The Relationship Between ‘Coming Out’ as Lesbian, Gay, or Bisexual and Experiences of Homophobic Behaviour in Youth Team Sports

Introduction Homophobia appears to be greater in sport settings than in others. However, little is known about whether lesbian, gay or bisexual (LGB) young people experience discriminatory behaviour in team sports because of their sexuality and whether coming out to sport teammates is associated with homophobic behaviour. Method This study used a sample (N = 1173; 15–21 years; collected in 2014–2015) from six countries (United States, United Kingdom, Canada, Australia, New Zealand, Ireland) to examine whether LGB youth who ‘come out’ to teammates experience homophobic behaviour. Results Close to half of the sample (41.6%) reported having been the target of homophobic behaviour (e.g. verbal slurs, bullying, assaults). Multivariate logistic regression models adjusting for age, gender, country and contact sport participation found that participants who ‘came out’ as being LGB to sports teammates were significantly more likely to report being a target of homophobic behaviour. There appeared to be a dose response with coming out to more people associated with a greater likelihood of experiencing homophobic behaviour. Conclusion The study results suggest a relationship between coming out as LGB and encountering homophobic behaviour in team sports. LGB experiences of homophobic behaviour appear common overall in this sample, but are greater in those who have come out to teammates. Policy Implication Sports administrators and governments need to develop programs and enforce policies that create safe sports environments where LGB youth can participate without encountering homophobic behaviour.


