Hidden homelessness: international evidence review

This report "Exploring Ways of Identifying and Counting Hidden Homeless Populations" presents an overview of the methods used internationally to identify or count people experiencing hidden forms of homelessness and the populations that may be likely to experience it. The report also considers the Scottish context and suggests areas where further research may be useful.

5. Population groups

5.1 Population groups often presented as hidden homeless in the literature

This section presents the population groups which were more commonly identified within the evidence as experiencing hidden homelessness. Six populations were consistently mentioned as experiencing hidden homelessness in the literature:

  • women and women who are parents
  • rural populations
  • young people
  • minority ethnic people
  • migrants, people who arrive to a country seeking asylum or that are refugees
  • LGBTI people

This does not mean that these are the only population groups that experience hidden homelessness, as there are gaps in the current research. However, the consistent international evidence found for each of these groups indicates that they have been found to be at risk of experiencing hidden homelessness.

This does not intend to be exhaustive description of homelessness for each group but to focus on the particular challenges to be included in homelessness counts and the methodological implications mentioned in the literature. Additionally, although the groups are presented as separate, reality is much more complex and they are not mutually exclusive. For example, a woman who is a mother can be from a minority ethnic group, or a LGBTI person can also be seeking asylum in the UK, hence the need to consider these groups intersectionally.

5.1.1 Households lead by women and women who are parents

This group is formed by people who identify as female and in some cases who also present in their parental role. The available evidence that is specific to women's homelessness is mostly qualitative and focuses on the homelessness pathways of individual women and their experiences or from small quantitative studies. There is a gap, which is also signalled in the literature, regarding large scale longitudinal enumeration studies focusing on women's homelessness. This is both due to the fact that some forms of homelessness often experienced by women are less visible- often due to personal safety concerns- or are not classified as such by some legal definitions (sofa surfing, living in overcrowded dwellings and survival sex) (Johnson et al., 2017; Pleace, 2016).

Women often appear to be less prevalent among homeless populations, particularly among people sleeping rough (Johnson et al., 2017). This has often led to the conclusion that women are less likely to experience homelessness than men. Nevertheless, it is noted in the literature that a portion of that underrepresentation of women in the homeless populations could be due to utilising other strategies to avoid (visibly) sleeping rough or in mixed homeless shelters.

There is consistency across the literature on the need for a gendered lens on homelessness to fully grasp the complexity of this phenomenon and to improve enumeration methods to better identify women experiencing homelessness. Bretherton (2020) emphasises that homelessness is a gendered experience meaning that gender is a key variable to fully understand homelessness.

Finally, there are contradicting views on the relevance of 'survival sex' or exchanging sex for accommodation as a strategy to secure shelter for women experiencing homelessness. In two studies conducted with women in Scotland with experience of homelessness, survival sex emerged as a prevalent subsistence strategy. 20% of female respondents had engaged in sex work to pay for accommodation compared to 3% of men and 28% of female respondents had spent the night with someone to get accommodation compared to 14% of men (Reeve, 2018, p. 171). However, Bretherton and Pleace (2018) note that there is not enough evidence of this being a widespread phenomenon. This could also be due to some inconsistencies in the literature where two groups may be overlapping in the data: women who exchange sex to find overnight accommodation and women who are sex workers who are also homeless.

Reasons why this group might be missed from counts

In the literature there are six main reasons presented why women's homelessness might not be adequately counted. Those are:

Definitions of homelessness which focus on rooflessness

Bretherton and Pleace (2018) conducted a review of the way women's homelessness in identified and counted across Europe (including the UK) and reached the conclusion that when the definition of homelessness is broader, particularly when concealed forms of homelessness are included, women appear in higher numbers (Bretherton & Pleace, 2018). The evidence notes that when homelessness data collection centres on unsheltered forms of homelessness is more likely to miss women experiencing homelessness (Bretherton, 2017; Johnson et al., 2017; Pleace, 2016). This is attributed to the fact that women tend to exhaust their informal options of accommodation, like staying with friends and family, before presenting to homeless services (Bretherton & Pleace, 2018; Engender, 2020b; Pleace, 2016). This leads to an undercount in methods like street counts and counts that use homeless shelter's administrative data.

