The role of gender in housing for individuals with severe mental illness: a qualitative study of the Canadian service context.

OBJECTIVE: This study was undertaken to examine the role of gender as it relates to access to housing among individuals with severe mental illness (SMI) in Canada.
DESIGN: An exploratory, qualitative approach was used to assess the perspectives of Canadian housing experts. The focus of inquiry was on the role of gender and associated intersections (eg, ethnicity) in pathways to housing access and housing needs for individuals with SMI.
SETTING: A purposeful sampling strategy was undertaken to access respondents across all Canadian geographic regions, with diversity across settings (urban and rural) and service sectors (hospital based and community based). PARTICIPANTS: -29 individuals (6 men and 23 women) considered to be experts in a housing service context as it pertains to SMI were recruited. On average, participants had worked for 15 years in services that specialised in the support and delivery of housing services to people with SMI. MEASURES: Semistructured interviews with participants focused on the role gender plays in access to housing in their specific context. Barriers and facilitators were examined as were intersections with other relevant factors, such as ethnicity, poverty and parenthood. Quantitative ratings of housing accessibility as a function of gender were also collected.
RESULTS: Participants across geographic contexts described a lack of shelter facilities for women, leading to a reliance on exploitative circumstances. Other findings included a compounding of discrimination for ethnic minority women, the unique resource problems faced in rural contexts, and the difficulties that attend access to shelter and housing for parents with SMI.
CONCLUSIONS: These findings suggest that, along with a generally poor availability of housing stock for individuals with SMI, access problems are compounded by a lack of attention to the unique needs and illness trajectories that attend gender.

To understand the role of gender in housing access and needs among individuals with severe mental illness (SMI).

Misperceptions at policy and service system levels regarding the need for housing as a function of gender are leading to circumstances that compound the impacts of mental illness.

Resource needs are particularly acute as additional points of marginalisation emerge from factors such as ethnic minority status and rural context.

Housing access problems have major impacts on the ability of individuals with SMI to raise children.

This study is among the first to directly address the interplay between housing access and gender among individuals with SMI.

The study accesses a broad, national range of expert perspectives.

Transferability to other national contexts outside Canada needs clarification as do potential biases attending respondents (ie, primarily female provider perspectives, no consumer perspectives represented).

While serving the purpose of an initial inquiry, further work is needed to confirm and expand upon the assertions of the service providers interviewed here.

The lack of adequate and affordable housing for individuals with severe mental illness (SMI) is repeatedly highlighted in research, public and policy discourses. It is a complex problem rooted in individual and structural factors. Individuals with SMI often present with a complex array of concurrent psychiatric and chronic medical conditions which are often compounded by poverty, social exclusion and substance misuse. Many structural factors also contribute to a lack of good-quality housing for people with SMI. Rent is often unaffordable, with costs typically taking up 70–90% of an individual's public benefits. This problem is exacerbated by Canada having just over 25 000 supported housing units available.1 As a result, over 500 000 Canadians with mental illness are inadequately housed and an estimated 120 000 are homeless.1 This is a critical problem as adequate housing is a key determinant of health2 and recovery.3 4

From research and practice perspectives, challenges arise in the very generic articulation of the housing problem for the SMI population. Considerations of the issue of housing have very seldom taken into account the intersections of resource and adversity that attend factors such as gender, race and ethnicity. As long as the specific mechanisms by which such factors and their intersections remain poorly understood, housing policies and programmes for people with mental illness will remain inefficient and ineffective.5 It is in light of this shortcoming in the SMI literature that the present study was undertaken to examine the role of gender in housing for Canadians with SMI. This initial qualitative exploration draws upon housing expert perspectives regarding service access and service needs as they intersect with gender, ethnicity and geographic context.

