With limited resources at their disposal, System Planners—such as the Calgary Homeless Foundation (CHF)—like to know how long various subgroups of persons experiencing homelessness will likely require various forms of housing support. With this in mind, Ali Jadidzadeh and I have co-authored a study that appears in Housing Studies—a leading housing journal. Titled Patterns of exits from housing in a homelessness system of care: The case of Calgary, Alberta, the study looks at the case of CHF-funded Housing First programs.
Here are 10 things to know:
- The study uses survival analysis and hazard models. To quote from the study: “[S]urvival analysis tells us when we can expect new housing units to become available for new tenants, and which programme types will have available units more quickly… [while] hazard analysis can tell us which tenants will be most likely to graduate, based on the individual characteristics of those tenants” (p. 7).
- The data in the study comes from Calgary’s Homelessness Management Information System (HMIS). Gathered between 1 April, 2012 and 31 March, 2015, the data pertain to people residing in Housing First programs funded by CHF. Thousands of people were involved in this data collection effort, including: persons living in Housing First programs who signed release of information forms; staff in Calgary’s Homeless-Serving System of Care who inputted the data; CHF staff who provided training and support to community on how to use HMIS; as well as CHF staff who then cleaned the data. The data in question were gathered on each person residing in Housing First every three months. An intake form was first completed by a case manager at intake into the program; then, another assessment form got completed every three months. There was also a form completed at exit from the program. These forms ask basic demographic information, as well as information about education, income, employment, history of family violence, use of health services and involvement in corrections. Most of the data gathered is based on self-reporting by an experienced case manager who receives accreditation. Blank copies of these forms can be accessed here.
- Key to the study is a concept known as graduation. Alberta’s provincial government provides homelessness funding to System Planners (such as CHF). To quote directly from the study: “In line with provincial programme guidelines in place during the period under consideration in the present study, a client is said to graduate from CHF-funded housing when they no longer require ‘housing support’ (i.e. case management). And in the case of temporary housing funded by CHF, a client is said to graduate once they complete programme requirements and move into a more permanent form of housing—either subsidized or unsubsidized…” (p. 3). [Note: subsequent to the period of study, the definition of graduation has changed for Calgary.] For a recent academic consideration of graduation, see this 2018 article.
- One of the study’s findings is that single adults without dependents require housing support longest, and families for the least amount of time. Put differently: single adults without dependents who have recently been homeless require social work (also referred to 'case management') support longer than other groups. I suspect a few factors may be at play here. First, single adults without dependents sometimes don’t have dependents because their children have been taken into child protection (possibly stemming from the parent’s challenges with mental health and/or substance use). It’s therefore intuitive that a person with such challenges would require social work support for a longer period of time. Also, in Alberta, singles without dependents receive less income assistance than other groups, making it more challenging to live independently (I encourage people to read the poverty chapter in this year’s Alberta Alternative Budget, which argues that the poverty gap for singles without dependents is much larger than for other groups).
- Women require social work support for longer periods than men (even when we control for employment and income). In fact, the study finds that men are 32% more likely to graduate than women. As noted in the study: “One possible reason for this is that women experiencing homelessness often find themselves in relationships with people who in turn jeopardize their housing stability…” (p. 20).
- Having a history of addictions does not appear to affect a client’s graduation rate. This is consistent with findings from the At Home/Chez Soi study, which found that formerly-homeless persons who consume large amounts drugs and/or alcohol maintain housing about as well as other formerly-homeless persons. This reaffirms the importance of the Housing First approach, which holds that a person should not have to go to a drug or alcohol treatment program as a precondition to receiving permanent housing.
- Older clients have lower graduation rates (meaning that it takes longer for them to move on to independence). Put differently: older people who have recently been homeless require social work support longer than other groups. The study notes: “Older clients having lower graduation rates should also not be surprising to many readers, as the health outcomes of seniors are poorer than those of younger clients” (p. 21). This is an especially important finding for System Planners across Canada, as older adults are making up an increasingly large share of the homeless population. In Calgary’s homeless shelter system, adults aged 55 and over, in 2017, accounted for 19% of bed spaces on any given night; in 2008, they accounted for just 9% of all bed spaces. These figures apply to single adults without dependents. This trend will likely continue for at least another decade or two.
- Findings pertaining to Indigenous peoples have already had ramifications on the ground. Indeed, the study finds that Indigenous peoples in the study needed support longer than non-Indigenous peoples. This holds even after controlling for income, education, and a history of family violence. Future research is needed that looks at factors that inhibit success among Indigenous peoples in Housing First [Note: Some research has already been undertaken on this in Edmonton. Check out this 2011 report and this 23-minute video]. These findings have also informed CHF’s engagement strategies with Indigenous peoples, including CHF”s hiring of an Indigenous advisor (since promoted to Director), a business case for two supportive housing buildings for Indigenous peoples (not yet funded) and future Indigenous-focused research.
- The study finds that having a source of income is positively correlated with graduation rates (i.e., it speeds up the move toward independence)—and this has already led to several changes in Calgary’s Homeless-Serving System of Care. This finding has helped inform an effort by CHF to identify, in collaboration with community partners, specific individuals in CHF-funded Housing First programs who, with some additional (short term) financial support, could likely graduate. This particular effort has been taking place for roughly one year; thus far, it has involved approximately 170 individuals with considerable success. The additional financial assistance provided varies by individual and is not intended to be permanent.
- In Calgary’s family homelessness sector, the study’s finding pertaining to income has led to the development of a new Adaptive Case Management (ACM) approach. ACM has a strong focus on providing short-term financial assistance to households in need (and it is discussed in detail in our Family System Planning Framework).
In Sum. This study finds that some groups move on from Housing First programs more quickly than others, and that some factors (such as a source of income) appear to accelerate graduation from Housing First. Because CHF embeds research into its day-to-day operations, we were well-positioned to start acting on findings well before the research was published.
For assistance with this blog post, I wish to thank Tim Aubry, Carla Babiuk, Victoria Ballance, Candice Giammarino, Ali Jadidzadeh (who also spent many hours cleaning this data so that if can be used for analysis by several stakeholders), Stephen Metraux, Shane Rempel and one anonymous source. Any errors are mine.
For a full copy of the article, please email me.