Kept in the Closet: Structural Stigma and the Timing of Sexual Minority Developmental Milestones Across 28 European Countries

Abstract

Structural stigma’s role in lesbian, gay, and bisexual (LGB) people’s attainment of identity development milestones remains unknown. In a sample of 111,498 LGB people (ages 15 to 65+) living across 28 European countries, associations were investigated between structural stigma measured using an objective index of discriminatory country-level laws and policies affecting LGB people and the timing and pacing of LGB self-awareness, coming out, and closet duration, and subgroup differences in these associations. On average, self-awareness occurred at age 14.8 years old (SD = 5.1), coming out occurred at 18.5 years old (SD = 5.7), and the closet was 3.9 years long (SD = 4.9); thereby highlighting adolescence as a key period for sexual identity development and disclosure. Greater structural stigma was associated with higher odds of never coming out, later age of coming out, and longer closet duration. Gender identity, transgender identity, and sexual identity moderated associations between structural stigma and these developmental milestones. Reducing structural stigma can plausibly promote sexual identity development among LGB populations, especially during adolescence when identity related milestones are often attained.

Introduction

For lesbian, gay, and bisexual (LGB) adolescents and adults, “the closet” refers to the developmental period between two major sexual minority developmental milestones: self-awareness of an LGB identity and first disclosure of this identity to someone else (Pachankis & Jackson, 2022). This developmental period, though variable in length, marks a nearly universal experience for sexual minority people during which they alone know of their sexual identity prior to disclosure to others. Although the closet has been long theorized to be a defining feature of sexual minority identity development (Cass, 1984), recent research has confirmed that (1) the closet tends to take place during the developmentally formative period of adolescence, at least in the United States (Bishop et al., 2020); (2) the closet is a reality in much of the world, including across Europe (Darakchi, 2023; Marhánková, 2021; Siraj, 2011; Torres & Rodrigues, 2022); (3) a majority of the adult population around the world remains in the closet for much of life (Pachankis & Bränström, 2019); and (4) the closet is associated with characteristic psychological adaptations (e.g., Pachankis & Hatzenbuehler, 2013). The current study builds upon this research by testing the hypothesis that the closet and its associated developmental milestones are shaped by the broader structural contexts in which LGB people live.

Structural stigma toward sexual minority people and communities refers to the geographically bound conditions of societal rejection of sexual minority people codified in discriminatory laws and policies (e.g., same-sex marriage bans, lack of employment nondiscrimination protections, absence of conversion therapy bans) and manifested in general population anti-sexual minority attitudes (Hatzenbuehler, 2016). Structural stigma toward sexual minority people is highly variable across geographies—whether at the level of states and municipalities (Everett et al., 2022; Perales & Todd, 2018) or world countries (Pachankis & Bränström, 2018; van der Star, Pachankis, et al., 2021)—with this variation shown to be associated with the mental health and well-being of sexual minority people living therein (Pachankis et al., 2021; Raifman et al., 2017). Two theoretical perspectives suggest potential pathways through which structural stigma may impact sexual minority people. First, originating in the US (Brooks, 1981) and later validated in European samples (Sattler & Lemke, 2019), minority stress theory posits that sexual minority people’s experiences of unique and chronic stressors—due the devaluation and stigmatization of their minoritized sexual identities—creates a hostile, stressful context that contributes to adverse consequences for health and well-being (Meyer, 2003). Structural stigma represents a key instantiation of minority stress (Hatzenbuehler, 2016; Meyer, 2003b). Minority stress theory has been further adapted to specify the ways in which the social environment can shape sexual minority identity development during adolescence (Goldbach & Gibbs, 2017). Second, legal socialization theory explains how law and policy guide individual behavior, interactions, and beliefs across the life course and diverse contexts (Trinkner & Reisig, 2021). Adolescence is a key period for legal socialization because this is when individuals often interact with laws for the first time and because adolescence is a period for formation of values and moral beliefs (Tyler & Trinkner, 2017). For sexual minority people, awareness of one’s own sexual identity occurs, on average, during adolescence (Bishop et al., 2020), the developmental period during which initial exposure to sexual minority laws and policies may also occur. Therefore, structural stigma toward sexual minority people may shape identity development, especially during adolescence.

Increasingly, researchers have provided evidence of the ways in which structural stigma shapes the everyday experiences of sexual minority people. For instance, sexual minority people living in US states or municipalities characterized as high in structural stigma (often operationalized as a comparatively higher number of laws and policies that deny or fail to protect the equal rights of the stigmatized) are more likely to report high motivation to conceal their sexual identity (Lattanner et al., 2021), and sexual minority people living in high structural stigma countries are more likely to report social isolation and internalized stigma (Pachankis et al., 2021). Relatedly, in a community study of sexual minorities in Italy, an unsuccessful attempt to pass legislation protecting sexual minority people from discrimination was associated with decreased sexual minority well-being (Rucco et al., 2023). Further still, sexual identity development may be dependent on political and cultural circumstances as a function of access to sexual minority community (Kowalska, 2011). However, whether structural stigma predicts sexual minority identity developmental experiences remains empirically untested.

Sexual minority people face at least two major developmental milestones related to the formation of their LGB identity—self-awareness as LGB and first disclosure of that identity to another person (i.e., “coming out”). The period between these events is highly variable, whereby for some people the two events may happen simultaneously and for others the time between the two events might last years (Bishop et al., 2020; Calzo et al., 2011) or even decades (Savin-Williams & Diamond, 2000). Still other sexual minority people might never disclose their sexual identity to another person (Pachankis et al., 2015), with some rejecting altogether the imperative to come out (McLean, 2007). Despite this diversity in sexual identity development, the timing and pacing of identity development processes and event, like self-awareness and coming out, are likely to vary based on the social environment in which LGB adolescents and adults live. Indeed, as noted above, psychological theories of minority stress (Meyer, 2003) and legal socialization (Tyler & Trinkner, 2017) position identity developmental processes as a function of the social environment (Goldbach & Gibbs, 2017; Pachankis & Jackson, 2022), but most empirical evidence regarding LGB identity development focuses on interpersonal influences, such as parental acceptance or rejection (D’Augelli et al., 1998).

Self-awareness and disclosure milestones frequently (Hall et al., 2021) and increasingly occur in adolescence (Darakchi, 2023; Russell & Fish, 2016), though little research has addressed how a broader structural context of stigma might impact adolescent sexual identity development. Institutionalization of anti-sexual minority cultural norms into structural stigma creates an environment wherein sexual minority adolescents may be unable to see how their sexual identities are shared with others, fit into society, and have cultural value, thus leaving them to feel like “second-class citizens” (Santaemilia, 2009). Indeed, growing up in countries with higher structural stigma is associated with more internalized homophobia (van der Star, Bränström, et al., 2021), and greater internalized homophobia is associated with less sexual identity disclosure (Sommantico et al., 2018). However, research investigating structural influences on identity-related outcomes has been limited to studies finding that structural stigma predicts identity concealment (i.e., gradient of disclosure or concealment after coming out; Pachankis & Jackson, 2022) among sexual minority adults. Consequently, further research is needed to understand how structural stigma might shape identity developmental processes such as the timing and pacing of identity developmental milestones known to be associated with sexual minority individuals’ health and life satisfaction (Russell & Fish, 2019; Vaughan & Waehler, 2010).