Gender and SMI

There are significant and pervasive differences between men and women with SMI that are present across the social determinants of health, illness onset and expression and the process of recovery. Taking schizophrenia as an example, women on average experience a later onset and less severe course of illness, though they suffer more depressive symptoms than men.6 A later onset of illness carries several important implications. Women typically have greater opportunities to establish a range of competencies and a higher degree of independence before the illness starts. This may account for the generally stronger skills that support independent living and longer periods of community tenure.7 8

Women with SMI have also been found to more actively seek social contact and support, which may have implications for less dependence on service providers and greater success in living independently.7 9 Men with SMI are, in turn, more likely to be homeless10 and more reliant upon health services,9 which may account for a further deterioration of independent living skills.11

Women with SMI also face several unique and prominent sources of adversity. Women are more likely to have experienced childhood abuse and trauma and symptoms of post-traumatic stress disorder.12 They also experience higher rates of coercive sexual encounters and sex trade involvement as a means of coping with poverty and homelessness.13 Coercion in relationships with male partners has also been extended to include influence upon the uptake of substance abuse and involvement in criminal activities.14 Furthermore, women have a greater likelihood of having been socialised into passive and otherwise dependent roles, which may be reified through service providers working within models of care that cultivate dependency.7 Such sociocultural factors are evident across a range of studies. For example, compared with women without SMI, those with SMI place a greater importance on having caring and compassionate providers than providers who encourage choice and independence.15 Women with SMI are also more likely to be perceived by family and other supports as more emotionally vulnerable and irrational in comparison to men.16 17

Pregnancy and child-rearing are an additional consideration for women with SMI. Typically framed in the context of enhancing risk, having a child has numerous implications around risks of medications in pregnancy and stressors associated with child-rearing as poverty and illness threaten custody.18 This is an increasingly relevant issue given the evidence that the fertility rate among Canadian women with schizophrenia is on the rise.19

Much less developed in the literature is attention to intersections of risk and resilience that attend the intersections of race, ethnicity, sexual and gender identity. Preliminary work suggests that these points of diversity have very important implications. While unique points of resilience emerge among these populations, recovery is typically complicated by multiple forms of discrimination and marginalisation and treatment needs that are seldom met adequately.20 21

Present study

Despite the consistent threads of evidence suggesting differences in mental illness aetiology, expression and recovery process as a function of gender, as a topic it is largely relegated to sidenote status in community mental health research and policy dialogues. This shortcoming is particularly evident in considerations of housing individuals with SMI. As such, this qualitative study was undertaken as an initial step to better articulate community service needs in Canada as a function of gender. Key informant interviews were undertaken with service providers across Canadian provinces and territories in both hospital and community service sectors and across both large urban, mid-sized cities and rural contexts.

Methods

Recruitment

In order to obtain a broad, national perspective on gender equity within the housing sector for people with SMI, service providers were sought in provinces and territories in all geographic regions. The recruitment strategy targeted balanced representation as a function of setting, with interviews conducted with providers working in the largest urban centres in Canada as well as representative small to mid-sized towns and cities. The representativeness of smaller centres was determined by a consensus of the urban providers (the inquiry starting point in each province/territory), seeking settings that were neither disproportionally over-resourced nor under-resourced. Balance was also sought in the respondent service sector, seeking representation from both hospital-based and community-based organisations.

Participant eligibility relied primarily on their being regarded as an individual who is highly knowledgeable regarding access to housing and community services for people with SMI. Recruitment started in large urban centres, with participants being recommended through sources such as key informants in the Mental Health Commission's recently completed consultation on housing and the At Home study. Large urban setting respondents were then asked to provide recommendations regarding experts in smaller settings. Recruitment continued until each region was adequately represented and qualitative themes had attained saturation with subsequent interviews revealing no new information. The study was reviewed and approved by the Toronto Center for Addiction and Mental Health institutional research ethics board (#183/2011).

Procedure

Interviews lasting an average of 45 min were conducted with all participants via telephone and were audio-recorded and transcribed verbatim. Demographic data collected for each participant included geographical location, type of service, number of years in current post, number of years working within mental health and housing sectors, gender and profession. The interview then moved on to explore participant perspectives and experiences pertaining to differences regarding access to, experiences within and needs for housing and associated resources as a function of gender.

The interview concluded with two items that were developed to quantify service provider perceptions regarding access to housing as a function of gender (eg, in a broad, general sense, how accessible is affordable housing for women/men with SMI in your city/town?). Participants were asked to score each item using a five-point Likert scale where 1 represented ‘completely inaccessible’ and 5 represented ‘completely accessible’. This scale was developed for this study as we were unable to identify an existing, validated tool for this purpose.

