Is LGBT homelessness different? Reviewing the relationship between LGBT identity and homelessness

Abstract

This article explores existing evidence concerning the causes and experience of homelessness among lesbian, gay, bisexual and trans (LGBT) people. A total of 88 sources, published after 2000, were included in the review. This represents the most comprehensive evidence review to date on the issue, mapping out what is known about LGBT homelessness and identifying gaps in knowledge for future research. With a few notable exceptions, the experiences of homeless LGBT people have been neglected in UK housing and homelessness literature. Despite this caveat, we found clear evidence that LGBT people are overrepresented among homeless populations. We argue, however, that the how and why of LGBT identity increasing a person’s chance of becoming homeless is complex and not yet well understood. The article therefore calls for a future research agenda that offers theoretically informed causal analyses to better understand the intersections of LGBT identity and homelessness.

Keywords:

• LGBT
• homelessness
• homeless pathways
• evidence review
• identity

Introduction

This article draws on the findings from an evidence review into homelessness among those who identify as lesbian, gay, bisexual and trans (LGBT).¹ The evidence review formed part of a larger externally funded study pulling together primary and secondary data sources on LGBT homelessness in the UK but which was unpublished at the time of writing. The research involved scrutiny of 88 sources published after 2000, representing the most comprehensive evidence review to date on the issue, mapping out what is known about LGBT homelessness and identifying gaps in knowledge for future research to address. We offer a critical examination of the evidence contained in one core strand of this review concerned with what we know (and do not know) about the causes of LGBT homelessness.

We begin the article by detailing the background to the study and our review methodology. In this first section we also highlight the methodological limitations of this international body of work. In section two, we present an overview of what the evidence base tells us about the scale of homelessness among LGBT populations, before moving on to a discussion of extant knowledge regarding the causes of homelessness among LGBT groups. We discuss how the experiences of LGBT people have been somewhat neglected by housing and homelessness scholars in the UK, but a relatively large body of international evidence (mainly quantitative) from the US and, to a lesser extent, Canada confirms a statistical relationship between LGBT identity and homelessness. The evidence clearly and consistently indicates that there is a greater prevalence of homelessness among LGBT people compared to those who do not identify as such. Related to this, quantitative studies which measure the prevalence of certain independent ‘predictors’ of homelessness also suggest that LGBT people are at greater risk of experiencing some of the known biographical factors that contribute to homelessness (for instance, childhood adversity, mental health problems, substance misuse). Family rejection (linked to non-acceptance of sexual or gender diversity) is often cited as the key reason for youth LGBT homelessness. Yet, we suggest, these pathways are currently under-researched, under-theorised and inevitably not well-understood.

The third section of the article moves beyond the frequently documented statistical associations between LGBT identity and homelessness which often belies the complex and multi-dimensional nature of the relationship. Here, we report on a smaller body of qualitative research which offers more explanatory accounts of the role of LGBT identity in people’s pathways into homelessness. Our paper concludes with the claim that the evidence base around the causes of LGBT homelessness is still in its infancy and so our collective understanding of how and why being LGBT might increase a person’s chance of becoming homeless is not yet well understood. Reflecting developments elsewhere within housing studies, we argue therefore for a future research agenda that offers theoretically informed causal analyses to better understand the intersections of LGBT identity and homelessness.

Background to the study and methodology

The article draws on the findings from an externally funded rapid evidence review of LGBT people’s experiences of homelessness undertaken by researchers at Sheffield Hallam University between January and March 2019. The review sought to summarise and synthesise existing evidence on the causes of homelessness, homelessness journeys, and the experiences of housing and support services for both LGBT and non-LGBT people. The aim of the review (and the wider study) was to develop an understanding of LGBT homelessness within a UK context, hence the discussion that follows is geared towards what the findings tell us about LGBT homelessness in the UK.

The study employed a ‘rapid’ rather than a ‘systematic’ methodology. This was considered the most appropriate methodology given the time and resource available. However, a number of standards used in ‘systematic’ evidence reviews were incorporated, including searching for grey literature. The review followed a thorough and rigorous process of searching, assessing and reviewing, but allowed for more subjective assessments of the utility of sources than systematic reviews. A multi-pronged search strategy was employed to identify evidence published after 2000 using a range of tools and databases. The database search was conducted using a combination of 52 different search terms (see Appendix for further details). We also issued a ‘call for evidence’ in which academics and organisations were invited to recommend high quality research studies and sources of evidence. Individuals were contacted directly and through relevant networks and mailing lists, including JISCMAIL.