Although structural stigma is theorized to broadly influence the health and well-being of all members of a stigmatized population (Hatzenbuehler, 2016), whether structural stigma influences identity-related processes similarly across all LGB subgroups remains relatively untested. At the same time, emerging research suggests that some subgroups of LGB people (e.g., gender non-conforming sexual minority men) might be more likely than others to experience adverse interpersonal experiences, such as victimization, in structurally stigmatizing contexts (Bränström et al., 2023), with potential implications for structural stigma’s association with sexual identity developmental milestones. Further, some LGB subgroups may be further impacted by additional sources of structural stigma such as structural transphobia characterized by laws and policies that fail to protect the rights of transgender people (Price et al., 2023). In addition, past studies have demonstrated differences in developmental milestones across gender and sexual identities. For instance, predominantly US samples of sexual minority adolescents and adults find that men become self-aware earlier but come out later than women (Bishop et al., 2020; Martos et al., 2015; Savin-Williams & Diamond, 2000). Regarding sexual identity differences in development, evidence from US national (Bishop et al., 2020), state (Calzo et al., 2011), and community studies (Martos et al., 2015) suggests that gay and lesbian adolescents and adults become aware of and disclose their sexual identity earlier than bisexual adolescents and adults. Researchers have proposed that these differences may be due to (1) biphobia from both heterosexual and sexual minority communities that delays sexual minority development (Bishop et al., 2020), (2) attraction to multiple genders causing bisexual young people to take longer to process and integrate their attractions and identity (Calzo et al., 2011), (3) opposite-gender attraction delaying recognition or importance of same-gender attraction (Martos et al., 2015), or (4) lower levels of sexual identity centrality and greater identity uncertainty for bisexual people (Dyar et al., 2015).

Developmental differences between transgender and cisgender sexual minority people are less clear and more seldomly studied than differences by gender and, to a lesser extent, sexual identity. In terms of transgender identities, qualitative research has highlighted the complex intersection of gender and sexual identity development and intersecting cultural and structural forces impacting this development (Hereth et al., 2020). In population-based surveys, many transgender adults report sexual minority identities, for instance 42% in Sweden (Bränström et al., 2022) and 83% in the United States (Carone et al., 2021). However, LGB identity milestones among transgender people remain poorly understood because quantitative studies examining sexual identity milestones have often excluded transgender respondents (Bishop et al., 2020; Martos et al., 2015). Overall, although research has established a link between some of the above identity subgroups and elements of LGB identity development, the role of structural stigma in identity development differences across LGB subgroups has not been examined.

Current Study

Whether structural stigma is associated with sexual minority identity development remains unknown. Therefore, this study addresses associations between structural stigma and identity-related developmental milestones, namely age of awareness as a sexual minority person, age of coming out, and duration of the closet among sexual minority adolescents and adults in Europe. First, differences were tested in the timing of these major developmental milestones across groups (i.e., gender identity, transgender identity, sexual identity). Differences were expected by gender identity (i.e., men becoming aware of their identity earlier but disclosing later than women) and sexual identity (i.e., gay and lesbian respondents becoming aware of their identity earlier than bisexual respondents). Differences by transgender identity were considered exploratory, and thus direction of any differences was not hypothesized. Second, it was investigated whether structural stigma was associated with whether one had come out, age of awareness and age of coming out, and the time between these two milestones (i.e., the closet). Specifically, the association was tested between structural stigma and odds of never coming out of the closet, and it was expected that the odds of never coming out would be higher in countries with high structural stigma. The association was then tested between structural stigma and age of awareness, age of coming out, and duration of the closet. It was expected that structural stigma would be associated with older age of awareness and coming out and a longer period of time in the closet (i.e., between first awareness and first coming out). Finally, gender identity, transgender identity, and sexual identity were explored as moderators of the associations between structural stigma and age of awareness, age of coming out, and duration of the closet. Because this aim is exploratory, directional hypotheses were not posited.

Methods

The European Union Lesbian, Gay, Bisexual, Transgender, and Intersex (EU-LGBTI) Survey was administered online between May and July 2019 by the European Agency for Fundamental Rights to monitor the fundamental rights affecting lesbian, gay, bisexual, transgender, and intersex (LGBTI) people across the 28 European Union (EU) member states in 2019 (i.e., Austria, Belgium, Bulgaria, Croatia, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, The Netherlands, Poland, Portugal, Romania, Slovakia, Slovenia, Spain, Sweden, and the United Kingdom). The survey was developed by a cross-European team of LGBTI topic experts and translated into 31 languages using forward translation of an English-language version, followed by back translation (European Union Agency for Fundamental Rights, 2020).

Eligibility criteria included: (1) residing in one of the 28 European countries in 2019, (2) being at least 15 years of age, and (3) identifying as lesbian, gay, bisexual, transgender, or intersex. Eligible respondents indicated understanding the study purpose and provided consent. Online recruitment occurred via newspapers, LGBTI organizations, online networks, and other types of social media targeted at LGBTI individuals and communities. Interested respondents were directed to a web-based survey located on a secure server. Additional details of survey development and methods are described elsewhere (European Union Agency for Fundamental Rights, 2020).

In total, 137,508 respondents across the 28 countries responded to the survey. To strengthen the ability to attribute variation in developmental milestones to variations in structural stigma, the sample was restricted to respondents who reported no migration between countries, thus excluding respondents with a history of migration (n=19,345; 14.1%). This approach rules out the possibility of differential migration by developmental milestone status (e.g., whereby people who disclosed at earlier ages might be more likely to move away from high structural stigma countries) and follows the approach used in previous studies of structural stigma (e.g., Pachankis et al., 2021). Analyses also excluded those who did not identify as lesbian, gay, or bisexual (n=1,637; 1.2%; e.g., heterosexual transgender respondents), and those who did not provide any developmental milestone data (n=5,028; 3.9%), resulting in a final analytic sample of 111,498 bisexual, gay, and lesbian respondents. The per-country sexual minority subsample size varied from 180 respondents in Luxembourg to 17,284 respondents in Spain.

Compared to the excluded respondents, the included sample was older and had a higher proportion of cisgender men and women than trans or non-binary respondents. Included respondents, on average, lived in countries with higher structural stigma and became aware of their identity as gay, lesbian, or bisexual at older ages.

Individual-Level Measures

Sexual and gender identity.

Respondents indicated their sexual identity from the options “lesbian,” “gay,’ “bisexual,” “heterosexual/straight,” and “other, please specify.” Respondents also indicated whether they identified as transgender, with transgender defined broadly (e.g., trans woman, trans man, non-binary, gender variant, polygender, agender, gender-fluid). Heterosexual transgender and intersex respondents were excluded from the current analyses because they were not asked questions about sexual minority identity development. The analytical variable for sexual identity included three groups based on three sexual identity categories eligible for inclusion in the survey (lesbian [referent], gay, bisexual).

Respondents reported their gender identity from the options “woman/girl,” “man/boy,” “trans woman/girl,” “trans man/boy, “non-binary or genderqueer or agender or polygender or gender-fluid,” and “other, do not identify with any of the named groups.” These categories were collapsed to create three gender categories for analyses: man (transgender and cisgender men), woman (transgender and cisgender women, [referent]), and non-binary or other gender identity.

A binary variable was created to compare transgender identity (“Are/were you a trans person?” selected “yes”, and/or “How would you describe yourself today?” selected “trans woman/girl,” “trans man/boy”, or “non-binary or genderqueer or agender or polygender or gender-fluid”) to cisgender identity (referent).

To make comparisons across sexual identity groups, sexual identity was combined with gender to create five gendered sexual identity categories: gay respondents, lesbian respondents, bisexual man respondents, bisexual woman respondents, and bisexual respondents with non-binary or other gender identities. Bisexual groups were fully stratified by gender; gay and lesbian groups were mixed-gender due to the inclusion of gay-identified and lesbian-identified non-binary respondents. Both gay and lesbian subgroups were not further stratified by gender due to limitations in gendered subgroup sample sizes (e.g., 97.7% of gay respondents were men and 92.0% of lesbian respondents were women).