Analysis

While not a complete grounded theory design, the analysis followed the rigorous thematic analysis procedure outlined by Charmaz,22 23 beginning with line-by-line coding of the data to pull out key, recurrent themes. Line-by-line coding was completed by one coder with category reports examined by a second research team member who also reviewed full transcripts and recordings of interviews. Coding was reviewed at several stages of the analysis, allowing for data to be contextualised, and different interpretations explored and resolved through consensus. Focused coding, where conceptual analysis took place, started after eight interviews when coherent and consistent linkages between categories and themes began to emerge. Emergent themes were shared in subsequent interviews for feedback and further exploration, which allowed for an ongoing process of member checking and elaboration of the analysis. Quantitative data were examined descriptively with mean differences being examined through independent samples t test analysis.

Results

Participants

Twenty-nine participants were recruited into the study (6 men and 23 women). All participants had worked within the mental-health sector, ranging 1–30 years, with 20 years being the average length of service. Most worked in either director level or direct service positions, at 45% and 31%, respectively, with the remainder working in middle management. On average, participants had worked for 15 years in services that specialised in the support and delivery of housing services to people with SMI. Most were regulated professionals in areas such as Social Work or Counselling and 59% were members of local advisory committees and boards related to housing. Table 1 provides a detailed summary of participant demographics.

Table 1

Geographical summary of participants differentiating by hospital or community sector

		City		Small–mid-sized town
Region	N	Hospital	Community	Hospital	Community
Canada	29	6	12	6	5
Western Canada*	11	2	4	3	2
Central Canada†	6	1	2	1	2
Atlantic‡	8	3	3	1	1
Northern§	4	0	3	1	0

*Western Canada includes Alberta, Manitoba, Saskatoon and British Columbia.

†Central Canada includes Ontario and Quebec.

‡Atlantic includes Prince Edward Island, New Brunswick, Nova Scotia, Newfoundland and Labrador.

§Northern includes Yukon Territories, North West Territories and Nunavut.

Ratings of accessibility

Quantitative measures of participant perceptions regarding global housing access revealed a mean score for women of 2.0/5 (SD=0.68) and for men of 2.3/5 (SD=0.86), a difference that was not statistically significant (t=1.53, p=0.13). Qualitative consideration of rating differences as a function of the service sector (hospital vs community) and setting (large urban vs town/small city) revealed minimal differences for both women (difference of ±0.1) and men (difference ±0.3).

Qualitative findings

Participants uniformly regarded housing as the basic source of ‘stability’ and a key-health determinant that is equally relevant for men and women with SMI. Numerous, generic themes recurred such as the comments on the low levels of housing stock, the limited availability and allocation of funding to develop and sustain adequate housing, and client difficulties in obtaining and maintaining adequate sources of income with which to cover rent and other basic expenses. Across service sectors and geographic regions, the context of supporting individuals with SMI was described as one of the pronounced financial strains.

Despite several cross-cutting similarities, these key-informant narratives of poverty, under-resourced housing service sectors, and stigma differed markedly as a function of gender, ethnicity and other points of intersection. Erroneous assumptions about service needs and stigmatisation were described as having a structural impact upon housing access. As will be described, this resulted in unique pathways of marginalisation, victimisation and poverty.

‘Structural’ adversity: cycles of invisibility, lack of access and vulnerability

Women were frequently referred to as the ‘invisible homeless’. Participants described how the perceived absence of homeless women and the perception of homeless men as more troublesome influenced policy, funding and service provision, resulting in women having to “turn to different systems than the formalized system” such as ‘couch surfing’ with families, friends and often with unknown men. While one provider found women's ability to seek alternative routes as ‘resourceful’, most voiced the opinion that the lack of shelter or emergency housing for women increased their vulnerability. For example, women with SMI would “trade sex for somewhere to stay” and “take somebody [in an exploitative context] in to help them pay the rent when they're feeling under pressure because of the costs.” For the majority, this was a question of powerlessness rather than one of resourcefulness.

Women's [with SMI] homelessness tends to be a little bit more invisible. We certainly have a number of women who have lived with their sister for a long time or crashed with unsavory males over a series of years. They have, in fact, been homeless they have just been temporarily couched. If you look at resource availability…they operate out of men's shelters, and aren't always in a safe environment for women. [Resources] certainly seem to be centered on male-accessible places…all these guys have all these shelters to choose from and if they get booted out of one after four weeks they just move over to the next one. We turn hundreds and hundreds of women away from our shelter because we just don't have the beds.