The initial search and call for evidence yielded 167 sources whose titles and abstracts were screened against inclusion criteria and then scored against the EPPI-Centre’s Weight of Evidence framework (Gough, 2007). The EPPI-Centre Weight of Evidence framework follows four dimensions: coherence and integrity; specific judgement on forms of evidence; specific judgement on relevance; and overall judgement. Studies are scored low, medium or high against each dimension. A total of 88 sources were included in the review after being scored medium to high in the Weight of Evidence framework, or as a result of containing a particularly useful insight. However, given the acknowledgement that the quality of evidence on LGBT research can be hampered due to sampling and data limitations (discussed further in this section) (Hudson-Sharp & Metcalf, 2016), a small number of studies which scored ‘low’ on the quality dimensions of the framework were included. Usually this was where the content of a source was highly relevant and so scored highly against this dimension of the framework and it concurred with other more robust sources. Of the 88 studies, 66 (75%) were international (non-UK). The body of work was dominated by evidence derived from quantitative data; only a small proportion of the evidence identified and judged to be ‘robust’ was qualitative in nature. An additional 14 non-LGBT specific sources were reviewed for comparative purposes, several of which were robust, recent evidence reviews that included many other sources.

A number of caveats and limitations to the evidence are important to note. Many sources present statistics but fail to provide an explanation of their evidential basis; they use different terminology, measurements, comparators, and are often talking about different sample populations (a useful discussion of these issues is provided in Ecker, 2016 and Ecker et al., 2019). Across the literature, definitions of homelessness are often contingent upon the sampling frames that are used in the studies. For example, some only sample homeless people in shelters, drop-in centres, the street, or schools. Some have a broader sampling frame not focused exclusively on a homeless population but rather e.g., LGBT people within a specified locale/community, school setting or healthcare service. A number of studies in the review also focused on specific sub-populations (e.g., gay males, men who have sex with men). This means that studies only capture a small proportion of the homeless population and questions remain about the generalisability of the findings. Lastly, definitions of homelessness can sometimes imply that it is a static experience, yet it is increasinlgy recognised that homelessness is better understood as a journey or trajectory that is shaped by geo-political contexts.

The terminology around sexuality and gender identity is fluid and changes over time and geography according to the social and political landscape, but commonly adopted terms vary most significantly between countries (at the same point in time). There is no standardised terminology across studies and, therefore, no uniform processes for determining the sexuality or gender identity of research participants. Regarding the former, the majority of studies used pre-determined options from which participants could self-identify their sexual or gender identity. For data analysis purposes, some quantitative studies reduce sexual identity categories into a dichotomous variable with heterosexual being one, and all of the other choices (gay, lesbian, bisexual, etc.) placed into a non-heterosexual category. Transgender individuals are sometimes removed from analyses or combined with LGB people because of their small numbers. Studies do not commonly disaggregate lesbian, gay, bisexual and trans samples. While more recent studies are more likely to differentiate or focus analytical attention on certain sexual and gender minority groups, there is a dearth of literature that focuses exclusively on the relationship between gender identity and homelessness, with trans people distinctly underrepresented in the evidence. Where trans people are included, they are rarely divided into the categories of ‘transgender male’ and ‘transgender female’ (less still other minority gender identities). Across all studies of this nature there is also the potential for response bias. Despite implementing consent procedures and methods for ensuring anonymity, participants may not have been willing, or felt safe, to disclose their sexuality or gender identity. It is also instructive to note that the tendency of most literature is to focus on LGBT homelessness within a youth context (adolescents and young adults, mainly between the ages of 16-25) (Yu, 2010). This is possibly because the average age of ‘coming out’ has decreased within the last ten years and being financially reliant on a parent or guardian may place this group at greater risk of homelessness. Furthermore, while the proliferation and wider awareness of sexual possibilities is abstractly available to all young people, this has not necessarily led to a corresponding extension of a social infrastructure to inform, support and validate sexual diversity (Prendergast et al., 2001; Dunne et al., 2002, pp. 110–111).

The scale and prevalence of LGBT homelessness

This section brings together the evidence on the prevalence of homelessness among LGBT people. It also collates the findings from a range of research studies that have attempted to measure the extent to which LGBT people are more likely than their non-LGBT counterparts to become homeless.

A number of studies from the US and, to a lesser extent, Canada have attempted to measure and provide estimates of the proportion of the homeless population who identify as LGBT, and while these studies contain methodological limitations (as discussed above), they consistently point to a disproportionate number of LGBT people who are experiencing (or have experienced) homelessness. These studies frequently indicate that LGBT people are more likely to experience homelessness than their heterosexual, cisgender peers. This claim is based on a body of methodologically rigorous research that indicates that LGBT people are disproportionately represented within homeless population groups (i.e., more homeless people identify as LGBT than do people in the general population, indicating a greater likelihood of experiencing homelessness) and more likely than heterosexual cisgender people to report experiencing homelessness within comparative studies.