Age.

Respondents reported age in years, which was then collapsed into bins to protect respondent confidentiality. Age bins were 15–17, 18–24, and continued in five-year increments until age 65 and older, for a total of 11 age bins across the lifespan. Age bins were ordered from youngest to oldest and treated as a continuous, mean-centered variable.

Sexual identity developmental milestones.

Respondents reported the timing of two sexual identity-related developmental milestones: (1) how old they were when they first became aware that they were lesbian, gay, or bisexual (age of awareness; “How old were you when you realized for the first time you are [lesbian/gay/bisexual]?”) and (2) how old they were when they first disclosed that they were lesbian, gay, or bisexual (age of coming out; “How old were you when you first told somebody you are [lesbian/gay/bisexual]?”). Respondents provided age in years or, if they had never disclosed their identity, they could select the option, “I have not told anybody.” In addition to age of awareness and age of coming out, two additional variables were calculated from responses to these questions. First, a binary variable for coming out indicated whether a respondent had never disclosed their sexual identity to anyone else (ever disclosed versus never disclosed). Second, for respondents who reported having ever disclosed their sexual identity to someone else, the duration of the closet was calculated as the number of years between age of awareness and age of coming out.

Country-Level Measures

Structural stigma.

Following prior research on LGB structural stigma (e.g., Pachankis et al., 2017), an index of LGB equal rights and policies was utilized to reflect country-level structural stigma toward sexual minority people and communities. This structural stigma index was created using the 2019 Rainbow Index from the International Lesbian, Gay, Bisexual, Trans and Intersex Association in Europe (ILGA Europe, 2019), which summarizes the state of LGBT-related laws and policies across 51 European countries.

The structural stigma index was created by summing across four domains of laws and policies (i.e., equality and non-discrimination, family, hate crime and speech, and asylum) within each country. Example laws and policies indexed in the current study include, among others, discrimination protections for sexual minority people in employment, education, and healthcare; bans on conversion therapy; marriage and partnership equality policies; parental rights and adoption protections; enumerated hate crime policies; and protections for sexual minority asylum seekers. Across law and policy domains, each country could be assigned up to 21 points for these distinct supportive protections. Each form of protection was assigned a single point except for same-sex marriage which was assigned a 2-point value to distinguish it from the more limited rights conferred by legal registered partnerships or legal cohabitation, which was assigned a 1-point value. In countries in which some but not all regions provided protections, a half point was awarded (e.g., regional conversion therapy bans in Spain). Each country’s score on this index was multiplied by −1 to convert to a measure of structural stigma (rather than support) and then standardized to facilitate interpretation in terms of standard deviation units. A summary of country-level scores is provided in Table 1.

Table 1.

Raw and Standardized Country-level Structural stigma Indices

	Structural Stigma Index
	Raw Sum	Standardized Score
Austria	14.5	−0.52
Belgium	19.0	−1.39
Bulgaria	6.0	1.12
Croatia	10.0	0.35
Cyprus	7.0	0.93
Czech Republic	7.0	0.93
Denmark	17.0	−1.00
Estonia	9.0	0.54
Finland	18.0	−1.19
France	15.0	−0.62
Germany	13.0	−0.23
Greece	8.0	0.73
Hungary	10.0	0.35
Ireland	14.0	−0.42
Italy	3.0	1.70
Latvia	4.0	1.50
Lithuania	8.0	0.73
Luxembourg	17.0	−1.00
Malta	19.0	−1.39
Netherlands	15.0	−0.62
Poland	3.0	1.70
Portugal	16.0	−0.81
Romania	7.0	0.93
Slovakia	8.0	0.73
Slovenia	11.0	0.15
Spain	16.0	−0.81
Sweden	16.0	−0.81
United Kingdom	20.0	−1.58

Income inequality.

Given its strong association with country-level structural stigma in prior research (Andersen & Fetner, 2008; Pachankis et al., 2015), country-level income inequality, operationalized as each country’s 2018 Gini coefficient (2017 for the United Kingdom based on available data; The World Bank, 2023), was used as a covariate. Higher values indicate greater income inequality.

Analytical Plan

Analyses were conducted in four parts. First, differences were investigated in developmental timing (age of awareness, age of coming out) and closet duration by gender identity, transgender identity, and sexual identity. Second, multilevel logistic regression was used to test the association between country-level structural stigma and odds of having never disclosed one’s sexual identity. Third, among respondents who reported ever having disclosed, multilevel linear regression was used to investigate the association between country-level structural stigma and (1) age of first awareness, (2) age of first coming out and (3) duration of the closet. Finally, gender identity, transgender identity, and sexual identity were investigated as moderators of the associations between country-level structural stigma and age of awareness, age of coming out, and closet duration.

To compare differences in mean developmental timing and closet duration across identity subgroups, t-tests were used for binary variables (transgender identity) and analysis of variance (ANOVA) was used for categorical variables (gender identity, sexual identity). Welch’s t-tests were used when variance between subgroups was unequal. Differences in mean age of awareness were compared in the full analytical sample. Differences in mean age of coming out and closet duration were compared in the subsample of respondents who reported ever having disclosed their sexual identity to someone else. To guard against false discovery in light of the number of analyses performed in this stage, a Benjamini-Hochberg correction was applied (Benjamini & Hochberg, 1995).

To test the association of structural stigma with odds of never having disclosed one’s sexual identity, multilevel logistic regression was used to account for the nested data structure (i.e., respondents within countries). Level 1 included individual characteristics (i.e., gender identity, transgender identity, sexual identity, and age), and Level 2 included country-level variables (i.e., Gini coefficient and structural stigma index). Fixed effects were estimated using the restricted maximum likelihood parameter estimation.

Among respondents who had ever disclosed, two-level linear regression models were fit predicting the association between country-level structural stigma and mean age of awareness of being LGB, mean age of coming out, and mean duration of the closet, using the same Level 1 and Level 2 variables as above. Fixed effects were estimated using the restricted maximum likelihood parameter estimation. To investigate whether the association of structural stigma and age of awareness, age of coming out, and closet duration differed as a function of identity subgroups, cross-level interactions were tested between structural stigma and identity subgroups (i.e., gender identity, transgender identity, sexual identity) in predicting each developmental milestone. To further probe moderation effects, simple slopes analyses were conducted to estimate the effect of structural stigma at each level of every identity subgroup for any interactions that were significant in moderation models. Simple slopes analyses were adjusted for age and Gini coefficient but no other identity subgroup variables. Simple slopes results were tested and reported for the range of actual structural stigma values in the current sample, from −1.58 (United Kingdom) to 1.70 (Italy and Poland).

Analyses were conducted in R version 4.2.2 (R Core Team, 2021) using lme4 (Bates et al., 2015) and lmerTest (Kuznetsova et al., 2017) packages and used maximum likelihood estimation. All regression models were adjusted for age.

Results

Table 2 presents sociodemographic characteristics of the full analytic sample. The sample included more men (52.9%) than women (41.5%) and non-binary (5.6%) respondents and was majority cisgender respondents (89.7%). Respondents were gay (44.8%), lesbian (20.3%), or bisexual (34.9%). Nearly two-thirds of bisexual respondents were women (65.4%). Most respondents were under age 30 (66.3%) and were in the ethnic majority of their country (95.8%). For the country-level structural stigma index, standardized scores ranged from −1.58 in the United Kingdom to 1.70 in Italy and Poland (Table 1).