The link between homelessness and domestic abuse

Across the literature there was a clear link established between women's experiences of homelessness and gender-based violence. Domestic abuse was seen as both a cause and a consequence of the homelessness episodes for some women, as well as a circumstance that shapes the behaviours of women while experiencing homelessness.

Domestic abuse is presented in the literature as a direct cause of homelessness in the parts of the UK. 1 in 5 women (21%) who have experienced domestic abuse in England became homeless at one point in their lives compared to 1% of those who hadn't (Scott & McManus, 2016, p. 6). Even when domestic abuse is not direct cause of a particular homelessness episode, gendered violence is still present throughout women's homelessness trajectories (Bretherton & Pleace, 2018).

With regards to enumeration, one of the challenges noted in the literature is that, in some countries, women who seek assistance with domestic abuse may not be recorded as homeless as they are supported by a refuge or dedicated charity and not by their local authority (Bretherton & Pleace, 2018).

Concealing strategies used due to physical safety concerns

There is evidence to suggest that women sleeping rough often use concealing strategies to avoid being targets of physical and sexual violence (Bimpson, Reeve, & Parr, 2020; Bretherton & Pleace, 2018; Pleace, 2016). This is by both concealing themselves from view and concealing their gender (Bretherton & Pleace, 2018). This hidden aspect of their homelessness- as with other groups that will be discussed later in this review- is often represented as a survival strategy, a way to remain safer in a potentially threatening situation. In this case point in time overnight (PiT) counts in streets or shelters may produce undercounts of this population (Bretherton, 2017).

Reeve (2018) highlights that women occupy public space in a different way to men. One study in Scotland found that 62% of women with experiencing of homelessness had slept rough out of a 144 participant sample (Reeve, 2018, p. 167). Of those women who had slept rough only 12% had been in contact with the rough sleeping team in their area to get support, which is linked to the following point (Reeve, 2018, p. 168).

Homelessness presentation

It was found that women may use sofa surfing more often than men as a strategy to avoid sleeping rough (Bretherton & Pleace, 2018). Homelessness counts do not always include ways to identify multiple households living in a single dwelling. This could potentially be one of the contributing factors to the undercount of women who are staying with friends and family to avoid being roofless.

Parental status

The parental status of women was presented as another challenge to their inclusion in homelessness counts. A paper published by the Centre for Homeless Impact notes that, in England, women who are categorised as 'single homeless' could be actually mothers whose children are temporarily in care of others (Bimpson et al., 2020). There is a risk that the parental status of women could become invisible once they live apart from their children, even if this is not a permanent situation.

Reeve (2018, p. 169) reported in their 2006 study that 30% of single homeless women that participated had mentioned having children under the age of 16, yet they were still recorded as 'single homeless'. Studies from both the UK and Australia signalled that there is a gap in understanding of family dynamics and how they influence women's homelessness (Bimpson et al., 2020; Johnson et al., 2017; Reeve, 2018).

5.1.2 Rural populations

What is considered a rural population varies according to each government/authority. However, there was consistency in the evidence around the challenges of identifying people experiencing homelessness in rural or more sparsely populated areas. The authors reviewed here argued in favour of the application of alternative methods to estimate the prevalence of homelessness in rural areas that are sensitive to rural contexts and are able to produce more robust and accurate representations of the issue.

Reasons why this group might be missed from counts

There are several factors presented across the literature about why rural homelessness is less frequently accounted for and less visible than urban homelessness. These include:

Measurement limitations

There was agreement across the literature that rural homelessness is often undercounted or misrepresented, and this may skew perceptions of its scale and have an impact on the interventions designed to tackle it (Gleason et al., 2022; Knopf-Amelung, 2013; David Robinson, 2004; Stroud & Pickett, 2017). Studies suggest that there is a circular issue in which homelessness is under researched in rural geographies because it is believed to have limited incidence and therefore, there is a lack of evidence of the contrary.

Cultural limitations

Robinson (2004), as well as (Cloke, Milbourne, & Widdowfield, 2001), introduce the concept of 'non-coupling' of the notions of homelessness and rurality. This means that often homelessness is not associated to rural spaces. Aspirational idealisations about rural spaces being privileged and free from the urban troubles create a barrier to raising awareness about social issues like homelessness (Cloke, Widdowfield, & Milbourne, 2000; David Robinson, 2004).

