Determinants of Public Opinion Toward Gender-Affirming Surgery in the United States

Abstract

Purpose:

National polling data indicate that Americans support the right of transgender persons to undergo gender-affirming surgery (GAS). It remains unknown whether public perceptions of GAS differ depending on patient subpopulations, anatomical site, or insurance coverage and whether the public widely believes that transgender people will regret GAS.

Methods:

We built a Qualtrics™ survey derived from an online validated 2017 Ipsos survey and distributed it to American adults through Amazon Mechanical Turk. Associations of demographic characteristics with perception of GAS were determined using multinomial logistic regression.

Results:

Respondents (n=312) were predominantly non-Hispanic White (69.2%), held a bachelor's degree (64.7%), and reported an annual income of $25,000 to $74,999 (64.4%). Approximately half of respondents identified as socially liberal (50.3%); 34.0% as socially conservative; and 15.7% as neither. Respondents supported a right to GAS independent of anatomy and insurance. Support for transgender children (62%) was less than for adult transgender men (84%) and women (83%). Despite supporting a right to GAS, respondents agreed that transgender adults (67%) and children (74%) would regret GAS. Education was the strongest predictor of support for GAS rights. Socially conservative respondents were significantly more likely than nonideological or liberal respondents to believe that transgender people would regret GAS.

Conclusion:

This large online sample of American adults with diverse ideologies demonstrated support for GAS independent of anatomical site and insurance. Support of GAS for transgender children is robust, although lower than support for adults. Despite broad support, most laypersons believe that transgender people would regret GAS.

Introduction

Over 1.4 million transgender and gender-diverse (TGD) people live in the United States, comprising 0.6% of the adult population¹ and an even larger percentage of U.S. adolescents.² Gender affirmation, the interpersonal process through which an individual receives recognition and support for their gender identity, is the most effective means of minimizing health disparities observed among TGD people.² A crucial aspect of gender affirmation for many TGD people is medical intervention, including hormone therapy and gender-affirming surgery (GAS). GAS encompasses any surgical procedure designed to affirm the gender identities of TGD people.³ Common procedures alter the breasts/chest (mastectomy, breast augmentation), genitals (phalloplasty, vaginoplasty), and nongenitals (facial feminization, voice surgery).⁴ Although variable levels of evidence exist for specific procedures, GAS is effective at improving the mental health of TGD individuals.^3,5–10

Despite demonstrated efficacy of GAS, access to specialized care remains limited. Public and private insurers have only changed coverage policies in the last decade to include gender-affirming care.¹¹ Nearly half of states explicitly prohibit Medicaid coverage of GAS.¹² Of insurers who consider genital GAS medically necessary, a majority include additional criteria for coverage beyond that dictated by the World Professional Association for Transgender Health (WPATH).¹³ In the 2015 U.S. Transgender Survey, 55% of respondents who sought coverage of GAS in the past year were denied, compared to 25% of those who sought hormone therapy.¹⁴ Beyond coverage questions, only 20 of 50 states have at least one surgeon that performs genital GAS.¹⁵

These persistent roadblocks have been exacerbated by a national political movement to ban gender-affirming care for minors, with 22 states introducing legislation to that effect in 2021.^16,17 Although a fraction of this legislation was enacted into law, this nascent movement portends that gender-affirming health care may become a protracted target for political restrictions. In 2022, state politicians have referred to GAS as “genital mutilation” and introduced bills that would make it a felony to provide gender-affirming care to TGD youth.¹⁸ Executive political pressure alone has successfully forced transgender clinic closures at academic centers, further limiting access to care.¹⁹

Despite the onslaught against TGD rights in the echelons of state government, a considerable two-thirds of Americans across political lines oppose bans of gender-affirming care for minors.²⁰ A 2017 Ipsos poll found that 70% of Americans support the right to GAS, specifically.²¹ Surface level public approval of LGBTQ+ rights, however, often masks more nuanced biases vulnerable to exploitation by anti-trans advocacy groups. A common example of this is the perception that TGD people will regret gender-affirming care.