[Community service provider, small town in Central Canada]

Women were described as being pushed into a vicious cycle of unstable housing and increasingly stressful environments, which led to poorer mental health, quality of life and little access to community services. This cycle was compounded in smaller towns, as women had to leave for larger cities where there was better access to mental health and housing services.

While men with SMI were described as having relatively greater access to housing and shelter, and experiencing less victimisation and sexual exploitation, they were in no way immune to the dynamics of poverty and marginalisation. They were described as typically being forced to live in ‘drug-infested’ areas due to stigma and poverty. Living in such neighbourhoods leads to high levels of anxiety among both men and women, and participants describe this as greatly increasing the risk of a relapse.

Intersection of gender and ethnicity

Belonging to an ethnic minority group acted as a further source of marginalisation, not just within hospital and community services but also within people's own ethnic groups. The evidence of ‘blind spots’, which fuelled gender and ethnic discrimination, were evident mostly in the experiences of women.

I'm thinking of one woman in particular who was from Pakistan and had been married most of her life but recently divorced and found herself on her own and with no place to live. I remember taking her to the shelter just before Christmas and she walked in her head scarf into this predominately male shelter waiting area and you could just feel it, it was very striking the degree to which she didn't fit in.

[Community service provider, urban area in Northern Canada]

For Aboriginal people living in rural Canada and the northern territories, the level of discrimination experienced by women with SMI was described as particularly heightened. Geography and location, intersecting with gender, ethnicity and intergenerational trauma, meant that women were trapped in situations compounded by lack of transportation and an inability to leave town.

We worked with a woman from a small Aboriginal community. Her non-Aboriginal husband got on the housing authority board and illegally had her name removed from home ownership papers just because the government could. She didn't win, because even lawyers are racist in the territories…so in small communities it is very controlled by men and they will choose men over Aboriginal women…Aboriginal women are the most marginalized.

[Community service provider, small town in Northern Canada]

In the Aboriginal service context, several participants noted a conflict between the community and hospital sectors, in that “psychiatrists, psychologists, and nurses”, who are considered “mainstream”, were not using a “culturally competent approach”, which resulted in a “racist” approach to care. These dissonances between community and hospital services created a discontinuity in care and impeded the recovery of Aboriginal women. Providers talked about how the “structural oppression” was internalised by women, who began to “talk about it like it's their fault”, and used words such as ‘loser’ and not having the “smarts or self control” to improve their lives.

Men, perceptions of risk and housing quality

While participants described the impact stigma had on both men and women, their narratives emphasised the manner in which it pervaded men's experiences. They described how the level of ‘trouble’ and ‘risk’ perceived by providers seemed to shape the transitioning and access to community housing for men with SMI. It was perceived as a “reality that men are more prone to violence” compared with women, who were often viewed as “easier tenants.”

More men end up without housing because as part of their illness they're maybe using more and tend to be more aggressive … landlords are more afraid. I mean we definitely have women that have been quite agitated but I think landlords tend to be more accepting of females.

[Hospital service provider, urban area in Western Canada]

Participants also described men as being subjected to more substandard levels of accommodation, based on the perception that men “don't care” about their homes and are content with living in “dorm-like situations”.

Women, children and families

There was a general perception across participant narratives that men with SMI seldom had family responsibilities. For men with children, this left them at a disadvantage—if they had a child, they would have even greater difficulty in accessing suitable housing than their female counterparts. The majority of the commentary in this topic, however, revolved around women's experiences. The presence of children created further complexity for women in the housing system, oscillating between facilitating and hindering housing access, recovery and community participation. Access was described as improved by some participants based on “the premise that women, as child bearers, will require family housing.” Having children was also described by some as facilitating greater access to services as their physical and mental health needs are considered a priority by providers. Others, however, described women with children as having greater difficulty in accessing care due to childcare responsibilities and fears that their circumstances of mental illness and victimisation may lead to their children being taken into custody. Additional challenges included having adolescent male children who are not allowed in female-only housing and shelter settings, and ‘catch-22’ scenarios that can make it nearly impossible for impoverished women to regain custody once it is lost.