Within US and Canadian literature, the proportion of (mainly young) homeless respondents identifying as LGBT ranges from 15 per cent (Leslie et al., 2002) to 44 per cent (Gattis, 2013), and it is estimated that LGBT youth comprise approximately 40 per cent of the clientele served by 354 agencies from across the US (Durso & Gates, 2012). In terms of the existing comparative data, a US study of over 6,000 school students found that LGB young people had an odds of reporting homelessness that was between four and 13 times that of their heterosexual peers (Corliss et al., 2011). Another large quantitative study of 1,839 high school students similarly found that lesbian, gay, bisexual, transgender, or questioning (LGBTQ) respondents were more likely to report staying the night in a public place or with a stranger (Rice et al., 2013). In this survey there was a statistically significant relationship between identifying as LGBTQ and any experience of homelessness in the past year. In another US study, sexual orientation was the only statistically significant correlate of different types of homelessness experiences (Rice et al., 2015).

It is important to take into account the fact that this body of empirical work almost exclusively focuses on the experiences of young people under 25, with only a few exceptions (e.g., Kattari & Begun, 2017; Flentje et al., 2016). This may have the effect of inflating LGBT homelessness statistics because young people are more likely to openly identify as LGBT, at least in countries which have seen changing social norms and demographics as well as changing policy and legislation (Mercer et al., 2013, Dahlgreen & Shakespeare, 2015) and/or be homeless (Watts et al., 2015). The dearth of research into homelessness among LGBT people during later adulthood and longitudinal research that explores developing sexual orientation and gender identity reflects a key gap in the international evidence base (Ecker, 2016; Ecker et al., 2019).

Within the UK literature, there are very few national or reliable large-scale measures of the number of LGBT people who are homeless. Reflecting the international evidence, those we did identify also pointed to an over-representation of LGBT people within their homeless population samples as well as a disproportionately high rate of homelessness within LGBT samples (although not as high as those identified within US studies). In research conducted by the charity DePaul (McCoy, 2018), 16 per cent of 712 young people living in temporary accommodation identified as LGBT. However, a headcount carried out by Porchlight (2015) in England in 2011, identified that double this number (32 per cent) of residents in their young persons’ services identified as LGBT, a significant over-representation compared to the national average. ONS figures from 2020 suggest that around 3.1 per cent of the UK population (aged 16 years and over) identify as LGB, an increase from 2.7 per cent in 2019 and almost double the percentage from 2014 (1.6 per cent) (ONS, 2022). There are no national figures for trans people but Stonewall estimate that around 1 per cent of the UK population are trans, which is also in line with Gender Identity Research & Education Society estimates of between 0.6 per cent and 1 per cent of the UK adult population (Government Equalities Office, 2018).

In terms of wider population samples, an online, self-selecting survey of over 5,000 LGBT people in Great Britain conducted in 2017 found that 18 per cent had experienced homelessness at some point in their lives. For trans respondents, this increased to one in four (25 per cent). LGBT women were more likely to have experienced homelessness than men (20 per cent and 15 per cent, respectively) (Bachman & Gooch, 2018). A local study in Brighton and Hove found similar rates of homelessness, with 22 per cent of the 819 LGBT respondents reporting experiences of homelessness. For trans people the figure was 36 per cent (Browne, 2007; Browne et al., 2007). There are no ‘robust’ comparative surveys of the general UK population that demonstrate high levels of methodological rigour and so it is not known whether these studies indicate high, low or average rates of homelessness among LGBT people in England compared to people not identifying as LGBT.

The evidence on the causes of homelessness: is LGBT homelessness different?

Within the general homelessness literature, there have been considerable attempts over the last twenty years or so to provide improved understanding of causation in studies of homelessness (Batterham, 2019). Drawing on and developing this body of scholarship, recent work by Bramley & Fitzpatrick (2018) concludes that a combination of individual, interpersonal and structural factors all play a role in homelessness, although their analysis underlines the centrality of poverty, especially in childhood, to the generation of homelessness.

Notwithstanding evidence about the influence of structural factors such as poverty on homelessness, there is also strong evidence that certain life histories place particular groups at higher risk of homelessness than others, including those with an experience of the care system, young people who have offended, people with childhood experiences of abuse and/or neglect, poor educational experiences and mental health issues (Watts et al., 2015). Literature focused on youth homelessness has consistently found that relationship breakdown between young people and their family to be a leading cause of homelessness for this group (Centrepoint, 2020). Adding nuance to this broad conclusion, it is suggested that the link between young age and homelessness is more accurately explained by the disproportionate experience of poverty among young people, rather than their youth, per se (Watts et al., 2015). As the most recent comprehensive evidence review on the causes of homelessness in the UK (Alma Economics, 2019) points out, many of these factors interact and are by no means mutually exclusive. Neither is each factor or combination of factors completely necessary or sufficient in causing homelessness - so that many who experience them may never become homeless.