This sentiment is commonly used as a thematic device by politicians and organizations advocating for restriction of gender-affirming care.^22,23 The power of this message is evidenced by its dissemination in the nonpartisan media, most notably during a 2021 CBS 60 Minutes segment²⁴ which emphasized the experience of “detransitioners,” individuals seeking to reverse the effects of gender-affirming care. In reality, regret after GAS is rare,^25–27 ∼1%, and is minimized further when WPATH preoperative eligibility criteria are followed.^28,29 The extent to which the general public conforms to this misinformation remains unknown.

In this study, we conduct an online survey of American adults to better quantify public opinion of GAS. This study aims to: (1) assess general respondents' knowledge and opinion of GAS under different circumstances, including patient age, gender identity, anatomy, and insurance coverage, (2) assess general respondents' opinion on whether TGD people who undergo GAS will regret their decision, and (3) provide detailed public opinion data that may support advocacy efforts to protect GAS access.

Methods

Mechanical Turk for public opinion research

Amazon Mechanical Turk (MTurk) is an online crowdsourcing platform.³⁰ MTurk is an established tool for public opinion research in the fields of political science, medicine, and public health^31–36 with comparable effectiveness and reliability to traditional survey methods.³⁴ Utilization of quality thresholds on MTurk workers, such as human intelligence task (HIT) approval rates, further improves data quality.³⁷ Worker inclusion criteria included a HIT approval rate minimum of 95%, age over 18, and location in the United States. Workers received $0.25 for completing the survey.

Survey

This study was exempted by the Washington University Institutional Review Board due to the deidentified nature of respondents from whom consent was voluntarily given.

Surveys were created with Qualtrics™ software (Provo, UT) and distributed through Amazon MTurk (Supplementary Appendix SA1). Questions were adopted verbatim or modified to the least degree possible with permission from previously validated sources.^21,38,39 Questions unavailable in validated sources were written de novo. Demographic information was collected for age, sex, gender identity, ethnicity, race, state of residence, education, income, religion, and social and economic ideology.

All respondents were provided a definition of “transgender people” from the Centers for Disease Control and Prevention before initiating the survey.⁴⁰ Perception of GAS was assessed across patient subpopulations, anatomical sites, and insurance coverage. Prior awareness of gender-affirming hormone therapy, GAS, GAS procedures, and opinion on whether GAS is reconstructive, cosmetic, and/or medically necessary were also assessed. Finally, respondents were asked about anticipated post-GAS regret. A 4-point Likert scale from “Strongly Disagree” to “Strongly Agree” was utilized. Response order was randomized.

Surveys were released in two batches on MTurk: a pilot batch to 20 workers on November 9, 2021 followed by the full release to 480 workers on November 12, 2021. Data collection was complete within 3 h of survey release. The median duration time to survey completion was 5 min and 4 sec with an absolute minimum duration of 1 min 26 sec. Each worker was given a unique ID to match them to a single response and verify completion. Ten survey responses with no matched worker were excluded. Eight respondents who indicated “prefer not to say” to demographics questions were excluded because these groups were too small to include in statistical modeling. Three additional respondents with missing responses were excluded. Thus, 487 of 508 responses were included in the preliminary analysis (Fig. 1).

FIG. 1.

Inclusion schema of study respondents.

On preliminary analysis, 175 respondents (35.9% of total) who responded to questions concerning gender identity or familiarity with transgender people in an illogical, or mutually exclusive, manner were excluded. This group included respondents who identified as transgender in question Q1 of the survey (Supplementary Appendix SA1) while simultaneously indicating a binary sex-gender concordant identity in questions D2 and D3. Those who indicated that they simultaneously knew and did not know transgender people were also excluded. Subanalysis of these combined respondents revealed their demographic composition to be distinct from the respondents who completed the survey in a logical manner (Supplementary Table S1).

Multinomial logistic regression revealed membership to the “Excluded” group to be an influential predictor of survey responses. These respondents were excluded from further analysis due to concerns over attentiveness and their understanding of the shared definition of “transgender” provided. The analysis of the remaining respondents (n=312) is reported in Results section (Fig. 1). The demographics and statistical analysis of the unfiltered study population (n=487) is available in the Supplement (Supplementary Tables S1–S5).

