 (0.42)* 0.99 (0.24)***  Moderator 0.10 (0.14) 0.09 (0.20) –0.80 (0.62) –0.10 (0.10) 0.11 (0.20)  Transgender candidate X moderator 0.56 (0.63) 0.12 (0.28) 0.21 (0.78) 0.02 (0.13) –0.42 (0.29) Note.—Reference level is a vote for the respondent’s own party; coefficients show effect on response relative to voting for own party. Estimates based on weighted data. Coefficients are from multinomials logit models with standard errors in parentheses. *p < .05; **p < .01; ***p < .001. View Large The other columns of table 2 present results from models that interacted the experimental condition with each of the measures hypothesized to moderate opposition toward a transgender candidate. Partisan differences are shown in the second column of table 2. Figure 3 shows the simulated first differences in responses when the party candidate is transgender (relative to a cisgender candidate), by party identity. Each bullet represents the change in probability of giving each response when the candidate is transgender, with brackets representing 95 percent confidence intervals around the estimates. Figure 3. View largeDownload slide First differences in vote choice given transgender nominee, by party ID. Bullets represent first differences in probability of each response given shift from cisgender to transgender nominee, with 95 percent confidence intervals. Simulated from model in second column of table 2. Figure 3. View largeDownload slide First differences in vote choice given transgender nominee, by party ID. Bullets represent first differences in probability of each response given shift from cisgender to transgender nominee, with 95 percent confidence intervals. Simulated from model in second column of table 2. Negative responses to transgender candidates were common among both Democrats and Republicans. Voters were less likely to support their party when the nominee is transgender, regardless of party (for Democrats, there was a predicted change in probability of support of –.26 [–.37, –.14]; for Republicans, –.38 [–.51, –.26]). While the estimated decline in support among Republicans was larger than that for Democrats, the uncertainty around the estimates precludes us from concluding that the two groups responded differently. Figure 4 presents predicted first differences in the probability of supporting a transgender nominee from one’s own party (relative to a cisgender nominee), across the full range of values of other moderating variables. Figure 4. View largeDownload slide First differences in probability of voting for own party given transgender nominee, by moderators. Ninety-five percent confidence intervals. Simulated from models in table 2. Figure 4. View largeDownload slide First differences in probability of voting for own party given transgender nominee, by moderators. Ninety-five percent confidence intervals. Simulated from models in table 2. Respondents had negative responses to a transgender nominee regardless of their position on most of these measures. Contrary to the earlier hypotheses, those who subscribed to egalitarian beliefs and those who did not; those who held traditional moral values and those who did not; liberals and conservatives; those with a high need for closure and those with a low need for closure; those who watched lots of TV and those who watched none—all were less likely to support a transgender candidate than a cisgender one. In some cases, values do appear to temper opposition to transgender candidates. Respondents who were not at all religious were not significantly less likely than the very religious to support a transgender nominee (a first difference in the probability of support of –.09 [–.29, .11]). The very religious, however, were substantially less likely to do so (–.43 [–.55, –.30]). The difference in responses from the least and most religious was statistically significant at the p < .05 level (.34 [.07, .61]), indicating that religiosity did increase opposition to transgender candidates, consistent with H3. The estimated first difference was also indistinguishable from zero for two other subsets of respondents: those who knew a transgender person (–.22 [–.47, .03]) and those with the most positive views of transgender people (.03 [–.35, .41]) were not significantly less likely to support a transgender nominee than a cisgender one. In neither case, however, are the differences between respondents at each end of the scale statistically significant, and so we do not overemphasize this finding. There was thus some suggestive evidence that the least religious, those who knew a transgender person, and those with positive views of transgender people did not discriminate against transgender candidates. However, the other variables failed to moderate opposition to transgender candidates, as hypothesized. Conclusions The visibility