In thinking about the causes of homelessness for LGBT people, it is important to bear in mind that people who identify as LGBT are subject to and experience the same structural constraints and personal problems that place non-LGBT people at risk of homelessness. If domestic abuse, poverty, or unemployment increases the overall risk of homelessness then they will surely increase an LGBT person’s risk of homelessness as well. This means that while the drivers of LGBT homelessness are likely to be the same as for those who do not identify as LGBT, they might also be related to minority gender and sexual identity in distinct ways. In trying to understand the relationship between minority gender and sexual identity and homelessness, then, the questions to pose might be: a) whether there are any factors over and above those identified in the generic homelessness literature (i.e. additional factors) that apply only to LGBT people; b) whether there are nuances and variations in the causal explanations of homelessness of relevance specifically to LGBT people (e.g. familial abuse is a homelessness trigger for younger people but in the case of LGBT young people, abuse may be related to their gender or sexual identity); and c) whether LGBT people are more likely to be affected by those structural or personal factors known to be triggers or causes of homelessness (for example, mental health issues or unstable employment).

Our review revealed that much of the existing evidence on the causes of LGBT homelessness is derived from quantitative studies. Such evidence is useful for establishing the comparative prevalence of different experiences and outcomes using robust statistical measures. These give us valuable indications about the likelihood of a link between being LGBT and becoming homeless. It is useful too that fourteen of these studies comprise comparative analyses of LGBT and heterosexual, cisgender groups. Nearly all of these (12) however were from the US or Canada, rather than the UK, and so the transferability of the results must be treated with caution. It is important to note that some of the evidence in our review focused on LGBT people’s daily experiences and circumstances while homeless. These studies in and of themselves do not provide an understanding of whether personal circumstances and experiences (e.g., drug dependency, mental ill-health) are behavioural adaptations to homelessness or pre-existing and potentially (direct or indirect) causal factors in people’s pathways into homelessness. They do, however, give some indication of the comparative differences in the extent to which LGBT people report risk factors that are known to increase vulnerability to homelessness (or episodes of). Notwithstanding the caveats, the evidence suggests that some of the factors that can cause homelessness are common experiences for LGBT people, and that LGBT people may be at greater risk, compared with non-LGBT people, of experiencing some of these potentially causal factors.

Individual-level homelessness ‘risk factors’ among LGBT people

Multiple factors contribute to homelessness for LGBT people (as they do for non-LGBT people), including structural inequalities, systemic inequalities, and individual-level problems (Ecker et al., 2019). It was beyond the scope of the review to fully explore LGBT people’s structural position in relation to all issues that might increase risk of homelessness (employment, class, poverty, education). However, within the homelessness studies reviewed, several patterns emerged that indicated a more pronounced risk of experiences that can lead to homelessness for LGBT people. The evidence also indicates that some distinct factors contribute to LGBT people’s homelessness or add nuance to issues understood more broadly as homelessness triggers or ‘risk factors’. The data on each of these (mostly personal and interpersonal) factors is considered below. This is by no means an exhaustive list, but it reveals some important differences between LGBT people and those who do not identify as LGBT that are likely to impact on their prevalence and experience of homelessness.

Family breakdown and parental rejection

An overriding position in the comparative and non-comparative literature, is that LGBT homelessness occurs through family conflict and breakdown, particularly when young people ‘come out’ to their families (Durso & Gates, 2012; Rhoades et al., 2018; Robinson, 2018; Shelton & Bond, 2017; Whitbeck et al., 2004; Yu, 2010). Reporting on the experiences of 295 homeless young people who had previous child welfare system involvement, among the 114 young people who identified as lesbian, gay, bisexual, transgender, or questioning (LGBTQ), Forge et al., (2018) found that 33 per cent were told to leave their homes compared to 19.7 per cent who were heterosexual and cisgender. In Whitbeck et al.’s (2004) study involving 428 homeless and runaway young people, gay adolescent males were five times more likely than heterosexual males to have left home because of a conflict regarding sexual orientation. Rhoades et al.’s (2018) national study of young users of a US LGBT crisis service similarly found that disclosing LGBT identity to parents and having experienced parental rejection because of LGBT identity were associated with increased odds of having experienced homelessness. Exploring this issue further, Whitbeck et al. (2001) found that such ‘conflict’ is actually often victimization and asserts that - for runaways - a substantial proportion of homeless LGBT youth leave family situations where they have experienced abuse. In a major UK study, the most common causes of homelessness among LGBT young people (as reported by housing providers) were parental rejection, familial physical, sexual, and emotional abuse, and familial aggression and violence. In the same study, 77 per cent of young people surveyed believed their sexual/gender identity to be a causal factor in their rejection from home (Albert Kennedy Trust, 2015). Shelton et al. (2018a), in one of the few studies on this issue that disaggregates trans participants, examined the experiences of transgender youth compared to the full lesbian, gay, bisexual, transgender, queer, and questioning (LGBTQ) sample and found that transgender youth reported higher frequencies of running away or being ejected from their family home, foster home or relatives’ home.

