LEGISLATING GENDER: DOES ‘X’ MARK THE SPOT FOR IMPROVING HEALTH AMONG GENDER-NONCONFORMING ADULTS? By Samuel Safford A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Sociology—Master of Arts 2023 ABSTRACT This paper uses data from the Behavioral Risk Factor Surveillance Systems (BRFSS) survey over the period 2014-2021 to analyze the impact of new state-level policies allowing individuals to identify as gender-neutral (‘X’ gender marker) on their government-issued driver’s license. In many states, this “third” gender category is not an option, and transgender and gender- nonconforming (TGNC) individuals face multiple barriers, including legislative restrictions, when seeking to adjust their gender identity and names on legal documents. A counterfactual framework exploits variation across states and years in enactment of gender-neutral driver’s license laws to show that a gender-neutral ‘X’ marker shows promise for attenuating inequality in overall physical health among gender-nonconforming adults. These findings highlight the potential benefit to physical health that allowing a gender-neutral option on legal documents might have for gender diverse individuals. This thesis is dedicated to my mom and dad, without whose support I would not be here. iii TABLE OF CONTENTS LIST OF TABLES ..........................................................................................................................v LIST OF FIGURES .......................................................................................................................vi INTRODUCTION ..........................................................................................................................1 BACKGROUND ............................................................................................................................3 Transgender and Gender-Nonconforming Health ..............................................................3 Gender Affirmation and Gender-affirming Policies ...........................................................4 Names, Gender Markers, and Identity Documents .............................................................5 METHODS .....................................................................................................................................9 Behavioral Risk Factor Surveillance System Data .............................................................9 State Level Characteristics and Gender-affirming Laws and Policies ..............................10 Regression Analysis ..........................................................................................................12 RESULTS .....................................................................................................................................14 Summary Statistics ............................................................................................................14 Regression Results ............................................................................................................21 DISCUSSION ...............................................................................................................................30 Limitations ........................................................................................................................31 CONCLUSION .............................................................................................................................33 REFERENCES .............................................................................................................................34 iv LIST OF TABLES Table 1: Descriptive overview (percentages) by gender identity .................................................15 Table 2: Descriptive overview (percentages) for gender-nonconforming adults by presence of GNID law ......................................................................................................................................18 Table 3: Estimated marginal effects from logistic regression predicting poor self-reported health for cisgender men and women (n = 1,532,206) ............................................................................22 Table 4: Estimated marginal effects from logistic regression predicting poor self-reported health for transmasculine and transfeminine adults (n = 5,115) ..............................................................24 Table 5: Estimated marginal effects from logistic regression predicting poor self-reported health for gender-nonconforming adults (n = 1,602) ..............................................................................26 v LIST OF FIGURES Figure 1: (Model 2) Predicted probabilities of reporting poor health by presence of GNID laws ...............................................................................................................................................28 Figure 2: (Model 3) Predicted probabilities of reporting poor health by presence of GNID laws ...............................................................................................................................................29 vi INTRODUCTION Despite only approximately half of a percent of the adult population in the United States identifying as transgender (Williams Institute 2022), four hundred and ninety anti-trans bills were introduced across the United States in the first three months of 2023 alone – three hundred and seventeen more than in all of 2022, a two hundred and fifty-seven percent increase (Trans Legislation Tracker 2023) – many of them seeking to ban gender-affirming care for transgender youth and adults under the age of twenty-five, thereby forcibly detransitioning them. Meanwhile, less than half of all states have any sort of protections for gender minority individuals on the books (Movement Advancement Project 2023a). In a time when transgender people are over four times more likely than cisgender people to be victims of violent crime (Flores at al. 2021), one in four black trans youth have reported a suicide attempt in the past year (The Trevor Project 2023), and trans and gender-nonconforming (TGNC) adults have significantly higher rates of poor physical health (Lagos 2018; Scheim et al. 2022) and mental distress (Cicero et al. 2020; Feldman et al. 2021), it is imperative that researchers, activists, and policymakers call for expanded access to gender-affirming care and policies. Gender-affirming healthcare and protective policies have proven positive implications for physical and mental health and healthcare access for TGNC youth and adults (Restar et al. 2020; King and Gamarel 2021; Gonzales et al. 2022; Green et al. 2022; Scheim et al. 2022; Tordoff et al. 2022). In order to add to this rapidly growing literature, I seek to examine whether policies that allow individuals to change their legal sex/gender marker to ‘X’ – so called gender-neutral ID laws (GNID) – have similar protective effects to policies that allow TGNC individuals to change their legal sex and name on identity documents (Crosby, Salazar, and Hill 2016; Le et al. 2016; Kidd, Dolezal, and Bockting 2018). In this study, I analyze data from the 2014-2021 1 Behavioral Risk Factor Surveillance System (BRFSS) and the Movement Advancement Project (MAP) to explore the potential for GNID laws to have a positive impact on poor self-rated physical health for gender-nonconforming (GNC) adults. The findings of this study have important public health implications for the promise of GNID laws and other similar policies that seek to be gender-affirming and protective for TGNC individuals. 2 BACKGROUND Transgender and Gender-Nonconforming Health Research has consistently shown that transgender and gender-nonconforming (TGNC) people have worse self-reported mental health (Burgwal et al. 2019; Crissman et al. 2019; Residner and Hughto 2019; Rimes et al. 2019; Cicero et al. 2020; Feldman et al. 2021) and physical health (Lagos 2018; Feldman et al. 2021; Scheim et al. 2022) than the cisgender population. This includes higher rates of disability (Smith-Johnson 2022) including HIV, heart disease, and emphysema (Feldman et al. 2021), greater frequency of mental distress (Crissman et al. 2019) including anxiety and depression (Lefevor et al. 2019), and self-harm, suicidality, and substance abuse (Lefevor et al. 2019; Rimes et al. 2019). GNC adults are also at higher risk of experiencing harassment, sexual abuse, and other traumatic events than cisgender and transmasculine and transfeminine people (Lefevor et al. 2019). Among the TGNC population, gender- nonconforming adults have significantly higher odds of reporting poor self-rated physical health (Lagos 2018; Burgwal et al. 2019) and mental health (Crissman et al. 2019; Rimes et al. 2019) than their transmasculine and transfeminine peers. TGNC people also face significant barriers in accessing healthcare (Residner and Hughto 2019; Cicero et al. 2020; Scheim et al. 2022), including gender-affirming treatment (Feldman et al. 2021). TGNC people are often hesitant to seek care, in part due to structural obstacles (Goldenberg et al. 2020) including lack of health insurance (Scheim et al. 2022) and negative cultural norms that foster bias among healthcare providers (Hsieh and shuster 2021; Loza et al. 2021). Academics have theorized that these disparities in health and access to healthcare can be explained in part by stigmatization (Link and Phelan 2001; Frost, Lehavot, and Meyer 2015; Stacey, Reczek, and Spiker 2022) and minority stress processes (Meyer 1995). Link and Phelan 3 (2001) conceptualize stigma as the “co-occurrence of its components – labeling, stereotyping, separation, status loss, and discrimination” (p. 363), and they note that stigmatization on one domain of people’s lives likely has dramatic bearing on additional domains such as housing, earnings, and health, to name but a few. Frost, Lehavot, and Meyer (2015) further note that prejudice-related stressors have unique negative consequences on health that persist beyond stressors unrelated to prejudice. Stigma against transgender people significantly limits access to vital resources and opportunities (Hughto, Reisner, and Pachankis 2015), and is additionally responsible for poor access to legal and medical resources (Lefevor et al. 2019). This stigma manifests in a variety of ways, including misgendering (McLemore 2018), and experiences of transphobia, including anti-trans violence (Lombardi 2009). Gender Affirmation and Gender-affirming Policies Stigma prevention coping interventions, including gender-affirming practices, policies, and care, can often help attenuate these stressors (Hughto, Residner, and Pachankis 2015). Research has shown that gender affirmation reduces disparities in healthcare access (Goldenberg et al. 2020; Scheim et al. 2022) and mental health among TGNC adults (Dorsen et al. 2022), including a reduction in suicidal ideation and psychological stress (Lelutiu-Weinberger, English, and Sandanapitchai 2020). Affirmation has been described by TNGC people as “being respected as a whole person” (Dorsen et al. 2022, p. 36), and can be as simple as using a TGNC individual’s preferred language, including chosen name and pronouns. One’s gender identity being respected and recognized may be considered a display of resilience, and is paramount in upholding transgender health (Lelutiu-Weinberger, English, and Sandanapitchai 2020). Medical gender affirmation, including gender-affirming care and surgeries, has consistently been shown to enrich the lives of TGNC individuals, in particular through marked 4 increases in mental health (Green et al. 2022; Tordoff et al. 2022; Lee and Rosenthal 2023; Swan et al. 2023), decreases in suicidality (Green et al. 2022; Tordoff et al. 2022), and increased feelings of internal validation (Dorsen et al. 2022). In addition to medical affirmation, social and legal gender affirmation have been linked to positive changes in mental health outcomes and healthcare utilization (King and Gamarel 2021; Scheim et al. 2022). Social channels can include incorporating inclusive language in the workplace (Perales, Ablaza, and Elkin 2022), external affirmation from friends, colleagues, and family (King and Gamarel 2021), and changing one’s name to one that better reflects one’s gender (Pollitt et al. 2021). Legal affirmation is primarily obtained through legal name changes (Hill et al. 2018; Restar et al. 2020; Loza et al. 2021) and gender-concordant gender markers on identity documents (Restar et al. 2020; Scheim, Perez- Brumer, and Bauer 2020; DeChants et al. 2022; Tan et al. 2022; Yee, Lind, and Downing 2022). Names, Gender Markers, and Identity Documents While research has shown that access to and utilization of gender-affirming medical care (Green et al. 2022; Tordoff et al. 2022; Lee and Rosenthal 2023; Swan et al. 2023) and non- discrimination policies and expanded access to gender-affirming medical care (Goldenberg et al. 2020; Gonzales, Tran, and Bennett 2022; Scheim et al. 2022l; Trusczczynski et al. 2022) lead to greatly increased physical and mental health for transgender individuals, recent studies have also highlighted the importance of gender-concordant documents and names. Goetz and Arcomano (2021), through interviews with trans individuals in the U.S. and Canada, found that almost all TGNC people interviewed were interested in legal name changes and updating their legal gender markers. Individuals who want to update their legal documents to be gender-concordant, but are unable to, have greater odds of attempting suicide in the last year (DeChants et al. 2022). TGNC 5 people who face barriers to obtaining gender-concordant IDs have higher levels of mental health problems (Tan et al. 2022) and higher rates of postponing medical care (Hill et al. 2018) than those with gender-concordant IDs. Legal name change has been shown to be an important structural intervention for gender diverse individuals, leading to increased socioeconomic stability (Hill et al. 2018), lower odds of harassment in public settings, and fewer housing-related issues due to gender identity (Loza et al. 2021). Perhaps most importantly, possession of gender- concordant IDs has been shown to lead to improved access to primary and transition-related health care (Hill et al. 2018) and being treated with more respect by doctors and other health care providers (Loza et al. 2021). In addition to access to legal name changes and gender-concordant gender markers, some GNC individuals have expressed an interest in ‘X’ gender markers for their IDs (Goetz and Arcomano). Support for this is not unanimous among GNC people (Goetz and Arcomano 2021) and activists (Saguy 2023), in part due to a fear of a lack of compatibility with other identity documents. However, while the use of identification documents as biopolitical documents is problematic, the ‘X’ gender marker can be used as a tool of resistance and radical self- determination for gender diverse individuals (Quinan and Oosthoek 2021). Despite concerns, nearly half of all states in America have begun allowing residents to change their gender markers to ‘X’ on their driver’s licenses and state ID cards (Movement Advancement Project 2023), and the Biden administration announced in April 2022 that people will be allowed to indicate their gender as ‘X’ on U.S. Passports (Block 2023). While previous researchers have studied the health benefits of gender-affirming policies for all transgender adults, showing promising attenuating effects of gender-concordant markers and preferred name on a variety of legal documents, this is the first study to consider the 6 implications for physical health of policies that allow gender-nonconforming adults to choose a gender-neutral ‘X’ marker on their IDs, an important contribution to the discourse on how to alleviate negative health consequences among this marginalized population. Using the Behavioral Risk Factor Surveillance System (BRFSS), this study advances the literature on the efficacy of potentially gender-affirming policies in helping diminish disparities in physical health among TGNC individuals. Keeping the patterns of inequality that transgender and GNC adults face in mind, and the potential impact of gender-affirming policies, I formulate three hypotheses. It stands to reason that the majority of the U.S. adult population will not seek to change the sex/gender marker on their licenses to ‘X’ unless it is a more accurate representation of their sex and/or gender identity, which can be tested through my first hypothesis. Hypothesis 1: Gender-neutral ID laws will have no statistically significant impact on the probability of reporting fair to poor health for the cisgender population. While it is uncertain whether transmasculine and transfeminine individuals are significantly more likely to identify with and prefer the male and female labels on identity documents, states that have begun to offer gender-neutral ‘X’ markers have noted that gender- nonconforming adults are the target audience for said policies, or were likely to be the primary group to take up this option. For example, California has labeled the gender-neutral gender marker as ‘nonbinary’ (California Courts Self-Help Guide 2023), and Washington has defined ‘X’ as “neither exclusively male or female” (Ingersoll Gender Center 2019). Due to this, and to the relatively small size of the gender-nonconforming population, it follows that there should be no significant effect of these policies on poor self-reported physical health for any group of 7 adults outside of those who are likely to take advantage of them, including other transgender adults experiencing health inequality. The logic for my second hypothesis follows the first. Hypothesis 2: Gender-neutral ID laws will have no statistically significant impact on the probability of reporting fair to poor health for the transmasculine and transfeminine population. Finally, to evaluate the efficacy of gender-neutral ID laws as channels of legal affirmation for GNC adults, I formulate my third hypotheses. This hypothesis serves to evaluate whether there are any detectable differences in self-reported poor physical health in the presence of GNID laws for GNC adults. Hypothesis 3: Gender-nonconforming adults in states with gender-neutral ID laws will have a lower probability of reporting poor health than gender-nonconforming individuals in states without said policies. 