Identity resilience, social support and internalised homonegativity in gay men

ABSTRACT

Identity resilience, a key concept in identity process theory (IPT), refers to individuals’ capacity to cope with threats to their identity. Identity resilience is based upon four identity principles: self-efficacy; self-esteem; positive distinctiveness; and, continuity. This study investigates whether identity resilience influences how much gay men internalise homonegativity. Given the insidious effects of internalised homonegativity upon psychological well-being, it is important to identify factors affecting its management. Greater identity resilience enables deployment of strategies that may reduce internalisation of homonegativity. These strategies include rejecting the salience of negative social representations of gay men and emphasising the availability of social support. While both these strategies affect how ‘out’ a gay man chooses to be, they are also linked to the experience of everyday discrimination. We predicted identity resilience would have both a direct negative association with internalised homonegativity and an indirect negative effect mediated by higher social support, lower everyday discrimination, and, less perceived negative representations of gay men and greater ‘outness’. Survey data from 333 gay men in the UK supported this model. Fostering the development of identity resilience and its attendant coping strategies may help in managing internalisation of homonegativity.

KEYWORDS:

• Identity resilience
• social support
• internalised homonegativity
• gay men
• identity process theory

Introduction

Globally, homonegativity (prejudiced attitudes and discriminatory behaviour towards sexual minorities) remains a significant societal challenge. In most Western societies, attitudes towards sexual minorities appear to be becoming more favourable but sexual minorities still report facing discrimination. After all, homonegativity occurs in many guises – ranging from overt denigration to subtle ‘micro-aggressions’ (Nadal et al., 2016). When facing societal stigma, sexual minorities may come to internalise this stigma as well as negative social representations of their sexual orientation, leading to a negative self-construal (Poštuvan et al., 2019). Internalised homonegativity is defined as ‘the individual’s direction of negative social attitudes [about their sexual orientation] towards the self’ and reflects a devaluation of the self and increased internal conflicts due to one’s sexual orientation (Meyer & Dean, 1998, p. 161).

There is much research into the negative social, psychological and health sequelae of the internalisation of homonegativity, including poor mental health, sexual risk-taking, relationship dysfunction, and antagonistic family relations (see Williamson, 2000). However, it is unclear why some people develop internalised homonegativity, what social and psychological factors can increase the risk of developing this negative self-schema, and whether this can be prevented. This study tests a model, derived from Identity Process Theory (IPT) (Breakwell, 2015a), of the system of factors predicting internalised homonegativity in an ethnically diverse sample of gay men in the United Kingdom. In particular, the effect of having a resilient identity on internalised homonegativity is examined. The study investigates the effects of homonegativity holistically, focusing not only on actual everyday discrimination but also on perceptions of negative social representations of homosexuality. Modelling the relationships between identity resilience, perceived negative social representations of homosexuality and experience of social support in shaping internalisation of homonegativity is a major innovative contribution of this study.

Homonegativity

Minority Stress Theory (Meyer, 2003) focuses on the insidious effects of homonegativity for health and wellbeing in gay men. Various types of homonegativity, such as victimisation, rejection from significant others and exposure to negative social representations of homosexuality, have been shown to increase the risk of depression, psychological distress, suicidal ideation and self-harm in sexual minorities (Jaspal et al., 2021; Rehman et al., 2020). In the present study, three specific types of homonegativity are examined: the perception that social representations of gay men in one’s own social environment are negative, everyday discrimination due to one’s sexual orientation, and feeling upset when recalling a significant coming out experience. ‘Homonegativity variables’ tap into three distinct dimensions of homonegativity, namely systemic social representations, personal experience, and affective reaction, that in turn are likely to stimulate the internalisation of homonegativity.

Exposure to homonegativity will vary across individuals even in the same social environment. Some may have long-term exposure to homonegativity due to dominant negative social representations in their social environment. Jolley and Jaspal (2020) found that gay men who are hypervigilant to discrimination are more likely to report its existence, possibly because previous experiences of discrimination have primed them to anticipate it in the future. Gay men who face discrimination may come to believe that people in their social context have negative and hostile views about their community (Castro et al., 2019).

Perceiving negative social representations of gay men in one’s social context is itself psychologically harmful as this may increase the state of hypervigilance observed in some gay men (Jolley & Jaspal, 2020; Meyer, 2003). Perceived negative social representations of gay men is in turn likely to accentuate the psychological and emotional effects of challenging experiences, such as coming out as gay to a significant other (Berg et al., 2016; Calvo et al., 2020; Wang et al., 2021). If one believes social representations of gay men are negative, one will experience more negative affect in recalling a negative coming out experience as the negative cognitions will reinforce the negative affect. It has been found that recalling even a singular, isolated negative coming out experience can threaten one’s sense of identity (Breakwell & Jaspal, 2021a; Ryan et al., 2015). Of course, coming out is a process, not just a single event. Coming out in different ways to different people over a lengthy period is not unusual. Throughout this process, the salient negative social representations of gay men will be operating, playing a role in the internalisation of homonegativity in an ongoing way.

Coming out models indicate that sexual minorities are motivated to assimilate, accommodate and enact their sexuality within their identity (e.g. Jaspal, 2019). Experiences of homonegativity (i.e. perceiving negative social representations, everyday discrimination and feeling upset upon recalling a negative coming out experience) represents a threat to the effective incorporation of being gay within the overall identity structure. Thus, there is the risk that exposure to homonegativity will adversely influence self-conceptualisation and evaluation in relation to one’s sexuality, precipitating internalised homonegativity.