Statistical analyses

Results were collected through Qualtrics and graphed with Microsoft Excel (Redmond, WA). Associations of individual-level characteristics (sex, age, ethnicity, education, income, social political ideology, religion, and personal familiarity with transgender people) with GAS attitudes were determined using multinomial logistic regression using IBM^® SPSS^® Statistics (Armonk, NY). Relative risk ratios are presented as 95% confidence intervals. Statistical significance was determined using a p-value <0.05.

Results

Demographic, socioeconomic, and ideological characteristics

A total of 312 respondents were included in the analysis (Table 1). Respondents were predominantly non-Hispanic White (69.2%), held a 4-year college degree (64.7%), and reported an annual income of $25,000 to $74,999 (64.4%). Half of respondents identified as socially liberal (50.3%), 34.0% identified as socially conservative, and 15.7% identified as neither.

Table 1.

Demographic, socioeconomic, and ideological characteristics of respondents (n=312)

	Frequency	Percentage
Sex
Male	160	51.3
Female	152	48.7
Identity
Male	159	51.0
Female	151	48.4
Transgender	2	0.6
Age
18–34	134	42.9
35–49	128	41.0
50+	50	16.0
Ethnicity
Hispanic	71	22.8
Non-Hispanic	241	77.2
Race
White	271	86.9
Non-White	41	13.1
Education
Less than a 4-year college degree	60	19.2
4-Year college degree	202	64.7
Advanced degree (e.g., MS, PhD, MD, JD)	50	16.0
Income
< $24,999	59	18.9
$25,000 to $74,999	201	64.4
> $75,000	52	16.7
Region
South	127	40.7
West	84	26.9
Northeast	40	12.8
Midwest	61	19.6
Religion
Roman Catholic	148	47.4
Other Christiansa	90	28.8
Non-religiousb	47	15.1
Non-Christian religious affiliatec	27	8.7
Social ideology
Liberal	157	50.3
Conservative	106	34.0
Neither conservative nor liberal	49	15.7
Economic ideology
Liberal	135	43.3
Conservative	118	37.8
Neither conservative nor liberal	59	18.9

^a “Other Christians” includes Christian (nondenominational), Protestant, Latter Day Saints, Orthodox.

^b “Non-religious” includes Agnostic, Atheist, Nothing in particular.

^c “Non-Christian religious affiliate” includes Jewish, Hindu, Muslim, Buddhist, Something else, Prefer not to say.

Familiarity with transgender people

All respondents were asked to read the following definition before answering survey questions that concerned transgender people:

“Some people describe themselves as transgender when they experience a different gender identity from their sex at birth. For example, a person born into a male body, but who feels female or lives as a woman would be transgender. Some transgender people change their physical appearance so that it matches their internal gender identity. Some transgender people take hormones and some have surgery. A transgender person may be of any sexual orientation—straight, gay, lesbian, or bisexual.”⁴⁰

A similar proportion of respondents reported having transgender friends, family, or acquaintances (46.4%) as those who reported rarely seeing or not personally knowing transgender people (51.0%). Two of 312 respondents (0.6%) reported personally identifying as transgender (Fig. 2).

FIG. 2.

Familiarity with transgender and gender-diverse people among respondents (n=312).

Attitudes toward GAS

A large majority of respondents agreed that transgender people are a natural occurrence (80.4%) and that they should be allowed to have GAS (86.5%) and gender-affirming hormone therapy (86.9%).

Similarly large majorities agreed that transgender men (84.3%) and transgender women (83.3%) should be allowed to have GAS. A smaller majority agreed that transgender children (62.2%) should be allowed to have GAS. Holding a 4-year college degree and a liberal social ideology were significantly associated with supporting GAS rights. Younger respondents were more likely to support a right for minors to undergo GAS (Table 2).

Table 2.