Introduction
Increasing participation in sport is a public health priority in many countries (Mansfield & Piggin, 2016). Participation in sport and particularly team sports provides unique psychosocial and physical health benefits to young people (Dohle & Wansink, 2013;Eime et al., 2013). Youth who play a team sport report higher levels of self-esteem, confidence, resilience, better social skills, and fewer depressive symptoms than those who do not play a team sport (Bailey, 2006;Dodge & Lambert, 2009;Vella et al., 2014). Accordingly, playing a team sport could be of particular benefit to lesbian, gay, and bisexual youth (LGB) who experience disproportionally high rates of depression, suicidality, and self-harm (Blais et al., 2015;Kann et al., 2018;Birkett et al., 2015;Lourie & Needham, 2016;Mereish & Poteat, 2015). However, recent population studies, such as large-scale longitudinal research (N ≈ 99,000) in Canadian high schools, have found that LGB youth are consistently less likely to report playing a team sport than heterosexual youth (Doull et al., 2018).
The most recent wave of data in the Canadian study shows that the largest disparity is between gay and heterosexual males, with gay males reporting active participation in team sports at half the rate of heterosexual peers (32.8% vs 67.6%). Similar disparities in sports participation rates have also been found in biennial national surveys conducted by the Centers for Disease Control and Prevention in American high schools (Kann et al., 2016(Kann et al., , 2018. The latest report found 61.2% of heterosexual males and 54.1% of females reported playing a team sport during the last 12 months compared with 40% of gay/bisexual males and 38% of lesbian/bisexual females. There are multiple reasons why a young person may choose to play, or not to play, team sports. However, low levels of participation among LGB youth would suggest that there are specific barriers discouraging this population from joining a team. Literature reviews and government reports in the United Kingdom (UK), Canada and Australia suggest that homophobia and homophobic behaviour are likely to be a primary participation deterrent (Brackenridge et al., 2007;Government of British Columbia, 2017;Greenspan et al., 2019a;Sport and Recreation Victoria, 2017;Sport England, 2018;UK Government, 2015). A recent UK Parliamentary Inquiry (Digital, Culture, Media, Culture, and Sport Committee, (UK), 2017) also called for action by sports governing bodies to end homophobia and homophobic behaviour because it had "serious concerns over the effects of low participation among LGB youth on their mental and physical health and well-being … It appears that young players and athletes sometimes feel (due to fear of being the target of discrimination) that they have to make the active choice between either coming out or continuing to participate in their chosen sport" (p. 9).
UK Parliamentarians highlighted the need for quantitative research examining the extent to which LGB youth experience homophobic behaviour in team sports and the impacts of that behaviour. A range of studies similarly note the need for published quantitative research investigating team sports experiences in LGB youth samples (Anderson et al., 2016;Brackenridge et al., 2007;Greenspan et al., 2017). However, the majority of research in LGB youth samples to date has adopted qualitative methods, which while crucial to understanding the issues, cannot establish the statistical relationships between the various factors at play. Such information is often important in shaping government policy.
Although there is a paucity of published research with LGB youth, it is reasonable to suggest that homophobic behaviour would be a factor in lower participation rates in light of quantitative evidence that homophobic behaviour in team sports has a range of negative impacts on LGB young adults (Pistella et al., 2020) including deterring them from participation (Baiocco et al., 2018). Quantitative studies (Herek, 2007(Herek, , 2015Herek & McLemore, 2013) in non-sport environments have also found that LGB youth avoid environments, or hide their sexuality, in settings where homophobic behaviour is common and where those that come out (reveal their sexuality) as LGB are likely to be the target of homophobic behaviour. For example, LGB youth are more likely to avoid school if they perceive homophobic attitudes and behaviour to be common and if they have been the target of this behaviour (Black et al., 2012).
Research has described homophobic attitudes and behaviour to be common in youth team sports (Brackenridge et al., 2007;Greenspan et al., 2019a;Morrow & Gill, 2003;O'Brien et al., 2013;Osborne & Wagner, 2007;Southall et al., 2011). Research in Canadian high-schools has also found that 89% of LGB students who entered locker rooms or school sport environments reported hearing homophobic language (e.g. fag, dyke) and nearly half (48%) heard this language 'frequently' or 'often' (Morrison et al., 2014). These findings are consistent with quantitative research conducted with physical education (PE) teachers and heterosexual athletes (Atteberry-Ash et al., 2018;Elze, 2003;Gill et al., 2010;Southall et al., 2011). Piedra et al. (2016) found that 92.7% of PE teachers had witnessed homophobic behaviour by students toward other students.
However, evidence from qualitative studies examining whether LGB youth are the target (victimised) by homophobic behaviour in sport settings is mixed. A range of studies have described youth team sports as environments where LGB identities are stigmatised and homophobic behaviour is common (Brackenridge et al., 2007). For example, a recent study of American youth (N = 71; 13-18 years) reported half of LGB participants had experienced sexuality-based harassment or assaults in the last year while playing sports and described "immense feelings of discomfort, and minimal ally ship or effective intervention from athletic staff even in the face of blatant homophobic and transphobic comments (from others)" (Greenspan et al., 2019a, p. 425). Male sports, particularly contact ball sports like rugby or American football, have been found to have highly masculine cultural milieu where individuals (e.g. gay people) who do not conform to traditional gender roles and norms are rejected and denigrated (Cunningham, 2019;Greenspan et al., 2019a;Osborne & Wagner, 2007;Steinfeldt et al., 2012).
Girls and women also report being stigmatised/stereotyped as lesbians and experience discrimination if they play traditionally male team sports (Greenspan et al., 2019a;Jeanes & Kay, 2007). In traditionally female sports (e.g. volleyball, netball), lesbians or bisexual women also report discrimination for not conforming to feminine appearance norms (Brackenridge et al., 2007;Krane, 2016). Homophobic behaviour in both male and female youth team sports appears to have multiple purposes including being used by athletes to distance themselves from homosexuality, signal their conformity to gender norms, facilitate bonding (e.g. through derogatory jokes about gay people), or just to insult others (Magrath, 2017;McCann et al., 2009;McCann et al., 2010).
However, detailed studies by Anderson, Adams, Magrath and colleagues suggest that homophobic attitudes and behaviours are no longer as common in team sport environments as the literature above would suggest (Adams & Anderson, 2012;Anderson et al., 2016;Magrath, 2017;Bush et al., 2012). They argue that this is due to changes in gender norms that underpin traditional attitudes and behaviours in these settings, with younger people in particular more accepting of diverse sexual identities. Anderson and colleagues ethnographic research suggests high school, university and community-based levels of sport are now "safe spaces" for gay and lesbian athletes with few reporting being the target of homophobic behaviour (Anderson et al., 2016, p. 147). Magrath (2017) acknowledges that the use of words such as 'fag' and derogatory jokes about gay people remain common in sport, but also suggests this "homosexually-themed" behaviour is not necessarily directed toward LGB athletes and is not considered by LGB athletes to be "homophobic" (p. 118). Instead of seeing this language as damaging to LGB youth, some studies suggest this language and phrases such as 'that's so gay' are used without malice or homophobic intent and provide a bonding mechanism between heterosexual and gay men. Anderson et al. (2016) suggest that low participation rates of gay men in sport may not be due to homophobia per se, but could be related to other factors including tastes and preferences as well as wider social factors. In summary, the literature presents conflicting findings regarding homophobia within sport. The current study seeks to provide further understanding of the experiences of LGB youth in team sports environments.
This brief research paper reports the results of an analysis of survey responses from both closeted, and openly LGB youth (N = 1173; ages 15-21) from six countries. We investigated whether these young people reported being the target of behaviour in team sport settings which they perceived to be homophobic (e.g. slurs, bullying, assaults). Our focus was on whether participants who had come out as LGB to their teammates were more, less or equally as likely as those who remained closeted to report they had been the target of homophobic behaviour. The aim of this paper is to provide evidence that can add to existing understanding of homophobic behaviour in sport and assist in informing future policy directions.

