A substantial literature focuses on homelessness and social programming (Lee, Tyler, & Wright, Reference Lee, Tyler and Wright2010; Shlay & Rossi, Reference Shlay and Rossi1992; Toro, Reference Toro2007; Trypuc & Robinson, Reference Trypuc and Robinson2009), yet major gaps exist. At present, knowledge and practices on homelessness tend to focus on young adults and young families, with less attention to older people (Beynon, Reference Beynon2009; Burns, Grenier, Lavoie, Rothwell, & Sussman, Reference Burns, Grenier, Lavoie, Rothwell and Sussman2012; Cohen, Reference Cohen1999; Crane & Warnes, Reference Crane and Warnes2001; Gonyea, Mills-Dick, & Bachman, Reference Gonyea, Mills-Dick and Bachman2010; McDonald, Dergal, & Cleghorn, Reference McDonald, Dergal and Cleghorn2004). However, with population aging (see Edmonston & Fong, Reference Edmonston and Fong2011), the number of older homeless people is expected to rise (Crane & Warnes, Reference Crane and Warnes2010; Culhane, Metraux, Byrne, Stino, & Bainbridge, Reference Culhane, Metraux, Byrne, Stino and Bainbridge2013). A strong evidence base is required to address the unique challenges of older homelessness. Footnote 1
Based on a comprehensive literature review, this article sets the stage for a research agenda that will inform national and provincial strategies; policies in housing, health, and social care; and community services for Canada’s older homeless population. After a brief comment on homelessness in Canada, we clarify the intersections of aging and homelessness and sketch out the existing knowledge that can be used to counter homelessness. Drawing on international and Canadian research, we review terminology, definitions, and distinctions in the literature; available statistics and the estimated prevalence of homelessness in Canada; pathways into homelessness in later life; and examples of heterogeneous experiences that exist among older homeless people.
Homelessness in Canada
Canada has a vast social geography, with regional or place differences affecting homelessness in unique ways. Federal, provincial, territorial, regional, municipal, and Aboriginal governments have all invested in homelessness reduction strategies (Gaetz, Donaldson, Richter, & Gulliver, Reference Gaetz, Donaldson, Richter and Gulliver2013). At the federal level, the government funds and supports targeted communities through the Homelessness Partnering Strategy (HPS), launched in 2007 and renewed in 2013 (Employment and Social Development Canada, 2014). Footnote 2 To date, responses to homelessness have focused primarily on crisis response delivered through shelters and emergency health care. Recently, however, communities across Canada have embraced a housing-first model that is intended to immediately provide housing, followed by other forms of support. Footnote 3 Results from At Home/Chez Soi – the major nationwide demonstration project that tested the housing first approach in five Canadian cities – suggests that this model can effectively reduce homelessness, and that it is less costly than emergency responses (Goering et al., Reference Goering, Veldhuizen, Watson, Adair, Kopp and Latimer2014). Footnote 4
Despite the number of initiatives taking place across the country, there is little attention to the unique needs of older homeless people, or to determining whether suggested approaches are effective for older people (e.g., Gaetz et al., Reference Gaetz, Donaldson, Richter and Gulliver2013; Goering et al., Reference Goering, Veldhuizen, Watson, Adair, Kopp and Latimer2014). An exception is Quebec’s national strategy on homelessness, which in addition to committing $52 million to homelessness (CBC News, 2014) considers the specific challenges, vulnerabilities, and needs of homeless people over age 50 (Gouvernement du Québec, 2014). Recognizing older people as a population at risk is a first step, but concerted policies regarding resource allocation and the development of services designed to meet older people’s needs, including housing programs for older people, are still missing across Canada.
Methodology: A Literature Review on Homelessness among Older People
This article reports the results of a comprehensive literature review of research published between 1978 and 2014. Locating literature on homelessness among older people required multiple stages of testing search terms and a manual review of printed titles. Footnote 5,Footnote 6 The successful search strategy, carried out in Web of Science databases, employed the terms older adult, senior, elder, elderly, old age, and late life to identify articles related to older people (1,106,339 results), combined with the terms homeless, homelessness, and unhoused (771 results). A manual review of the 771 articles deemed 163 articles to be relevant. A final pool of 140 articles was selected for review once duplicates and book reviews were removed.