In addition to the comparative data cited above, studies of LGBT populations indicate that parental rejection is associated with experiences of homelessness. A study by Shelton et al. (2018b), for instance, found that 70 per cent of 442 lesbian, gay, bisexual, transgender, queer, and questioning (LGBTQ) survey respondents (who were between the ages of 18-26) reported being ejected from/asked to leave their family home by their parents. Likewise, among the 52 people who self-identified as LGB in a study of 268 homeless youth (aged 14-22), ‘arguments with parents’ was the most frequently cited reason for leaving home (51 per cent) along with verbal abuse (34 per cent) (Gangamma et al., 2008).

Child abuse and neglect

Data collated from our review indicated that LGBT people have a comparatively greater probability of having experienced childhood adversity. This includes being more likely to report involvement with child protection services (Forge et al., 2018; Gaetz et al., 2016; Van Leeuwen et al., 2006). In Forge et al.’s (2018) study, those with prior child welfare system-involvement comprised 43 per cent of the full sample of 693 homeless and runaway youth (Atlanta Youth Count). Of these, 8.8 per cent were trans and 29.8 per cent were lesbian, gay, bisexual, or questioning (LGBQ). Linked to the point above, data sources also suggest that LGBT young people are more likely to have experienced abuse as a child (Frederick et al., 2011). In Gaetz et al.’s (2016) study, transgender youth and lesbian, gay, bisexual, queer, and/or two-spirit² (LGBQ2S) youth were much more likely to have experienced all types of abuse as a child. Likewise, in Forge et al.’s (2018) study which drew on data from 295 young people experiencing homelessness who had previous child welfare system involvement, two-thirds of young people who were lesbian, gay, bisexual, transgender, or questioning (LGBTQ), compared to over one half of youth who were heterosexual and cisgender reported experiencing child abuse. In a (non-comparative) UK study with a sample of 44 LGBT homeless young people aged 16-25, Cull et al. (2006) found that 26 per cent had been physically abused as children and 30 per cent had been sexually abused or sexually assaulted. It is noteworthy that Mountz et al. (2018) found increased care placement disruption among trans youth compared to LGB youth.

Poor educational experiences

The evidence review indicated that homeless LGBT young people are more likely to have poor educational experiences which may be a factor contributing to homelessness among LGBT people (Bidell, 2014; Forge et al., 2018). In studies, high percentages of LGBT youth reported a disrupted education as a result of discrimination or bullying, and higher than average numbers of LGBT homeless young people also left school early. In one study, 67 per cent of homeless young people reported that they had been bullied at school and this appeared to be linked to truanting and exclusion (Cull et al., 2006). In Gaetz et al. (2016), while 46 per cent of respondents (homeless youth in Canada) had ‘often’ been bullied at school, this rose to 63 per cent for lesbian, gay, bisexual, transgender, queer, and/or two-spirit (LGBTQ2S) homeless youth in the sample.

Mental health problems

Non-LGBT specific studies frequently find high prevalence of mental ill health among participants and point to the relationship between mental health and homelessness (Reeve & Batty, 2011; Fitzpatrick et al., 2013; Seria-Walker, 2018). Comparatively poor mental health among LGBT homeless populations is potentially a contributory factor in LGBT people’s ability to find and maintain stable housing. In the Australian Trans Pathways study (Strauss et al., 2017), almost three-quarters of participants had at some point been diagnosed with depression; 72 per cent with anxiety disorder; 80 per cent had self-harmed; and 48 per cent had attempted suicide (Strauss et al., 2017). In another large-scale study (Kidd et al., 2017), among the 310 young homeless people who reported sexual or gender minority identities (within a larger sample of 1103 young people accessing Canadian homelessness services), there was a 70 per cent suicide attempt rate compared with 39 per cent for straight and cisgender participants. High rates of mental ill-health among young people are found to be linked to homophobic, biphobic or transphobic (HBT) bullying, LGBT-hostile family environments, navigating heterosexual and cisgender norms as well as factors unrelated to sexual orientation or gender identity (McDermott et al., 2016).

Drug use

The evidence in the review indicated that homeless people who identify as LGBT are more likely to have used substances (Cochran et al., 2002; Van Leeuwen et al., 2006; Gattis, 2013; Kidd et al., 2017), to have used a variety of substances (Cochran et al., 2002; Frederick et al., 2011; Gattis, 2013), to use them more frequently and for longer periods (Cochran et al., 2002, Clatts et al., 2005; Gattis, 2013) and to use hard drugs (Moon et al., 2000; Kattari et al., 2017). Frederick et al. (2011), for instance, found that sexual minority youth reported using a greater number of different types of drugs in the prior 12 months than heterosexual-identified youth; and sexual minority young women were more likely to report daily drug use.