Sports Participation
Participants were asked if they had played on a youth sport team and which sports they played. They were provided with an initial list of 18 team sports (e.g. soccer, rugby, basketball, Gaelic football, lacrosse) and/or could provide the team sport they played via a text box. Because previous research (e.g. Southall et al., 2011) has found higher levels of homophobia in male contact ball sports, we created a dichotomous variable to indicate (0 = No, 1 = Yes) whether participants who had specifically played a contact ball sport (i.e. Rugby Union, Rugby League, American Football, or Australian Football).

Came Out to Teammates
Participants who played on a sport team were asked if they had come out to their teammates. Specifically, they were asked if they had 'come out' to: no one, one or two people, some people, everyone.' This measure was adapted from the biennial surveys of American LGB high-school students (Kosciw et al., 2015). To avoid problems with the existing measure whereby distinctions between two response options for indicating how many people they had come out to would be difficult 'one or two' and 'some' people, we collapsed these two responses and associated data to form a single category called 'partially out' to people. Accordingly, three responses (scores) were used for analysis (0 = Not out to anyone, 1 = Partially out, 2 = Out to everyone).

Experienced Homophobic Behaviour
The outcome variable was whether participants had been the target of homophobic behaviour in team sports. While there are a number of measures of general bullying, there was no existing measure to specifically assess whether an LGB person had experienced (been a target) of homophobic behaviour within team sports. To assess this, participants were asked if homophobic behaviour had been directed toward them in a team sport environment. Examples of the homophobic behaviours were provided ("e.g. jokes, humour, assaults, bullying, slurs") in the survey. Participants responded yes or no to this question. For analysis, yes and no responses were coded as 1 and 0, respectively.

Procedure
Participant recruitment and data collection were conducted between June 2014 and January 2015. Data was collected via a 10-min online survey conducted by Nielsen Sports on behalf of key partners including government, non-government/sport and corporate organisations. These organisations included the Australian Government, International Gay Rugby, Bingham Cup Sydney 2014, International Federation of Gay Games, You Can Play (charity), ACON (health charity) and the Commonwealth Bank of Australia.
Participants were asked to participate in a "study into the lesbian, gay, bisexual community's experiences, primarily in organised and competitive team sport which will greatly assist in helping the wider community understand issues affecting the LGB community." The study was promoted using professional sport athletes who encouraged people to take the survey and 'share their story' regardless of sexuality or sport experience.
The survey link was also shared by mainstream sporting organisations (e.g. World Rugby, Cricket Australia), mainstream sport media outlets including ESPN (global), EuroSport (UK), TSN (Canada), Sky Sports (Australia) and by LGBT+ media outlets including Pink News (UK), Star Observer (Australia) and Outsports (USA). Links to the study were also paid 'promoted' on Facebook by the LGBT+ charities Stonewall (UK) and Sydney Gay and Lesbian Mardi Gras. Finally, cards advertising the study and iPads with direct links to the survey were provided to sportspeople and spectators at two international sporting events: The Gay Games (Cleveland) and the Sydney Bingham Cup (world-cup of gay and inclusive rugby). The use of multiple methods, and channels, to recruit LGB people, follows guidelines for conducting research with this marginalised population from the American National Academy of Medicine (National Academy of Medicine, 2011). Informed consent was obtained from all individual participants included in the study. Human Research Ethics Committee clearance was sought, but deemed not to be needed for this secondary analysis of the dataset provided, which contained no identifying information.