Summarizing the literature to arrive at a general understanding of homelessness among older people was challenging due to variations in methodologies, samples, and research foci. Studies concentrated on homeless women or men, particular trajectories (e.g., substance use, violence, mental health, etc.), and tended to be city or region specific. Further, the heterogeneity of the homeless population made it difficult to separate the impacts of age in combination with “race”, class, gender, health status, geography, service availability, and so forth (Aubry, Farrell, Hwang, & Calhoun, Reference Aubry, Farrell, Hwang and Calhoun2013; Rothwell & Mott, Reference Rothwell and Mott2013). Results were also constrained by a paucity of literature on the intersecting locations of homelessness in late life, as well as by reported challenges of access and maintained contact over time (Greenwood, Schaefer-McDaniel, Winkel, & Tsemberis, Reference Greenwood, Schaefer-McDaniel, Winkel and Tsemberis2005; O’Connell et al., Reference O’Connell, Roncarati, Reilly, Kane, Morrison, Swain and Jones2004). This review is a best attempt to outline the state of knowledge in the field and to identify gaps, in order to stimulate an agenda that is inclusive of the needs of diverse groups of older people who are homeless (or at risk of becoming so).
Defining Aging and Homelessness
Definitions and categories of homelessness vary among sources and between programs. The Canadian Homelessness Research Network (2012) comprehensively describes homelessness as “the situation of an individual or family without stable, permanent, appropriate housing, or the immediate prospect, means and ability of acquiring it” (p. 1). Their definition specifies four groups of homeless people: the unsheltered, the emergency sheltered, the provisionally sheltered, and those at risk of homelessness. Other sources use homeless in a narrower, more literal sense, referring to those who live in temporary shelters or on the streets. People who tend to avoid shelters and outreach services are often referred to as “rough sleepers” (Crane & Warnes, Reference Crane and Warnes2000; Johnsen, Cloke, & May, Reference Johnsen, Cloke and May2005; O’Connell et al., Reference O’Connell, Roncarati, Reilly, Kane, Morrison, Swain and Jones2004). Footnote 7
Distinctions are made between transitional or temporary, episodic or cyclical, and chronic homelessness (Aubry et al., Reference Aubry, Farrell, Hwang and Calhoun2013; Culhane & Metraux, Reference Culhane and Metraux2008; Echenberg & Jensen, Reference Echenberg and Jensen2008; Kuhn & Culhane, Reference Kuhn and Culhane1998; Rothwell & Mott, Reference Rothwell and Mott2013). Some researchers define the chronic category as three months of homelessness (Trypuc & Robinson, Reference Trypuc and Robinson2009), whereas others categorize people who are repeatedly homeless for more than one year as chronically homeless (McDonald et al., Reference McDonald, Dergal and Cleghorn2004). Against this backdrop, there are two notable trends where older people are concerned. First, people who are chronically homeless often use shelters as a means of housing rather than an emergency service. Second, chronically homeless people tend to be older, persistently unemployed, and are more likely to be disabled or use substances (Mott, Reference Mott2012).