Internalised homonegativity

Identity process theory (IPT) (Breakwell, 2001, 2015a, 2015b) posits that the following two psychological processes are central to identity construction, development and maintenance:

• assimilation-accommodation whereby the individual actively incorporates novel elements, such as their sexual orientation, into their identity structure

• evaluation, which refers to the process of appending meaning and value to the novel identity element on the basis of experience and the social representations to which they are exposed. For instance, being gay may be construed as a flaw in identity and induce feelings of shame, which reflects internalised homonegativity (Williamson, 2000).

Consistent with models of coming out (e.g. Kranz & Pierrard, 2018), the assimilation-accommodation and evaluation processes are enacted when the individual becomes aware of their sexual orientation and begins to recognise this as an aspect of the identity structure. A key tenet of IPT is that these identity processes rely, in large part, upon social representations of the new element in order to establish its meaning and value to be incorporated into the identity (Jaspal, 2019; Jaspal & Cinnirella, 2010). This will differ in accordance with social context – in gay affirmative contexts, gay men will have greater access to more positive social representations of their sexual orientation and, thus, be at lower risk of internalised homonegativity than in less gay affirmative social contexts.

As a negative self-schema, internalised homonegativity constitutes a means of making sense of stigma encountered in one’s social context and the incorporation of it as self-stigmatisation within the identity structure. Thus, the individual appends negative valence to their newly assimilated identity element – they assimilate-accommodate it but believe that it is undesirable, wish it could change and feel ashamed of it. There is an established empirical link between internalised homonegativity and identity threat, poor mental health outcomes, relationship dysfunction, and poor family relations in gay men (Breakwell & Jaspal, 2021a; Herek & Garnets, 2007; Williamson, 2000). Qualitative and quantitative research using IPT suggests that internalised homonegativity may arise from the inability of the assimilation-accommodation and evaluation processes to produce a positive sense of self (Jaspal, 2019; Maatouk & Jaspal, 2021). In short, internalised homonegativity is an important outcome of the assimilation-accommodation and evaluation processes of identity among gay men, with many potential consequences, not least regarding willingness to disclose one’s sexual orientation to others by ‘coming out’.

Outness and social support

Coming out is not a binary phenomenon and may be best thought of in terms of a continuum or an ongoing process (Mohr & Fassinger, 2000). Although people can often identify and recall an isolated experience of coming out, involving a significant other such as their parents, in practice they come out repeatedly over the life course. For instance, one may be ‘out’ to one’s family and friends but find oneself having to do so again in the workplace. Furthermore, individuals vary in the extent to which they are out – some may disclose their sexual identity to another person but never discuss the topic again, while others come out and are able to seek support from sympathetic others. It is important to distinguish between different levels of outness since an individual who comes out but never discusses the topic again may have less access to social support and may come to view their sexuality as taboo. It is important to note that coming out is not necessarily always a precursor to receiving social support since some coming out experiences can be negative and lead individuals subsequently to conceal their sexual identity (Breakwell et al., 2021).

It is noteworthy that gay men perceive different levels of ‘psychological safety’ in relation to coming out. In some contexts and to some people, coming out may seem entirely appropriate but, in others and to others, it may be construed as risky. Gay men who perceive negative social representations of gay men to be dominant in their own social environment (see Breakwell, 2014), that is, those who believe that most heterosexual people stigmatise, denigrate and reject them for being gay are less likely to be open (or ‘out’) about their sexual orientation (Moradi et al., 2010). They may feel less safe and express more caution about coming out and opt to conceal their identity as a means of self-protection.

It is also reasonable to view coming out as a form of identity enhancement and, in some cases, coping. After all, decades of research on the coming out process shows that it can facilitate feelings of identity authenticity, self-acceptance and a positive sense of self (Cain, 1991; Ragins, 2004; Wells & Kline, 1987). It can provide exposure to positive social representations of one’s sexual orientation which may displace the negative social representations previously attached to this identity element. In fact, Jaspal and Breakwell (2021) found that those who reported greater identity threat due to racial discrimination were more likely to self-identify as British, suggesting that this might constitute a strategy for coping and, more specifically, for deflecting or undermining the ‘power’ of discrimination.

Gay men with increased access to social support are more likely to come out than those with less access (Legate et al., 2012) possibly because this diminishes the risks perceived to be attendant upon coming out, such as rejection or discrimination from others. Furthermore, gay men with higher levels of social support are likely to be less prone to hypervigilance and, thus, may perceive fewer negative social representations of gay men in their social context. Moreover, they are less likely to be perturbed by a negative coming out experience since they will have access to effective resources to cope. Pre-existing social support may also mitigate the negative effects of a possible negative coming out experience.

It is also possible that some people construe coming out as a means of coping with stressors, such as everyday discrimination – after all, by disclosing their identity to others, they increase the likelihood of accessing social support and exposure to more affirmative social representations of their sexual orientation (Sommantico et al., 2018). Indeed, the quest for support from others and broader change are viewed as coping strategies in IPT (Breakwell, 2015a; Jaspal & Breakwell, 2021). Of course, when coming out is enacted as a coping strategy, the individual will do so strategically – to specific individuals perceived to be more accepting of them and incrementally (i.e. after a positive experience). For instance, it has been found that gay men often report coming out to their mothers first because they anticipate support from them (Miller & Boon, 2000).