8 METHODS Behavioral Risk Factor Surveillance System Data To explore the potential benefit of gender-neutral ID laws, I utilize data from Behavioral Risk Factor Surveillance System (BRFSS) surveys, which are administered by the Centers for Disease Control and Prevention (CDC). The BRFSS is a nationally representative health survey of noninstitutionalized U.S. adults, conducted continuously throughout the year by each U.S. state’s public health department using household-based probability sampling and random digit dialing of both cell phones and landlines (Centers for Disease Control 2013). This paper utilizes data from the 2014-2021 releases of the survey, which include a small sample of respondents in 2022. Beginning in 2014, the BRFSS included an optional sexual orientation and gender identity module (SOGI) that allows respondents to self-identify as transgender. If respondents indicate that they do identify so, they may select between three options: male-to-female transgender (transfeminine), female-to-male transgender (transmasculine), or transgender, gender-nonconforming (GNC). Across the seven and a half years of available data since its implementation, over 1.6 million respondents have submitted answers to this question. A categorical measure of gender identity was constructed from their answers, with 0 = a respondent identifies as cisgender male (n = 679,403), 1 = cisgender female (n = 855,177), 2 = transmasculine (n = 2,463), 3 = transfeminine (n = 2,664) and 4 = GNC (n = 1,613), for a total of 1,541,320 individuals. Individuals who answered ‘don’t know/not sure’ or who refused to answer are discarded from the sample, but could potentially be imputed in future studies (Lagos 2018). From the BRFSS, I construct one dependent variable. This variable of interest, fair to poor health, is a dichotomous outcome variable constructed based upon a respondent’s self-rated 9 physical health over the past 30 days. Answers classified as “poor” and “fair” are combined as one, with the reference of good health consisting of individuals who answered “good,” “very good,” and “excellent.” This construction of an indicator for poor health has shown to be a robust predictor of mortality in populations irrespective of socioeconomic status (DeSalvo et al. 2006; Frankenberg and Jones 2004; Gorman and Sivaganesan 2007; Lagos 2018). Sociodemographic controls include age, racial and ethnic identity (non-Hispanic white, non-Hispanic Black, Hispanic of any race, and all other non-Hispanic respondents of other races), level of educational attainment (did not graduate high school, high school diploma/GED, some college, and college degree), employment status, and household income equal to or greater than $50,000 per year. To account for other household factors, indicators for whether a respondent has a child or is married or partnered is included, as partnerships have been shown to be good measures of social support (Rendall et al. 2011; Santini et al. 2015; Hult-Lundstad 2018), and parenthood has been shown to have mixed costs and benefits on an individual’s health (Carson, Adamo, and Rhodes 2018; Simon and Caputo 2019; Metzger and Gracia 2023). Finally, a categorical measure of whether the respondent is a current or former user or does not currently use tobacco cigarettes is included, as this has been shown in previous literature to be an important contributing factor in poor health (Case and Paxson 2005; Preston and Wang 2006). State Level Characteristics and Gender-affirming Laws and Policies The policy variable of interest, gender-neutral ID law, is a dichotomous measure created using internet searches of gender-neutral driver’s license policies and data from the Movement Advancement Project (2023). Between 2015 and 2023, 23 states have begun allowing their residents to choose between M, F, or X on their driver’s licenses, beginning with Oregon. Of the 23 states with such a policy, 14 states contributed data on transgender residents to the BRFSS 10 between 2014 and 2021, including Arkansas, California, Colorado, Connecticut, Hawaii, Massachusetts, Minnesota, New Jersey, New Mexico, Nevada, Pennsylvania, Rhode Island, Utah, Virginia, Vermont, and Washington. While not every state’s policy is identical in how one might obtain the new gender marker (e.g., it is free in California, but in Michigan there is an additional processing fee), their implementation is similar. Specifically, each policy allows an individual to change the sex or gender marker on their driver’s license from male (M) or female (F) to gender-neutral (X). To date, not every state that allows individuals to obtain this new marker on their driver’s license also allows the ‘X’ option for other legal documents. Future studies should examine the potential benefit of these more expansive policies, or at least consider them as control variables. Since the political climate of each state might have some influence on individual health among its inhabitants, I construct two additional measures, including a dichotomous measure of whether the current state governor is a Democrat, and one categorical measure of whether Democrats hold a majority in none, one, or both state legislative houses. Furthermore, to account for the influence of other gender-affirming policies on health, I additionally construct three categorical measures of strength of policy, including negative, neutral, and positive, following the Movement Advancement Project’s grading scale. The three policies include how easy it is to change one’s legal gender on their driver’s license and birth certificate, as well as whether a state requires the publication of a name change announcement (Movement Advancement Project 2023). For driver’s licenses, states that require proof of surgery, court order, or amended birth certificates are classified as negative, states with unclear or unknown written policies, no standard procedure, or that require provider certification are classified as neutral, and states with easy-to-understand forms and no provider certification are considered positive. For birth 11 certificates, states that do not allow for one to amend their gender marker or require proof of ‘sex reassignment’ surgery are classified as negative, states with unclear, unknown, or unwritten policies or that may require a court order to change gender markers are classified as neutral, and those that issue new certificates with no surgery or court order requirements are classified as positive. Name change policies are considered negative if states require individuals to publish a name change announcement, neutral if they have unclear policies but allow individual court discretion, and positive if there are no publication requirements. Regression Analysis Marginal effects were obtained using binary logistic regression models to examine how disparities in self-rated physical health differed for adults in states with gender-neutral ID laws and those without. For all samples, three models were run. First, the base model assessed health disparities by gender-neutral ID laws, gender identity, and sociodemographic characteristics including sexuality, age, race/ethnicity, educational attainment, and survey year. For the second model, additional socioeconomic controls including employment status, household income, household composition, whether the individual was insured, and smoking history are also included. Finally, in order to account for the impact of other gender-affirming policies that may exist in states that have gender-neutral ID laws, I also control for three additional types of laws that determine how easy it is to 1) change one’s legal name, 2) alter one’s gender marker from ‘M’ to ‘F’ and vice versa on legal documents, and 3) change one’s legal name and/or one’s gender marker on a birth certificate. I estimate all three models using a split-sample method separately for cisgender men and women, transmasculine and transfeminine people, and finally for gender-nonconforming adults to test my hypotheses. While using a simple interaction term of gender-neutral ID laws and 12 gender identity would allow me to explore these questions, split-sampling allows me to examine heterogeneity in the relationship between predictors and health outcomes, which are likely to be very complex and may vary dramatically across each different group of adults. The evidence for this last assumption is based primarily on the significant differences in sociodemographic factors across the