Violence and sexual exploitation

The evidence suggests that the sexual victimisation of homeless LGBT people also differs from the experiences of heterosexual homeless people. According to a number of studies, LGBT homeless people are significantly more likely to be subject to sexual victimisation or violence (Cochran et al., 2002; Tyler, 2008; Gaetz et al., 2016; Forge et al., 2018). Over half (51.1 per cent) of lesbian, gay, bisexual, transgender, or questioning (LGBTQ) youth in Forge et al.’s (2018) research were victimised while they were homeless compared to one third (34 per cent) of youth who were heterosexual or cisgender. A quantitative survey involving 712 young people (aged 16 to 25) who were receiving help from homelessness organisations in the UK found that 12 per cent of all survey respondents said they had engaged in sexual activity in exchange for a place to stay. This proportion rose considerably for LGBT young people of whom nearly a quarter had engaged in sexual activity for a place to stay (McCoy, 2018).

Explaining the role of LGBT identity in homelessness pathways

Much of the work discussed above points to statistical associations between LGBT identity and homelessness. Yet such links, while indicative, are not able to establish a causal relationship and therefore explain what it is about minority gender identity and sexual orientation that might make somebody more at risk of experiencing homelessness (Bramley & Fitzpatrick, 2018). An identification of ‘risk factors’ offers little that is meaningful in terms of helping us to understand (and therefore prevent) pathways to homelessness (Batterham, 2019; Somerville, 2013). Instead, they present clusters of potentially relevant social factors - often presented somewhat speciously as static and discrete – without explanation of how they are mediated and experienced within the lives of individuals. The trajectory of LGBT people into homelessness and the role of pre-existing disadvantage, trauma and unmet need, common within LGBT people’s life-histories, is largely underexplored.

While pathways to homelessness among LGBT groups are not well-understood, a modest number of studies have begun to shed light both on the complexity of these pathways and the interaction between different personal and interpersonal factors. Although family conflict/disclosure of sexual identity is often cited as the key cause of youth homelessness, a small body of robust qualitative evidence offers a more nuanced analysis of the role of family conflict in LGBT people’s pathways to homelessness. These studies demonstrate that although sexual or gender identity may often play a role in homelessness, the influence of sexual and gender identity in homelessness can take different forms beyond familial rejection after ‘coming out’ (Prendergast et al., 2001; Dunne et al., 2002).

An ethnographic study with 14 gay and bisexual young men by Castellanos (2016), one of the few studies that focuses specifically on minority ethnic LGBT homeless people, found that gay and bisexual (GB) youth homelessness was not merely the result of conflict over sexuality for most. Although sexuality played a role in exacerbating family conflict and creating housing instability, over half had left home before disclosure of sexual orientation. Furthermore, most of them had returned home after disclosure of sexual orientation occurred, maintained family connections, and received financial support from their families after leaving home. Three main pathways to homelessness emerged. First, some young gay men were placed in state systems of care (child welfare and the juvenile judicial system) before disclosure of sexual orientation; upon leaving those settings, some returned home but family conflict restarted and precluded them from remaining there. In the second pathway, the participants were forced to leave or left home due to severe conflict over homosexuality, masculinity, and gender presentation. Most common was a third pathway where disclosure of sexual orientation exacerbated pre-existing family conflict resulting in the young person leaving or being forced out of the family home. We see in each of these examples, family conflict related to participants’ sexual identity but the relationship between sexual identity and homelessness was not straightforward. Castellanos (2016) also makes the point that not all LGBT youth become homeless, even in the presence of family conflict.

Cull et al.’s (2006) study which investigated the needs and experiences of LGBT homeless youth in Brighton and Hove (UK), also sheds light on the link between youth homelessness and family conflict. Here, some young people were evicted from the family home, or chose to leave, because their parents were intolerant of the young person’s sexual identity; others chose to leave because they presumed their parents would have a negative reaction if they did know about the young person’s sexual identity. Some young people chose to leave home because they felt isolated in the place where they lived and wanted to move to a place where LGBT lives were more visible.

O’Connor & Molloy (2001) also found that identifying as LGBT was not always a direct cause of parental rejection (and hence, homelessness) and argue that most situations are much more complex. Indirectly, the trauma felt from homophobic, biphobic, or transphobic bullying, discrimination, or sexual exploitation in other spheres may manifest in poor mental health and behaviour in the home which eventually leads to family conflict and alienation. This acknowledgement is supported by ethnographic research by Tunåker (2015) who found that, in some cases, ‘coming out’ to the family did directly cause homelessness. However, most situations were more complex with sexual orientation and gender identity subtly and intricately linked to housing situations. It was more common for young people to have arguments with parents or guardians that were unrelated to sexual orientation or gender identity but arose out of living in an oppressive housing situation that led to emotional problems and contributed towards the young person’s decision to leave home.

