Harm reduction through housing first: an assessment of the Emergency Warming Centre in Inuvik, Canada

While homelessness in developed nations is documented historically, its significance as a social problem gained significant attention in the 1980s [15]. Since then, homelessness has grown to be a perplexing and vexatious problem for policy makers, and an embarrassment to nations and communities. Within the Canadian context, homelessness is at crisis levels [7]. While urban homelessness may have been the impetus for attention, one that initially occupied the foreground of research and policy development, attention to rural homelessness emerged as its own significant problem in the 1990s [29]. Rural homelessness may share key elements with its urban counterpart, but it remains on the periphery of attention as it contradicts common conceptions of health and social well-being often associated with life in rural communities. Significantly, rurality may actually add to and exacerbate the problem of homelessness in terms of health and well-being. This is particularly true in the Canadian Arctic, where life is either viewed as uninhabitable or viewed as communal, mutually supportive and in harmony with nature [35, 43, 44]. Arguably, neither image is accurate, but the rurality and remoteness of the area coupled with the extreme weather mean that there are fewer services to draw on, more acute housing needs and increased difficulty of bringing in supplies including food and building material. This exacerbates issues of homelessness as availability of affordable housing generally cannot keep pace with demand.

To be sure, the causes of homelessness are numerous, complex and not easily solved. Substance abuse, mental illness, physical and emotional abuse, marital breakdown, loss of employment, transition from institutionalized care, and economic factors such as, lack of affordable and/or available housing, and economic restructuring figure prominently in research on homelessness [10, 27]. Some authors have explored socio-structural explanations and found that the demise of social housing programmes in Canada in the 1990s is clearly linked to a surge in homeless populations, both urban and rural [5, 7, 34]. Similarly, in the UK, May et al. [26] trace the growth in numbers of the single persons experiencing homelessness through the 1980s and 1990s with the dismantling of the welfare state and then with the increasing governmentality of neo-liberal and social interventionist policies to tackle chronic street homeless in the 1990s and early 2000s. Similarly, Milbourne and Cloke [30] later suggest that the last few decades have seen increasing complexity emerge in homelessness in Australia as the demographics of those experiencing homeless shift to encompass a wider variety of people impacted by socio-economic changes in health, housing and poverty. In New Zealand, issues of street homelessness have been relatively hidden. In comparing Auckland in New Zealand to Vancouver and Edmonton in Canada, Collins [7] highlights the significant difference in welfare and housing policies of the two countries. Where Canada experienced a significant increase in homelessness in the wake of cutbacks to social housing in the early 1990s and early 2000s, Collins’ [7] research suggests that New Zealand has been somewhat sheltered from homelessness because of social housing together with cultural practices that protect socially excluded family members. However, more recent neo-liberal policies and increasing housing cost may render a different picture in the near future, as intimated by Anderson and Collins [2] in a comparative study exploring indigenous homelessness between Canada, New Zealand, and Australia.

Addressing large scale socio-structural challenges in a global economic context could possibly be one of the most difficult challenges for developed nations in recent history. Neo-liberal policies related to housing crises and the demise of social welfare policies have clearly taxed communities and individuals [24] leading to an expanding population of hard to house people, those who are chronically, cyclically, temporarily or transitionally homeless [10]. The global economic collapse in 2008 propelled homelessness to new levels as nations scrambled to balance budgets through austerity measures which negatively affected social welfare policies [12, 15, 32].

Housing First is grounded in a harm reduction philosophy. It is an approach that assumes helping people from where they currently live rather than from artificial, unrealizable and agency-derived goals of abstinence and sobriety results in better long-term results [17].

Research supports the application of Housing First models in urban and rural Canadian contexts [18, 23, 41], and in Australia, Great Britain, New Zealand and USA [28], particularly for homeless persons dealing with substance abuse issues. While context may dictate subtle variations, harm reduction principles stipulate that safe, stable housing, client centred support, highly integrated care teams, such as assertive case management, and transitional/supportive housing options lead to lower levels of service consumption, higher levels of treatment seeking behaviour, lower levels of substance abuse and stable housing over the long term [23, 28, 41].

This paper details the findings from an evaluation of the effectiveness of the Emergency Warming Centre (EWC) in Inuvik during the winter months in 2014–2015. This harm reduction project, loosely based on a Housing First model, provides the context for exploring comparative approaches for addressing homelessness and concurrent disorders in rural areas.