Identity resilience

Recent developments in IPT (Breakwell, 2020a, 2020b; Breakwell & Jaspal, 2021a, 2021b) have focused upon the concept of identity resilience. Identity resilience is achieved when individuals perceive their identity configuration to be characterised by a high overall combined rating of their self-efficacy, self-esteem, continuity and distinctiveness. Identity resilience is a psychological reflection of individuals’ subjective belief in their capacity to understand and overcome challenges, their self-worth and value, certainty of who they are and will remain despite inevitable changes, and their self-construal as positively distinctive from others. It is not domain-specific in that the total identity of the individual (not just one element, for example, one’s sexual orientation), seen across the life course, will determine one’s level of identity resilience. In IPT, identity resilience is seen as a stable self-schema, akin to a trait (Breakwell et al., 2021). Over the life course, the individual will develop a general sense of their identity resilience. This perception of identity resilience is grounded in many social phenomena and experiences, such as group memberships, education, exposure to cultures, religion, as well as in individual characteristics (some of which are also partially determined by social experiences), such as personality traits, intellectual capacity, or physical abilities. While the ways in which identity resilience is expressed may vary according to context, identity resilience itself is a stable characteristic of the individual and not generally context-specific (even though a particularly traumatic experience, and some patterns of experiences, may over time result in revisions in the individual’s level of identity resilience).

It has been found that the extent to which one possesses identity resilience will significantly influence the individual’s capacity to cope with stressors that have the potential to threaten identity, such as adverse social and psychological experiences in relation to their stigmatised sexual identity. Each of the four components of the Identity Resilience Index (Breakwell et al., 2021) has been shown individually to facilitate effective and adaptive coping responses to stressors (Brewer, 1991; Dumont & Provost, 1999; Sadeh & Karniol, 2012). Recent research shows that identity resilience is protective against negative affective experiences (Breakwell & Jaspal, 2021a, 2021b). Thus, identity resilience may enable the individual to resist the aversive social representations characteristic of internalised homonegativity. It also is possible that gay men with higher identity resilience may be able to resist such aversive social representations and, thus, to avoid internalising them in their self-evaluation. Gay men with higher identity resilience also may be more likely to feel empowered to seek and adopt adaptive coping strategies, such as seeking social support from others.

A theoretical model predicting internalised homonegativity

Drawing principally on IPT and relevant research into identity resilience, homonegativity and internalised homonegativity in gay men, we propose a theoretical model (Figure 1) and test the following specific hypotheses:

H1: Following Breakwell et al. (2021) who suggest that identity resilience may bolster the effective operation of identity processes and resistance to threat, we propose that identity resilience will be negatively associated with internalised homonegativity, controlling for the effects of the homonegativity variables and outness.

H2: Identity resilience will be negatively associated with perceived negative social representations of gay men but positively associated with receiving of social support.

H3: Since anticipated hostility from others may preclude coming out (Moradi et al., 2010; Sabat et al., 2014), we hypothesise that perceived negative social representations of gay men will be negatively associated with outness.

H4: Building on the empirical observation that social support is associated with decreased odds of victimisation in sexual minorities (Ybarra et al., 2015), we hypothesise that social support will be negatively associated with perceived negative social representations of gay men and everyday discrimination. Furthermore, consistent with work showing the buffering effect of social support on psychological adversity (Fingerhut, 2018), we hypothesise that social support will be negatively associated with feeling upset when recalling a significant coming out experience. As a facilitator of coming out (Legate et al., 2012), social support will be positively associated with outness.

H5: It has been found that the actual experience of discrimination is associated with the anticipation of further stigma in sexual minorities (Castro et al., 2019) due partly to a hypervigilance effect (Jolley & Jaspal, 2020). Therefore, everyday discrimination should be associated with perceived negative social representations of gay men. Consistent with previous findings (Walch et al., 2016), everyday discrimination should be positively related to internalised homonegativity. However, research using identity process theory (e.g. Breakwell, 2015a; Jaspal & Breakwell, 2021a) indicates that coming out may act as a strategy for coping with discrimination and, thus, everyday discrimination should be positively associated with outness.

H6: Negative affect experienced by some gay men upon coming out can lead to increased risk of internalised homonegativity (Williamson, 2000). Therefore, we predict that feeling upset when recalling a negative coming out experience will be positively associated with internalised homonegativity.

H7: As a possible facilitator of exposure to more positive social representations of one’s sexual orientation (e.g. Sommantico et al., 2018), we predict that outness will be negatively associated with internalised homonegativity.

Figure 1. Theoretical model of the factors predicting internalised homonegativity.

Method

Ethics

This study was approved by Nottingham Trent University’s Schools of Business, Law and Social Sciences Ethics Committee (REF: 2020/227). All participants provided electronic consent before completing the study.

Participants

An ethnically diverse sample of 333 cisgender gay men in the UK was recruited to participate in a questionnaire study of the psychological aspects of coming out as gay on Prolific, an online participant recruitment platform. There were two eligibility criteria: being aged 18 or over and self-identifying as a gay man. Furthermore, an ethnically diverse sample was requested in order to avoid the biases associated with having an exclusively White British participant sample. Data were collected in February 2021. Table 1 provides the main characteristics of the sample.

Table 1. Socio-demographic characteristics of the sample.