of transgender people in the media and political debate has increased dramatically in recent years. That visibility has not been accompanied by uniformly positive attitudes toward transgender people, rights, and candidates. Our data reveal that a non-trivial proportion of the public has negative views of transgender people, perceiving them as less trustworthy, happy, and moral than cisgender people. While there is majority support for transgender workplace protections, military service, and student protections, the results show a more divided public on the issue of gender-neutral bathrooms, as we might expect given lower support for “body-centric” policies (Miller et al. 2017). And finally, the experiment reveals stronger opposition to transgender candidates than that suggested by self-reports of willingness to support a candidate (Haider-Markel et al. 2017). Beyond these descriptive findings, the results offered support for—and illuminate the limitations of—several broader theories of public opinion. Broad values helped explain opinions about specific groups and issues. In particular, moral traditionalism predicted negative views of transgender people and opposition toward transgender rights. These results echo earlier studies showing moral traditionalism’s relationship with attitudes toward LGB rights (Brewer 2008). The other values studied (egalitarianism, religiosity, ideology, and party) had similarly expected relationships with support for transgender rights (but not with views of transgender people). As hypothesized, more egalitarian, more liberal, less religious, and more Democratic respondents expressed greater support for transgender rights. Likewise, there was support for theories linking need for cognitive closure to political attitudes. Respondents who had a high need for closure were less supportive of transgender rights, indicating the political consequences of seemingly nonpolitical psychological characteristics. In contrast, there was no evidence supporting interpersonal contact theory. Consistent with Flores (2015), respondents who knew a transgender person were not significantly more likely to view transgender people positively nor be more supportive of their rights. Only 11 percent of respondents reported knowing a transgender person, however, making it difficult to discern any true impact of contact. Experimental studies in which exposure to transgender people is manipulated (and hence more prevalent than in probability sampling-based survey research) may offer a greater opportunity to discern contact’s actual effects (see Flores et al. 2018). There was support for parasocial contact theory. Consistent with the notion that positive media portrayals can cultivate more positive attitudes toward marginalized groups, there was a positive association between TV consumption (but not other forms of media exposure) and views of transgender people. Absent a larger-scale content analysis of media portrayals of transgender people and measures of more specific media consumption, we are hesitant to say more here. Still, there is indication that respondents with greater exposure to media portrayals of transgender people viewed them more positively. Indeed, views of transgender people as a group are central to understanding attitudes toward transgender rights. The more positive a person’s views of transgender people, the more they tended to support protecting their rights. Akin to research on support for LGB rights (Brewer 2003), those with more positive views of transgender people were more supportive of policies benefiting them. This latter finding suggests one potential trajectory for public opinion on these issues. Brewer (2003, 2008) shows that increasingly positive perceptions of gay people resulted in an upswing in support for gay rights during the 1990s. This change reflected more positive media portrayals of LGB people (Garretson 2015). Given the media’s increasing (and mostly sympathetic) attention to transgender people, we might expect a similar pattern of change to occur for transgender issues. The finding that greater TV viewing cultivated more positive perceptions of transgender people suggests that such a mechanism may come into play. Throughout, these effects are arguably more limited than expected given the centrality of such variables to public opinion on other issues. Given the relative newness of the issue for many citizens, the relationship between their predispositions and transgender rights may take time to crystallize. However, the emergence of significant relationships in these early data suggests that public opinion on transgender rights already exhibits predictable structure. At the same time, it is possible that social desirability bias may inflate the apparent support for transgender people and rights reported here. The