Attempts to define homelessness among older people exist, but are more limited. There is general acceptance that homelessness among older people is increasing (Crane & Warnes, Reference Crane and Warnes2010; Culhane et al., Reference Culhane, Metraux, Byrne, Stino and Bainbridge2013), but differences in life trajectories and health status present challenges to defining the parameters of an older homeless population (McDonald, Dergal, & Cleghorn, Reference McDonald, Dergal and Cleghorn2007). While 65 – the dominant age of retirement – is the most widely accepted marker of old age, it is deficient where homelessness is concerned. Older people who are homeless tend to exhibit mental and physical health characteristics that are more consistent with non-homeless people who are approximately 10 years older than they are (Cohen, Reference Cohen1999; Gonyea et al., Reference Gonyea, Mills-Dick and Bachman2010; Hibbs et al., Reference Hibbs, Benner, Klugman, Spencer, Macchia and Mellinger1994; Hwang et al., Reference Hwang, Lebow, Bierer, O’Connell, Orav and Brennan1998; Morrison, Reference Morrison2009; Ploeg, Hayward, Woodward, & Johnston, Reference Ploeg, Hayward, Woodward and Johnston2008). A study of older homeless people in Toronto found that those over 50 subjectively considered themselves “old” (McDonald et al., Reference McDonald, Dergal and Cleghorn2004), reflecting the general trend of considering homeless people over 50 as “older” (Cohen, Reference Cohen1999; Garibaldi, Conde-Martel, & O’Toole, Reference Garibaldi, Conde-Martel and O’Toole2005; Gonyea et al., Reference Gonyea, Mills-Dick and Bachman2010; McDonald et al., Reference McDonald, Dergal and Cleghorn2007; Ploeg et al., Reference Ploeg, Hayward, Woodward and Johnston2008; Shinn et al., Reference Shinn, Gottlieb, Wett, Bahl, Cohen and Baron Ellis2007). People who live on the streets also have higher rates of early mortality than the general population (Cohen, Reference Cohen1999; Hibbs et al., Reference Hibbs, Benner, Klugman, Spencer, Macchia and Mellinger1994; Hwang et al., Reference Hwang, Lebow, Bierer, O’Connell, Orav and Brennan1998; Morrison, Reference Morrison2009), with the average age of death for a homeless person in Canada cited as 39 years (Trypuc & Robinson, Reference Trypuc and Robinson2009). Given these findings, we suggest that 50 is an appropriate age threshold for research and programming for older people.
Statistics and Estimated Prevalence of Homelessness among Older People
Prevalence statistics normally provide the foundation to develop a research and policy agenda, but the incidence of homelessness – specifically homelessness among older people – is difficult to attain (Link et al., Reference Link, Susser, Stueve, Phelan, Moore and Struening1994; Mott, Moore, & Rothwell, Reference Mott, Moore and Rothwell2012). Homelessness is a global issue, and an estimated 100 million people are homeless worldwide (United Nations Organization, 2005). Yet Canada does not gather comprehensive data on homelessness (Trypuc & Robinson, Reference Trypuc and Robinson2009). The data collected by Statistics Canada is based on the number of people living in shelters (Statistics Canada, 2012) – and this makes it difficult to estimate the number of people who are unsheltered, provisionally sheltered, or who are at risk of homelessness. Different understandings of the living circumstances that constitute homelessness, the realities of transitions between places (i.e., lack of fixed address), and varied counting methods further complicate this issue. Some studies use a point prevalence count to estimate the number of homeless people at a specific time. To arrive at their results, researchers conduct a survey of shelter users and count the number of individuals in “homeless hotspots” on one night. In other studies, researchers use a period prevalence count to estimate the homeless population over a given duration (Hulchanski, Reference Hulchanski2000). The use of administrative data, such as the recorded number of people using a shelter over a given length of time, is an example of the period prevalence method. Footnote 8
Despite limitations, prevalence estimates suggest that Canada’s homeless population ranges from 150,000 to 300,000, with particular groups, such as Aboriginal populations, greatly over-represented (Goering et al., Reference Goering, Veldhuizen, Watson, Adair, Kopp and Latimer2014; Laird, Reference Laird2007; Patrick, Reference Patrick2014). Footnote 9 Approximately 20,170 individuals (0.05% to 0.06% of the population) lived in shelters between 2001 and 2011 (Statistics Canada, 2012), and in 2008 there were 1,128 shelters in Canada (Echenberg & Jensen, Reference Echenberg and Jensen2008). Data suggests that Toronto has the largest number of homeless people in the country, but cities in Alberta also have significant rates of homelessness (Gaetz et al., Reference Gaetz, Donaldson, Richter and Gulliver2013). Information collected by point prevalence methods indicates that there were approximately 5,086 homeless people on a single night in Toronto in 2008 (representing 0.19% of the city population); 1,602 on one night in Vancouver in 2012 (representing 0.27% of the city population); and 3,190 on one night in Calgary in 2012 (representing 0.29% of the city population) (Gaetz et al., Reference Gaetz, Donaldson, Richter and Gulliver2013). Although comparable data for Montreal are not currently available, Montreal is expected to conduct its first point in time count in 2015.