Table 2.

Sample Demographic Frequencies

	n	%
Gender Identity 1
Man	59032	52.9
Woman	46273	41.5
Non-Binary+	6193	5.6
Transgender Identity
Transgender	11512	10.3
Cisgender	99986	89.7
Sexual Identity 2
Lesbian	22583	20.3
Gay	49955	44.8
Bisexual	38860	34.9
Woman	25375	22.8
Man	10176	9.1
Non-Binary+	3322	3.0
Ethnicity
Ethnic Minority	4676	4.2
Ethnic Majority	106822	95.8
Age
15–17	16171	14.5
18–29	57867	51.9
30–39	18630	16.7
40–49	10818	9.7
50–59	5979	5.4
60 or older	2032	1.8

Note. n = 111,498. Non-binary+ = non-binary and other identities.

¹Gender identity categories man and woman are inclusive of cisgender and transgender respondents.

²“Heterosexual/Straight” and all other sexual identities were excluded from the sexual minority subsample based on parent study eligibility.

On average, respondents first realized they were lesbian, gay, or bisexual at age 14.8 (SD = 5.1). Most respondents (93.9%; n = 104,692) reported having disclosed their sexual identity (i.e., come out) to someone else. Among those who disclosed, the average age of coming out was 18.5 (SD = 5.7) years old, and the average duration of the closet (i.e., time between awareness and first coming out) was 3.9 (SD = 4.9) years.

Do Identity Subgroups Differ in Age of Awareness, Never Coming Out, and Age of Coming Out?

Differences across subgroups in developmental timing and closet duration are summarized in Table 3; differences of at least six months (0.5 years) are summarized in text. Men (M = 14.1) and non-binary respondents (M = 14.3) were, on average, younger at age of awareness than women (M = 15.7). Gay respondents (M = 13.7) were, on average, the youngest at age of awareness, followed by bisexual non-binary respondents (M = 14.9), and then bisexual women (M = 15.5), bisexual men (M = 15.6), and lesbian respondents (M = 15.9). Differences in age of awareness were less than six months or non-significant for transgender compared to cisgender respondents.

Table 3.

Differences in Developmental Milestone Timing and Pacing by Sexual and Gender Identity Subgroups

	Age of Awareness				Age of Disclosure				Closet Duration
	n = 111,498				n = 104,692				n = 104,692
	M	SD	Test Statistic	p	M	SD	Test Statistic	p	M	SD	Test Statistic	p
Gender
Woman	15.7	5.3	1381.0	<.0001	18.1	5.5	573.0	<.0001	2.4	4.0	3935.0	<.0001
Man	14.1	4.9			19.0	5.8			5.1	5.3
Non-Binary+	14.3	5.3			16.9	5.2			2.8	4.1
Transgender			6.7ϯ	<.0001			10.8ϯ	<.0001			18.8	<.0001
Transgender	15.1	6.5			17.8	6.8			3.0	5.9
Cisgender	14.7	4.9			18.6	5.5			4.0	4.9
Sexual Identity			998.1	<.0001			361.6	<.0001			2328.0	<.0001
Lesbian	15.8	5.9			18.7	6.0			2.9	4.6
Gay	13.7	4.7			19.0	5.6			5.3	5.3
Bisexual	15.5	5.0			17.7	5.4			2.5	4.0
Woman	15.5	4.7			17.4	4.9			2.0	3.3
Man	15.6	5.6			18.8	6.6			3.6	5.2
Non-Binary+	14.9	5.3			17.1	5.4			2.5	4.0

Note. M = mean. SD = standard deviation. Non-binary+ = non-binary and other identities.

^ϯIndicates Welch’s t-test conducted to account for unequal variance between groups. Test statistics are F-values for ANOVA (three or more groups) or t-values for t-tests (two groups). All significance tests were adjusted for multiple tests of comparison using the Benjamini-Hochberg correction.

Respondents who ever came out as LGB were, on average, younger at age of awareness (M = 14.7) than respondents who had never come out (M = 16.2). Compared to the out group, the group who was not out included more adolescents (i.e., age 15–17; non-disclosure = 25.1%; disclosure = 13.7%), and more bisexual men (25.3%; 8.1%) and bisexual women (31.9%; 22.2%). Indeed, among adolescents 10.6% were not out, and more bisexual men (17.0% and bisexual women (8.6%) were not out compared to gay (4.0%) and lesbian (3.0%) respondents.

Among the subsample who ever came out, there were 6-month or greater differences in mean age of coming out by gender identity, transgender identity, and sexual identity. Non-binary respondents disclosed their sexual identity at the youngest age (M = 16.9), followed by women (M = 18.1) and then men (M = 19.0). Transgender respondents disclosed their sexual identity earlier (M = 17.8) than cisgender respondents (M = 18.6). Bisexual women and bisexual non-binary respondents disclosed at younger ages (M = 17.4; M = 17.1) than lesbian respondents (M = 18.7), bisexual men (M = 18.8), and gay respondents (M = 19.0).

Does Closet Duration Differ by Identity Subgroups?

Among the subsample who ever came out, there were 6-month or greater differences in closet duration by gender identity, transgender identity, and sexual identity, as shown in Table 3. Closet duration was shorter for women (M = 2.4) and non-binary respondents (M = 2.8) than for men (M = 5.1), and shorter for transgender respondents (M = 3.0) than for cisgender respondents (M = 4.0). Across sexual identity groups, closet duration was shortest for bisexual women (M = 2.0) and longest for gay respondents (M = 5.3).

Is Country-Level Structural Stigma Associated with Whether People Come Out?

For every standard deviation increase in country-level structural stigma, odds of never having come out as LGB were 1.33 times higher, when accounting for individual characteristics and country-level income inequality, as shown in Table 4. Of the total variance in odds of never coming out, 4.2% was attributable to between-country variation.

Table 4.

Adjusted Associations of Country-Level Structural Stigma with Never Coming Out, Age of Awareness, Age of Coming Out, and Duration of The Closet

	Never Out				Age of Awareness				Age of Coming Out			Duration of The Closet
	n = 111,498				n = 104,692				n = 104,416			n = 104,416
	β	SE	OR	p	β	SE	p		β	SE	p	β	SE	p
Intercept	−3.97	0.35	0.02	<.0001	17.35	0.53	<.0001		18.71	0.49	<.0001	1.38	0.64	0.0416
Individual Level Factors
Gender Identity
Woman (referent)
Man	0.74	0.04	2.09	<.0001	−0.73	0.06	<.0001		0.32	0.06	<.0001	1.04	0.05	<.0001
Non-Binary+	−0.38	0.07	0.69	<.0001	−1.71	0.09	<.0001		−1.36	0.09	<.0001	0.35	0.08	<.0001
Transgender Identity
Cisgender (referent)
Transgender	0.34	0.05	1.41	<.0001	0.72	0.07	<.0001		0.72	0.07	<.0001	0.03	0.06	0.6805
Sexual Identity
Lesbian (referent)
Gay	−0.45	0.06	0.64	<.0001	−1.96	0.07	<.0001		−1.12	0.07	<.0001	0.85	0.06	<.0001
Bisexual	1.13	0.05	3.11	<.0001	0.25	0.04	<.0001		0.29	0.04	<.0001	0.06	0.04	0.1549
Age	0.05	0.01	1.05	<.0001	0.64	0.01	<.0001		1.47	0.01	<.001	0.84	0.01	<.0001
Country Level Factors
Gini	0.02	0.01	1.02	0.1300	−0.05	0.02	0.0104		0.00	0.02	0.8486	0.05	0.02	0.0249
Structural Stigma	0.29	0.05	1.33	<.0001	0.03	0.06	0.6493		0.21	0.06	0.0019	0.19	0.0	0.0283

Note. SE = standard error. OR = odds ratio. Non-binary+ = non-binary and other identities (agender, polygender, gender-fluid, genderqueer).