board between cisgender men and women, transmasculine and transfeminine adults, and GNC adults as seen in Table 1, including LGBQ identity, race, level of education, household income, and partnership status. This could also be examined by fully interacting all independent variables with indicators for GNID laws and GNC identity, which yields qualitatively similar results, and is available upon request. All results were obtained in Stata version 16 using the svy function to adjust for stratified and clustered sampling and to cluster standard errors at the state level. Survey weights were not utilized due to potential issues with the BRFSS-provided raked weights (Todoroff and Hajat 2019; Cicero et al. 2020; Lett and Everhart 2022), and thus results are not representative nationally, as this approach treats the BRFSS as an unweighted cluster stratified random sample of the United States (Lett and Everhart 2022). To account for time- invariant differences in years and states, fixed effects for both are included as well. 13 RESULTS Summary Statistics Summary statistics for cisgender men, cisgender women, and transmasculine, transfeminine, and GNC individuals are provided in Table 1. Of all respondents in the sample, only 1,613 individuals identified as GNC, representing roughly 0.1 percent of the sample population. In terms of self-rated physical health, gender-nonconforming individuals are significantly more likely than all other groups to have reported fair to poor health in the past month. 27.46% of GNC adults report poor health, compared to 17.05% of cisgender men, 18.23% of cisgender women, 24.00% of transmasculine people, and 24.59% of transfeminine people. This overview shows that GNC adults are significantly more likely to report fair to poor health than their cisgender peers, and for some outcomes, their transgender peers. Additional potential risk factors are also more prevalent among GNC adults, including a much larger proportion of the individuals in the sample reporting some form of LGBQ+ identity, which has been commonly documented to contribute to worse health outcomes (King et al. 2008; Ayhan et al. 2020; Rees, Crowe, and Harris 2020). GNC adults in the sample were also more likely to be younger, less likely to be insured, and more likely to have a lower household income than their cisgender peers, all potential contributing factors for their worse reported health outcomes. Table 2 presents summary statistics for GNC individuals, separated by whether they reside in a state with a GNID law at the time they were surveyed. GNC adults in states with gender-neutral ID laws are significantly less likely to report poor health, as well as more likely to identify as LGBQ, are younger on average, are more likely to be college graduates and employed, and earn significantly more than their GNC peers in states without GNID laws. These disparities across GNC people separated only by a difference in one policy are stark, and likely 14 Table 1: Descriptive overview (percentages) by gender identity Poor self-rated health Identity document laws Gender neutral ID law Driver's license gender markers Negative law Neutral law Positive law Birth certificate gender markers Negative law Neutral law Positive law Name change reqs. Negative law Cisgender Men (n = 679,403) Cisgender Women (n = 855,177) Transmasculine (n = 2,463) Transfeminine (n = 2,664) Transgender, GNC (n = 1,613) 17.05 [16.96, 17.14] 18.23 [18.15, 18.31] 24.00 [22.31, 25.68] 24.59 [22.95, 26.22] 27.46 [25.28, 29.64] 14.32 [14.23, 14.41] 12.92 [12.85, 12.99] 14.70 [13.30, 16.10] 12.24 [10.99, 13.48] 21.02 [19.03, 23.01] 18.18 [18.09, 18.27] 8.76 [8.69, 8.83] 19.51 [19.42, 19.59] 28.70 [26.92, 30.49] 8.26 [8.20, 8.32] 8.12 [7.04, 9.20] 17.57 [16.12, 19.01] 7.02 [6.05, 7.99] 73.06 [72.95, 73.16] 72.23 [72.14, 72.33] 63.17 [61.27, 65.08] 75.41 [73.78, 77.05] 16.49 [14.68, 18.30] 10.29 [8.81, 11.78] 73.22 [71.05, 75.38] 29.81 [29.70, 29.92] 12.71 [12.64, 12.79] 57.48 [57.37, 57.60] 30.82 [30.72, 30.92] 13.49 [13.42, 13.57] 55.69 [55.58, 55.79] 36.26 [34.36, 38.16] 13.44 [12.09, 14.79] 50.30 [48.33, 52.28] 30.11 [28.36, 31.85] 11.75 [10.53, 12.98] 58.15 [56.27, 60.02] 24.43 [22.33, 26.53] 13.21 [11.55, 14.86] 62.37 [60.00, 64.74] 16.32 [16.23, 16.40] 16.59 [16.51, 16.67] 13.93 [12.56, 15.29] 16.55 [15.14, 17.97] 19.47 [17.53, 21.40] 15 Table 1 (cont'd) Neutral law Positive law Sociodemographics LGBQ+ Race/ethnicity White, non-Hispanic Black, non-Hispanic Hispanic Other race, non-Hispanic Age in years Education No or some high school High school graduate/GED Some college College graduate 52.50 [52.38, 52.62] 31.19 [31.08, 31.30] 51.45 [51.35, 51.56] 31.95 [31.85, 32.05] 47.30 [45.33, 49.27] 38.77 [36.84, 40.70] 53.38 [51.48, 55.27] 30.07 [28.33, 31.81] 48.98 [46.53, 51.42] 31.56 [29.29, 33.83] 4.24 [4.20, 4.29] 4.48 [4.44, 4.52] 27.08 [25.32, 28.84] 35.70 [33.88, 37.52] 58.03 [55.62, 60.44] 78.83 [78.73, 78.93] 78.58 [78.50, 78.67] 70.77 [68.97, 72.56] 68.73 [66.97, 70.49] 68.13 [65.86, 70.41] 6.47 [6.41, 6.53] 6.78 [6.72, 6.84] 7.92 [7.86, 7.98] 8.35 [8.29, 8.41] 6.40 [6.34, 6.45] 6.67 [6.62, 6.72] 9.66 [8.50, 10.83] 9.62 [8.46, 10.79] 9.95 [8.76, 11.13] 54.29 [54.25, 54.33] 56.56 [56.53, 56.60] 49.81 [49.04, 50.58] 6.50 [6.44, 6.55] 27.51 [27.40, 27.62] 26.24 [26.13, 26.34] 39.76 [39.64, 39.87] 5.99 [5.94, 6.04] 26.18 [26.09, 26.28] 29.14 [29.05, 29.24] 38.68 [38.57, 38.78] 11.26 [10.01, 12.51] 34.23 [32.35, 36.10] 28.58 [26.79, 30.36] 25.93 [24.20, 27.67] 9.91 [8.77, 11.05] 10.10 [8.95, 11.24] 11.26 [10.06, 12.46] 51.17 [50.48, 51.86] 14.46 [13.12, 15.80] 36.12 [34.29, 37.95] 26.51 [24.84, 28.20] 22.90 [21.30, 24.50] 7.50 [6.21, 8.79] 11.59 [9.62, 12.70] 13.21 [11.55, 14.86] 43.00 [42.03, 44.97] 10.01 [8.54, 11.48] 30.85 [28.59, 33.11] 27.86 [25.67, 30.05] 31.28 [29.01, 33.55] 16 Table 1 (cont'd) Currently employed Household income Less than $25,000 $25,000 to $49,000 $50,000 or more Household composition Any children Currently married/partnered Currently insured Smoking status Never smoker Former smoker Current smoker 57.41 [57.29, 57.53] 45.45 [45.35, 45.56] 45.92 [43.95, 47.89] 47.94 [46.04, 49.83] 51.02 [48.58, 53.47] 28.47 [28.37, 28.58] 20.76 [20.66, 20.86] 50.77 [50.65, 50.88] 24.99 [24.89, 25.09] 60.64 [60.53, 60.76] 91.77 [91.70, 91.84] 51.68 [51.56, 51.80] 32.85 [32.74, 32.96] 15.47 [15.37, 15.55] 38.28 [38.18, 38.39] 21.38 [21.30, 21.47] 40.33 [40.23, 40.44] 26.66 [26.56, 26.75] 52.77 [52.67, 52.88] 94.17 [94.12, 94.22] 60.66 [60.56, 60.76] 25.81 [25.71, 25.90] 13.54 [13.46, 13.61] 47.06 [45.08, 49.03] 22.66 [21.00, 24.31] 30.29 [28.47, 32.10] 26.63 [24.89, 28.38] 45.03 [43.06, 46.99] 89.24 [88.02, 90.47] 58.83 [56.89, 60.78] 22.90 [21.24, 24.46] 18.27 [16.74, 19.80] 48.24 [46.34, 50.13] 22.22 [20.64, 23.80] 29.54 [27.81, 31.28] 20.61 [19.07, 22.15] 43.32 [41.44, 45.20] 88.40 [87.18, 89.62] 53.27 [51.37, 55.16] 26.76 [25.08, 28.45] 19.97 [18.45, 21.49] 46.50 [44.06, 48.93] 22.88 [20.82, 24.93] 30.63 [28.37, 32.88] 25.29 [23.17, 27.42] 41.60 [39.19, 44.01] 89.46 [87.96, 90.96] 59.21 [56.81, 61.61] 22.50 [20.46, 24.54] 18.29 [16.40, 20.18] Note: 95% confidence intervals are shown in brackets. 17 Table 2: Descriptive overview (percentages) for gender-nonconforming adults by presence of GNID law Poor self-rated health Identity document laws Driver's license gender markers Negative law Neutral law Positive law Birth certificate gender markers Negative law Neutral law Positive law Name change publication requirements Negative law Neutral law Positive law State political climate Democratic party legislative control No majority in either legislative house Democratic majority in one house Democratic majority in both houses States With No GNID Law (n = 1,274) States With GNID Law (n = 339) 28.65 [26.16, 31.14] 23.01 [18.51, 27.51] 20.88 [18.64, 23.11] 10.83 [9.12, 12.54] 68.29 [65.73, 70.85] 30.06 [27.54, 32.58] 13.11 [11.25, 14.96] 56.83 [54.11, 59.55] 0.00 [0.00, 0.00] 8.26 [5.31, 11.20] 91.74 [88.80, 94.69] 3.24 [1.35, 5.14] 13.57 [9.91, 17.23] 83.19 [79.18, 87.19] 23.70 [21.37, 26.04] 50.55 [47.80, 53.30] 25.75 [23.34, 28.15] 3.54 [1.56, 5.52] 43.07 [37.77, 48.37] 53.39 [48.06, 58.73] 58.99 [56.30, 61.68] 12.84 [11.01, 14.67] 28.17 [25.71, 30.63] 16.47 [12.47, 20.47] 19.46 [15.19, 23.73] 64.07 [58.90, 69.24] 18 p 0.04 < 0.001 < 0.001 < 0.001 < 0.001 Table 2 (cont'd) Democratic governor Sociodemographics LGBQ+ Race/ethnicity White, non-Hispanic Black, non-Hispanic Hispanic Other race, non-Hispanic Age in years Education No or some high school High school graduate/GED Some college College graduate Currently employed Houshold income Less than $25,000 $25,000 to $49,000 $50,000 or more Household composition Any children Currently married/partnered 52.76 [50.03, 55.50] 73.95 [69.22, 78.68] 53.30 [50.55, 56.04] 75.81 [71.13, 80.39] < 0.001 < 0.001 67.90 [65.33, 70.46] 69.03 [64.08, 73.97] 0.37 8.08 [6.59, 9.58] 11.07 [9.34, 12.79] 12.95 [11.11, 14.80] 44.93 [48.82, 46.03] 10.63 [8.93, 12.33] 31.42 [28.86, 33.97] 28.43 [25.94, 30.91] 29.53 [27.02, 32.04] 49.14 [46.39, 51.89] 48.49 [45.84, 51.34] 22.29 [20.00, 24.58] 29.12 [26.63, 31.62] 5.31 [2.91, 7.71] 11.50 [8.09, 14.92] 14.16 [10.43, 17.89] 35.76 [33.94, 37.57] 7.69 [4.84, 10.55] 28.70 [23.85, 33.55] 25.74 [21.06, 30.42] 37.87 [32.67, 43.07] 58.11 [52.83, 63.39] 38.64 [33.43, 43.85] 25.07 [20.44, 29.71] 36.28 [31.14, 41.43] < 0.001 0.02 0.003 0.004 25.51 [23.11, 27.91] 42.86 [40.13, 45.58] 24.48 [19.88, 29.08] 36.87 [31.71, 42.04] 0.70 0.05 19 Table 2 (cont'd) Currently insured Smoking status Never smoker Former smoker Current smoker 88.54 [86.79, 90.29] 92.92 [90.18, 95.66] 57.85 [55.13, 60.56] 23.23 [20.91, 25.56] 18.92 [16.76, 21.07] 64.31 [59.18, 69.43] 19.76 [15.50, 24.02] 15.93 [12.01, 19.84] 0.02 0.10 Note: 95% confidence intervals are shown in brackets. hint at some underlying unmeasured difference in the effects of individual gender-affirming policies that cannot be detected through the positive/neutral/negative classification I have utilized, (Du Bois et al. 2018; Goldenberg et al. 2020; Mezey 2020; Gonzales, Tran, and Bennett 2022; Truszczynski et al. 2022) individuals’ access to healthcare (Kattari et al. 2015; Hughto et al. 2016; Bakko and Kattari 2020), or level of support in each state for transgender and gender identity issues (Pew Research Center 2022). Using the Movement Advancement Project’s gender identity scores, only Arkansas had both a negative score and a GNID policy, while Pennsylvania and Utah had neutral scores, and Virginia had a low score as well as GNID laws. Of states without GNID laws, the overwhelming majority had low or negative gender identity scores. In addition to differences in gender-affirming policies, states with GNID laws might also share significant differences in political climate, demographics, and socioeconomic status than states without GNID laws. States with GNID laws were significantly more likely to have Democratic Governors (73.95% vs. 52.76), Democratic party control in both legislative houses (64.07% vs. 28.17%), were more likely to have positive driver’s license gender marker policies (91.74% vs. 68.29%), positive birth certificate gender marker policies (83.19% vs. 56.83%), as well as positive name change publication requirements (53.39% vs. 25.75%). These significant 20 differences in policy environments, as well as differences in demographic and economic characteristics of the GNC adults in states with and without GNID laws discussed above, are important to keep in mind when situating my final results in context, as they might have a greater impact on differences in self-rated physical health than individual policies such as GNID laws. Regression Results The primary results from each model are presented in Tables 3, 4, and 5, separated across each subpopulation and between three columns for each model. When comparing Akaike information criterion (AIC) and Bayesian information criterion (BIC) across all models, the third was found to be the best fit, leading me to utilize model three to analyze my three hypotheses. There are no significant effects of gender-neutral ID laws for poor self-rated physical health for cisgender men and women (ME = -0.001, p = 0.324) and transmasculine and transfeminine adults (ME = - 0.037, p = 0.122), leading me to accept the first and second hypotheses that there is no overall effect of said laws for these populations. There is, however, significant heterogeneity in the relationship between predictors and health outcomes. Specifically, race and household composition are significant predictors of higher (lower) percentage point increases (decreases) in the probability of poor self-reported health respectively for cisgender adults compared to TGNC adults holding all else constant, while LGBQ identity predicts a significantly larger percentage point increase in the probability of poor health for transfeminine and transmasculine and GNC people than for cisgender men and women. Hypotheses three is also accepted when using the third model (ME = 0.124, p = 0.002). Interestingly, when not including additional gender-identity policy controls in model 2, GNID laws showed no evidence of having a significant attenuating effect (ME = -0.054, p = 0.151) in 21 Table 3: Estimated marginal effects from logistic regression predicting poor self-reported health for cisgender men and women (n = 1,532,206) Model 1 Model 2 Model 3 Gender neutral ID law Driver's license gender markers Neutral law Positive law Birth certificate gender markers Neutral law Positive law Name change publication requirements Neutral law Positive law 0.001 [-0.002, 0.004] -0.001 [-0.004, 0.001] Negative law (reference) -0.002 [-0.005, 0.001] - - - - - - - - -0.005 [-0.011, 0.001] -0.004 [-0.009, 0.002] Negative law (reference) - - 0.003 [-0.001, 0.006] -0.002* [-0.005, -0.0003] Negative law (reference) - - -0.004 [-0.008, 0.0004] 0.005 [-0.004, 0.005] Gender identity Cisgender man (reference) Cisgender woman LGBQ Race/ethnicity Black, non-Hispanic Hispanic Other race, non-Hispanic Age 65+ 0.006*** [0.005, 0.007] -0.006*** [-0.007, -0.005] -0.006*** [-0.007, -0.005] 0.044*** [0.041, 0.047] 0.037*** [0.035, 0.040] 0.023*** [0.020, 0.026] White, non-Hispanic (reference) 0.037*** [0.035, 0.040] 0.023*** [0.020, 0.026] 0.037*** [0.035, 0.040] 0.051*** [0.049, 0.054] 0.037*** [0.034, 0.039] 0.051*** [0.049, 0.054] 0.037*** [0.034, 0.039] 0.051*** [0.049, 0.054] 0.037*** [0.034, 0.039] -0.023*** [-0.025, -0.022] -0.023*** [-0.025, -0.022] -0.023*** [-0.025, -0.022] 22 Table 3 (cont'd) Education High school graduate/GED Some college College graduate Currently employed Household income > $50k Household composition Any children Currently married/partnered Currently insured Smoking status Former smoker Current smoker Model 1 Model 2 Model 3 No or some high school (reference) -0.099*** [-0.102, -0.096] -0.099*** [-0.102, -0.096] -0.099*** [-0.102, -0.096] -0.126*** [-0.129, -0.123] -0.126*** [-0.129, -0.123] -0.126*** [-0.129, -0.123] -0.182*** [-0.195, -0.179] -0.182*** [-0.185, -0.179] -0.182*** [-0.195, -0.179] - - - - - - - -0.141*** [-0.142, -0.140] -0.141*** [-0.142, -0.140] -0.055*** [-0.057, -0.054] -0.055*** [-0.057, -0.054] -0.031*** [-0.033, -0.030] -0.032*** [-0.033, -0.030] -0.030*** [-0.031, -0.028] -0.030*** [-0.031, -0.028] 0.010*** [0.008, 0.012] Never smoker (reference) 0.054*** [0.053, 0.057] 0.010*** [0.008, 0.012] 0.054*** [0.053, 0.056] 0.097*** [0.095, 0.099] 0.097*** [0.095, 0.099] Note: Estimates are weighted. 95% confidence intervals are shown in brackets. *p < .05; **p < .01; ***p < .001 the probability of reporting poor health for GNC adults. While neutral and positive driver’s license and birth certificate gender marker policies have no significant impact on poor health for any subpopulation, name change publication policies that are neutral lead to a 1 percentage point reduction in the probability of reporting poor health for transmasculine and transfeminine people, and positive name change policies lead to a significant 13.1 percentage point decrease and 20.1 percentage point increase in the probability of reporting poor health for transfeminine and 23 Table 4: Estimated marginal effects from logistic regression predicting poor self-reported health for transmasculine and transfeminine adults (n = 5,115) Model 1 Model 2 Model 3 Gender neutral ID law Driver's license gender markers Neutral law Positive law Birth certificate gender markers Neutral law Positive law Name change publication requirements Neutral law Positive law Gender identity Transfeminine LGBQ Race/ethnicity Black, non-Hispanic Hispanic Other race, non-Hispanic Age 65+ -0.034 [-0.083, 0.015] -0.037 [-0.085, 0.010] Negative law (reference) -0.010 [-0.067, 0.048] - - - - - - - - 0.016 [-0.119, 0.150] 0.004 [-0.118, 0.127] Negative law (reference) - - 0.040 [-0.036, 0.117] -0.004 [-0.056, 0.047] Negative law (reference) - - -0.010* [-0.118, -0.003] -0.131* [-0.231, -0.031] Transmasculine (reference) -0.006 [-0.031, 0.018] -0.006 [-0.030, 0.017] -0.007 [-0.030, 0.017] 0.075*** [0.046, 0.103] 0.021 [-0.019, 0.062] 0.046** [0.018, 0.074] White, non-Hispanic (reference) 0.012 [-0.027, 0.051] 0.044** [0.016, 0.072] 0.013 [-0.025, 0.052] -0.042*** [-0.082, -0.003] 0.026* [-0.019, 0.071] -0.028 [-0.069, 0.013] 0.013 [-0.029, 0.056] -0.027 [-0.068, 0.014] 0.016 [-0.027, 0.058] 0.026 [-0.019, 0.071] -0.045** [-0.072, -0.018] -0.044** [-0.071, -0.018] 24 Table 4 (cont'd) Education High school graduate/GED Some college College graduate Currently employed Household income > $50k Household composition Any children Currently married/partnered Currently insured Smoking status Former smoker Current smoker Model 1 Model 2 Model 3 No or some high school (reference) -0.082*** [-0.122, -0.043] -0.136*** [-0.180, -0.092] -0.081*** [-0.121, -0.042] -0.184*** [-0.229, -0.140] -0.118*** [-0.159, -0.078] -0.118*** [-0.159, -0.077] -0.253*** [-0.297, -0.209] -0.153*** [-0.195, -0.110] -0.151*** [-0.194, -0.108] - - - - - - - -0.156*** [-0.180, -0.131] -0.156*** [-0.181, -0.132] -0.101*** [-0.127, -0.075] -0.101*** [-0.127, -0.085] -0.053*** [-0.082, -0.025] -0.053*** [-0.081, -0.025] -0.007 [-0.032, 0.017] -0.007 [-0.032, 0.017] 0.009 [-0.027, 0.045] Never smoker (reference) 0.087*** [0.058, 0.115] 0.009 [-0.028, 0.045] 0.008*** [0.059, 0.115] 0.083*** [0.053, 0.113] 0.083*** [0.053, 0.113] Note: Estimates are weighted. 