Additionally, homelessness tends to be minimised by local residents of rural spaces as it can be perceived as a threat to the reputation of their place (Cloke et al., 2000). This can mean that homeless people in rural spaces find their behaviours over-policed and may have to use strategies to make themselves invisible to avoid challenging the views of their own community (Cloke et al., 2000; Knopf-Amelung, 2013).

Resourcing limitations

Another argument present in the literature is that fewer temporary accommodation alternatives are available in rural areas (Gleason et al., 2022). Lack of sufficient support service provision may make moving away from rural areas the most viable option to exit homelessness (Cloke et al., 2000). However, people's desire to remain in their area might delay presentation as homeless and enhance the use strategies to conceal their homelessness (Gleason et al., 2022). This makes homelessness effectively less visible in rural settings (Cloke et al., 2000).

Geographical limitations

Lastly, the dispersed populations across wide geographical areas can contribute to the invisibility of rural homelessness (Knopf-Amelung, 2013).

Point in Time counts

Overnight street (PiT) counts tend to undercount rural homeless populations as they rely on the visibility of people. This is not in line with the way homelessness presents in rural areas (Gleason et al., 2022). In addition, reaching a representative sample of the homeless population in a rural area would require multiple counting teams conducting a simultaneous count. This would not only be costly but also may not be possible give the limited voluntary resources present in smaller communities in rural areas (Hall, 2017).

Service-based counts

Service-based counts are often considered as an alternative to overnight PiT counts for rural areas as the data collection period is longer and collection is done through support services, allowing for better chances to identify and count homelessness in rural populations (Hall, 2017). However, the availability of services in the area can impact the outcome and may lead to an undercount when services are more sparce (Knopf-Amelung, 2013). There is the additional consideration that service-based counts will only identify and count those who have reached out to services and, as previously discussed, some people in rural areas may choose to remain invisible to the local agencies to avoid being moved out of the area (Cloke et al., 2000).

5.1.3 Young people

Compared to the other populations considered in this section, there was a greater proportion of publications focussed on youth homelessness within the literature. Across the literature reviewed in this section the age range considered in each study varied anywhere from 13 to 26 years. This was pointed out by some of the authors as a challenge for the direct comparison of data from multiple sources (J. Smith, 2013). There was an intersection in the evidence between of youth, LGBTI and minority ethnic homelessness.

Reasons why this group might be missed from counts

Keeping appearances for fear of stigmatisation

A qualitative study with young people experiencing hidden homelessness in Canada reported that one of the factors contributing to the lower visibility of this population was the need to maintain appearances and keep their homelessness hidden from their social circle to maintain their social standing. Their social circle included employers, teachers, and friends, among others. It was also done to avoid being taken advantage due to their precarious situation (Gausvik, 2015).


A study in the US and another in Canada found that young people with experience of homelessness recalled overall good experiences in family shelters. However, this was not the case for adult homeless shelters for which they recalled negative experiences and expressed reluctance to engage with them once they were considered adults (Auerswald, Lin, Petry, Laura, & Hyatt, 2013; Gausvik, 2015). This could be one of the reasons some 'older' young people might be reluctant to approach support services and, therefore, not be identified as homeless.

Transient nature

Homeless youth tend to be transient, moving between locations during the course of a day and often traveling from one city to another (Clarke, 2016; Gausvik, 2015). The literature links this to the process of most young adults transitioning from their nuclear home or guardian institution into adult independent living after mandatory school attendance finishes. Nevertheless, the literature also highlights that this might not be a sufficient explanation and there is a need for further research into the pathways in and out of homelessness of young people (Clarke, 2016).

Intermittent nature of homelessness

The evidence reviewed here argues that young people have different homelessness patterns than adults and are often intermittently homeless (Clarke, 2016). This includes either returning home relatively quickly or cycling on and off the street (Morgan, 2013). This type of dynamic homelessness pattern requires a suitable longitudinal approach to be identified and counted. A cross-sectional approach might produce unreliable results for this population.

Fear of criminalisation/institutionalisation

Two studies of young people experiencing homelessness in the US, one with ethnic minorities and another on sofa surfing youth found that young people experiencing homelessness were often reluctant to approach support services due to fear of criminalisation/ institutionalisation. Fear of involvement of child protection services is specifically mentioned in the literature (Curry et al., 2017; Petry et al., 2022).