Is Country-Level Structural Stigma Associated with Timing of Developmental Milestones or Closet Duration?

Among respondents who ever came out, country-level structural stigma was not associated with age of awareness but was associated with later age of first coming out, when accounting for individual and country-level covariates. For every standard deviation increase in country-level structural stigma, age of coming out was 2.5 months later and closet duration was 2.3 months longer, as shown in Table 4. Of the total variance in age of disclosure and duration of the closet, 2.5% and 1.7% were attributable to between-country variation, respectively.

Does the Association Between Structural Stigma and Age of Awareness Vary as a Function of Identity Subgroups?

Associations of country-level structural stigma with age of awareness varied as a function of gender identity, transgender identity, and sexual identity as shown by significant subgroup*structural stigma interactions in Table 5. Further probing with simple slopes analyses revealed a significant association of structural stigma with age of awareness for non-binary respondents, transgender respondents, and lesbian and gay respondents as shown in Figure 1 in the first panel. Non-binary respondents became aware of their LGB identity an estimated 7.6 months earlier in the highest (Poland and Italy, structural stigma = 1.70) compared to the lowest (United Kingdom, structural stigma = −1.58) structural stigma countries. Transgender respondents became aware of their LGB identity an estimated 1.3 years (15.7 months) earlier in the highest compared to lowest stigma countries. Lesbian respondents became aware of their LGB identity an estimated 5.8 months earlier in the highest compared to lowest stigma countries. In contrast, gay respondents became aware of their LGB identity an estimated 7.7 months later in highest compared to lowest stigma countries. Effect estimates for all simple slopes analysis are found in Table 6.

Table 5.

Association of Structural Stigma with Age of Awareness, Age of Coming Out, and Duration of The Closet Moderated by Identity Subgroups

	Age of Awareness			Age of Coming Out			Duration of the Closet
	n = 104,692			n = 104,692			n = 104,692
	β	SE	p	β	SE	p	β	SE	p
Intercept	17.35	0.55	<.0001	18.72	0.53	<.0001	1.39	0.64	0.0403
Individual Level Factors
Gender Identity
Woman (referent)
Man	−0.73	0.06	<.0001	0.33	0.06	<.0001	1.05	0.05	<.0001
Non-Binary+	−1.60	0.09	<.0001	−1.18	0.09	<.0001	0.41	0.09	<.0001
Transgender Identity
Cisgender (referent)
Transgender	0.57	0.07	<.0001	0.52	0.07	<.0001	−0.03	0.07	0.6637
Sexual Identity
Lesbian (referent)
Gay	−1.95	0.07	<.0001	−1.12	0.07	<.0001	0.84	0.06	<.0001
Bisexual	0.26	0.04	<.0001	0.30	0.04	<.0001	0.06	0.04	0.1417
Age	0.65	0.01	<.0001	1.48	0.01	<.0001	0.84	0.01	<.0001
Country Level Factors
Gini	−0.05	0.02	0.0132	0.00	0.02	0.8934	0.05	0.02	0.0257
Structural Stigma (SS)	−0.12	0.07	0.1155	−0.01	0.07	0.9364	0.12	0.08	0.1592
Moderation
Man*SS	0.07	0.05	0.1879	0.20	0.05	0.0002	0.13	0.05	0.0085
Non-Binary+*SS	0.29	0.09	0.0012	0.51	0.09	<.0001	0.24	0.08	0.0039
Transgender*SS	−0.43	0.07	<.0001	−0.54	0.07	<.0001	−0.13	0.06	0.0405
Gay*SS	0.26	0.06	<.0001	0.30	0.06	<.0001	0.03	0.06	0.6125
Bisexual*SS	0.06	0.04	0.1429	0.01	0.04	0.7722	−0.05	0.04	0.1443

Note. SE = Standard Error. Non-binary+ = non-binary and other gender identities (agender, polygender, gender-fluid, genderqueer). SS = Structural Stigma. Reference is cisgender lesbian woman.

Figure 1. Simple Slopes Analysis Plotting Effect of Country-Level Structural Stigma Across Identity Subgroups.

Dashed lines indicate non-significant slope at p<.05. Plotted estimates based on models adjusted only for Gini coefficient and age. Non-significant interactions from full models in Table 4 were not included in simple slopes analysis. Axes range for age of awareness, age of coming out, and duration of closet in years are based on mean +/−1 standard deviation.

Table 6.

Simple Slopes Analysis for Effect of Structural Stigma across Identity Subgroups

	Age of Awareness			Age of Coming Out			Closet Duration
	β	SE	p	β	SE	p	β	SE	p
Gender Identity
Woman	−0.12	0.07	0.0931	−0.03	0.06	0.5983	0.09	0.08	0.2802
Man	0.10	0.07	0.1674	0.38	0.06	<.0001	0.29	0.08	0.0018
Non-binary+	−0.19	0.09	0.0294	0.03	0.08	0.7054	0.23	0.10	0.0216
Transgender Identity
Cisgender	−0.02	0.07	0.8179	0.22	0.06	0.0007	0.25	0.09	0.0084
Transgender	−0.40	0.08	<.0001	−0.22	0.07	0.0028	0.17	0.09	0.0735
Sexual Identity
Lesbian	−0.15	0.07	0.0468	−0.03	0.07	0.6551	0.13	0.08	0.1440
Gay	0.20	0.07	0.0080	0.48	0.07	<.0001	0.29	0.08	0.0016
Bisexual	−0.08	0.07	0.2899	0.00	0.07	0.9715	0.08	0.08	0.3479

Note. Estimates adjusted for age and Gini coefficient.

Does the Association Between Structural Stigma and Age of Coming Out Vary as a Function of Identity Subgroups?

Associations of country-level structural stigma with age of coming out varied as a function of gender identity, transgender identity, and sexual identity, as shown by significant subgroup*structural stigma interactions in Table 5. Further probing with simple slopes analyses revealed a significant association of structural stigma with age of coming out for men, cisgender and transgender respondents, and gay respondents as shown in Figure 1 in the second panel. Men became came out an estimated 1.2 years (14.9 months) later in the highest compared to the lowest structural stigma countries. For cisgender and transgender respondents, the effects of structural stigma on age of coming out were in opposite directions. Cisgender respondents came out an estimated 8.8 months later in the highest compared to lowest stigma countries. In contrast, transgender respondents came out as LGB an estimated 8.7 months earlier in highest compared to lowest stigma countries. Gay respondents came out an estimated 1.6 years (18.9 months) later in highest compared to lowest stigma countries.

Does the Association Between Structural Stigma and Closet Duration Vary as a Function of Identity Subgroups?

Associations of country-level structural stigma with closet duration varied as a function of gender identity and transgender identity, as shown by significant subgroup*structural stigma interactions in Table 5. Interactions between structural stigma and sexual identity were not significant. Further probing with simple slopes analyses revealed a significant association of structural stigma with closet duration for men, non-binary respondents, cisgender respondents, and gay respondents as shown in Figure 1 in the third panel. For both men and non-binary respondents, higher structural stigma was associated with longer closet duration. Compared to the lowest stigma countries, in the highest stigma country men were in the closet an estimated 11.2 months longer and non-binary respondents were in the closet an estimated 9.0 months longer. Cisgender respondents were in the closet an estimated 9.7 months longer in the highest compared to lowest structural stigma countries.