So when a woman applies for housing, if they don't have guardianship they can only apply for a single even though they need a 3-bedroom to get their kids back…so maybe she will get a one bedroom and is living in an overcrowded situation again…living in that overcrowded place plays into the mental health of the parent and the children and that creates the tension and starts to stir the pot. Then throw some alcohol on that, and then starts the domestic violence, then she flees the violence, and the partner stays in the house and she has to start the process all over again.

[Community service provider, urban area in Northern Canada]

This problem was described as unfolding differently within some Aboriginal communities. For example, grandmothers were described as “rescuing” their grandchildren while mothers were “couch surfing” and “people (in these communities) take in their adult children and adult grandchildren”.

Discussion

This study of housing expert perspectives was undertaken to describe the housing service context for Canadians with SMI as a function of gender. It is an initial attempt to articulate these issues in the Canadian context and advance the small but growing knowledge base related to gendered experiences of serious mental illness.

This study highlighted several problems. When considered in the broadest sense, housing access problems were considered equivalent regardless of the gender and geographic context, and the types of barriers faced varied greatly. Biased perspectives regarding housing needs were described as leading to women with SMI being forced into exploitative and victimising circumstances. Even when accommodating women, the emphasis on male-focused services was described as being alienating and potentially risky for women to access. These findings are reflective of the literature that, while noting greater success with independent living among women,7 8 also describes much higher rates of coercive sexual encounters and sex trade involvement.13 Indeed, it brings forward the question of necessity in the greater levels of independence observed among women and whether, for many, the purported observations of independence may overlie contexts of exploitation and victimisation.

Additionally, while the many stresses that attend parenting by women with SMI have been documented,18 this study noted several implications for housing-related services. These included ambivalence about the net impacts for mothers with SMI. Some participants cited enhanced outreach and support extended to mothers and others describing mothers as having less time to access services or actively avoiding services for fear of losing custody. Fear of losing custody of children was associated with women being forced into higher-risk living circumstances. Women who had lost custody were also described as struggling with catch-22 scenarios in which they required adequate housing to regain custody but could not obtain it unless they had custody. The issues attending custody and parenting are, however, quite complex and were not fully addressed in this study. It is an area that would greatly benefit from further inquiry that could more intensively examine the intersections (gender, ethnicity, geography) and specific contexts (family vs individual homelessness) therein.

Men were also described as facing significant problems accessing adequate housing, albeit with some different factors involved. This was reflected in the finding that men were comparably rated with women in terms of overall access to housing services. Men with SMI were understood to have difficulties in accessing stable housing due to their being perceived as being more prone to disturbance and violence. Furthermore, there existed a perception that men cared less about the quality of housing and shelter spaces, resulting in their being provided with lower quality and, more frequently, large dormitory-type spaces. These findings are reflected in the observation of greater rates of homelessness among men with SMI.10

While considered to be relatively rare, child custody was considered to be a pronounced challenge for men as supports for families were described as being almost exclusively geared towards single parent women. This problem extended to women with teenaged male children who, if they appeared too adult, were likely to be sent to separate shelters due to concerns that their presence might disturb other female residents.

Ethnicity and service sector size were generally described as compounding the housing access problems faced by women. Consistent with previous work,21 it was observed that racialised women faced several forms of discrimination in male-oriented shelter and housing systems. Provider concerns for Aboriginal women revolved around culturally inappropriate care and the pronounced lack of access to housing and limited mobility that attend poverty in remote communities—risks that are enhanced when political and service leadership structures are predominantly male and equitable policy can be undermined through patriarchal implementation. Such observations align with international debate and advocacy regarding human rights violations against Aboriginal women,24 violations that need to be understood within ongoing practices and effects of colonialism. Conversely, some providers spoke of small Aboriginal communities providing greater support to women with SMI, with grandparents and other family members playing a greater role in care for children and housing when needed.

While the transferability of these findings will require further study that integrates a consideration of client perspectives and outcomes and extends beyond potential biases in the sample (eg, sample selection biases and the majority of participants being women), they point to a need for a systematic examination of inequities in the Canadian housing service sector for people with SMI. This group of Canadian housing experts observed structural forms of discrimination that were compounded by poverty and mental illness within the context of an ongoing pervasive difficulty in accessing adequate housing. As emphasised by Canada's mental health strategy,25 until gender and other key intersections such as ethnicity are better addressed, our services will continue to ignore major determinants of illnesses such as schizophrenia.26 27