Age	M = 33.01	SD = 12.10	Minimum = 18	Maximum = 74
Ethnicity	White British	South Asian	Black African	Black Caribbean	Mixed (White/Black)	Mixed (White/Asian)	Mixed (Other)	White Other
	n = 210 (63.1%)	n = 52 (15.6%)	n = 6 (1.8%)	n = 8 (2.4%)	n = 26 (7.8%)	n = 22 (6.6%)	n = 1 (0.3%)	n = 8 (2.4%)
Religion	No religion	Christianity	Judaism	Hinduism	Islam	Sikhism	Other
	n = 248 (74.5%)	n = 59 (17.7%)	n = 7 (2.1%)	n = 2 (0.6%)	n = 7 (2.1%)	n = 1 (0.3%)	n = 9 (2.7%)
Relationship	Single	Monogamous	Open	Engaged	Married	Civil partnership	Other
	n = 168 (50.5%)	n = 22 (6.6%)	n = 72 (21.6%)	n = 16 (4.8%)	n = 17 (5.1%)	n = 26 (7.8%)	n = 12 (3.6%)
Income	<£10,000	£10,000 to £19,999	£20,000–29,999	£30,000–39,999	£40,000–49,999	£50,000–59,999	£60,000>
	n = 83 (24.9%)	n = 49 (14.7%)	n = 62 (18.6%)	n = 47 (14.1%)	n = 33 (9.9%)	n = 22 (6.6%)	n = 37 (11.1%)
Job status	Employed	Self-employed	Student	Retired	Unemployed	Other
	n = 181 (54.4%)	n = 35 (10.5%)	n = 52 (15.6%)	n = 16 (4.8%)	n = 45 (13.5%)	n = 4 (1.2%)
Education	GCSE/O-Levels*	A-/ AS-Levels**	Undergraduate	Postgraduate	Apprenticeship	Other
	n = 48 (14.4%)	n = 67 (20.1%)	n = 147 (44.1%)	n = 66 (19.8%)	n = 2 (0.6%)	n = 3 (0.9%)

* The General Certificate of Secondary Education (GCSE) refers to an academic qualification in a particular subject usually at age 15/16 and is taken in England, Wales and Northern Ireland. Until 1988, Ordinary Level (O-Level) was the equivalent academic qualification.

** The Advanced Level (A-Level) is a subject-based qualification that follows GCSE, usually taken at age 17/18 after 2 years of study, in the United Kingdom. The Advanced Subsidiary Level (AS-Level) is taken after 1 year of study, prior to progression to A-Level. A-Levels precede university-level education.

Procedure

The questionnaire was administered and completed online. Participants completed measures of identity resilience, social support, perceived negative social representations of gay men, everyday discrimination, feeling upset when recalling a significant coming out experience, outness and internalised homonegativity. They also provided demographic information, including age, gender, sexual orientation, religion, educational attainment, occupational status and income. They provided electronic consent, were debriefed and were paid a token amount for participating in the study. The study took approximately 20 minutes to complete and there were two attention checks in the study, which all of the participants passed. There were no missing data.

Measures

Identity resilience

The Identity Resilience Index (Breakwell et al., 2021), comprising 16 items with responses on a 5-point scale (1 = strongly disagree to 5 = strongly agree), was used. Sample items are: ‘On the whole, I am satisfied with myself’ and ‘There is continuity between my past and present’. The Identity Resilience Index has four subscales referring to self-esteem, self-efficacy, continuity and positive distinctiveness, respectively. However, Breakwell et al. (2021) propose that the sum of all items can be used as a measure of identity resilience and thus the measure is used in this way in this study. Possible scores ranged from 16 to 80. A higher score indicated higher identity resilience (α = 0.80; 95% confidence intervals: Lower bound 0.76, upper bound 0.83).

Social support

The Interpersonal Support Evaluation List (Cohen et al., 1985), comprising 12 items with responses measured on a 4-point scale (1 = definitely false to 4 = definitely true), was used. Sample items are: ‘When I need suggestions on how to deal with a personal problem, I know someone I can turn to’ and ‘I don’t often get invited to do things with others’. In an Australian study of gay and lesbian people (Alba et al., 2020), scale score reliability was excellent (α = 0.90). Possible scores ranged from 12 to 48. A higher score indicated greater availability of social support (α = 0.91; 95% confidence intervals: Lower bound 0.90, upper bound 0.927).

Perceived negative social representations of gay men

An adapted version of the Attitudes Towards Gay Men Short-Version Scale (Siebert et al., 2014), comprising nine items with responses on a 5-point scale (1 = strongly disagree to 5 = strongly agree), was used. The clause ‘Most people think that’. was added to each item to measure the extent to which one believes that each view is widely endorsed by society. In addition, one item indicating public indifference and three items to capture perceived views about coming out as gay were added. The scale is outlined in full in the Supplementary. Possible scores ranged from 9 to 45. A higher score indicated higher perceived negative social representations of gay men (α = 0.87; 95% confidence intervals: Lower bound 0.85, upper bound 0.89).