Sensitivity Analyses

Given the age differences among gender and sexual identity groups (e.g., on average, bisexual respondents were younger than gay and lesbian respondents, non-binary respondents were younger than women), analyses were replicated testing identity subgroup differences in age of milestones and duration of the closet with adjustment for age. Results and interpretation of these adjusted analysis did not meaningfully differ from those reported in Table 3.

To confirm whether model results were interpreted consistently regardless of categorical variable reference groups, all regression models were replicated with rotated gender identity referent (woman, man, non-binary) and sexual identity referent (lesbian, gay, bisexual). Results of these models confirmed the results and interpretation of reported models with lesbian woman as the reference group; that is, gay respondents and men were the groups with the strongest associations between structural stigma later age of awareness and disclosure.

Additional analyses initially examined moderation by ethnic minority status. However, because the present study did not include information regarding specific ethnic minority group membership and due to the small proportion of respondents who endorsed an ethnic minority status (<5%), the final analyses focused on differences by gender identity, transgender identity, and sexual identity rather than also ethnic minority status.

Discussion

Despite theoretical accounts that sexual minority development is shaped by environmental context, few prior studies have examined associations between structural stigma and sexual minority developmental milestones. To address this gap, the present study examined associations between country level-structural stigma and age of LGB identity awareness, age of first coming out, and duration of the closet in a large international sample of LGB adolescent and adults from 28 European countries. Results of this study provide evidence that country-level structural stigma (operationalized as national laws and policies failing to protect the rights and welfare of LGB people) was associated with the timing and pacing of LGB-related developmental milestones. Specifically, greater country-level structural stigma was associated with later age of coming out and longer time spent in the closet (i.e., the period between self-awareness and coming out) for LGB survey respondents living in that country. Evidence was also found that although developmental milestones occur, on average, during adolescence, timing of awareness and coming out and duration of the closet vary, to some extent, according to gender identity, transgender identity, and sexual identity. Associations of country-level structural stigma with milestone timing and pacing also varied as a function of gender identity, transgender identity, and sexual identity, such that structural stigma was associated with (1) earlier self-awareness for non-binary and transgender respondents and later self-awareness for cisgender and gay respondents; (2) earlier coming out for transgender respondents and later coming out for men, cisgender, and gay respondents; and (3) longer duration of the closet for men, non-binary respondents, and gay respondents. Together, these results provide among the first evidence that structural stigma is associated with LGB developmental milestones and demonstrate variations in this association across identity subgroups among LGB adolescents and adults living in 28 countries.

Structural Stigma and Timing and Pacing of Developmental Milestones

Overall, country-level structural stigma was associated with odds of never coming out, age of coming out, and duration of the closet, but not age of awareness, such that odds of never coming out were higher, timing of coming out was delayed, and pacing of milestones was extended in higher structural stigma countries. Notably, country-level structural stigma was not associated with age of awareness. This may be due in part to the differing impact of structural stigma on internal versus interpersonal developmental processes. Internally, self-awareness may constitute a more immutable, less context-dependent feature of sexual minority identity development compared to decision-making about coming out to another person. In other words, perhaps only when identity development enters an interpersonal space—often during adolescence—social influences such as structural stigma exert measurable impact. Despite the lack of effect of structural stigma on age of awareness, country-level structural stigma did in fact predict the odds of never coming out. Although LGB respondents became aware of their LGB identity, on average, at the same age across countries with varying levels of structural stigma, whether or not they ever then disclosed their identity to someone else was highly variable by structural stigma. Indeed, the odds of never coming out was an estimated 1.4 times higher for every standard deviation increase in structural stigma. Recent estimates suggests that a majority of sexual minority people globally may remain highly concealed as a function of structural stigma (Pachankis & Bränström, 2019), but the present study is the first study to empirically link structural stigma to one’s likelihood of remaining completely closeted. The influence of structural stigma may alter LGB identity development by delaying or impeding disclosure to others. Notably, most LGB respondents in this study became aware of their sexual identity in early adolescence, the same developmental period when socialization through interactions with others and environment, including the legal environment, emerges and becomes especially impactful (Tyler & Trinkner, 2017). Developmental adaptation of the minority stress theory for adolescence emphasizes the interpersonal contexts relevant to LGB adolescent development such as family and peers (Goldbach & Gibbs, 2017). Results of the current study further extend contextual considerations by demonstrating the role of structural stigma in adolescent sexual minority identity development.

Timing of coming out and duration of the closet both varied as a function of structural stigma such that LGB adolescents and adults in the highest stigma countries (Poland and Italy) came out, on average, around eight months later than those who resided in the lowest stigma country (United Kingdom), thus spending about eight months longer in the closet. This elongated period of the closet may delay or deny young LGB people resources linked to outness such as community connectedness (Roberts & Christens, 2021), access to LGB resources (Vaccaro & Newman, 2017), and support from others (Chang et al., 2021; Reyes et al., 2023), which may be pivotal to their well-being. Indeed, social isolation serves as a risk factor for negative mental health among LGB adolescents (Garcia et al., 2020) and increases among sexual minority men as a function of country-level structural stigma, perhaps as a function of the closet (Pachankis et al., 2021). Further, these delayed milestones occur for many young LGB respondents during adolescence when development is focused on peer relations (Brown & Larson, 2009) and a transition to greater autonomy (Zimmer-Gembeck & Collins, 2008), which may be challenged or interrupted by delayed or elongated LGB development (Russell & Fish, 2019). Thus, structural stigma might influence LGB identity development directly through the timing of identity milestones and indirectly through mechanisms such as social isolation and diminished access to community supports (Kowalska, 2011). Notably, the closet and coming out were operationalized in the current study based on respondents’ first report of identity disclosure to another person. Although the majority of sexual minorities across all age groups reported a mean age of first coming out during adolescence or young adulthood, it is common for sexual minority people to disclose their identities across their lifespan and further research is needed to understand how structural stigma impacts additional decision-making regarding identity disclosure in distinct contexts (e.g., work, home, school) and to distinct people (e.g., parents, peers, romantic partners) (Doyle & Barreto, 2022; Pachankis & Jackson, 2023).

Subgroup Differences in Developmental Milestone Timing and Pacing

Examination of timing and pacing of developmental milestones in this diverse, international study of LGB adolescents and adults revealed meaningful differences by gender identity, transgender identity, and sexual identity. In the current study, gendered patterns in development were similar to previous studies in the US showing that men become self-aware earlier, but come out later, leading to a longer period of time in the closet relative to women (Savin-Williams & Diamond, 2000). The current study expands these findings by investigating gender differences beyond the binary of men and women to provide evidence that non-binary people become aware of their LGB identities at about the same age as men but disclose earlier than either men or women and thus experience the closet earlier than women but with similar closet duration to that of women. Previous research has demonstrated gender differences in LGB developmental milestones (Bishop et al., 2020; Martos et al., 2015), but little research has investigated why these gender differences in timing and pacing exist. Historical context could play a role in gay and bisexual men delaying coming out during the height of the HIV pandemic due to enduring stigma against gay and bisexual men (Bishop et al., 2020). This explanation, however, does not explain why age of awareness is earlier for men and non-binary respondents than women and why non-binary respondents disclose earlier than women. The rigidity of masculinity, and the higher connection between gender role ideology and homophobia among heterosexual men compared to heterosexual women (Polimeni et al., 2000) could partially explain why young men realize their sexual minority identity earlier than women, who may be permitted more flexibility in their sexuality (Baumeister, 2000) and their childhood gender expression (Lytton & Romney, 1991). These gender differences are further complicated by the overlap with developmental differences by sexual identity. Prior studies have found that bisexual respondents become self-aware and disclose their sexual identity at later ages than lesbian and, especially, gay respondents (Bishop et al., 2020; Martos et al., 2015). The present study found mixed support for these prior findings when contrasting bisexual respondents with gay and lesbian respondents: bisexual respondents became self-aware of their LGB identity later than gay respondents but disclosed earlier than lesbian respondents. When bisexual respondents were disaggregated by gender, bisexual men were found to proceed through milestones more slowly than other bisexual and lesbian respondents but more quickly than gay respondents; whereas bisexual women were found to proceed through milestones more quickly than any other group. These patterns across gender and sexuality point to a need for future research to explain both their respective and intersecting drivers (Nagoshi et al., 2023).