strong opposition to transgender candidates even among those predisposed to support them (i.e., Democrats, egalitarians, and those low in moral traditionalism) suggests potential conflict between public and private opinions on these issues. Although values and psychological traits helped explain attitudes toward transgender people and rights, they were much less useful in explaining attitudes toward transgender candidates. The solid opposition toward transgender politicians uncovered here was in many cases not lessened by variables we would expect to have a moderating effect. There was some suggestion that the least religious, those who knew a transgender person, and those with positive views of transgender people were less negatively predisposed toward a transgender candidate. Perhaps more striking, though, is the fact that opposition did not vary systematically by egalitarianism, moral traditionalism, ideology, NFCC, or media consumption. This suggests deep opposition even among groups that we would expect to be positively predisposed, and the significant obstacles that transgender candidates face. Of course, there are limitations to these conclusions: the experiment excluded those who do not identify with either major party, who may respond differently to transgender candidates than partisans do. Further, voters may react differently to experimental stimuli that include visual images or more detailed candidate information. Certainly, when actual transgender candidates run for office, they come with rich biographical and policy profiles that will shape public opinion as well, making this a particularly fruitful area for future research. It is difficult to speculate about how the public’s attitudes on this newly salient issue might affect future policy directions. On the one hand, the data showed sizeable support for several transgender rights issues, suggesting a more tolerant public than might be expected given recent restrictive laws. On the other hand, the same variables that have polarized political debate in general predicted attitudes on transgender rights too, suggesting that greater elite attention to the issue might lead to more divided responses among the public (Zaller 1992). Given the links between descriptive and substantive representation found elsewhere in the literature (e.g., Haider-Markel 2010), perhaps most striking was the widespread opposition to transgender candidates. These results suggest that any near-term change in policy will likely be the result of sympathetic allies rather than the election of transgender politicians themselves. Supplementary Data Supplementary data are freely available at Public Opinion Quarterly online. The authors thank the editors and reviewers for useful comments and suggestions, and Benjamin E. Bagozzi for extremely helpful guidance regarding the analysis of the candidate experiment. The data for this study were collected by the University of Delaware’s Center for Political Communication, with funding from the William P. Frank Foundation. All errors remain the authors’ own. Appendix: Question wording Need for cognitive closure, egalitarianism, and moral traditionalism. Now, please tell me how strongly you agree or disagree with each of the following statements. The first one is: [INSERT ITEM; RANDOMIZE ITEMS]—do you strongly agree, somewhat agree, somewhat disagree, or strongly disagree with this statement? I don’t like situations that are uncertain. I dislike questions that could be answered in many different ways. I dislike it when a person’s statement could mean many different things. One of the big problems in this country is that we don’t give everyone an equal chance. If people were treated more equally in this country we would have fewer problems. We should be more tolerant of people who choose to live according to their own moral standards even if they are very different from our own. Modern lifestyles are contributing to the breakdown of society. MEDIA CONSUMPTION On an average day, about how many hours do you watch television, including watching media on a computer, tablet, or phone? How often do you watch [INSERT; RANDOMIZE]: regularly, sometimes, hardly ever, or never? National network evening news programs such as ABC World News, CBS Evening News or NBC Nightly News News magazine shows such as 60 Minutes, 20/20, or Dateline The Fox News Cable Channel MSNBC CNN The Late Show with Stephen Colbert on CBS Last Week Tonight with John Oliver Shows on the cable network E! such as Keeping Up with the Kardashians and I Am Cait How often do you read a daily newspaper: regularly, sometimes, hardly ever, or never? Do you use the Internet or email at least occasionally? [IF “YES”] How often do you get news online: regularly, sometimes, hardly ever, or never? Familiarity with term “transgender.” Next, how