Outness

The Outness Inventory (Mohr & Fassinger, 2000), comprising 11 items with responses measured on an 8-point scale (0 = not applicable; 1 = person definitely does not know about sexual orientation status to 7 = person definitely knows about sexual orientation status and it is openly talked about), was used. The scale measures the extent to which an individual’s sexual orientation is known by and openly discussed with people, such as ‘new straight friends’, ‘work peers’, ‘mother’, ‘father’, ‘leaders of religious community’. The Outness Inventory has three subscales referring to outness to family, the world and religion, respectively. However, Mohr and Fassinger (2000) propose that the sum of all items can be used as a measure of overall outness and thus the measure is used in this way in this study. In an American study of gay, lesbian and bisexual people (Riggle et al., 2017), the scale score reliability was excellent (α = 0.90). Possible scores ranged from 0 to 77. A higher score indicated a higher level of outness (α = 0.82; 95% confidence intervals: Lower bound 0.79, upper bound 0.85).

Everyday discrimination

The Everyday Discrimination Scale (Williams et al., 1997), comprising nine items with responses on a 6-point scale (0 = never to 5 = almost every day), was used to measure the frequency of discriminatory events due to one’s sexual orientation. Sample items are: ‘Because of your sexual orientation, you are treated with less courtesy than other people are’ and ‘ … people act as if they’re better than you are’. Possible scores ranged from 0 to 45. In a previous UK study of gay men (Jolley & Jaspal, 2020), scale score reliability was good (α = 0.87). A higher score indicated more frequent everyday discrimination due to one’s sexual orientation (α = 0.90; 95% confidence intervals: Lower bound 0.88, upper bound 0.91).

Feeling upset by the coming out experience recalled

The following four items were used to measure the extent of being upset by the recalled coming out experience: ‘To what extent did this experience upset you at the time?’, ‘To what extent does remembering this experience upset you currently?’, ‘To what extent does remembering this experience upset you in general?’ and ‘How often does remembering this experience upset you in general?’ Responses were measured on a 5-point scale (1 = not at all to 5 = extremely). The four items were summed. Possible scores ranged from 4 to 20. higher scores denoting greater upset due to recalling a significant coming out experience (α = 0.87: 95% confidence intervals: Lower bound 0.848, upper bound 0.893).

Internalised homonegativity

The Internalised Homophobia Scale (Herek et al., 2009), comprising 9 items with responses on a 5-point scale (1 = not at all true of me to 5 = very true of me), was used. Sample items are: ‘I wish I weren’t gay’ and ‘I have tried to become more sexually attracted to women’. Possible scores ranged from 9 to 45. In an American study of sexual minorities (Heiden-Rootes et al., 2020), the scale score reliability was satisfactory (α = 0.75). A higher score indicated greater internalised homonegativity (α = 0.88, 95% confidence intervals: Lower bound 0.862, upper bound 0.9).

Data analysis

SPSS and AMOS version 20 were used to conduct the analyses. The plan of analysis was predicated on the hypotheses and the most effective ways of testing them. First, Pearson-product moment correlations were conducted to measure relationships between the continuous variables. Second, a hierarchical multiple regression was conducted with a bootstrap at 1000 samples to test which variables predict internalised homonegativity. Third, a path analysis bootstrapped at 1000 samples was conducted to test the theoretical model presented in Figure 1. Path analysis that has mediation pathways requires errors to be normally distributed. Furthermore, assumptions for this type of analysis also include linearity (Normal Probability Plot), homoscedasticity (Plot of residuals versus predicted value), independence (Durbin-Watson statistic) of residuals, the presence of outliers (Cook’s distance < 1, N = 333) and the absence of multicollinearity (Variance Inflation Factor (VIF) < 2). All of these assumptions were tested for the purpose of path analysis and no problems were found.

Results

Descriptive statistics

Table 2 provides a full summary of the descriptive statistics for the main variables of interest.

Table 2. Descriptive statistics for the key variables of this study.

Continuous variables	Mean	SD	Minimum	Maximum
Identity resilience	52.89	8.79	20	74
Social support	36.23	7.84	12	48
Perceived negative social representations	25.35	5.71	12	40
Everyday discrimination	8.71	7.44	0	41
Outness	38.33	14.82	3	77
Feeling upset upon recall	8.68	4.21	4	20
Internalised homonegativity	15.56	6.80	9	41

Correlations

Identity resilience was positively associated with social support (r = 0.34, p < 0.001) (as predicted in H2) and outness (r = 0.23, p < 0.001) but negatively associated with perceived negative social representations of gay men (r = −0.24, p < 0.001) (as predicted in H2), everyday discrimination (r = −0.13, p < 0.05) and, as predicted in H1, internalised homonegativity (r = −0.23, p < 0.001). Social support was positively associated with outness (r = 0.46, p < 0.001) but negatively associated with perceived negative social representations of gay men (r = −0.26, p < 0.001), everyday discrimination (r = −0.20, p < 0.001), feeling upset upon recall (r = −0.21, p < 0.001), all of which supported H4, and internalised homonegativity (r = −0.29, p < 0.001). Perceived negative social representations of gay men was positively associated with everyday discrimination (r = 0.32, p < 0.001) (as predicted in H5), feeling upset upon recall (r = 0.20, p < 0.001) and internalised homonegativity (r = 0.21, p < 0.001) but, as predicted in H3, negatively associated with outness (r = −0.28, p < 0.001). Everyday discrimination was positively associated with feeling upset upon recall (r = 0.14, p < 0.05) and internalised homonegativity (r = 0.16, p < 0.001) (as predicted in H5). However, contrary to H5, everyday discrimination and outness were not significantly correlated (p > 0.05). Outness was negatively associated with internalised homonegativity (r = −0.36, p < 0.001), supporting H7. Feeling upset upon recall was positively associated with internalised homonegativity (r = 0.21, p < 0.001), supporting H6. Table 3 provides a full summary of the correlations between the main continuous variables.