Relative to cisgender respondents, transgender respondents reported becoming aware of their LGB identity at later ages. These differences may stem from the close connection between sexual identity development and gender identity development: because sexual identity is often (though not always) predicated on one’s gender, sexual identity and gender identity development are deeply intertwined for many people (Hereth et al., 2020). For transgender people, gender identity development is often complicated by structural anti-transgender stigma (Bränström & Pachankis, 2021; Price et al., 2023), which encourages people to repress gender identities that do not match sociocultural expectations for the sex they were assigned at birth, even at the level of their own awareness (Hereth et al., 2020). Accordingly, to the extent that sexual and gender identity development are intertwined, sexual identity awareness might occur on a later timeline for transgender people relative to cisgender people. Alternately, cisgender and transgender people may initially explore and label their sexual identities at similar ages— but transgender people may subsequently experience substantial changes in how they understand and express their gender identities, which may prompt further exploration, and possible relabeling, of their sexual identities (Hereth et al., 2020; Katz-Wise & Hyde, 2017). Thus, being transgender may ultimately contribute to a richer understanding of one’s sexual identity, yet milestones may appear delayed when assessed only for one’s current sexual identity label as in the present study. These proposed explanations could be assessed in future studies by collecting milestone data for gender identity development alongside past and present sexual identity labels.

Structural Stigma and Differences in Developmental Milestones Across Identity Subgroups

The current study provides evidence that country-level structural stigma is associated with LGB developmental milestone timing and pacing; however, the association between structural stigma and these experiences was stronger among some subgroups than others. Higher structural stigma was associated with longer duration of the closet for men and non-binary respondents. For non-binary respondents, higher structural stigma was associated with earlier age of awareness; in contrast, for men, higher structural stigma was associated with later age of coming out. Thus, in higher structural stigma countries the closet began earlier for non-binary respondents and extended later for gay respondents. Two factors may be at play here. First, men and non-binary respondents, on average, become self-aware and come out earlier than women, potentially making them more susceptible to the influence of structural stigma on any subsequent identity milestones (Russell & Fish, 2019), especially those involving a public identity. Second, men, women, and non-binary people face distinct challenges and advantages at the intersection of multiple systems of oppression (Collins, 2015) when coming out, with potentially distinct interactions with structural stigma. Perhaps the stronger association of structural stigma with age of coming out for men, compared to women, is due in part to men’s earlier age of awareness. Specifically, sexual minority adolescents who progress through developmental milestones earlier may experience greater time to be influenced by the adverse environmental circumstances surrounding sexual minority individuals. Therefore, it could be that for men—who on average came out a year and half earlier than women—earlier age of awareness confers more time to be influenced by structural stigma, including in ways that impact subsequent identity milestone attainment.

Structural stigma was associated with delayed timing and extended pacing of milestones for men, cisgender respondents, and gay respondents. However, the association of structural stigma with milestone attainment was in the opposite direction for non-binary participants such that non-binary individuals became aware of their sexual identities at earlier ages in higher structural stigma countries. Perhaps structural stigma toward sexual minority individuals also reflects gender-based structural stigma (Scheer et al., 2022) such that non-binary individuals became aware of their sexual identities earlier because gender non-conforming experiences are treated as particularly problematic in high structural-stigma settings (Hostetler & Herdt, 1998). Further investigation is needed to understand how experiences of structural sexism, gender binarism, and transphobia may intersect with sexual minority structural stigma to explain patterns of sexual minority structural stigma across gender identity, transgender identity, and sexual identity.

Structural stigma also interacted with transgender identity to predict developmental milestones. Specifically, at low levels of structural stigma, transgender respondents had later predicted ages of awareness and coming out, whereas cisgender respondents had a later age of coming out and longer closet duration. Above, it was suggested that later age of awareness among transgender (vs. cisgender) respondents could result, in part, from the fact that many transgender people explore and relabel their sexual identities over the course of gender identity development (Hereth et al., 2020; Katz-Wise et al., 2016; Katz-Wise & Hyde, 2017). However, this research has primarily occurred in low structural stigma contexts in the US, and it is plausible that sexual identity exploration and relabeling is less common for transgender people in higher-stigma contexts. For instance, a transgender person’s sexual identity exploration may by impeded when structural stigma makes it difficult to find respectful sexual and romantic partners (Gamarel et al., 2020; Murchison et al., 2022). Another possible explanation relates to childhood gender nonconformity. The prevalence of childhood gender nonconformity is considerably higher among transgender people than among cisgender LGB people (Gordon et al., 2017; Reisner et al., 2014). Among LGB people, greater childhood gender nonconformity is associated with earlier timing of sexual identity milestones (Bishop et al., 2020), and researchers have speculated that this occurs because gender nonconforming children may be stereotyped as LGB by their parents and peers (Bishop et al., 2020). This stereotyping process may be heightened in high-structural stigma contexts given parents’ potentially greater worries about gendered norms in such contexts, thus potentially explaining why structural stigma was inversely associated with age of sexual identity awareness and disclosure among transgender (but not cisgender) and non-binary respondents. Regardless of why transgender respondents experienced both milestones earlier and non-binary respondents became aware earlier in contexts with higher (vs. lower) structural stigma, these seemingly paradoxical findings suggest a developmental scenario in which transgender and non-binary LGB people become aware of, and, for transgender people disclose, their LGB identities in higher structural stigma environments at younger ages, during which they might have fewer coping resources, less developed cognitive capacity, and fewer protections for coping (Russell & Fish, 2019).

Implications

The present evidence that structural stigma is associated with sexual identity developmental milestone timing and pacing suggests routes of intervention. Despite a gradient from low to high structural stigma across the 28 included countries, no country had legislated all protective policies indexed by the International Lesbian and Gay Association (ILGA Europe, 2019). Relatively lower structural stigma appears to foster an environment where some sexual minority young people can disclose their identity at earlier ages, thus spending less time isolated in the closet. Reducing structural stigma in all countries may create a safer environment for sexual minority development in the future.