Some studies have also challenged general homelessness literature that tends to find a mutually reinforcing relationship between mental ill-health, substance use and homelessness (i.e., acting as both cause and consequence and reinforcing one another). Qualitative evidence about the experiences of homeless LGBT people indicates that mental ill-health and substance use might be more common as consequences of homelessness. In Clatts et al.’s (2005) research with homeless young men who have sex with men, they suggest that both drug use and sex work are better understood as behavioural adaptations to homelessness, rather than causes of it. This finding is corroborated by Rosario et al. (2012) who found that ‘runaway/throwaway’ LGB youths initiated substance misuse either simultaneously with or after their initial episode of homelessness. Similarly, in Lankenau et al. (2005) nine out of 10 men became homeless before starting to undertake sex work. Therefore, rather than causing homelessness, it is likely that negative outcomes or events such as substance use and sexual work are a consequence of homelessness, and a way of coping with the stressors of homelessness, at least for some subgroups.

Likewise, sexual or gender identity is not always necessarily the driving force in people’s life histories (Prendergast et al., 2001). Matthews & Poyner (2017), for example, found evidence of LGBT people being made homeless due to familial rejection, but this was not a common experience among their 20 respondents and they concluded that ‘sexual or gender identity is often not the root cause of someone’s homelessness’ (2017, p. 1). Others within this tradition emphasise that sexual or gender identity is rarely a single, determining causal factor, and that becoming homeless is not always a direct consequence of sexual orientation (Ecker, 2016; Ecker et al., 2019; Prendergast et al., 2001; Tunåker, 2015; World Habitat, 2018). This echoes the general homelessness literature which points to the interaction between structural and individual factors in causing homelessness, rather than one single determinant (Bramley & Fitzpatrick, 2018; Johnsen & Watts, 2014).

Related to this, a small body of literature challenges the assumption that LGBT homeless people are ‘victims’ because of their sexual orientation or gender identity. Rather than being seen as a negative correlate of homelessness, sexual and gender identity is instead recognised as a positive driving force in people’s homelessness journeys. Prendergast et al. (2001) identified how ‘difference’ in the guise of sexual identity can enable new forms of inclusion to emerge and bring advantages such as freedom to explore sexual orientation, entry into the gay ‘scene’, and access to support. In their research, Matthews et al. (2019) also emphasise the agency expressed by their participants when they described their move to homelessness. In a variety of extremely difficult situations, it was a choice which allowed another part of their identity – their sexual orientation or gender identity – to flourish. While critiquing an overemphasis on public health implications and risk associated with LGBT homelessness within North American literature, they also draw attention to the important point that this research comes from a context where the welfare safety net is very poor. As other scholars have argued, contemporary homelessness is shaped by historical geographies (Mitchell, 2011). Homelessness in the US has evolved to take a specific ‘style’ given the significant erosion of an already weak post-World War II welfare state (Mitchell, 2011). In the UK (and particularly Scotland), the negative outcomes identified in the US may be mitigated, to an extent, by the broader welfare entitlement system, homelessness safety nets and supported housing options, though evidence suggests elements of such North American styles of homelessness are also becoming apparent in other national contexts, including the UK and other European countries (Cloke et al., 2010; Thörn, 2011). Some countries’ policies are also more LGBT-friendly than others. Wilson (2013) argues that European welfare states, in sharing a broad consensus on the importance of social rights to citizenship, and providing higher levels of human security, are or have the potential to be more lesbian and gay friendly than the US. The link between welfare systems and LGBT-friendliness of a state is complex, and also largely depends on the relationship between state, church and welfare (Wilson, 2013).

Perhaps one of the most ground-breaking studies on LGBT identity in homelessness pathways so far is that by Prendergast et al. (2001) as it is the first that begins to conceptualise the complex mediating role that sexual orientation plays in homelessness journeys. In their analysis of the life-histories of 19 homeless LGB young people, Prendergast et al. (2001, pp. 66–67) identify a range of experiences, factors and behaviours, which together contributed towards the complex pathways that led to homelessness. Their participants’ stories shared common themes of disadvantage and exclusion that echoed the wider literature on homelessness such that: ‘in many major respects our sample appear very like other homeless young people today’. However, while LGBT people’s homelessness pathways may appear similar to people who do not identify as LGBT, there are clear ways in which LGBT people’s experiences differ and their gender or sexual identity interacts in complex ways with their homelessness. Prendergast et al. (2001) emphasise that when explaining the influence of sexual orientation on the causes (and experiences) of homelessness, it is necessary to understand sexual orientation as it is embedded in the material circumstances and lived experience of people’s everyday lives. In their study, Prendergast et al. (2001) suggest that over time sexual orientation played a significant role in their participants’ homelessness journeys but that this varied across the sample. They grouped their sample into four categories to illustrate this point: ‘homeless young people who happened to be gay’; ‘those for whom homelessness brought opportunities to challenge an assumed heterosexual identity’; ‘those for whom being gay had greatly influenced their homeless status’; and ‘those who were homeless because they were gay’.