They should be allowed to have surgery to change their physical appearance so that it matches their identity (n=312)

	Female-to-male		Male-to-female		Children (<18 years)
	Disagree	Agree	Disagree	Agree	Disagree	Agree
	RRR (95% CI)	RRR (95% CI)	RRR (95% CI)	RRR (95% CI)	RRR (95% CI)	RRR (95% CI)
Sex (ref: Male)
Female	1	2.121 (1.071–4.198)	1	1.785 (0.920–3.463)	1	0.935 (0.569–1.538)
Ages (ref: 50+)
18–34	1	2.045 (0.828–5.050)	1	1.487 (0.626–3.528)	1	2.730 (1.350–5.520)
35–49	1	1.289 (0.531–3.128)	1	1.955 (0.787–4.854)	1	2.483 (1.220–5.055)
Ethnicity (ref: Hispanic)
Non-Hispanic	1	0.824 (0.367–1.847)	1	0.459 (0.188–1.120)	1	0.633 (0.345–1.163)
Education (ref: less than a 4-year degree)
4-Year college degree	1	2.908 (1.223–6.917)	1	2.852 (1.220–6.665)	1	1.953 (1.015–3.755)
Advanced degree (e.g., MS, PhD, MD, JD)	1	2.891 (0.905–9.231)	1	1.822 (0.623–5.330)	1	2.054 (0.885–4.765)
Income (ref: $25,000 to $74,999)
< $25,000	1	0.707 (0.299–1.670)	1	0.445 (0.196–1.011)	1	0.972 (0.510–1.854)
> $75,000	1	1.774 (0.679–4.634)	1	0.808 (0.344–1.900)	1	1.176 (0.592–2.334)
Religion (ref: Non-religious)
Roman Catholic	1	0.350 (0.101–1.205)	1	0.515 (0.159–1.668)	1	1.885 (0.863–4.118)
Other Christian	1	0.397 (0.117–1.347)	1	0.511 (0.158–1.648)	1	0.845 (0.384–1.862)
Non-Christian affiliation	1	3.407 (0.347–33.419)	1	1.836 (0.309–10.903)	1	1.381 (0.485–3.935)
Social ideology (ref: Conservative)
Liberal	1	2.857 (1.362–5.991)	1	3.993 (1.914–8.333)	1	1.079 (0.617–1.888)
Neither conservative Nor liberal	1	1.043 (0.415–2.618)	1	1.686 (0.665–4.272)	1	0.588 (0.277–1.251)
Familiarity (ref: Don't know a transgender person)
Know a transgender person	1	0.683 (0.349–1.336)	1	0.860 (0.444–1.665)	1	0.877 (0.533–1.445)

Bolded cells indicate results with p-values <0.05.

CI, confidence interval; RRR, relative risk ratio.

Two-thirds of respondents agreed that transgender adults (66.7%) and children (73.7%) who undergo GAS will regret their decision. Personally knowing a transgender person, possessing a liberal or nonideological social ideology, and female sex were significantly associated with a decreased likelihood to endorse that transgender patients would regret GAS. The strengths of these associations were diminished for pediatric patients relative to adults. Younger and Christian respondents were more likely to endorse that transgender people would regret GAS (Table 3).

Table 3.

Transgender (adults/children) who have surgery to change their physical appearance will come to regret their decision (n=312)

	Adults		Children
	Disagree	Agree	Disagree	Agree
	RRR (95% CI)	RRR (95% CI)	RRR (95% CI)	RRR (95% CI)
Sex (ref: Male)
Female	1	0.414 (0.232–0.737)	1	0.575 (0.326–1.015)
Ages (ref: 50+)
18–34	1	3.921 (1.723–8.290)	1	1.123 (0.491–2.571)
35–49	1	2.609 (1.158–5.875)	1	0.968 (0.421, 2.224)
Ethnicity (ref: Hispanic)
Non-Hispanic	1	0.627 (0.310–1.266)	1	0.943 (0.479–1.857)
Education (ref: less than a 4-year degree)
4-Year college degree	1	0.888 (0.417–1.891)	1	0.577 (0.264–1.262)
Advanced degree (e.g., MS, PhD, MD, JD)	1	0.735 (0.280–1.933)	1	0.520 (0.197–1.372)
Income (ref: $25,000 to $74,999)
< $25,000	1	0.744 (0.361–1.532)	1	1.155 (0.560–2.379)
> $75,000	1	0.932 (0.425–2.044)	1	1.267 (0.568–2.826)
Religion (ref: Non-religious)
Roman Catholic	1	5.970 (2.492–14.304)	1	2.616 (1.154–5.932)
Other Christian	1	2.856 (1.177–6.390)	1	2.308 (0.975–5.464)
Non-Christian affiliation	1	1.256 (0.408–3.866)	1	2.558 (0.808–8.095)
Social ideology (ref: Conservative)
Liberal	1	0.149 (0.071–0.312)	1	0.304 (0.147–0.627)
Neither conservative nor liberal	1	0.196 (0.077–0.499)	1	0.136 (0.055–0.333)
Familiarity (ref: Don't know a transgender person)
Know a transgender person	1	0.509 (0.287–0.902)	1	0.596 (0.342–1.037)