familiar are you with the term transgender: Do you know what this term means, have you heard of it but are not sure what it means, or have you never heard of the term transgender before? Support for transgender rights Please tell me how strongly you favor or oppose each of the following. The first one is: [INSERT ITEM; RANDOMIZE ITEMS]—do you strongly favor, favor, oppose, or strongly oppose this? Laws that protect transgender people from job discrimination Allowing transgender people to serve openly in the U.S. military Requiring public buildings like courthouses to have gender-neutral restrooms for patrons to use Laws that protect transgender students from discrimination in schools Interpersonal contact Now, thinking about the people you know, do you have a close friend or family member who is transgender? Views of transgender people Next, please tell me how well each of the following describes most people. The first one is: [INSERT ITEM; RANDOMIZE ITEMS]—does this describe most people very well, somewhat well, not too well, or not well at all? Trustworthy Moral Happy Next, please tell me how well each of the following describes transgender people. The first one is: [INSERT ITEM; RANDOMIZE ITEMS IN SAME ORDER AS IN PREVIOUS QUESTION]—does this describe transgender people very well, somewhat well, not too well, or not well at all? Trustworthy Moral Happy Candidate experiment The precise wording of the candidate question depended on the respondent’s own party identification. Respondents identifying as Republicans or leaning toward the Republican Party were asked one of the following eight conditions; Democratic respondents received identical questions but the party affiliations of the candidates were reversed. The order of the candidates was rotated in each condition. Next, we’d like to ask how you would vote in a future election for public office. If the Republicans nominated a man and the Democrats nominated a man, who would you be more likely to vote for? the Republicans nominated a woman and the Democrats nominated a woman, who would you be more likely to vote for? the Republicans nominated a man and the Democrats nominated a woman, who would you be more likely to vote for? the Republicans nominated a woman and the Democrats nominated a man, who would you be more likely to vote for? the Republicans nominated a transgender man and the Democrats nominated a man, who would you be more likely to vote for? the Republicans nominated a transgender woman and the Democrats nominated a man, who would you be more likely to vote for? the Republicans nominated a transgender man and the Democrats nominated a woman, who would you be more likely to vote for? the Republicans nominated a transgender woman and the Democrats nominated a woman, who would you be more likely to vote for? Religiosity How religious do you consider yourself to be—very religious, somewhat religious, not too religious, or not at all religious? Ideology In general, would you describe your political views as very conservative, somewhat conservative, moderate, somewhat liberal, or very liberal? Party identity In politics today, do you consider yourself a Republican, Democrat, or Independent? [IF “DEMOCRAT”/“REPUBLICAN”] Would you call yourself a strong [Democrat/Republican] or not a very strong [Democrat/Republican]? [IF “INDEPENDENT”] As of today, do you lean more to the Republican Party or more to the Democratic Party? Age What is your age? Education What is the highest level of school you have completed or the highest degree you have received? Race Are you, yourself, of Hispanic or Latino origin, such as Mexican, Puerto Rican, Cuban, or some other Spanish background? What is your race? Are you white, black, Asian, or some other race? Gender Would you describe yourself as transgender, or not? Footnotes 1. See Online Appendix Table A5. 2. The gender of both candidates was randomized within conditions. Gender was not associated with vote choice, and so we collapse the conditions here. The Online Appendix includes details of all conditions and presents conjoint analysis showing no effect of the candidates’ gender. References Agiesta , Jennifer . 2016 . “ Poll: 6-in-10 Oppose Bills Like the North Carolina Transgender Bathroom Law .” CNN . Available at http://www.cnn.com/2016/05/09/politics/poll-transgender-bathroom- law-north-carolina/. Allport , Gordon W . 1954 . The Nature of Prejudice . Reading, MA : Addison-Wesley . American Association for Public Opinion Research . 2016 . Standard Definitions: Final Dispositions of Case Codes and Outcome Rates for Surveys , 9th ed . AAPOR . Bond , Bradley J. , and Benjamin L. Compton . 2015 . “ Gay On-Screen: The Relationship Between Exposure to Gay Characters on Television and Heterosexual Audiences’ Endorsement of Gay Equality .” Journal of Broadcasting and Electronic Media 59 : 717 – 32 . Google Scholar CrossRef Search ADS Bramlett , Brittany H . 