Conclusion

Our review of the evidence suggests that people who identify as LGBT have an increased risk of homelessness and are more likely to experience the social and biographical factors that can play a part in individuals’ homelessness pathways. The findings of our review also suggest that the pathways to homelessness among those who identify as LGBT are qualitatively different to those who do not identify as LGBT. Yet there are several limitations to the extant evidence underpinning these claims and key knowledge gaps.

First, the experiences of homeless LGBT people are much more visible in the North American literature, having been somewhat neglected in the UK mainstream academic literature on homelessness. A second major gap concerns how the literature focuses on certain sub-groups to the neglect of others. It gives an overwhelming focus to gay homeless males, with homeless trans, lesbians and bisexuals distinctly underrepresented. These are crucial gaps as some studies indicate that homelessness journeys among LGBT groups are differentiated in important ways according to sexuality and gender identity. These suggest that intersections of gender and sexuality may channel homeless people away from or towards certain vulnerabilities and behaviours. Frederick et al.’s (2011) study, for instance, found that sexual minority young women differ more from sexual minority young men and heterosexual young women than was understood previously. Protective factors also exist for some groups in some places. This might be because many of the visible gay communities in urban centres cater primarily to gay men, and that lesbian and bisexual women often lack access to supportive social spaces. Moreover, bisexually identified individuals report ostracism and biphobia from both heterosexual and gay/lesbian communities. The limited extant evidence on trans homelessness also indicates that this group may be particularly marginalised. The evidence base is also dominated by experiences of young LGBT homeless people (generally under the age of 25). While this body of literature provides important knowledge and understanding, it also overlooks how homelessness among middle-aged and older LGBT people may present distinct needs and experiences. And third, there are very few studies with LGBT people from minority ethnic or religious backgrounds who may be more profoundly excluded, discriminated against, or victimised from both within and without their ethnic and religious communities. There is a critical need for future studies to take an intersectional approach – to extend the reach of their sampling and depth of analysis – to expand current understandings of LGBT homelessness.

An additional significant gap identified in the literature relates to the analytical and conceptual frameworks employed to explain the causes of LGBT homelessness. This is not quite keeping pace with the development of causal frameworks in more general homelessness literature (Fitzpatrick et al., 2011) which acknowledge the complex interplay of individual, interpersonal and structural factors and the need to understand the different ways in which these factors operate or work within a pathway that leads to homelessness (Somerville, 2013). Related to this, the quantitative nature of much of the existing evidence on the causes of LGBT homelessness renders it less able to interrogate why and how gender identity and sexual orientation influence homelessness journeys. There is a need for more high-quality studies that employ qualitative methodologies to generate nuanced explanatory understandings of the ways that sexual orientation and gender identity might cause homelessness. Such studies should also shift the (currently predominant) focus from individual risk factors and family characteristics, drawing attention instead to the structural and systemic issues as well as social and economic conditions which can lead to homelessness. Underlying economic and structural disadvantage among LGBT people, as it relates to gender and sexual identity and impacts on homelessness risk, is rarely discussed in the homelessness literature.

The present article makes a key contribution to understandings of homelessness pathways for LGBT people and strengthens the calls of fellow housing scholars in the UK and US who argue for a more theoretically informed understanding of how LGBT identities intersect with homelessness. This is in the interests of doing justice to factors in homelessness pathways such as poverty, ethnicity, race and other forms of social stratification, and with regard to the complexity of the identity and belonging issues at stake (Matthews et al., 2019; Robinson, 2018; Wheeler et al., 2017).

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Notes on contributors

Lindsey McCarthy

Lindsey McCarthy is a Research Fellow at CRESR with a track record of research covering homelessness and marginality; service encounters, support journeys and policy/practice responses for those facing multiple disadvantage; housing solutions for households in housing need; and experiences of ‘home’ and homemaking for marginalised groups.

Sadie Parr

Sadie Parr is a Senior Research Fellow at CRESR interested in the needs and experiences of multiply disadvantaged families and individuals, how they are governed and the role of the state in their lives.

Notes

1 It is recognised that the term LGBT comprises a range of different sexualities and gender identities and so is being used in this article as a descriptor for diverse populations. Although the term LGBT is used throughout the article, when specific studies are referred to, the sample terminology for that study is used.

2 In Gaetz et al. (2016) ‘two-spirit’ refers to Indigenous persons who are LGBTQ2S or gender-variant.