95% confidence intervals are shown in brackets. *p < .05; **p < .01; ***p < .001 transmasculine and GNC people respectively. The negative impact on self-reported health for positive name change policies among GNC individuals is unexpected, as research has shown that chosen name use among TGNC people may be associated with sizable reductions in negative health outcomes (Pollitt et al. 2021), and may be related to the relatively small sample size of GNC individuals in states with GNID laws (n = 339), significant differences among the policies relating to legal name changes, or some other unmeasured difference or data limitations. Furthermore, name change publication policies that are neutral (ME = -0.096, p = 0.042) 25 Table 5: Estimated marginal effects from logistic regression predicting poor self-reported health for gender-nonconforming adults (n = 1,602) Model 1 Model 2 Model 3 Gender neutral ID law Driver's license gender markers Neutral law Positive law Birth certificate gender markers Neutral law Positive law Name change publication requirements Neutral law Positive law LGBQ Race/ethnicity Black, non-Hispanic Hispanic Other race, non-Hispanic Age 65+ Education High school graduate/GED Some college -0.061 [-0.136, 0.014] -0.054 [-0.128, 0.020] Negative law (reference) -0.124** [-0.201, -0.047] - - - - - - - - -0.140 [-0.367, 0.087] -0.119 [-0.336, 0.098] Negative law (reference) - - -0.00001 [-0.141, 0.141] 0.089 [-0.006, 0.184] Negative law (reference) - - 0.015 [-0.104, 0.134] 0.201** [0.050, 0.352] 0.156*** [0.111, 0.202] 0.030 [-0.057, 0.116] 0.141*** [0.095, 0.187] White, non-Hispanic (reference) 0.046 [-0.043, 0.135] 0.145*** [0.099, 0.191] 0.046 [-0.042, 0.137] 0.023 [-0.047, 0.094] 0.082* [0.010, 0.155] 0.100** [0.036, 0.164] 0.030 [-0.041, 0.100] 0.064 [-0.004, 0.132] 0.039 [-0.026, 0.103] 0.025 [-0.044, 0.095] 0.059 [-0.010, 0.127] 0.044 [-0.021, 0.108] No or some high school (reference) -0.169*** [0.0259, -0.078] -0.118** [-0.204, -0.032] -0.120** [-0.206, -0.035] -0.198*** [-0.291, -0.105] -0.134** [-0.224, -0.044] -0.132** [-0.222, -0.043] 26 Table 5 (cont'd) College graduate Currently employed Household income > $50k Household composition Any children Currently married/partnered Currently insured Smoking status Former smoker Current smoker Model 1 Model 2 Model 3 -0.296*** [-0.385, -0.208] -0.215*** [-0.303, -0.126] -0.217*** [-0305, -0.130] - - - - - - - -0.093*** [-0.138, -0.048] -0.087*** [-0.132, -0.042] -0.095*** [-0.142, -0.048] -0.095*** [-0.142, -0.048] -0.049 [-0.142, -0.048] -0.047 [-0.096, 0.003] 0.054* [0.009, 0.099] 0.051* [0.007, 0.096] 0.004 [0.009, 0.099] Never smoker (reference) 0.080** [0.027, 0.134] 0.002 [-0.065, 0.069] 0.083** [0.029, 0.137] 0.137*** [0.079, 0.196] 0.134*** [0.082, 0.198] Note: Estimates are weighted. 95% confidence intervals are shown in brackets. *p < .05; **p < .01; ***p < .001 or positive (ME = -0.131, p = 0.010) lead to significant percentage point decreases in poor self- rated physical health for transmasculine and transfeminine adults, although there is no significant gender-affirming impact from neutral and positive birth certificate and driver’s license gender marker policies. Overall, the marginal effects are qualitatively identical for all sociodemographic controls between the three models, which is supported by their relatively similar AIC and BIC scores. Finally, Figures 1 and 2 present the average predicted probabilities derived from models two and three respectively for fair to poor health. Consistent with findings from marginal effects from the second model, there are no significant differences in the probability of reporting poor health for cisgender adults (0.175 vs. 0.178), transfeminine and transmasculine people (0.210 vs. 27 Figure 1: (Model 2) Predicted probabilities of reporting poor health by presence of GNID laws 0.248), and GNC people (0.233 vs. 0.287) residing in states with GNID laws compared to those without. Using the third model, however, GNC adults in states with GNID laws have a significantly lower probability of reporting a health disadvantage (0.182) than GNC adults in untreated states (0.306). Cisgender adults (0.175 vs. 0.177) and transgender adults (0.234 vs. 0.244) do not have significantly different probabilities, however. The significant disparity in results across models is further shown with these predicted probabilities, and potentially suggest that GNID laws and other identity document policies are correlated with each other or with some other omitted variable. Further research should consider different ways of measuring gender- affirming policies that might play an important role in alleviating poor health among TGNC individuals when considering the merit of gender-neutral ID laws as gender-affirming policies. 28 Figure 2: (Model 3) Predicted probabilities of reporting poor health by presence of GNID laws 29 DISCUSSION This study uses a large sample of cisgender, transgender, and gender-nonconforming adults to examine disparities in health across each population, and to investigate whether gender-neutral ID laws that allow individuals to alter their sex/gender marker on their driver’s license to ‘X’ serve as a gender-affirming policy that might have any attenuating effect on poor self-reported physical health for GNC adults. I examined differences in demographic and socioeconomic characteristics across gender identities as well as between GNC adults who live in states with and without GNID laws. All three of my hypotheses are supported by an examination of descriptive statistics and analytical modeling. Consistent with previous research, TGNC individuals are significantly more likely to report poor health and decreased socioeconomic stability than their cisgender peers. GNC adults in states with GNID also report significant differences in demographic characteristics and poor health compared to GNC individuals in states without GNID laws. When controlling for important demographic and economic factors as well as state-level gender-affirming policies, I find that GNID laws are associated with a 12.4 percentage point reduction in the probability of reporting poor physical health for GNC adults. Neither cisgender men and women nor transmasculine and transfeminine adults had any significant associated decreases in self-reported poor physical health, supporting my assumption that GNID laws are likely to be targeted towards and beneficial to GNC adults. These findings are also supported by plotted predicted probabilities obtained from the third model. This may suggest that GNID laws serve as gender-affirming policies that prevent stigma similar to other gender-affirming practices (Hughto, Residner, and Pachankis 2015; Lelutiu-Weinberger, English, and Sandanapitchai 2020; Dorsen et al. 2022), policies (Goldenberg et al. 2020; Gonzales, Tran, and Bennett 2022; Scheim 30 et al. 2022l; Trusczczynski et al. 2022), and access to other gender-concordant documents (Hill et al. 2018; Loza et al. 2021; Goetz and Arcomano 2021). While there is disagreement among GNC adults as to the importance or necessity for gender-neutral gender markers (Goetz and Arcomano 2021), my findings suggest that GNID laws that allow GNC adults to have at least one gender-concordant document might have important implications for self-rated physical health in spite of expressed concerns (Saguy 2023). Limitations It is important to note that these findings come with a number of important limitations that might have inflated the association between GNID laws and improvements in self-rated physical health among GNC adults. This study only examines the association between said policies and self- rated physical health, not causal relationships; since the Behavioral Risk Factor Surveillance System surveys are cross-sectional and not longitudinal, it is impossible to establish that there is a causal relationship between the enaction of GNID laws and any subsequent change in self- reported physical health. Health and demographic data are self-reported, which might lead to my results being affected by poor memory recall or social desirability bias. Previous studies have also noted that BRFSS data suffers from potential mismeasurement and underreporting that similar surveys of TGNC populations also face (Tordoff, Andrasik, and Hajat 2019; Cicero et al. 2020; Lett and Everhart 2022), leading to potentially skewed results. Issues with the BRFSS- provided raked weights (Lett and Everhart 2022) and the fact that only 43 states adopted the optional SOGI module between 2014 and 2021 also do not allow my results to be representative nationally. The ability of the 2014-15 BRFSS’ gender identity questions to provide information about respondents’ sex has also come under scrutiny when identifying transgender respondents 31 because survey interviewers initially assess respondents’ sex based on their interpretation of the timbre of a respondent’s voice as opposed to asking about sex assigned at birth (Lagos 2019). Additionally, since BRFSS does not ask respondents whether they identify as gender- nonconforming unless they identified as transgender during the interview, and the survey does not allow gender-nonconforming individuals to select multiple gender identities that may reflect their true experience with gender, this study makes claims only about those gender- nonconforming respondents who explicitly identify as transgender and who identify more closely with a gender-nonconforming identity than with another label that was not included. Finally, LGBTQ legislation and anti-trans laws are rapidly changing, and their complicated nature make it almost impossible to operationalize as variables, making it difficult to control for all negative, neutral, and positive policies across the 43 states included in my sample. 