Analyses
Because we were interested in the experiences of young LGB people who had played team sport, all statistical analyses were conducted on LGB team sports participants only (N = 975, 83% of total sample; male N = 611, 78% of all males). Cross-tabulations and Chi-square values were calculated to test for differences between males and females on variables of interest. Spearman's correlation coefficients were calculated to establish bivariate relationships between variables.
Multivariate logistic regression models tested the relationship between experiencing homophobic behaviour and demographic variables (gender, age, country), contact ball sports participation and coming out to teammates. Table 1 displays frequencies, proportions and significant differences on variables of interest by gender. Female participants (92.9%; N = 364) were more likely than male participants (78.2%; N = 611; X 2 (1) = 39.784, p < .001) to reporting playing on a sport team. Most participants reported hiding their sexuality from some or all of their teammates. The majority of participants reported being out to 'no one' on their sport team, with just 20% of the sample reporting being out to 'everyone.' Males were significantly more likely than females to report playing a contact ball sport. Males were also more likely than females to report they had experienced homophobic behaviour in team sports environments. Nearly three-fifths of participants who came out to everyone (57.7%; N = 109), and nearly half who partially came out (46.6%; N = 118) reported they had been the target of homophobic behaviour in team sports, whereas two-fifths (40.4%; N = 203) of participants who came out to no one reported being a target of this behaviour. Table 2 displays Spearman's correlation coefficients for the variables of interest. Significant relationships were only found between gender, coming out to teammates and experiencing homophobic behaviour. Participants who came out were significantly more likely to report being the target of this behaviour.

Results
As can be seen in Table 3, after accounting for other variables in the model, only gender (being male) and coming out to teammates were associated with experiencing homophobic behaviour. Participants who partially came out to their teammates had 1.5 times higher odds, and those who came out to everyone had 2.2 times higher odds, of reporting they had been the target of homophobic behaviour, compared with participants who did not come out to anyone. Finally, the odds of experiencing homophobic behaviour were 2.1 times higher for males than females.