Table 3. Correlations between the main variables of interest.

	1	2	3	4	5	6
1.Identity resilience
2.Social support	.34**
3.Perceived negative social representations	−.24**	−.26**
4.Everyday discrimination	−.13*	−.20**	.32**
5.Outness	.23**	.46**	−.28**	.02
6.Feeling upset upon recall 7.Internalised homonegativity	−.09 -.23**	−.21** -.29**	.20** .21**	.14* .16**	−.05 -.36**	.21**

*p < .05 **p < .001

Multiple regression model predicting internalised homonegativity

A hierarchical linear regression was conducted to examine which variables accounted for the most variance in internalised homonegativity (see Table 4). The variables of identity resilience, everyday discrimination, outness, and feeling upset by recalling a significant coming out experience, which were hypothesised to have direct effects on internalised homonegativity, were inserted as predictors, and internalised homonegativity was inserted as the criterion variable. The variables were entered in 4 steps, with identity resilience being added in the final step.

Table 4. Hierarchical multiple regression model predicting internalised homonegativity.

Predictors	R^2	F	p	β	sr	t	p	95% CI
Model 1	0.13	49.43	<0.001
Outness				−0.36	−0.36	−7.03	<0.001	−0.212, −0.119
Model 2	0.17	33.45	<0.001
Outness				−0.35	−0.35	−6.98	<0.001	−0.206, −0.116
Feeling upset upon recall				0.20	0.20	3.92	<0.001	0.158, 0.477
Model 3	0.19	25.54	<0.001
Outness				−0.36	−0.35	−7.13	<0.001	−0.208, −0.118
Feeling upset upon recall				0.18	0.18	3.53	<0.001	0.126, 0.445
Everyday discrimination				0.14	0.14	2.87	0.004	0.042, 0.222
Model 4	0.20	20.99	<0.001
Outness				−0.33	−0.32	−6.44	<0.001	−0.196, −0.104
Feeling upset upon recall				0.17	0.17	3.38	<0.001	0.114, 0.431
Everyday discrimination				0.13	0.13	2.54	0.011	0.026, 0.207
Identity resilience				−0.13	−0.12	−2.48	0.014	−0.176, −0.020

Outness was entered into Step 1, because this has been found to be a particularly strong correlate of internalised homonegativity, and explained 13% of the variance in internalised homonegativity. At Step 2, outness and feeling upset upon recalling a coming out experience explained 17% of the variance in internalised homonegativity. Feeling upset upon recalling a coming out experience was added in at this stage because, as an affective reaction associated with one’s sexuality, it was deemed to be important to internalised homonegativity. R-square change was 0.04 and F-change was 15.33 (p < 0.001). At Step 3, outness, feeling upset upon recall and everyday discrimination explained 19% of the variance in internalised homonegativity. Everyday discrimination was included at this stage because it has been found to shape gay men’s reactions to their own sexual orientation. R-square change was 0.02 and F-change was 8.26 (p = 0.004). At Step 4, outness, feeling upset upon recall, everyday discrimination and identity resilience explained 20% of the variance in internalised homonegativity. Identity resilience was added in last to determine the additional contribution made by a key identity variable which, unlike the others, is not directly associated with sexuality. R-square change was 0.02 and F-change was 6.15 (p = 0.014).

The regression model was statistically significant. Of all predictors, outness was the most powerful, followed by feeling upset upon call, everyday discrimination, and identity resilience. These results show that identity resilience explains additional variance in internalised homonegativity when controlling for the effects of everyday discrimination, outness, and feeling upset by recalling a significant coming out experience and thus support H1.

Path analysis

A path analysis (Figure 2) was performed with a bootstrap at 1000 samples with the main predictor of identity resilience; the mediators (social support, perceived negative social representations of gay men, everyday discrimination, feeling upset by recalling a significant coming out experience, and outness) to predict the dependent variable of internalised homonegativity. Model fit was excellent with χ2 (7, 333) = 6.71, p > 0.05, a Root Mean Square Error of Approximation (RSMEA) of 0.000, a Tucker-Lewis Index (TLI) of 1.003 and a Comparative Fit Index (CFI) of 1.000

Figure 2. Path analysis predicting internalised homonegativity.

.

Identity resilience had a statistically significant negative association with internalised homonegativity with a β = −0.13, S.E. = 0.04; p = 0.011. There were also mediation effects.

Identity resilience was negatively associated with perceived negative social representations [β = −0.15, S.E. = 0.04; p = 0.005], supporting H2, which in turn was negatively associated with outness with a β = −0.22, S.E. = 0.13; p < 0.001. This supports H3. Outness was negatively associated with internalised homonegativity [β = −0.33, S.E. = 0.02; p < 0.001], supporting H7. Perceived negative social representations of gay men was positively associated with feeling upset upon recall [β = 0.15, S.E. = 0.04; p = 0.006], which in turn was positively associated with internalised homonegativity [β = 0.17, S.E. = 0.08; p < 0.001], supporting H6.