The age structure of the homeless population is even more difficult to assess, but data suggest that approximately 6 per cent of the visible homeless population in Canada are over age 65 (Stuart & Arboleda-Flórez, Reference Stuart and Arboleda-Flórez2000) and 9 per cent are over 55 (Social Planning and Research Council of BC, 2005). Data suggest that Toronto has a large proportion of older homeless people, with the 2013 point in time count reporting that 29 per cent of the homeless population in Toronto is age 51 and older (City of Toronto, 2013). In terms of shelter use, people older than 55 are considered to represent 14 per cent to 28 per cent of shelter users (Stergiopoulos & Herrmann, Reference Stergiopoulos and Herrmann2003). Although older people are a minority in the homeless population – perhaps due to their higher rates of early mortality – they are known to spend more time in shelters than their younger homeless counterparts (Serge & Gnaedinger, Reference Serge and Gnaedinger2003). With few viable housing alternatives for older people and over-crowding in acute hospitals, there is pressure on shelters to accept older and unwell patients who can no longer care for themselves, and to fill the gap in convalescent care (Serge & Gnaedinger, Reference Serge and Gnaedinger2003). Little is known, however, about patterns of shelter use among certain sub-groups of the older homeless population such as older Aboriginal people or older immigrants.
Older People’s Pathways into Homelessness
Identifying trajectories into homelessness provides insight into the needs and challenges of older people who are homeless. Research indicates that a gradual decline and/or trigger event(s) (Gonyea et al., Reference Gonyea, Mills-Dick and Bachman2010; Shinn et al., Reference Shinn, Gottlieb, Wett, Bahl, Cohen and Baron Ellis2007), as well as various individual and structural factors, contribute to homelessness among older people. The following discussion attempts to untangle the complex interconnections between structural conditions, cumulative circumstances, risk factors, and trigger events. Our goal is to give readers a better sense of the diverse conditions that operate and are associated with homelessness among older people.
Macro-level forces that disadvantage particular groups of older people are considered to increase the risk of homelessness. Although Canada funds supports and services for older people, Footnote 10 structural issues associated with homelessness in later life include (a) inadequate affordable housing; (b) fewer available jobs; (c) poverty; and (d) policies that limit the access of particular populations to health, disability, and pension benefits (Gaetz et al., Reference Gaetz, Donaldson, Richter and Gulliver2013; Lee et al., Reference Lee, Tyler and Wright2010; Tully & Jacobson, Reference Tully and Jacobson1994). Since the 1990s, the rising cost of housing has also resulted in more Canadians living below the low-income cut-off in urban and rural areas (Skaburskis, Reference Skaburskis2004). Although the impacts of the recent global recession and economic crisis have yet to be adequately researched, literature suggests that poverty among older people is a growing concern. Asset poverty research, for example, shows that 28 per cent of people age 66 and older do not have sufficient financial assets to survive at the low-income threshold for three months (Rothwell & Haveman, Reference Rothwell and Haveman2013).