Discussion
There had been no quantitative research examining whether coming out as LGB to sport teammates is associated with being more or less likely to be a target of homophobic behaviour in a sport setting. The present study examined this question in a multi-country sample and found that coming out in a team sport setting was associated with a greater odds ratio for experiencing homophobic behaviour. There is some evidence for a dose-response relationship whereby coming out to more teammates is associated with greater odds of having encountered homophobic behaviour. More broadly, a large proportion of the sample reported having been the target of this behaviour in sport, regardless of whether they had come out to teammates or not.
There had been some suggestion in the research literature (Southall et al., 2011) that participating in contact ball sports might be associated with more experiences of homophobic behaviour. Here, we did not find that playing a contact ball sport to be associated with homophobic behaviour in either bivariate or multivariate analyses. However, consistent with previous studies (Brackenridge et al., 2007;O'Brien et al., 2013), males in our study were found to have higher odds ratios for being the target of homophobic behaviour than females. It is also worth noting that the majority of young people in the study reported concealing their sexuality from their teammates. The results of our study raise the question of whether LGB youth in team sports may be less likely to come out to others than in non-sport settings. For example, US  − .14** .05 .13** .10* *p < .01, **p < .001 a Gender: 0 = F, 1 = M research in LGB high-school students (Kosciw et al., 2015) found that 21.6% of students had not come out to anyone at their high school, whereas the present study found that 53.2% of participants had not come out to anyone in their team. Although the unadjusted odds of having encountered homophobic behaviour in the United Kingdom and New Zealand appear to be higher and lower, respectively, when compared with the United States (Table 3), after adjusting for other factors, there was no significant difference in the odds of encountering homophobic behaviour. Future research examining whether coming out is more or less common in sport vs. non-sport settings would be of interest, as would research examining differences across countries and different sports. This paper provides new evidence suggesting that homophobic behaviours remain problematic in youth team sport settings. The results of our study are consistent with the findings of quantitative studies conducted with LGB high-school students (Elze, 2003;Greenspan et al., 2019a;Greenspan et al., 2019b;Morrison et al., 2014), LGB young adults (Baiocco et al., 2018;Pistella et al., 2020), physical education teachers (Piedra et al., 2016) and heterosexual athletes (Gill et al., 2010;MacDonald, 2018;O'Brien et al., 2013;Southall et al., 2011;Steinfeldt et al., 2012). These studies have reported homophobic attitudes and behaviour to be common, particularly in male team sports environments. However, the findings differ to the conclusion of several studies that suggest LGB people who come out in team sports do not experience discrimination, and, therefore, homophobic behaviour is not an ongoing issue (Anderson et al., 2016;Magrath, 2017). This may be due to differences in the sports and or clubs examined by these researchers. It is also potentially due to differences between studies in what is understood by homophobic behaviour, with Anderson et al. (2016) suggesting that slurs and homophobic language may not be indicative of homophobic intent and therefore do not constitute homophobic behaviour.
Collectively, the results of the present and other studies suggest that homophobic behaviour is commonly encountered in youth sports environments (Greenspan et al., 2019a;Shaw, 2019). Whether some of this behaviour represents negative attitudes and antipathy toward LGB youth is uncertain, as some authors suggest that it may simply be due to unthinking casual normative language and behaviour that is not necessarily intended to be derogatory or hurtful (Anderson et al., 2016). Regardless, research on the effects of stigma suggests that such behaviour does negatively impact LGB youth, and accordingly, they may be more likely to try to avoid environments, such as sport, due to the presence of this homophobic behaviour (Herek, 2007(Herek, , 2015. This may also explain the large disparities in team sports participation between LGB youth and their heterosexual peers reported by recent population studies (Doull et al., 2018;Kann et al., 2018).
Being the target of homophobic behaviour is harmful and associated with higher rates of suicide, self-harm and depression among LGB youth (Birkett et al., 2015;Burton et al., 2013;Lourie & Needham, 2016), whereas playing team sports in a safe and supportive environment, free of discriminatory behaviour, has been shown to provide LGB youth with many psychosocial benefits (Blais et al., 2015). Ending discriminatory behaviours and encouraging more LGB youth to play team sports has the potential to help close the gap across a wide range of negative psychosocial and health disparities between LGB youth and their peers (Greenspan et al., 2019a).
The results provide empirical support to calls by other scholars (Greenspan et al., 2019b), LGB advocates (Englefield et al., 2016), educators (GLSEN, 2013 and policy makers in various countries (Digital, Culture, Media, Culture, and Sport Committee, (UK), 2017;Shaw, 2019) for action on homophobic behaviour by sport organisations. The leaders of large sport organisation in the United Kingdom, United States, Australia and New Zealand have also made public commitments to end homophobia in their sports (Home Office, 2011;Portwood, 2015;Shaw, 2019;World Rugby, 2015). The findings of this study suggest that these commitments may not have been followed by meaningful action to end this discriminatory behaviour (Shaw, 2019;Storr et al., 2018). The findings also support the need for sport administrators, clubs and coaches to put in place policies and procedures for sanctioning homophobic language and behaviour. The current data, and research by Anderson et al. (2016) suggest that this may not be occurring.
There are limitations to the study. Although the study followed recommended sampling approaches for difficult to reach and highly stigmatised populations (Kosciw et al., 2015;Kull et al., 2016; National Academy of Medicine, 2011), the sample is purposive in nature and this limits generalisability of the findings. The sampling approach used may also have resulted in recruitment of LGB youth who were more likely to have encountered homophobic behaviour and thus were more likely to want to share their experiences. However, the results suggest that there was considerable heterogeneity in the sample for age, gender, those encountering homophobia and/or those participating in contact ball sports. Although it is reasonable to suggest 'coming out' precedes experiences of homophobic behaviour, the correlational design of the study means that causal inferences cannot be made on the relationship. The measure of homophobic behaviours used did not distinguish between different types of homophobic behaviour. As such, we were unable to examine if specific types of homophobic behaviour are more or less affected by coming out. The lack of psychometrically developed and tested measures in this area is a challenge for the field to address. Large-scale, longitudinal studies using representative samples and multi-dimensional scales are needed to address many of these limitations. The absence of rigorous longitudinal research speaks to the expense of such studies, and the absence of funding to support them.
Notwithstanding the limitations, this study investigated for the first time whether coming out to team mates would be associated with more or less experiences of homophobic behaviour by LGB young people. In doing so, the research has addressed an important gap in the literature and provided much needed evidence to guide policy makers and sports administrators interested in increasing participation in sport and creating a more inclusive sporting environment for all. Stronger policy stances that engage the public and provide better protections and rights are associated with improved attitudes towards LGB populations (Ofosu et al., 2019). Worryingly, the results suggest that being openly LGB may result in encountering more homophobia. This may, to some extent at least, explain why there are so few openly gay, lesbian, or bisexual athletes in high profile sports. There are multiple possible reasons why sportspeople may engage in homophobic behaviour, including the possibility that such behaviour reflects an unconsidered adherence to current social norms amongst other players or fans, rather than being driven by negative attitudes towards LGB people per se, an argument advanced in McCormack et al. (2016) examination of homophobic language use amongst young men. Clearly, more research is needed to better understand why LGB youth continue to encounter homophobic behaviour and language in sport settings.
The present results, alongside a large body of research detailing the extent and nature of homophobia in sport, provide policy makers and sport administrators with supportive evidence in which to develop strategies aimed at reducing homophobic behaviour in sport. Although there is some evidence that education may be effective in reducing homophobia (Baams et al., 2017), stronger regulatory principals and actions are needed to address prejudices around sexuality. In so doing, better regulations and policies can enhance the health and well-being of LGB youth through their increased and/or ongoing participation in team sport.