Bolded cells indicate results with p-values <0.05.

A large majority of respondents agreed that transgender people have the right to surgically alter their face and vocal cords (84.6%), breasts/chest (86.2%), and genitals and reproductive organs (82.4%). Possessing a 4-year college degree, a liberal social ideology, and female sex were associated with support of GAS across anatomical sites. Younger respondents were more likely to support genital GAS. Non-Catholic Christians were less likely to support facial GAS (Table 4).

Table 4.

They should be allowed to have surgery to alter the following structures so their physical appearance matches their gender identity (n=312)

	Face and vocal cords		Breasts/chest		Genitals and reproductive organs
	Disagree	Agree	Disagree	Agree	Disagree	Agree
	RRR (95% CI)	RRR (95% CI)	RRR (95% CI)	RRR (95% CI)	RRR (95% CI)	RRR (95% CI)
Sex (ref: Male)
Female	1	2.308 (1.146–4.648)	1	2.409 (1.164–4.983)	1	2.349 (1.211–4.555)
Ages (ref: 50+)
18–34	1	1.652 (0.687–3.973)	1	0.932 (0.355–2.443)	1	2.586 (1.061–6.304)
35–49	1	2.357 (0.929–5.978)	1	1.265 (0.462–3.464)	1	1.562 (0.660–3.697)
Ethnicity (ref: Hispanic)
Non-Hispanic	1	0.619 (0.263–1.459)	1	0.515 (0.199–1.333)	1	0.778 (0.355–1.708)
Education (ref: less than a 4-year degree)
4-Year college degree	1	3.236 (1.344–7.789)	1	3.718 (1.506–9.178)	1	3.376 (1.473–7.736)
Advanced degree (e.g., MS, PhD, MD, JD)	1	2.609 (0.810–8.401)	1	1.613 (0.540–4.819)	1	2.040 (0.737–5.643)
Income (ref: $25,000 to $74,999)
< $25,000	1	0.562 (0.234–1.350)	1	0.678 (0.268–1.716)	1	0.858 (0.364–2.022)
> $75,000	1	0.755 (0.318–1.794)	1	0.814 (0.335–1.977)	1	1.050 (0.446–2.471)
Religion (ref: Non-religious)
Roman Catholic	1	0.261 (0.060–1.136)	1	0.323 (0.086–1.213)	1	0.928 (0.314–2.749)
Other Christian	1	0.204 (0.048–0.875)	1	0.338 (0.090–1.264)	1	0.833 (0.284–2.444)
Non-Christian affiliation	1	0.527 (0.087–3.193)	1	1.017 (0.157–6.591)	1	1.418 (0.330–6.094)
Social ideology (ref: Conservative)
Liberal	1	3.611 (1.676–7.781)	1	2.740 (1.264–5.940)	1	4.339 (2.093–8.995)
Neither conservative nor liberal	1	1.183 (0.477–2.932)	1	1.283 (0.480–3.430)	1	2.204 (0.874–5.557)
Familiarity (ref: Don't know a transgender person)
Know a transgender person	1	0.755 (0.389–1.545)	1	1.048 (0.515–2.130)	1	1.031 (0.537–1.978)

Bolded cells indicate results with p-values <0.05.