In this context, individuals may experience a gradual decline into homelessness through precarious employment, diminishing finances leading to poverty, poor mental and physical health, decreasing social connections (Morris, Judd, & Kavanagh, Reference Morris, Judd and Kavanagh2005; Shinn et al., Reference Shinn, Gottlieb, Wett, Bahl, Cohen and Baron Ellis2007), psychiatric conditions (Barak & Cohen, Reference Barak and Cohen2003), or alcoholism (Crane, Reference Crane1999; Dietz, Reference Dietz2009). Lower levels of education (Rank & Williams, Reference Rank and Williams2010), precarious work history, and incarceration (Kushel, Evans, Perry, Robertson, & Moss, Reference Kushel, Evans, Perry, Robertson and Moss2003; Metraux & Culhane, Reference Metraux and Culhane2006) are also associated with a greater risk of homelessness. Those who experience higher levels of victimization and poverty when younger are also more likely to be homeless later in life (Browne & Bassuk, Reference Browne and Bassuk1997; Koegel, Melamid, & Burnam, Reference Koegel, Melamid and Burnam1995; North, Smith, & Spitznagel, Reference North, Smith and Spitznagel1994; Stein, Leslie, & Nyamathi, Reference Stein, Leslie and Nyamathi2002; Toro, Reference Toro2007), as are those who experience traumatic life changes, especially if the individual has limited social and family networks (Morris et al., Reference Morris, Judd and Kavanagh2005). Aboriginal people and lesbian, gay, bisexual, transgendered, and queer people (LGBTQ) are also over-represented in the homeless population (Addis, Davies, Greene, MacBride-Stewart, & Shepherd, Reference Addis, Davies, Greene, MacBride-Stewart and Shepherd2009; Patrick, Reference Patrick2014).
At the individual level, people who experience vulnerabilities may lack the personal, economic, or social resources to cope with emergency situations. In turn, events such as housing loss; death of a spouse, relative, or close friend who may have provided care; domestic violence; and/or family breakdown may trigger homelessness (Crane & Warnes, Reference Crane and Warnes2005; Gonyea et al., Reference Gonyea, Mills-Dick and Bachman2010). Such situations may be increasingly complex for older people. For example, a Toronto study found that 70 per cent of people over age 50 became homeless between the ages of 41 and 60 as a result of family breakdown, eviction, and/or a loss of employment (McDonald et al., Reference McDonald, Dergal and Cleghorn2004).
Older people typically experience one of two types of homelessness: they are either homeless throughout their lives and continue this pattern as they age (i.e., chronic homelessness), Footnote 11 or they become homeless for the first time in later life (i.e., late-life homelessness). The literature suggests that the second pathway is increasingly common. Research conducted with older homeless people in the United States, England, and Australia found that two thirds had not experienced homelessness earlier in life, while the other third had been homeless before (Crane et al., Reference Crane, Byrne, Fu, Lipmann, Mirabelli and Rosa-Bartelink2005). Footnote 12 In addition to representing a new sub-population of homeless people, older people typically experience longer periods of homelessness than younger people because they are less likely to reintegrate into the workforce (Caton et al., Reference Caton, Dominguez, Schanzer, Hasin, Shrout and Felix2005). Concerns about the rising numbers of people who are homeless for the first time in late life in Canada (McDonald et al., Reference McDonald, Dergal and Cleghorn2007), and internationally (Caton et al., Reference Caton, Dominguez, Schanzer, Hasin, Shrout and Felix2005; Crane et al., Reference Crane, Byrne, Fu, Lipmann, Mirabelli and Rosa-Bartelink2005), underscore the importance of identifying and responding to populations who are already homeless, as well as those who are at risk.
Gaps in Knowledge: Sub-populations in the Older Homeless Population
Here we summarize, and extrapolate beyond our literature review, to draw attention to under-researched trends with regards to gender, over-represented sub-populations, geography, health, substance use, and the unique needs of older people who are homeless. Some of the findings presented here are constrained by a limited (or in some cases non-existent) body of research. Greater attention to the impacts of age in combination with life course inequalities produced by “race”, class, gender, ability, health status, and geography is urgently needed to set an inclusive agenda, and to design and implement programs that reduce and ultimately eliminate homelessness among diverse groups of older people in Canada.