The research took a fundamentally community-based research approach working with the local community [31, 33] employing a participatory action research (PAR) methodology. The researchers obtained ethical review from the Royal Roads Research Ethics Board, and approval for the research in the Northwest Territories was granted by the Aurora Research Institute. Initially, the research involved qualitative and quantitative research; however, the small sample rendered the results of quantitative aspect of the research inconclusive. Consequently, this paper presents the findings from the qualitative component of the research. Respondents were recruited to participate by the researchers who visited the EWC shortly after opening in October 2014. A sample of nine out of a possible 20 “guests”, as they were called, completed 1 h interviews and participated in focus groups at two points in time: October 2014 and April 2015. Interviews with two additional guests were completed at either the pre- or post-period, but they were left out of the analysis as they did not complete both of the interviews or focus groups. Guests were asked about their health, well-being, lifestyle and support systems, their thoughts on the impact of the centre and their perspectives on the community and community-based supports. While an interview guide was used to start questioning, probes were used to encourage guests to elaborate their answers (see Appendixes 1 and 2 for the April 2015 interviews schedules with guests and stakeholders). This approach allowed for more authentic, rich and arguably more accurate responses from guests [25]. Three interviewers were used in each round, with two remaining consistent. Each guest was given a $25 gift card for completing the October component and another $25 gift card for the April aspect of the research, which were redeemable at a local store.

In addition to guests, key stakeholders of the EWC, including volunteers, staff members, board members and founding members, participated in focus groups and 1-h-long interviews in October and April. Stakeholders were contacted in August 2014 and asked if they would be willing to support and participate in the research. The interviews and focus groups with stakeholders were aided by an interview guide with questions on the management and functioning of the centre and how attendance at the centre affected guests. The interview guides were created and validated by the research team using findings of previous research done in the rural homelessness and housing first initiatives. In total, seven interviews were held with staff and key stakeholders. Four focus groups were held, with between four and 12 participants. The same themes formed the basis of questions asked, but these followed a more emergent path. Participants included a mixture of guests, staff, board members, founding members and other interested community members.

Situated within a symbolic interaction framework, which highlights the intersubjective nature of social reality [36], a social constructivist approach to research guided that data analysis. Homeless persons with concurrent disorders are marginalized from the broader community, but they represent a distinct community unto themselves. Accordingly, the data analysis lends itself to an interpretive framing which emphasizes equality. Harris [14] argues that social constructionists are compelled to engage in research that can improve equality of relations and social condition. Thus, the content analysis of the interview and focus group data are informed by interpretive, analytical frames that provide participants an expressive voice [25] with a view to promote social change. Based on the constructivist approach used here, the content analysis from the interview and focus group data with guests and stakeholders were used to identify emergent themes.

The results provided some specific information regarding the effectiveness of the Emergency Warming Shelter in Inuvik, including suggestions for future endeavours. This section summarizes key findings before outlining the emergent themes and subthemes. For Inuvik, this and previous research indicate there is a clear migration of homeless men and women into Inuvik from surrounding areas. Earlier research by Young and Moses [43] supports the claim that services for this population are fragmented, inadequate or ill prepared to cope with the special needs of this population. It supports the observation that mental health and addiction issues are common within the population experiencing homelessness and that mental health issues are often undiagnosed and/or untreated. This is partially due to a lack of service. Similarly, there is a severe lack of supportive housing designed to support a population with specific and often numerous issues. For example, there is an evidence of multiple intersections of violence and trauma, including post-traumatic stress disorder, intergenerational trauma and experience of residential schools, which exacerbate the potential for mental health and addiction issues.

The evaluation of the EWC in Inuvik provides evidence of the potential efficacy of the Housing First wet shelter model for homeless persons with concurrent disorders. As a logical extension of harm reduction approaches to addiction, discussions with guests of the centre suggested they intended to reduce alcohol consumption or had already started taking steps to reduce alcohol intake. Arguably, the EWC had an overall positive effect on many guests in terms of changing alcohol consumption behaviour. Findings from the research suggest guests’ levels of social functioning improved over the pre- and post-test interviews. However, there was a high level of anxiety given the impending closure of the EWC. There were indications that the centre provided a level of habituation, which may have left guests more vulnerable during the period when it was not open.