Two-thirds of respondents agreed that GAS is medically necessary (68.3%) and that it should be covered by private health insurers (76.0%), Medicaid (68.6%), Medicare (67.5%), and the VA (71.8%). Possessing a 4-year college degree was the strongest predictor of support for coverage of GAS across insurers. Female sex, younger age, liberal social ideology, and lower income were associated with support of coverage by some but not all insurers. Non-Hispanic respondents were less likely to support GAS coverage under Medicare or VA insurance (Table 5).

Table 5.

Surgery to change the physical appearance of transgender people so that it matches their internal gender identity should be covered by the following health insurers (n=312)

	Private health insurers		Medicaid for low-income Americans		Medicare for senior Americans (65 years and older)a		U.S. Department of Veterans Affairs for active and retired military personnel
	Disagree	Agree	Disagree	Agree	Disagree	Agree	Disagree	Agree
	RRR (95% CI)	RRR (95% CI)	RRR (95% CI)	RRR (95% CI)	RRR (95% CI)	RRR (95% CI)	RRR (95% CI)	RRR (95% CI)
Sex (ref: Male)
Female	1	1.924 (1.079–3.341)	1	1.591 (0.936–2.704)	1	1.481 (0.874–2.509)	1	1.772 (1.018–3.084)
Ages (ref: 50+)
18–34	1	2.438 (1.119–0.5310)	1	1.731 (0.818–3.663)	1	2.583 (1.232–5.413)	1	1.534 (0.686–3.427)
35–49	1	2.398 (1.091–5.268)	1	1.316 (0.624–2.775)	1	1.400 (0.677–2.895)	1	0.818 (0.374–1.788)
Ethnicity (ref: Hispanic)
Non-Hispanic	1	0.526 (0.252–1.097)	1	0.644 (0.336–1.236)	1	0.460 (0.234–0.906)	1	0.484 (0.238–0.985)
Education (ref: less than a 4-year degree)
4-Year college degree	1	2.809 (1.355–5.825)	1	2.973 (1.503–5.884)	1	2.213 (1.117–4.383)	1	2.425 (1.186–4.958)
Advanced degree (e.g., MS, PhD, MD, JD)	1	2.107 (0.830–5.350)	1	2.358 (0.987–5.630)	1	2.557 (1.048–6.237)	1	3.114 (1.208–8.026)
Income (ref: $25,000 to $74,999)
< $25,000	1	1.371 (0.628–2.990)	1	1.267 (0.620–2.588)	1	1.041 (0.518–2.093)	1	2.304 (1.020–5.205)
> $75,000	1	0.920 (0.434–1.953)	1	0.660 (0.332–1.310)	1	0.555 (0.279–1.103)	1	0.660 (0.327–1.333)
Religion (ref: Non-religious)
Roman Catholic	1	1.742 (0.703–4.313)	1	1.671 (0.733–3.810)	1	2.251 (0.999–5.071)	1	2.113 (0.902–4.954)
Other Christian	1	0.926 (0.377–2.273)	1	0.806 (0.351–1.848)	1	1.302 (0.573–2.962)	1	1.270 (0.535–3.013)
Non-Christian affiliation	1	1.199 (0.366–3.924)	1	0.901 (0.309–2.628)	1	1.549 (0.521–4.602)	1	1.374 (0.445–4.244)
Social ideology (ref: Conservative)
Liberal	1	2.367 (1.239–4.521)	1	1.332 (0.732–2.423)	1	1.709 (0.945–3.092)	1	1.467 (0.783–2.748)
Neither conservative nor liberal	1	1.049 (0.468–2.348)	1	0.541 (0.252–1.164)	1	0.986 (0.452–2.154)	1	0.569 (0.260–1.247)
Familiarity (ref: Don't know a transgender person)
Know a transgender person	1	1.217 (0.684–2.165)	1	1.199 (0.708–2.030)	1	1.432 (0.846–2.422)	1	1.447 (0.834–2.511)

Bolded cells indicate results with p-values <0.05.

^a n=1 respondent excluded from analysis due to missing response.