32 CONCLUSION This study is one of the first to examine the association between gender-neutral ID laws and self- reported physical health among GNC adults. My findings support previous research that has shown that transgender and gender-nonconforming adults report worse physical health than their cisgender peers (Lagos 2018; Scheim et al. 2022), and contribute to the literature by showing stark differences in demographic and economic characteristics among GNC adults in states with and without GNID laws. I also find some promising evidence that GNID laws are associated with a significant reduction in the probability of reporting poor physical health for GNC adults. My study faces a number of structural limitations currently present in the limited data that include TGNC adults, as well as complications from how difficult it is to track and operationalize rapid changes in LGBTQ+ related policies and the current flurry of anti-trans laws that have been circulating through state legislative bodies. The literature would benefit from future qualitative research targeting GNC individuals with and without gender-neutral ID markers to understand why or why not they serve to affirm one’s gender. Additional data would allow researchers to further explore better ways of measuring and controlling for additional gender-affirming policies that might have conflated or confounded my results, as well as what the rate of uptake is for GNID policies both among the GNC population and across other gender minority individuals. Finally, this research could be utilized to develop a conceptual framework to understand how broad social policies serve to directly influence embodied experiences of marginalization. 33 REFERENCES Ayhan, C. H. B., Bilgin, H., Uluman, O. T., Sukut, O., Yilmaz, S., & Buzlu, S. (2020). A systematic review of the discrimination against sexual and gender minority in health care settings. International Journal of Health Services, 50(1), 44-61. Bakko, M., & Kattari, S. K. (2020). Transgender-related insurance denials as barriers to transgender healthcare: differences in experience by insurance type. Journal of General Internal Medicine, 35, 1693-1700. Burgwal, A., Gvianishvili, N., Hård, V., Kata, J., García Nieto, I., Orre, C., ... & Motmans, J. (2019). Health disparities between binary and non binary trans people: A community- driven survey. International Journal of Transgenderism, 20(2-3), 218-229. Block, M. (2022, April 1). The White House announces moves to gender neutral passports. NPR. https://www.npr.org/2022/04/01/1090192947/the-white-house-announces-moves-to- gender-neutral-passports California Courts Self-Help Guide. (2023). Update your gender marker or sex identifier on your identity documents. https://selfhelp.courts.ca.gov/gender-recognition/update-gender- marker-ID-documents Carson, V., Adamo, K., & Rhodes, R. E. (2018). Associations of parenthood with physical activity, sedentary behavior, and sleep. American Journal of Health Behavior, 42(3), 80- 89. Case, A., & Paxson, C. (2005). Sex differences in morbidity and mortality. Demography, 42, 189–214. Centers for Disease Control. (2013). The BRFSS data users guide. https://www.cdc.gov/brfss/data_documentation/pdf/UserguideJune2013.pdf Cicero, E. C., Reisner, S. L., Merwin, E. I., Humphreys, J. C., & Silva, S. G. (2020). The health status of transgender and gender nonbinary adults in the United States. PloS One, 15(2), e0228765. Cicero, E. C., Reisner, S. L., Merwin, E. I., Humphreys, J. C., & Silva, S. G. (2020). Application of behavioral risk factor surveillance system sampling weights to transgender health measurement. Nursing Research, 69(4), 307. Crissman, H. P., Stroumsa, D., Kobernik, E. K., & Berger, M. B. (2019). Gender and frequent mental distress: Comparing transgender and non-transgender individuals' self-rated mental health. Journal of Women's Health, 28(2), 143-151. 34 Crosby, R. A., Salazar, L. F., & Hill, B. J. (2016). Gender affirmation and resiliency among black transgender women with and without HIV infection. Transgender Health, 1(1), 86- 93. DeChants, J. P., Price, M. N., Green, A. E., Davis, C. K., & Pick, C. J. (2022). Association of updating identification documents with suicidal ideation and attempts among transgender and nonbinary youth. International Journal of Environmental Research and Public Health, 19(9), 5016. DeSalvo, K. B., Bloser, N., Reynolds, K., He, J., & Mutner, P. (2006). Mortality prediction with a single general self-rated health question. Journal of General Internal Medicine, 21, 267–275. Dorsen, C. G., Leonard, N., Goldsamt, L., Warner, A., Moore, K. G., Levitt, N., & Rosenfeld, P. (2022). What does gender affirmation mean to you? An exploratory study. Nursing Forum, 57(1), 34-41. Du Bois, S. N., Yoder, W., Guy, A. A., Manser, K., & Ramos, S. (2018). Examining associations between state-level transgender policies and transgender health. Transgender Health, 3(1), 220-224. Feldman, J. L., Luhur, W. E., Herman, J. L., Poteat, T., & Meyer, I. H. (2021). Health and health care access in the US transgender population health (TransPop) survey. Andrology, 9(6), 1707-1718. Frankenberg, E., & Jones, N. R. (2004). Self-rated health and mortality: Does the relationship extend to a low income setting? Journal of Health and Social Behavior, 45, 441–452. Frost, D. M., Lehavot, K., & Meyer, I. H. (2015). Minority stress and physical health among sexual minority individuals. Journal of behavioral medicine, 38, 1-8. Goetz, T. G., & Arcomano, A. C. (2021). “X” marks the transgressive gender: A qualitative exploration of legal gender-affirmation. Journal of Gay & Lesbian Mental Health, 1-15. Goldenberg, T., Reisner, S. L., Harper, G. W., Gamarel, K. E., & Stephenson, R. (2020). State policies and healthcare use among transgender people in the US. American Journal of Preventive Medicine, 59(2), 247-259. Gonzales, G., Tran, N. M., & Bennett, M. A. (2022). State policies and health disparities between transgender and cisgender adults: Considerations and challenges using population-based survey data. Journal of Health Politics, Policy and Law, 47(5), 555- 581. Gorman, B. K., & Sivaganesan, A. (2007). The role of social support and integration for understanding socioeconomic disparities in self-rated health and hypertension. Social Science & Medicine, 65, 958–975. 35 Green, A. E., DeChants, J. P., Price, M. N., & Davis, C. K. (2022). Association of gender- affirming hormone therapy with depression, thoughts of suicide, and attempted suicide among transgender and nonbinary youth. Journal of Adolescent Health, 70(4), 643-649. Hendricks, M. L., & Testa, R. J. (2012). A conceptual framework for clinical work with transgender and gender nonconforming clients: An adaptation of the Minority Stress Model. Professional Psychology: Research and Practice, 43(5), 460. Hill, B. J., Crosby, R., Bouris, A., Brown, R., Bak, T., Rosentel, K., ... & Salazar, L. (2018). Exploring transgender legal name change as a potential structural intervention for mitigating social determinants of health among transgender women of color. Sexuality Research and Social Policy, 15, 25-33. Holt-Lunstad, J. (2018). Why social relationships are important for physical health: A systems approach to understanding and modifying risk and protection. Annual Review of Psychology, 69, 437-458. Hsieh, N., & Shuster, S. M. (2021). Health and health care of sexual and gender minorities. Journal of Health and Social Behavior, 62(3), 318-333. Ingersoll Gender Center. (2019). Third gender markers on WA state IDs: Community questions and answers. https://ingersollgendercenter.org/thirdgendermarker-qa/ Kattari, S. K., Walls, N. E., Whitfield, D. L., & Langenderfer-Magruder, L. (2015). Racial and ethnic differences in experiences of discrimination in accessing health services among transgender people in the United States. International Journal of Transgenderism, 16(2), 68-79. Kidd, J. D., Dolezal, C., & Bockting, W. O. (2018). The relationship between tobacco use and legal document gender-marker change, hormone use, and gender-affirming surgery in a United States sample of trans-feminine and trans-masculine individuals: implications for cardiovascular health. LGBT Health, 5(7), 401-411. King, W. M., & Gamarel, K. E. (2021). A scoping review examining social and legal gender affirmation and health among transgender populations. Transgender Health, 6(1), 5-22. King, M., Semlyen, J., Tai, S. S., Killaspy, H., Osborn, D., Popelyuk, D., & Nazareth, I. (2008). A systematic review of mental disorder, suicide, and deliberate self harm in lesbian, gay and bisexual people. BMC Psychiatry, 8, 1-17. Lagos, D. (2018). Looking at population health beyond “male” and “female”: Implications of transgender identity and gender nonconformity for population health. Demography, 55(6), 2097-2117. Lagos, D. (2019). Hearing gender: Voice-based gender classification processes and transgender health inequality. American Sociological Review, 84(5), 801-827. 