From guests’ perspectives, the qualitative data provide an encompassing interpretation of the EWC. The importance of the centre for health and overall well-being was a constant theme. A safe and warm place to stay that provided some meals, a place to store one’s personal belongings, shower and laundry facilities provided respite from the life of being homeless and addicted. Connecting guests with health and social services available in the community was not in the centre’s mandate, and few guests accessed resources. Clearly, the results demonstrate a need for more comprehensive, coordinated and inclusive services. This was reiterated by data collected from the centre staff. Respondents relayed that some basic structures were in place to support the guests and staff, but these were relatively rudimentary. In time, these policies and processes could be further developed, including staff training.

The data provide some insights into the effectiveness of the Housing First models in rural and/or northern contexts. Research from several jurisdictions underscores these observations. Regarding harm reduction strategies in Canada, Krause et al. [23] observe that Housing First approaches are correlated with improved health outcomes, reductions in substance use, increased health seeking behaviours and more prosocial activities. Waegemakers Schiff and Turner’s [41] research on rural homelessness in Canada echoes these findings. Moreover, MacIntyre’s [28] research on the effectiveness of 11 wet shelters, those not requiring sobriety, in Canada, England, Ireland, New Zealand and the US, identifies the positive effects of Housing First approaches on clients’ well-being. In all cases, however, Housing First can be considered a necessary but insufficient response to the problems experienced by homeless persons with concurrent disorders. Other necessary ingredients for serving this population include programming that addresses the myriad issues associated with homelessness and concurrent disorders. Housing First programmes based on harm reduction require a client centred approach, intensive case management that is responsive to individual clients’ needs, continuous support for clients, respectful and trained staff, interagency collaboration between service providers and access to community programming and social activities [23, 40, 41]. MacIntyre summarizes this approach succinctly when she states that wet shelters work because “…leaders of organizations providing these services and the staff who, on a daily basis, offer a mix of compassion, realism and professional support to people who desperately in need” (2009, p.3).

This research provided an evaluation of the Inuvik EWC operating between October 2014 and May 2015. The EWC was loosely based on a harm reduction approach as it offered meals and a safe, warm place to sleep for homeless or hard to house persons with concurrent disorders who could not access other sleeping accommodations because they were under the influence of alcohol or drugs, or because of their behaviour. Admittedly, the research does have shortcomings—it is based on a small sample of homeless persons and stakeholders. As well, the centre was not designed to provide any services other than food and nighttime winter shelter, and as such, does not fit a true Housing First model of intervention. In addition, although guests can be inebriated when they enter the centre, they cannot drink on the premises or after they have checked in for the night; therefore, the EWC is not designed to be a wet shelter. These aspects alone should prove fatal to the centre’s operation, yet without the centre, there is a possibility that the guests of the centre would have been at serious risk of illness, injury and death. Although the data provide qualified support for the EWC, it is clear that it served its function of keeping guests safe and alive. That the RCMP reported fewer admissions to their cells is an evidence of a positive effect on the EWC in terms of the appropriate use of police services.

Overall, the results from this research corroborate much of the extant literature on Housing First in term of the elements necessary for successful Housing First approaches to homelessness based on harm reduction. This research established a starting point in rural, northern contexts by identifying the complex interplay of complications between homelessness, mental illness and addiction in harsh and remote environments, particularly in the wake of massive global economic changes affecting resource extraction in northern Canada. This also provides reason to challenge the status quo notion of rural homelessness articulated by Cloke and Milbourne who suggested that “…it remains the case that rurality can also be intertwined with political conservatism, moral individualism and cultural tendencies to blame the victim” ([6], p. 273). A holistic systems approach that recognize and address multiple and intersecting issues that lead to, and keep people within, homelessness are much more effective in terms of long-term strategies. Harm reduction strategies such as wet shelters and the Housing First model offer a more holistic approach, but they come with public criticism. They require clear vision, community consultation, support and education, and strong allied support systems of transition. In addition to continued and larger research projects on homelessness and concurrent disorders in rural locales, future research should examine the most effective strategies used to promote and develop housing first strategies in communities lacking the infrastructure and expertise to implement harm reduction approaches to homelessness and concurrent disorders.