5.1.4 Minority ethnic people

The evidence discussed in this section refers to mainly three minority ethnic communities: Aboriginal populations in Canada; Gypsy/Traveller communities in Ireland and the UK; and Black and Minority Ethnic communities across the UK (Netto et al., 2004). The term 'Aboriginal peoples' is used in these studies to refer to the three legally defined culture groups in Canada: M├ętis, Inuit, and First Nations.

In 2020 the Scottish Government published the Housing needs of minority ethnic groups: evidence review which found evidence that supports that there is an over-representation of some minority ethnic communities in homelessness applications in Scotland. There is agreement as well across the literature on the disproportionate representation of ethnic minorities in the homeless populations in the UK, US, Ireland and Canada (Bramley et al., 2022; Netto et al., 2004; David Robinson & Coward, 2003; Shankley & Finney, 2020). Some of the reasons given for this overrepresentation were:

  • not enough appropriate (culturally and materially) accommodation available
  • higher levels of deprivation
  • presence of added barriers to access housing support such as:
  • o discrimination from service providers
  • o support services not being culturally appropriate
  • o not perceiving themselves as homeless
  • o lack of awareness of available services (The Scottish Government, 2021)

Although we have some understanding of the elevated prevalence of ethnic minorities in homelessness there is still work to do to understand if there is a proportion of that group experiencing homelessness that is concealed and missing from counts.

Some studies linked overcrowding to hidden homelessness for ethnic minorities in Ireland, UK and Canada (Addison et al., 2022; Belanger, 2013; Morgan-Williams, 2020; Shankley & Finney, 2020). This would mean that people of certain ethnic minorities may live in overcrowded accommodation to avoid, for example, sleeping rough. This was also linked with people not seeing themselves as homeless due to understanding homelessness solely as rooflessness. However, further research is needed to provide clarity on that link and to understand if there are specific forms in which hidden homelessness presents for ethnic minorities.

Reasons why this group might be missed from counts

When speaking about ethnic minorities it is relevant to note that this is a broad term which encompasses different communities with distinct cultures and circumstances that vary across different countries. Nevertheless, there are some barriers to enumeration that consistently appeared throughout the literature:


A publication by the Cork and Kerry Regional Traveller Accommodation Working Group reported that some Gypsy/Traveller people might not identify themselves as such to local authorities or landlords for fear of discrimination (Morgan-Williams, 2020). There are also, at the time this report was published, no specific homelessness services for Gypsy/Travellers in Cork and Kerry, and mainstream support services are often not sensitive to the needs of Travellers (Morgan-Williams, 2020).

This could also be related to the historical criminalisation of nomadism in UK and the Republic of Ireland, but more research is needed to understand these links (Drummond, 2007). A recent evidence review by the Scottish Government found that there is a considerable body of evidence that the Gypsy/Traveller population continues to face high levels of discrimination and harassment in Scotland (The Scottish Government, 2020).

A report by the Centre for Homeless Impact also highlighted that discrimination from housing service providers and overall systemic racism in US and Canada as an access barrier to support and, therefore, the inclusion of Gypsy/Travellers in homelessness counts. This report also noted a gap in evidence on the impact of racism and discrimination on homelessness outcomes for ethnic minorities in the UK (Finney, 2022).

Overcrowding is not always included within the definition of homelessness used in counts

When overcrowding is not considered as a form of homelessness there is a risk of undercounting ethnic minorities. Overcrowding has appeared consistently in the literature as one the most prevalent forms of hidden homelessness among ethnic minorities in the UK, particularly Gypsy/Traveller communities (Addison et al., 2022; Bramley et al., 2022; Shankley & Finney, 2020; The Scottish Government, 2021). The evidence review conducted by Addison et al. (2022) suggests that minority ethnic groups are disproportionately affected by overcrowding in Wales across all age groups and in both urban and rural settings. Additionally, there is evidence to suggest that overcrowding particularly affects Bangladeshi, Pakistani and Black African households in the UK (Bramley et al., 2022; Shankley & Finney, 2020).