2012 . “ The Cross-Pressures of Religion and Contact with Gays and Lesbians, and Their Impact on Same-Sex Marriage Opinion .” Politics & Policy 40 : 13 – 42 . Google Scholar CrossRef Search ADS Brewer , Paul R . 2003 . “ The Shifting Foundations of Public Opinion About Gay Rights .” Journal of Politics 65 : 1208 – 20 . Google Scholar CrossRef Search ADS Brewer , Paul R . 2008 . Value War: Public Opinion and the Politics of Gay Rights . Lanham, MD : Rowman and Littlefield . Casey , Logan S. , and Andrew Reynolds . 2015 . Standing Out: Transgender and Gender Variant Candidates and Elected Officials Around the World . LGBTQ Representation and Rights Initiative at the University of North Carolina at Chapel Hill . Cook , Elizabeth A. , Ted G. Jelen , and Clyde C. Wilcox . 1992 . Between Two Absolutes: Public Opinion and the Politics of Abortion . Boulder, CO : Westview . Craig , Stephen C. , Michael D. Martinez , James G. Kane , and Jason Gainous . 2005 . “ Core Values, Value Conflict, and Citizens’ Ambivalence about Gay Rights .” Political Research Quarterly 58 : 5 – 17 . Dasgupta , Nilanjana , and Shaki Asgari . 2004 . “ Seeing Is Believing: Exposure to Counterstereotypic Women Leaders and Its Effect on the Malleability of Automatic Gender Stereotyping .” Journal of Experimental Social Psychology 40 : 642 – 58 . Google Scholar CrossRef Search ADS Feldman , Stanley . 1988 . “ Structure and Consistency in Public Opinion: The Role of Core Beliefs and Values .” American Journal of Political Science 32 : 416 – 40 . Google Scholar CrossRef Search ADS Flores , Andrew R . 2015 . “ Attitudes Toward Transgender Rights: Perceived Knowledge and Secondary Interpersonal Contact .” Politics, Groups, and Identities 3 : 398 – 416 . Google Scholar CrossRef Search ADS Flores , Andrew R. , Donald P. Haider-Markel , Daniel C. Lewis , Patrick R. Miller , Barry L. Tadlock , and Jami K. Taylor . 2018 . “ Challenged Expectations: Mere Exposure Effects on Attitudes About Transgender People and Rights .” Political Psychology 39:197–216. doi: 10.1111/pops.12402 Garner , Andrew . 2013 . “ Ambivalence, the Intergroup Contact Hypothesis, and Attitudes about Gay Rights .” Politics & Policy 41 : 241 – 66 . Google Scholar CrossRef Search ADS Garretson , Jeremiah J . 2015 . “ Exposure to the Lives of Lesbians and Gays and the Origin of Young People’s Greater Support for Gay Rights .” International Journal of Public Opinion Research 27 : 277 – 88 . Google Scholar CrossRef Search ADS Gazzola , Stephanie B. , and Melanie A. Morrison . 2014 . “ Cultural and Personally Endorsed Stereotypes of Transgender Men and Transgender Women: Notable Correspondence or Disjunction ?” International Journal of Transgenderism 15 : 76 – 99 . Google Scholar CrossRef Search ADS Gerbner , George , Larry Gross , Michael Morgan , and Nancy Signorielli . 1984 . “ Political Correlates of Television Viewing .” Public Opinion Quarterly 48 : 283 – 300 . Google Scholar CrossRef Search ADS Golebiowska , Ewa A . 2001 . “ Group Stereotypes and Political Evaluation .” American Politics Research 29 : 535 – 65 . Google Scholar CrossRef Search ADS Green , Donald P. , Bradley Palmquist , and Eric Schickler . 2002 . Partisan Hearts and Minds: Political Parties and the Social Identity of Voters . New Haven, CT : Yale University Press . Haider-Markel , Donald P . 2010 . Out and Running: Gay and Lesbian Candidates, Elections, and Policy Representation . Washington, DC : Georgetown University Press . Haider-Markel , Donald P. , Patrick R. Miller , Andrew R. Flores , Daniel C. Lewis , Barry L. Tadlock , and Jami K. Taylor . 2017 . “ Bringing ‘T’ to the Table: Understanding Individual Support of Transgender Candidates for Public Office .” Politics, Groups, and Identities 5 : 399 – 417 . Google Scholar CrossRef Search ADS Herrick , Rebekah , and Sue Thomas . 1999 . “ The Effects of Sexual Orientation on Citizen Perceptions of Candidate Viability .” In Gays and Lesbians in the Democratic Process , edited by Ellen D. B. Riggle and Barry L. Tadlock , pp. 170 – 91 . New York : Columbia University Press . Human Rights Campaign . 2016 . “Anti-Transgender Legislation Spread Nationwide, Bills Targeting Transgender Children Surge.” Available at http://hrc-assets.s3-website-us-east-1.amazonaws.com//files/assets/resources/HRC-Anti-Trans-Issue-Brief-FINAL-REV2.pdf. Jost , John T. , Jack Glaser , Arie W. Kruglanski , and Frank J. Sulloway . 2003 . “ Political Conservatism as Motivated Social Cognition .” Psychological Bulletin 129 : 339 – 75 . Google Scholar CrossRef Search ADS PubMed King , Gary , Christopher J. L. Murray , Joshua A. Salomon , and Ajay Tandon . 