Gender
Research suggests that men outnumber women about four to one among all homeless people (Cohen, Reference Cohen1999), but the gender gap is thought to be narrower among older people (McDonald et al., Reference McDonald, Dergal and Cleghorn2007). The gender difference in estimated prevalence reflects that men are more likely to use shelter services, and are thus more visible in the homeless population (Rich & Clark, Reference Rich and Clark2005). By contrast, the prevalence of older homeless women is likely under-reported, particularly among those leaving abusive situations (Kosor & Kendal-Wilson, Reference Kosor and Kendal-Wilson2002). With homeless women less visible, it is difficult to provide precise information on gender differences. However, research does point to divergent paths of men and women. On the one hand, men are more likely to be homeless or precariously housed throughout their lives (Hecht & Coyle, Reference Hecht and Coyle2001), with their homelessness often connected to loss of employment (McDonald et al., Reference McDonald, Dergal and Cleghorn2004), mental health problems, or addiction (Peressini, Reference Peressini2007). Older homeless women, on the other hand, often experience poverty as a result of family circumstances, the structure of the pension system (Rahder, Reference Rahder2006), as well as trigger events such as family breakdown (McDonald et al., Reference McDonald, Dergal and Cleghorn2004), eviction (Hecht & Coyle, Reference Hecht and Coyle2001), and abuse (Kosor & Kendal-Wilson, Reference Kosor and Kendal-Wilson2002). Footnote 13 Other studies suggest that the number of women over age 55 forced to leave their homes because of physical and sexual violence is increasing (Grossman & Lundy, Reference Grossman and Lundy2003), and that when homeless, women experience much higher risks of abuse and victimization than men (Dietz & Wright, Reference Dietz and Wright2005; Wenzel, Leake, & Gelberg, Reference Wenzel, Leake and Gelberg2001). Older women’s housing needs are also complicated by a national shortage of shelters and social housing units for abused women (Rahder, Reference Rahder2006).
Over-represented Populations
Research conducted on earlier parts of the life course confirms that a disproportionate number of Aboriginal people; lesbian, gay, bi-sexual, transgendered, and queer (LGBTQ) people; and immigrants are homeless or at risk of homelessness in Canada (Fiedler, Schuurman & Hyndman, Reference Fiedler, Schuurman and Hyndman2006; Gaetz, Reference Gaetz2006; Goering et al., Reference Goering, Veldhuizen, Watson, Adair, Kopp and Latimer2014; Patrick, Reference Patrick2014). Footnote 14 Although older people are rarely considered in this literature on over-represented sub-populations, we expect that members of these groups will face unique challenges as they age (Brotman, Ferrer, Sussman, Ryan, & Richard, Reference Brotman, Ferrer, Sussman, Ryan, Richard, Orel and Fruhauf2014), and may experience difficulties exiting homelessness in later life. Footnote 15 Aboriginal people are noted to be over-represented by a factor of 10 in the Canadian homeless population (Hwang, Reference Hwang2001) – a trend that is associated with historical and current practices of colonialism, systemic discrimination and exclusion, and lack of affordable housing (Patrick, Reference Patrick2014). The one available study on older Aboriginal people who are homeless (an unpublished master’s thesis; see Lange, Reference Lange2010) highlights the unique challenges experienced by older Aboriginal people, including risks of homelessness when they move to cities for medical care. In such cases, older Aboriginal people tend to fall through the cracks of existing service structures because it is unclear whether governments or band organizations are responsible for their care (Lange, Reference Lange2010). Similarly, although research on earlier periods of the life course finds that trajectories to homelessness are related to family breakdown among LGBTQ youth (Abramovich, Reference Abramovich, Gaetz, O’Grady, Buccieri, Karabanow and Marsolais2013; Corliss, Goodenow, Nichols, & Austin, Reference Corliss, Goodenow, Nichols and Austin2011), we were unable to find literature on homelessness among older LGBTQ people. Footnote 16
Contrasting with the dearth of research on older Aboriginal or LGBTQ people, a very small literature does exist on the intersections of age, homelessness, and immigration. The over-representation of immigrants in Canada’s homeless population is attributed to their increased likelihood of poverty and housing insecurity, as well as their lower pension contributions (Springer, Webber, & Lum, Reference Springer, Webber and Lum2011). These findings have been extended to consider older people, with a Toronto study finding that 55 per cent of recently homeless older people were born outside of Canada, compared with 29 per cent of individuals who were chronically homeless (McDonald et al., Reference McDonald, Dergal and Cleghorn2007).