The measure of structural stigma in the current study is framed primarily as an index of the absence of protective laws with higher scores indicating fewer protections and more structural stigma. In contrast, low-structural stigma countries have structural stigma scores reflecting greater presence of protective laws. Thus, the structural influence on LGB development could be conceptualized as both structural stigma that harms development and structural support that protects development. To reduce structural stigma, country-level legislation may need to be reformed in two ways: (1) elimination or reform of laws that bar sexual minorities from access to equal rights and protections (e.g., same-sex marriage bans) and (2) development of laws that protect sexual minority rights (e.g., enacting nondiscrimination laws). Reforming laws that directly impact sexual minority adolescents may be especially critical to identity development because major developmental milestones typically occur during this period. For example, conversion therapy bans and non-discrimination protections in education, two laws and policies included in this study’s structural stigma index, may have direct application to sexual minority adolescents. Furthermore, the current measure of structural stigma relies on an international evaluation of several different policies and laws protecting the rights of sexual minority Europeans and thus reflect an overall context of structural stigma; however, this measure does not explicitly evaluate youth-specific sexual minority supportive laws and policies. Results of the current study show that, on average, LGB Europeans become aware of their LGB identity and first come out during adolescence. However, monitoring of LGB protective policies, including in this study, often focuses on LGB adult rights, such as same-sex marriage, parental rights, and workplace protections (e.g., The European Region of the International Lesbian, Gay, Bisexual, Trans & Intersex Association, 2019) with less enumerated focus on youth-specific policy (Hatzenbuehler et al., n.d.). Future research should consider youth-specific structural stigma to understand how policies intended to support and protect LGB youth may shape development more or less directly than the broader structural context of general and adult-specific protections. Results of research focused on youth-specific laws may also provide necessary evidence to address whether advocates and policy makers in higher structural stigma countries might prioritize specific these laws and policies to protect development.

Limitations and Future Directions

Despite strengths of the current analysis across 28 countries and more than 100,000 LGB adolescents and adults, results should be interpreted within the study’s limitations. First, the cross-sectional nature of the study precludes disentangling age versus cohort differences. Although LGB people across generations may share commonalities in developmental milestones, there are likely key developmental differences due to their historical context, individual experiences, and response to structural stigma (Hammack, 2005). Still, despite reductions in structural stigma over time, the relative level of structural stigma between countries is fairly stable; thus, countries with low structural stigma today were also the countries with relatively lower structural stigma in the past (Flores, 2019). This relatively stability somewhat hinders the ability to assess the influence of changing country-level structural stigma contexts and any age variation in exposure to differing country-level structural stigma within the same country, despite a few notable country-level exceptions (Bränström & Pachankis, in press). Second, respondent age was necessarily collapsed into age bins for participant confidentiality concerns. This grouping may have prevented uncovering potentially important variance had the more typical 1-year age increment been used. Third, the current sample was not a probability sample and should not be interpreted as representative of LGB adolescents or adults in the EU or in any single country. Both national samples (Gates, 2011) and census data (Sharfman & Cobb, 2022) indicate that bisexual woman may be the largest gendered sexual orientation group in the LGB community; therefore, in the current sample gay men may have been overrepresented. It is unknown whether recruitment through LGBTI organizations and social media contributed to the gender and sexual identity distributions of the sample. Fourth, the current study excludes those who might exclusively identify with sexual identity labels such as asexual, pansexual, and queer. Future studies are needed to represent all sexual minority identities, especially younger sexual minority people, who are more likely to use these labels (Bochicchio et al., 2023; Watson et al., 2020). Fifth, many sexual minority people are unwilling to reveal their sexual identity on surveys (Ferlatte et al., 2017), perhaps especially in high structural stigma contexts (Pachankis et al., 2015). Thus, the sample might be biased such that those whose development is most affected by structural stigma might be the least likely to participate. However, such differential selection by structural stigma and development would bias results toward the null. Sixth, the present study did not include any measure of health or well-being. Although several studies have linked developmental timing of LGB milestones to mental health (Katz-Wise et al., 2017), substance use (Fish & Pasley, 2015), and suicide (Dirkes et al., 2016) and structural stigma to LGB mental health (Pachankis et al., 2021; Perales & Todd, 2018) and life satisfaction (Pachankis & Bränström, 2018), further research is needed to understand how structural stigma might impact health and well-being including through developmental milestone timing and pacing. Such research would also be able to clarify the adaptive versus maladaptive implications of this timing and pacing (D’Augelli & Grossman, 2001), perhaps depending on the structural environment in which each person comes out. Seventh, sexual identity milestones were measured as discreet events, however, the experience of the “the closet” is complicated by the reality that most sexual minority people will disclose their sexual identity many times, including in subtle ways, across contexts and the life course (Maliepaard, 2018). Future research should consider the ways that structural stigma impacts decisions to disclose identity across a variety of contexts and life stages.

Finally, results suggest that multiple social identities and positions (e.g., gender identity, gender modality, sexual identity) are relevant to LGB sexual identity development. Principles of intersectionality (Collins, 2015; Combahee River Collective, 1978; Crenshaw, 1991) suggest that these differences are shaped by various forms of structural stigma (e.g., sexism, cisgenderism, racism), which are largely inextricable from one another, such that fully understanding the impact of stigma often requires considering multiple social identities and positions at once. For example, sexism, cisgenderism, and anti-LGB stigma might combine to produce unique sexual identity development experiences for transgender sexual minority women when compared with cisgender sexual minority women and with transgender people of other genders. Given statistical power concerns and a lack of directional a priori hypotheses, test interaction terms were note tested among multiple social identities and positions in the present study. Instead, qualitative research is suggested as a next step. For instance, in-depth interviews with a purposive sample of cisgender and transgender people of various genders could be used to describe how an environment of high structural stigma shapes sexual identity development across gender and transgender identities, identifying both commonalities and differences across groups (Abrams et al., 2020).

Conclusion

Although theories on the closet hypothesize that the closet and its associated developmental milestones are shaped by broader structural contexts, this is among the first studies to empirically test these association. Across 28 countries and more than 100,000 LGB adolescents and adults living in Europe, this study provided evidence of the connection between country-level structural stigma and LGB developmental milestones. Timing of milestones and duration of the closet differed across identity subgroups with gay respondents, on average, becoming aware of their LGB identity earlier and coming out later than lesbian respondents and bisexual non-binary respondents, thereby exhibiting the longest duration of the closet across gendered sexual identity groups. Structural stigma was associated with age of disclosure and duration of the closet, but not with age of awareness. Association of structural stigma with milestone timing and pacing varied as a function of gender identity, transgender identity, and sexual identity. For the majority of LGB people, these major developmental milestones and processes occur, at least in part, in adolescence. Thus, creating a safer structural environment through affirmative laws and policies may have the greatest benefit for adolescents, for whom identity awareness and opportunities for disclosure in social relationships represent emergent and salient developmental tasks. Finally, this study provides some of the first evidence that structural stigma is associated with LGB developmental milestones, suggesting that not only can structural stigma influence individual development but also the public visibility known to recursively eradicate structural stigma. Reducing structural stigma at its source in unjust laws and policies therefore has potential to ensure the future visibility and developmental flourishing of the LGB population across countries.

Biographies

Eric K. Layland is an Assistant Professor at the University of Delaware in Newark, Delaware. His major research interests include LGBTQ+ adolescent and family development, intersectional stigma and health disparities, and LGBTQ+ mental health and substance use interventions.

Richard Bränström is an Associate Professor at the Karolinska Institutet in Stockholm, Sweden. His main research interest is the study of how social, psychological and contextual factors influence the health of sexual and gender minority individuals.

Gabriel R. Murchison is a Post-Doctoral Associate at the Yale School of Public Health in New Haven, Connecticut. His major areas of research include sexual assault, intimate partner violence, mental health, and substance use among adolescents and young adults, especially in the LGBTQ+ community.

John E. Pachankis is the Susan Dwight Bliss Professor of Public Health at the Yale School of Public Health in New Haven, Connecticut. He conducts psychiatric epidemiological studies and intervention research to understand and improve the mental health of sexual and gender minority populations in the US and around the world.

Data Sharing Declaration

The individual-level data that support the findings of this study are available from the European Union Agency for Fundamental Rights, but restrictions apply to the availability of these data.