2004 . “ Enhancing the Validity and Cross-Cultural Comparability of Measurement in Survey Research .” American Political Science Review 98 : 191 – 207 . Google Scholar CrossRef Search ADS Kruglanski , Arie W. , and Donna M. Webster . 1991 . “ Group Members’ Reactions to Opinion Deviates and Conformists at Varying Degrees of Proximity to Decision Deadline and of Environmental Noise .” Journal of Personality and Social Psychology 61 : 212 – 25 . Google Scholar CrossRef Search ADS PubMed Lewis , Gregory B . 2011 . “ The Friends and Family Plan: Contact with Gays and Support for Gay Rights .” Policy Studies Journal 39 : 217 – 38 . Google Scholar CrossRef Search ADS Lewis , Gregory B. , and Marc A. Rogers . 1999 . “ Does the Public Support Equal Employment Rights for Gays and Lesbians ?” In Gays and Lesbians in the Democratic Process , edited by Ellen D. B. Riggle and Barry L. Tadlock , pp. 118 – 45 . New York : Columbia University Press . Mastro , Dana , Elizabeth Behm-Morawitz , and Michelle Ortiz . 2007 . “ The Cultivation of Social Perceptions of Latinos: A Mental Models Approach .” Media Psychology 9 : 347 – 65 . Google Scholar CrossRef Search ADS McCann , James A . 1997 . “ Electoral Choices and Core Value Change: The 1992 Presidential Campaign .” American Journal of Political Science 4 : 564 – 83 . Google Scholar CrossRef Search ADS Miller , Patrick R. , Andrew R. Flores , Donald P. Haider-Markel , Daniel C. Lewis , Barry L. Tadlock , and Jami K. Taylor . 2017 . “ Transgender Politics as Body Politics: Effects of Disgust Sensitivity and Authoritarianism on Transgender Rights Attitudes .” Politics, Groups, and Identities 5 : 4 – 24 . Google Scholar CrossRef Search ADS Newport , Frank . 2016 . “ Disentangling Attitudes Toward Transgender Bathroom Use .” Gallup , May 20. Available at http://www.gallup.com/opinion/polling-matters/191774/disentangling-attitudes-toward-transgender-bathroom.aspx. Norton , Aaron T. , and Gregory M. Herek . 2013 . “ Heterosexuals’ Attitudes Toward Transgender People: Findings from a National Probability Sample of U.S. Adults .” Sex Roles 68 : 738 – 53 . Google Scholar CrossRef Search ADS Olson , Laura R. , Wendy Cadge , and James T. Harrison . 2006 . “ Religion and Public Opinion About Same-Sex Marriage .” Social Science Quarterly 87 : 340 – 60 . Google Scholar CrossRef Search ADS Pettigrew , Thomas F. , and Linda R. Tropp . 2006 . “ A Meta-Analytic Test of Intergroup Contact Theory .” Journal of Personality and Social Psychology 90 : 751 – 83 . Google Scholar CrossRef Search ADS PubMed Pew Research Center . 2012 . “ Assessing the Representativeness of Public Opinion Surveys .” Pew Research Center , May 15. Available at http://www.people-press.org/2012/05/15/assessing-the-representativeness-of-public-opinion-surveys/. Roets , Arne , and Alain Van Hiel . 2011 . “ Item Selection and Validation of a Brief, 15-Item Version of the Need for Closure Scale .” Personality and Individual Differences 50 : 90 –9 4 . Google Scholar CrossRef Search ADS Rössler , Patrick , and Hans-Bernd Brosius . 2001 . “ Do Talk Shows Cultivate Adolescents’ Views of the World? A Prolonged‐Exposure Experiment .” Journal of Communication 51 : 143 – 63 . Google Scholar CrossRef Search ADS Schiappa , Edward , Peter B. Gregg , and Dean E. Hewes . 2006 . “ Can One TV Show Make a Difference? Will & Grace and the Parasocial Contact Hypothesis .” Journal of Homosexuality 51 : 15 – 37 . Google Scholar CrossRef Search ADS PubMed Stoutenborough , James W. , Donald P. Haider-Markel , and Mahalley D. Allen . 2006 . “ Reassessing the Impact of Supreme Court Decisions on Public Opinion: Gay Civil Rights Cases .” Political Research Quarterly 59 : 419 – 33 . Google Scholar CrossRef Search ADS University of Illinois Springfield Survey Research Office . 2015 . Examining Attitudes towards the “T” in LGBT: Public Support for Transgender Rights and Supportive Policies . Available at https://transgenderinclusivepolicysurvey.wordpress.com. Webster , Donna M. , and Arie W. Kruglanski . 1994 . “ Individual Differences in Need for Cognitive Closure .” Journal of Personality and Social Psychology 67 : 1049 – 62 . Google Scholar CrossRef Search ADS PubMed Wilcox , Clyde , Paul R. Brewer , Shauna Shames , and Celinda Lake . 2007 . “ If I Bend This Far I Will Break? Public Opinion about Same-Sex Marriage .” In The Politics of Same-Sex Marriage , edited by Craig A. Rimmerman and Clyde Wilcox , pp. 215 – 242 . Chicago : University of Chicago Press . Zaller , John R . 1992 . The Nature and Origins of Mass Opinion . Cambridge : Cambridge University Press . Google Scholar CrossRef Search ADS © The Author(s) 2018. Published by Oxford University Press on behalf of the American Association for Public Opinion Research. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Public Opinion QuarterlyOxford University Press

Published: May 15, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off