Identity resilience was positively associated with social support [β = 0.34, S.E. = 0.05; p < 0.001], supporting H2, which in turn was positively associated with outness with a β = 0.44, S.E. = 0.09; p < 0.001 and then internalised homonegativity. There were also mediation effects. Social support was negatively associated with perceived negative social representations of gay men [β = −0.16, S.E. = 0.04; p = 0.003], which in turn was associated with feeling upset upon recall and then internalised homonegativity. This supports H4. Social support was also negatively associated with everyday discrimination [β = −0.20, S.E. = 0.05; p < 0.001], which in turn was positively associated with internalised homonegativity [β = 0.13, S.E. = 0.05; p < 0.001], supporting H5. Everyday discrimination was also positively associated with outness with a β = 0.18, S.E. = 0.10; p = 0.002 and then internalised homonegativity. Everyday discrimination was positively associated with perceived negative social representations with a β = 0.27, S.E. = 0.04; p < 0.001, supporting H5. Perceived negative social representations was negatively associated with outness [β = −0.22, S.E. = 0.13; p < 0.001] which in turn was negatively associated with internalised homonegativity with a β = −0.33, S.E. = 0.02; p < 0.001. Social support was negatively associated with feeling upset when recalling a negative coming out experience [β = −0.17, S.E. = 0.03; p = 0.003], which in turn was positively associated with internalised homonegativity.

Discussion

The results indicate that the self-schema of identity resilience is both directly and indirectly associated with decreased internalised homonegativity. Their association is mediated by various social psychological factors, including accessibility of social support, perceived negative social representations of gay men, everyday discrimination, feeling upset when recalling a significant coming out experience, and outness.

More specifically, the homonegativity variables of everyday discrimination and feeling upset when recalling a significant coming out experience were associated with significantly greater internalised homonegativity, suggesting that, over time, these experiences might stimulate a negative evaluation of one’s sexual identity. Conversely, outness was inversely associated with internalised homonegativity, indicating that, by disclosing their sexual identity to sympathetic others, gay men may be exposing themselves to more affirmative social representations of their sexual orientation. It is thought that this in turn may facilitate a more positive evaluation of one’s sexual identity. Yet, the risk factors for these stressors and coping strategies appear to be explained in part by the individual’s pre-existing level of identity resilience.

Identity resilience

While we found only a modest negative correlation between identity resilience and internalised homonegativity (r = −0.23), identity resilience explains additional variance in internalised homonegativity even after the effects of both the homonegativity variables and outness are taken into account. This may suggest that identity resilience, constructed across the life course and across many domains of life (not just sexuality), may be protective against developing internalised homonegativity. Given that identity resilience reflects the individual’s subjective belief in their self-esteem, continuity, positive distinctiveness and self-efficacy (Breakwell et al., 2021), it is plausible that having a higher level of identity resilience will be negatively correlated with internalised homonegativity. With higher baseline levels of identity resilience, gay men may be better positioned to resist the internalisation of negative social representations of their sexual orientation.

Similarly, identity resilience was negatively correlated with the perception of negative social representations of gay men. This is not to suggest that gay men fail to recognise that there are negative social representations where they exist but rather that they are less likely to believe the majority of heterosexual people hold these negative representations of them if they themselves possess higher levels of identity resilience.

Identity resilience was positively associated with social support, indicating, as predicted, that greater confidence in their sense of self may be leading gay men to believe that they will not be rejected by others and can therefore ask others for support – not only in the domain of sexuality but across other dimensions of identity. Indeed, social support is often limited not because it is unavailable but rather because it is not sought due to fear of stigma and rejection (McDavitt et al., 2008). Our findings suggest that those with higher identity resilience may be more willing to accept the social and psychological risks associated with support-seeking, such as refusal, denigration or rejection. This supports the notion that identity resilience facilitates the effective operation of defensive coping tactics (in this case, the receipt of social support) used by the individual (Breakwell et al., 2021). However, more empirical evidence will be needed to corroborate this observation.

Social support, homonegativity and outness

Social support was negatively associated with all three homonegativity variables: perceiving negative social representations of gay men, everyday discrimination and feeling upset when recalling a significant coming out experience. It is likely that having access to a support network will decrease the likelihood of using maladaptive self-protection measures, such as hypervigilance for evidence of discrimination, following exposure to actual discriminatory events (Breakwell, 2015a).

Using social support generally constitutes an effective and sustainable strategy for coping with stress-inducing events, such as coming out (Sattler et al., 2016). Therefore, it is not surprising that having accessible social support was negatively correlated with feeling upset upon recalling a significant coming out experience. Having support available may encourage gay men to discuss, reflect upon and (re-)evaluate the meanings of their coming out experience, thereby enabling them to come to terms with challenging coming out experiences. In addition, having social support available was associated with higher levels of outness. It is reasonable to assume that general social support from others might empower gay men to engage in the potentially risky practice of disclosing their sexual orientation.

Since our study included no longitudinal evidence, it is possible that being out facilitates access to greater social support rather than social support facilitating outness. However, we predicted, and the path analysis demonstrated, that social support appears to be predicting outness, suggesting that having greater social support may empower people to come out. After all, self-disclosure requires the individual to have a sympathetic other available to them (Breakwell, 2015a). Indeed, anticipated social support has been found to be predictive of self-disclosure experiences (Legate et al., 2012; Sabat et al., 2014). It is important to note that not all coming out experiences necessarily lead to the gaining of social support since some gay men report facing hostility upon coming out (Breakwell & Jaspal, 2021a), which may preclude coming out again in the future to others (Sabat et al., 2014). Nevertheless, feeling generally supported by significant others may increase the likelihood of being out, as indicated in our model. This argument is also consistent with recent developments in identity process theory that suggest that the availability of social support is likely to determine the choice of coping strategy (Jaspal, 2018, 2019). Accessible social support is associated with higher identity resilience. Thus, it is plausible that higher levels of self-esteem, self-efficacy, continuity and positive distinctiveness will equip the individual with the psychological resources to cope with the risk of coming out and of facing the possibility of an undesirable reaction.