Prior awareness of gender-affirming care

There was strong prior awareness of gender-affirming hormone therapy (88.8%) and GAS (87.5%). Lower awareness was reported for individual GAS procedures. Among masculinizing surgeries, mastectomy (54.8%) and hysterectomy (47.4%) were the most commonly known by respondents (Fig. 3A). Among feminizing surgeries, breast augmentation (53.2%) and facial feminization (45.8%) were the most commonly known (Fig. 3B).

FIG. 3.

Awareness of specific GAS procedures: (A) masculinizing GAS; (B) feminizing GAS (n=312). GAS, gender-affirming surgery.

The most commonly cited sources of information to affect respondents' GAS opinions were social media platforms (43.9%), television (41.3%), a medical professional (40.7%), personal experiences (38.8%), and print media (37.5%) (Fig. 4).

FIG. 4.

Sources of information which affect opinion of GAS (n=312).

Discussion

Among a large, ideologically diverse online sample, we found American adults broadly supportive of GAS across patient population, surgery type, and insurance coverage. This study adds to the body of literature which indicates majority support for TGD rights among the American public, despite national campaigns to restrict these rights on multiple fronts.⁴¹ Support of GAS rights is strongest among those with higher educational attainment. However, we identify two vulnerabilities in public support of GAS: reduced rates of support for GAS in minors and a widely held belief that TGD people who have GAS will regret their decision.

Educational attainment was the strongest predictor of support of diverse GAS rights. This is consistent with prior research demonstrating that Americans with a 4-year college degree are more likely than those with less than a 4-year college degree to support LGBT nondiscrimination protections.⁴¹ We observed the strongest correlation between support of GAS rights among those with a 4-year college degree, but not those with professional degrees. This observation is likely due to issues of statistical power, not a decreased predilection to support GAS rights among the most educated.

Familiarity with TGD people has also been associated with greater support for TGD rights.²¹ Among respondents, personal familiarity with TGD people did not predict greater support for GAS among different patient groups, anatomical sites, or insurance coverage. Familiarity was, however, strongly associated with a decreased belief in post-GAS regret. More consistent predictors of GAS support were female sex and possessing a liberal social ideology.

Across social ideological lines, there was overwhelming support for TGD peoples' right to GAS (liberal 91.1%, conservative 79.2%, neither 87.8%). This support for GAS was independent of gender transition, surgery type, and insurance coverage, a novel finding that expands upon prior national polling demonstrating uniform opposition among Democrats, Republicans, and Independents against legislation that would prohibit gender-affirming care for minors.²⁰

Still, there exist two vulnerabilities in support for GAS rights within our data. First, we found less support across ideological lines of GAS for TGD minors (62.2%) than for TGD adults (84.3% for trans men and 83.3% for trans women). The appropriate provision of gender-affirming care to TGD minors is fraught with legal, ethical, and political considerations that go beyond those encountered in the care of TGD adults. This inherent complexity to pediatric gender-affirming care makes it vulnerable to political restrictions, including those bills introduced during the 2021 and 2022 legislative sessions which restrict TGD minors from accessing gender-affirming care.

Previous WPATH standards recommended that nonmastectomy GAS not be carried out until patients reach the legal age of majority, 18 years in the United States, except on a case-by-case basis.⁴ New standards are in line with those recommendations after preliminary proposals lowered the minimum age recommendations for GAS to 15 years for chest masculinization, 16 years for breast augmentation and facial surgery, and 17 years for metoidioplasty, orchiectomy, vaginoplasty, and hysterectomy.⁴² In light of relatively weak support for adolescent GAS and heightening political campaigns to ban gender-affirming care from minors, any future lowering of age barriers for GAS in professional guidelines may spur additional counteractions with the potential to adversely affect the provision of nonsurgical gender-affirming care.