The lack of available research on the aging of over-represented sub-populations reflects the widespread invisibility of marginalized groups and a failure to acknowledge inequalities that can result in homelessness. To better understand and address the needs of over-represented groups, responses to homelessness must account for the relationship between systemic factors such as discrimination, racism, and colonization; structural barriers and problems of access; and experiences of marginalization across the life course and in later life.
Geographic Location
The Canadian literature on homelessness among older people primarily notes geographic trends and urban-rural differences (e.g., McDonald et al., Reference McDonald, Dergal and Cleghorn2004). The majority of homeless people live in large cities (Statistics Canada, 2001) where services such as shelters are located, Footnote 17 with shelter use reportedly higher in Quebec, Alberta, Ontario, British Columbia, and Manitoba than in other provinces and territories (Statistics Canada, 2001). At the same time, context and place-based issues are considered to impact experiences of homelessness among older people (Abbott & Sapsford, Reference Abbott and Sapsford2005). While homelessness is typically considered an urban problem, those living outside urban or resource-based areas may draw on different strategies to meet their needs, or face additional challenges accessing services and support (North et al., Reference North, Smith and Spitznagel1994). With homelessness experienced differently among communities and across the country, geographic and place-based issues that may surface for diverse groups of older people (e.g., those new to homelessness in late life) require further investigation.
Health Issues
Health problems experienced across the life course are both a risk factor for, and an outcome of, homelessness. People with mental health or addiction problems are more likely to become homeless (Bhui, Shanahan, & Harding, Reference Bhui, Shanahan and Harding2006). At the same time, people who lack stable housing face threats to their mental and physical health (Bhui et al., Reference Bhui, Shanahan and Harding2006; Power & Hunter, Reference Power and Hunter2001; Schanzer, Dominguez, Shrout, & Caton, Reference Schanzer, Dominguez, Shrout and Caton2007), and older homeless people are considered to face greater physical and mental health disadvantages than younger groups (Dennis, McCallion, & Ferretti, Reference Dennis, McCallion and Ferretti2012; Gonyea et al., Reference Gonyea, Mills-Dick and Bachman2010; Kellogg & Horn, Reference Kellogg and Horn2012; Lipmann, Reference Lipmann2009; Martins, Reference Martins2008; Ploeg et al., Reference Ploeg, Hayward, Woodward and Johnston2008; Quine, Kendig, Russell, & Touchard, Reference Quine, Kendig, Russell and Touchard2004). Garibaldi et al. (Reference Garibaldi, Conde-Martel and O’Toole2005) found that those over age 50 were 3.6 times more likely than younger homeless people to suffer from a chronic medical problem, while Kim, Ford, Howard, and Bradford (Reference Kim, Ford, Howard and Bradford2010) noted that the likelihood of having mental health problems doubles for homeless people over the age of 42.
Specific health issues have also been documented for older people who are homeless. McDonald et al. (Reference McDonald, Dergal and Cleghorn2004) found that the most frequently reported ailments among older homeless people in Toronto were vision, arthritis, dental problems, and back problems, while Kellogg & Horn (Reference Kellogg and Horn2012) found that hypertension, cardiac disease, lung disease, diabetes, and arthritis are prevalent among older homeless people in the United States. There are also gender differences in health issues, with women reporting greater difficulties with arthritis and bladder control, and men more likely to have back and skin problems (McDonald et al., Reference McDonald, Dergal and Cleghorn2004). Older homeless men – particularly those who lose their jobs between ages 60 and 65 – are also at higher risk of suicide (Greater Vancouver Shelter Strategy, 2013). In some circumstances, health conditions are already present when one becomes homeless; in other cases, they manifest or worsen during periods of homelessness (Horn, Reference Horn2008; Hwang et al., Reference Hwang, Lebow, Bierer, O’Connell, Orav and Brennan1998).