We found that the perception of negative social representations of gay men was inversely correlated with outness. Gay men may be, in part, basing their decisions about outness on what they believe are the prevailing social representations of gay men (Hunter, 2007). The decision to refrain from coming out may therefore constitute a self-protection measure against exposure to negative social representations of their sexual orientation and the discriminatory behaviours that these social representations precipitate. Put simply, by concealing their sexual orientation, gay men may think themselves to be less susceptible to the stigma that they perceive to be prevalent in their social context.

Notably, we also found a positive association between everyday discrimination due to one’s sexual orientation and extent of outness, which may seem counter-intuitive (cf., Bry et al., 2017; Quinn, 2017). It could, of course, be claimed that by being more out, one risks more discrimination. Given the nuanced nature of coming out and the observation that people may come out to varying degrees and (strategically) to some but not to others, our alternative explanation should be considered. It is, in fact, consistent with the notion of coming out as a self-protection measure – when gay men experience everyday discrimination first-hand (in contrast to just thinking that social representations of gay men are negative), they may be susceptible to identity threat, which in turn will stimulate action to try to reduce threat (Breakwell, 2015a; Jaspal & Breakwell, 2021). This may include coming out to sympathetic others who are in turn able to provide support – notably, more affirmative responses to their sexual orientation. Although everyday discrimination is harmful for identity and is directly associated with increased internalised homonegativity, through outness (used as a coping strategy) gay men can achieve, proactively, greater self-protection.

Identity process theory attempts to explain how people proactively attempt to cope with adversity, such as exposure to stigmatising social representation. It suggests that being out may facilitate exposure to more favourable representations of one’s sexual orientation and thereby allow the individual to achieve greater feelings of identity authenticity and validation from others on the basis of their sexual orientation (Vannini & Franzese, 2008). All this can facilitate the development of a more favourable self-schema in relation to one’s sexual orientation (i.e. less internalised homonegativity). This aligns with research using identity process theory which shows that people may strategically identify with, and align themselves with, groups and individuals in order to achieve a more positive sense of self (Bonaiuto et al., 1996; Jaspal & Breakwell, 2021). Furthermore, it should be acknowledged that outness (though measured in terms of a state at a given point in time) is actually an ongoing process. Internalised homonegativity may be occurring at any point during the trajectory of coming out. However, through providing exposure to positive social representations of one’s sexuality, being out may actually mitigate the development or perseverance of internalised homonegativity (Breakwell, 2014).

Limitations and future directions

This study has several limitations that should be addressed in future research. First, although the sample was ethnically diverse, no single ethnic minority group was large enough to facilitate comparisons in internalised homonegativity between specific ethnic groups. This would be a worthwhile empirical endeavour for future research. Second, the cross-sectional survey method used indicates the level of association between variables in the posited model. It is not a test of the direction, or indeed existence of, causal relationships. For instance, while identity process theory would suggest that social support may facilitate coming out, it is possible that coming out promotes feelings of social support. Experimental and cohort sequential longitudinal research will help clarify possible causal links. Third, this study focused on only a limited number of possible mediators of the relationship between identity resilience and internalised homonegativity. This might also explain the relatively small amount of variance in internalised homonegativity explained by our regression model. Furthermore, it is noteworthy that many of the correlation coefficients between key variables were small to medium, suggesting that the antecedents of internalised homonegativity are complex and not conclusively determined in our study. Elaborating the theoretical model that we proposed, researchers may measure other stressors, such as childhood adversity, bullying and relationship dysfunction, as possible moderators of this relationship. Qualitative work that can further clarify how these constructs operate in conjunction will be valuable.

Conclusions

This study provides support for a theoretical model, derived from identity process theory, predicting internalised homonegativity in gay men. The results indicate that identity resilience has both direct and indirect associations with internalised homonegativity. Identity resilience appears to be related to this negative self-evaluation through the mediators of awareness and assimilation of negative social representations, social support, everyday discrimination, feeling upset recalling one’s coming out experience and degree of outness. Clearly, there remains a societal need to continue to combat discrimination against sexual minorities. However, the findings also indicate the merit in focusing social and counselling interventions on building identity resilience in gay men experiencing or at risk of internalised homonegativity. This may attenuate the feeling, facilitated by hypervigilance, that negative social representations of gay men are widespread in one’s social context, facilitate outness (identity enactment) and, crucially, reduce internalised homonegativity and its insidious effects on psychological wellbeing.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data can be accessed by contacting the corresponding author.

Supplementary material

Supplemental data for this article can be accessed here.

Additional information

Notes on contributors

Rusi Jaspal

Rusi Jaspal is Pro Vice-Chancellor (Research and Knowledge Exchange) and a Professor of Psychology at the University of Brighton. He is the author of 'The Social Psychology of Gay Men' (Palgrave, 2019).

Glynis M Breakwell

Dame Glynis Breakwell is a professor of social psychology whose research focusses upon identity processes, reactions to risk and the dynamics of mistrust.