A second vulnerability of public support of GAS is perceived postsurgical regret. In our survey population, despite overwhelming support for GAS rights, large majorities of respondents agreed that TGD adults and minors would regret GAS. Socially conservative and Christian respondents were significantly more likely to endorse this belief. On the surface, it is perplexing why individuals who support GAS access would simultaneously believe that TGD people would regret GAS. There is evidence, however, of a current of antipathy toward LGBT people in the United States that evades detection by most conventional pollsters.⁴³ This is evidenced by large percentages of Americans who worry about exposing their children to TGD people or allowing TGD people to use the restroom of their choice.²¹ A similar observation, perhaps, is that Americans are much more positive about lesbian and gay people than about the morality of homosexuality or homosexual sex specifically.⁴⁴ It is, thus, less surprising that the public would support TGD peoples' rights while simultaneously condemning their personal decisions to undergo medical care for gender affirmation. Countering the myth of post-GAS regret could be consequential in mitigating support for future restrictive actions.

Education outreach has the potential to bridge the remaining divide between public opinion of GAS and the current scientific medical consensus. Specific populations to be targeted include men, those without a college degree, social conservatives, and Christians. Considerable support for the right to undergo gender-affirming care even among these populations should be seen as an opportunity to further reduce stigma. One potential solution could be adapting the popular opinion leader (POL) intervention which has been used effectively in the goals of HIV transmission prevention and stigma mitigation.^45,46 In this intervention, community members are recruited and trained by health advocacy groups as POLs to educate their social networks and combat misinformation. Specific efforts should be made to promote the positive outcomes of guideline-directed gender-affirming care among TGD adolescents and highlight the low rate of post-GAS regret among adult TGD individuals.

Limitations

The demographics of our survey population align with previous MTurk studies, which under-represent racial and ethnic minorities, the less-educated, conservative, and older respondents relative to the U.S. adult population.⁴⁷ Our population was especially underrepresented in those with a less than a bachelor's degree: only 19.2% of our population compared to an estimated 65.2% of the adult population.³⁸ The extent of these differences precluded iterative proportional fitting.

Importantly, our population was ideologically diverse. Prior studies have shown political ideology, specifically separated into the domains of social and economic ideology, to be highly predictive of opinions across a wide array of topical issues.^48,49 Nearly identical divisions of political opinion and psychological disposition are observed between conservatives and liberals whether recruited from MTurk or the mass public.⁵⁰ The diversity of our respondents in these domains allows for greater generalizability of our results than from sociodemographic domains alone. Moreover, as we highlighted the importance of education and social liberalism in predicting diverse support for GAS rights, we suspect that our findings in a representative population would demonstrate decreased support for adolescent GAS and increased belief in post-GAS regret. Our findings of these vulnerabilities in sentiment toward TGD people would, thus, remain powerful.

Conclusion

This large online sample of American adults with diverse political ideologies demonstrates high GAS support across patient subpopulations, surgery types, and insurance coverages, although overall less support for GAS for TGD youth than adults. We demonstrate that there is a commonly held belief that TGD people who undergo GAS will regret their decision, despite evidence in the medical literature that post-GAS regret is a rarity. Providers of and community advocates for gender-affirming care should work to address gaps in public knowledge surrounding GAS and postsurgical regret.

Abbreviations Used

CI

confidence interval

GAS

gender-affirming surgery

HIT

human intelligence task

MTurk

Mechanical Turk

POL

popular opinion leader

RRR

relative risk ratio

TGD

transgender and gender-diverse

WPATH

World Professional Association for Transgender Health

Authors' Contributions

Conceptualization, J.S.B., K.P., G.B.S., J.M.S., A.K.S.-W.; Data curation, J.S.B.; Formal analysis, J.S.B.; Methodology, J.S.B., K.P., G.B.S., J.M.S., A.K.S.-W.; Project administration, A.K.S.-W.; Supervision, A.K.S.-W.; Visualization, J.S.B., K.P., A.K.S.-W.; Writing—original draft, J.S.B.; Writing—review and editing, K.P., G.B.S., J.M.S., A.K.S.-W.

Author Disclosure Statement

The authors have no conflicts of interest to disclose. J.M.S. is a consultant for 3M and is Co-Founder of Lifesprout, Inc.

Funding Information

No funding was received for this article.

Cite this article as: Burton JS, Pfeifauf K, Skolnick GB, Sacks JM, Snyder-Warwick AK (2023) Determinants of public opinion toward gender-affirming surgery in the United States, Transgender Health X:X, 1–13, DOI: 10.1089/trgh.2022.0119.