36 Le, V., Arayasirikul, S., Chen, Y. H., Jin, H., & Wilson, E. C. (2016). Types of social support and parental acceptance among transfemale youth and their impact on mental health, sexual debut, history of sex work and condomless anal intercourse. Journal of the International AIDS Society, 19, 20781. Lee, J. Y., & Rosenthal, S. M. (2023). Gender-affirming care of transgender and gender-diverse youth: Current concepts. Annual Review of Medicine, 74. Lefevor, G. T., Boyd-Rogers, C. C., Sprague, B. M., & Janis, R. A. (2019). Health disparities between genderqueer, transgender, and cisgender individuals: An extension of minority stress theory. Journal of Counseling Psychology, 66(4), 385–395. https://doi.org/10.1037/cou0000339 Lelutiu-Weinberger, C., English, D., & Sandanapitchai, P. (2020). The roles of gender affirmation and discrimination in the resilience of transgender individuals in the US. Behavioral Medicine, 46(3-4), 175-188. Lett, E., & Everhart, A. (2022). Considerations for transgender population health research based on US national surveys. Annals of Epidemiology, 65, 65-71. Link, B. G., & Phelan, J. C. (2001). Conceptualizing stigma. Annual Review of Sociology, 27(1), 363-385. Lombardi, E. (2009). Varieties of transgender/transsexual lives and their relationship with transphobia. Journal of Homosexuality, 56, 977–992. Loza, O., Beltran, O., Perez, A., & Green, J. (2021). Impact of name change and gender marker correction on identity documents to structural factors and harassment among transgender and gender diverse people in Texas. Sexuality, Gender & Policy, 4(2), 76-105. McLemore, K. A. (2018). A minority stress perspective on transgender individuals’ experiences with misgendering. Stigma and Health, 3(1), 53–64. https://doi.org/10.1037/sah0000070 Metzger, S., & Gracia, P. (2023). Gender differences in mental health following the transition into parenthood: Longitudinal evidence from the UK. Advances in Life Course Research, 56, 100550. Meyer, I. H. (1995). Minority stress and mental health in gay men. Journal of Health and Social Behavior, 36, 38–56. Mezey, S. G. (2020). Transgender policymaking: The view from the states. Publius: The Journal of Federalism, 50(3), 494-517. Movement Advancement Project. (2023). Identity Document Laws and Policies. https://www.lgbtmap.org/equality-maps/identity_document_laws 37 Perales, F., Ablaza, C., & Elkin, N. (2022). Exposure to inclusive language and well-being at work among transgender employees in Australia, 2020. American Journal of Public Health, 112(3), 482-490. Pew Research Center. (2022). How Americans view policy proposals on transgender and gender identity issues, and where such policies exist. https://www.pewresearch.org/fact- tank/2022/09/15/how-americans-view-policy-proposals-on-transgender-and-gender- identity-issues-and-where-such-policies-exist/ Pollitt, A. M., Ioverno, S., Russell, S. T., Li, G., & Grossman, A. H. (2021). Predictors and mental health benefits of chosen name use among transgender youth. Youth & Society, 53(2), 320-341. Preston, S. H., & Wang, H. (2006). Sex mortality differences in the United States: The role of cohort smoking patterns. Demography, 43, 631–646 Quinan, C. L., & Oosthoek, D. (2021). Trans and Non-binary Identities and a Politics Beyond Recognition: On the Possibility of the X. In Advances in Trans Studies: Moving Toward Gender Expansion and Trans Hope (Vol. 32, pp. 93-107). Emerald Publishing Limited. Reisner, S. L., & Hughto, J. M. (2019). Comparing the health of non-binary and binary transgender adults in a statewide non-probability sample. PLoS One, 14(8), e0221583. Rees, S. N., Crowe, M., & Harris, S. (2021). The lesbian, gay, bisexual and transgender communities' mental health care needs and experiences of mental health services: An integrative review of qualitative studies. Journal of Psychiatric and Mental Health Nursing, 28(4), 578-589. Rendall, M. S., Weden, M. M., Favreault, M. M., & Waldron, H. (2011). The protective effect of marriage for survival: A review and update. Demography, 48, 481–506. Restar, A., Jin, H., Breslow, A., Reisner, S. L., Mimiaga, M., Cahill, S., & Hughto, J. M. (2020). Legal gender marker and name change is associated with lower negative emotional response to gender-based mistreatment and improve mental health outcomes among trans populations. SSM-Population Health, 11, 100595. Rimes, K. A., Goodship, N., Ussher, G., Baker, D., & West, E. (2019). Non-binary and binary transgender youth: Comparison of mental health, self-harm, suicidality, substance use and victimization experiences. International Journal of Transgenderism, 20(2-3), 230- 240. Santini, Z. I., Koyanagi, A., Tyrovolas, S., Mason, C., & Haro, J. M. (2015). The association between social relationships and depression: A systematic review. Journal of Affective Disorders, 175, 53-65. Scheim, A. I., Perez-Brumer, A. G., & Bauer, G. R. (2020). Gender-concordant identity 38 documents and mental health among transgender adults in the USA: A cross-sectional study. The Lancet Public Health, 5(4), e196-e203. Scheim, A. I., Baker, K. E., Restar, A. J., & Sell, R. L. (2022). Health and health care among transgender adults in the United States. Annual Review of Public Health, 43, 503-523. Simon, R. W., & Caputo, J. (2019). The costs and benefits of parenthood for mental and physical health in the United States: The importance of parenting stage. Society and Mental Health, 9(3), 296-315. Smith-Johnson, M. (2022). Transgender adults have higher rates of disability than their cisgender counterparts: Study examines rates of disability among transgender adults and cisgender adults. Health Affairs, 41(10), 1470-1476. Stacey, L., Reczek, R., & Spiker, R. (2022). Toward a holistic demographic profile of sexual and gender minority well-being. Demography, 59(4), 1403-1430. Swan, J., Phillips, T. M., Sanders, T., Mullens, A. B., Debattista, J., & Brömdal, A. (2023). Mental health and quality of life outcomes of gender-affirming surgery: A systematic literature review. Journal of Gay & Lesbian Mental Health, 27(1), 2-45. Tan, K. K., Watson, R. J., Byrne, J. L., & Veale, J. F. (2022). Barriers to possessing gender- concordant identity documents are associated with transgender and nonbinary people's mental health in Aotearoa/New Zealand. LGBT Health, 9(6), 401-410. Tordoff, D., Andrasik, M., & Hajat, A. (2019). Misclassification of sex assigned at birth in the behavioral risk factor surveillance system and transgender reproductive health: A quantitative bias analysis. Epidemiology, 30(5), 669-678. Tordoff, D. M., Wanta, J. W., Collin, A., Stepney, C., Inwards-Breland, D. J., & Ahrens, K. (2022). Mental health outcomes in transgender and nonbinary youths receiving gender- affirming care. JAMA network open, 5(2), e220978-e220978. Truszczynski, M., Truszczynski, N., Estevez, R. I., & Elliott, A. E. (2022). Does policy matter? The impact of state and city anti-discrimination policy on the discrimination experiences of trans and nonbinary people. Sexuality Research and Social Policy, 19(4), 1786-1794. White Hughto, J. M., Reisner, S. L., & Pachankis, J. E. (2015). Transgender stigma and health: A critical review of stigma determinants, mechanisms, and interventions. Social Science and Medicine (1982), 147, 222-231. White Hughto, J. M., Murchison, G. R., Clark, K., Pachankis, J. E., & Reisner, S. L. (2016). Geographic and individual differences in healthcare access for US transgender adults: A multilevel analysis. LGBT Health, 3(6), 424-433. 39 Yee, K., Lind, B. K., & Downing, J. (2022). Change in gender on record and transgender adults’ mental or behavioral health. American Journal of Preventive Medicine, 62(5), 696-704. 40