In Scotland, Bangladeshi and African households have a 28% rate of overcrowding compared to 8% for 'White Scottish' households and 6% for' White: Other British' households (The Scottish Government, 2021, p. 42). A qualitative study by the Department for Levelling Up, Housing and Communities (DLUHC) also found overcrowding to be an issue for households of Asian descend in UK (Department for Levelling Up, 2022).

Not perceiving themselves as homeless

A report by Morgan-Williams (2020) for the Regional Gypsy/Traveller Accommodation Working Group (RTAWG) of Cork and Kerry pointed out that the way homelessness is defined can lead to undercounting of homelessness in the Gypsy/Traveller population in Ireland. This report argues that the representation of situations of overcrowding as 'accommodation sharing', rather than housing insecurity or hidden homelessness, can contribute to this community also not perceiving themselves as homeless (Morgan-Williams, 2020).

Another study noted that there are differences in the meanings attached to homelessness between minority ethnic communities in Scotland (Netto et al., 2004). For example, many minority ethnic people living in over-crowded accommodation, due to inability to access other housing options, may not see themselves as being homeless (Netto et al., 2004). People who do not see themselves as homeless are less likely to consider approaching their local authorities for housing support, which makes them less likely to be counted.

5.1.5 Migrants, people who are seeking asylum or are refugees

This section focuses on analysing the literature on ways of identifying and counting migrants, refugees and asylum seekers experiencing concealed types of homelessness. Although these groups may be also understood as minority ethnic in their new context, they have the distinctive characteristic of experiencing a different legal situation relating to citizenship/right to remain and different access to local resources (social and economic) (Hermans et al., 2020).

Studies from the UK suggested that migrants, asylum seekers and refuges are more likely to experience homelessness than the general UK population (Flatau, Smith, Carson, & et al., 2015; David Robinson, Reeve, & Casey, 2007). People who have arrived in the UK seeking asylum make up the 0.6% of the total UK population (Oxford University Migration Observatory, 2022). Depending on the circumstances in which migrants enter a country they might not have the same recourse to public funds as a local and, therefore, may not be entitled to the same support (Gray, Rodriguez-Guzman, Argodale, & Bartholdy, 2021). Citizens Advice estimated that by the end of 2019 there were 1.376 million people in UK that had no recourse to public funds (Oxford University Migration Observatory, 2020).

A report by the European Observatory of Homelessness found that among in countries like Denmark, Finland and the UK, asylum seekers, refugees and irregular migrant who had not been granted asylum were not present in homelessness services in large numbers or visible in their records (I. Baptista, Benjaminsen, Busch-Geertsema, Pleace, & Striano, 2016). This makes counts through local authorities or housing support services less likely to identify and count them.

Overall, the literature reviewed here highlights the need for bespoke and wide-reaching strategies to produce robust counts of the homeless migrant population. Given the complexities of capturing data about this group, the literature supports the use of a combination of methods that includes a combination of street counts, national surveys, and/or administrative data from multiple service types (Hermans et al., 2020; Netto et al., 2004).

Reasons why this group might be missed from counts

Factors that contribute to the hidden homelessness of migrants, asylum seekers and refugees are multiple and closely linked to each country's laws regarding movement in and out of their territories (Kissoon, 2010). Patterns of migration, legal requirements and, therefore, pathways into and out of homelessness faced by migrants change over time (Fiedler, Schuurman, & Hyndman, 2006; Hermans et al., 2020). Any homelessness data collection strategy needs to be responsive to these changes.

Limited access to homelessness support

The evidence suggests that limited access to housing support is due to complex but interrelated circumstances of those who migrate. Focusing on the UK, people who have migrated here might have no recourse to public funds depending on their status. Furthermore, asylum seekers and refugees in UK are often placed in temporary accommodation that is dispersed across a territory where they have no connections or local social networks to turn to in case of need (Netto et al., 2004). Additionally, while obtaining their refugee status, they are often required to navigate a number of complex systems, including a new housing system with very little social and economic capital (Shankley & Finney, 2020).

In relation to this, a study published in 2004 on homelessness of minority ethnic people (including recent migrants) in Scotland noted that having limited access to information due to language barriers is also an obstacle to accessing support and, consequently, being counted through service providers records or routine homelessness data collections (Netto et al., 2004).

Concerns over disclosure

In the evidence it was noted that service providers often found themselves considering how ethical it was for them to count people in irregular situations. The term 'functional ignorance' was used to refer to the practice of ignoring the legal or illegal immigration status of people who approach support services for help to avoid turning them away (Hermans et al., 2020). The disclosure of personal information of residence status for migrants presents a challenge to consistent enumeration of this population. This could be a contributing factor to the lower presence of migrants, asylum seekers and refugees in homelessness services or local authority data.

5.1.6 LGBTI people

The literature on LGBTI hidden homelessness reviewed here had a substantial overlap with the literature on youth hidden homelessness. The evidence highlighted the overrepresentation of LGBTI people in the homeless population and particularly of LGBTI youth across multiple countries (Ecker, 2016; Gausvik, 2015; Norman-Major, 2018; Norris & Quilty, 2021; Sanders, Whelan, Murcia, & Jones, 2022). There was therefore a gap in research on adult LGBTI homelessness.

Evidence points at the complexity of LGBTI homelessness as it is related to complex circumstances (Norman-Major, 2018). Some of the structural drivers of LGBTI homelessness mentioned in the literature, and specifically for LGBTI youth homelessness, were discrimination, poverty and inequality (Ecker, 2016; Norman-Major, 2018; Norris & Quilty, 2021). In addition, the intersection of these circumstances with personal experiences of being in care; the presence of family conflict, violence and instability and the rejection of caregivers because of a person's sexual orientation or gender identity have been identified as potential triggers of a homeless episode (Norris & Quilty, 2021).

Reasons why this group might be missed from counts


One of the difficulties for this group to be included in homelessness figures is directly related to the discrimination faced by LGBTI people (Ecker, 2016; Ecker, Aubry, & Sylvestre, 2019; Norris & Quilty, 2021). Norris & Quilty (2021) noted that LGBTI homeless youth do not often use mainstream homeless services in the US for fear of experiencing discrimination and they remain unrecorded in homeless statistics. There is a risk that when counts are conducted through mainstream service providers LGBTI people may not want to present as such for fear of discrimination.

Another study in the US and Canada found that LGBTI people tend to have specific patterns of support service use (Ecker, 2016). For example, LGBTI people in the US and Canada made more frequent use of sexual health clinics, food programmes and mental health services than non-LGBTI people experiencing homelessness (Ecker, 2016). As this was a known pattern, this study included these kinds of services to increase the chances of this population which proved a successful strategy.

Inconsistencies in data collection

Inconsistencies in the way the gender and sexual identity are recognised and defined can skew results and complicate compiling and comparison of data across service providers (Ecker et al., 2019; Norris & Quilty, 2021). The main risk is that inconsistent categories and groupings can lead to an undercount of parts of the LGBTI community and limit the understanding of their specific circumstances when homeless. It can also lead to small sample sizes which are an issue for achieving necessary power in statistical analysis (Ecker, 2016).

A qualitative study by Norris and Quilty (2021) with LGBTI people experiencing homelessness in Ireland took an inclusive approach to definitions of gender and sexual identity which helped capture better quality data on this population. To avoid excluding people from the LGBTI community they used a broad definition of gender identity. This helped include as many people as possible from across the spectrum of LGBTI identities in the sample. This meant using the expansive abbreviation of LGBTQI+ meaning lesbian, gay, bisexual, transgender, queer, intersex and other sexual orientations and categories not accurately described by the previous terms used in research to cater to the continuous expansion of our understanding of gender and sex orientation.

Concerns over privacy

The previously mentioned study in Ireland also noted that LGBTI homelessness is sometimes unrecorded in official homelessness counts because service providers have concerns about the ethics and privacy implications of collecting data on homeless people's gender identity and sexuality (Norris & Quilty, 2021). Fears of pressuring people to 'come out' or disclose information they do not feel comfortable sharing can lead to lower response rates of questions relating to gender and/or sexual identity in surveys administered by support services staff.

Not perceiving themselves as homeless

A study with homeless LGBTI youth in Ireland found that there was a widespread perception that being homeless is exclusively being roofless and this complicated their self-identification as homeless when spent the night sofa surfing or in hostels (Norris & Quilty, 2021).


Email: socialresearch@gov.scot

Back to top