Homelessness is also considered to have long-term effects on health and aging (Brown, Kiely, Bharel, & Mitchell, Reference Brown, Kiely, Bharel and Mitchell2012; Waldbrook, Reference Waldbrook2013). Those who are homeless are more likely than are older people in the general population to suffer from geriatric syndromes such as functional impairment, frailty, depression, visual impairment, and urinary incontinence (Brown et al., Reference Brown, Kiely, Bharel and Mitchell2012). Where perceived health is concerned, formerly homeless older women, for example, express that the physical state of homelessness, lifelong socioeconomic disadvantage, trauma and stress, substance use, and neglect of health needs all contribute to poorer health in later life (Waldbrook, Reference Waldbrook2013).
Substance Use
Drug and alcohol use is often associated with homelessness (Blazer & Wu, Reference Blazer and Wu2009; Dietz, Reference Dietz2009; Khandor & Mason, Reference Khandor and Mason2008; Kuhn & Culhane, Reference Kuhn and Culhane1998), but the literature on substance use among older people is inconclusive. Some research finds that substance use patterns differ between age cohorts and decrease with age (Blazer & Wu, Reference Blazer and Wu2009; Cohen, Reference Cohen1999). Other studies note that where younger and older homeless people are equally likely to report alcohol abuse (Dennis et al., Reference Dennis, McCallion and Ferretti2012; Dietz, Reference Dietz2009; Hecht & Coyle, Reference Hecht and Coyle2001), older people are less likely to report drug use (Hecht & Coyle, Reference Hecht and Coyle2001). Conversely, some studies suggest that drug use among older people has been increasing and is expected to continue on an upward trajectory (Beynon, Reference Beynon2009; Proehl, Reference Proehl2007). In the United States, Garibaldi et al. (Reference Garibaldi, Conde-Martel and O’Toole2005) find that those over age 50 are 2.4 times more likely to be dependent on heroin than those under 50. Higher rates of drug use than previous generations are attributed to a cohort effect: people tend to maintain drug habits throughout their lives, and greater co-morbidity as a result of prolonged drug or alcohol use is expected (Beynon, Reference Beynon2009). Despite debates, it is clear that the paucity of relevant information and services on substance use among older homeless people leaves an already vulnerable population at greater risk (Blazer & Wu, Reference Blazer and Wu2009; Proehl, Reference Proehl2007).
Unique Needs of Older Homeless People
Older and younger homeless people have shared needs that include housing, income, food, and health care. However, research finds that older people who are homeless also have unique needs with regards to safety and access to health and social services. Footnote 18 Older homeless people are more likely than their younger counterparts to have mental and physical health concerns, and may require access to specialized medical care beyond what is available in shelters (Power & Hunter, Reference Power and Hunter2001). Living without a home can be especially challenging in later life, making older people’s housing needs particularly urgent (Abbott & Sapsford, Reference Abbott and Sapsford2005). Interviews with health care providers illustrate that mental health conditions can create challenges where continued engagement with older homeless people is concerned, and that memory problems for example, may cause some older people to forget appointments (Cohen, Onserud, & Monaco, Reference Cohen, Onserud and Monaco1992; Horn, Reference Horn2008; Proehl, Reference Proehl2007). Older homeless people’s reports of discriminatory treatment and stigmatization in health care settings also demonstrate the need for medical staff to become more sensitive in their responses (Lipmann, Reference Lipmann2009; Martins, Reference Martins2008; Quine et al., Reference Quine, Kendig, Russell and Touchard2004).
Difficulties navigating government services can also be a barrier to accessing supports and services. Many older homeless people do not receive the full amount of government assistance for which they qualify (Ploeg et al., Reference Ploeg, Hayward, Woodward and Johnston2008), and language has been identified as a significant barrier to accessing housing and support services among older people who are homeless (McDonald et al., Reference McDonald, Dergal and Cleghorn2007). There are also concerns regarding the appropriateness of services available for older homeless people. One Canadian project identified a gap in services for homeless people aged 50 to 65, with clients reporting frustration because neither the services offered, nor the programs created for the general homeless population, suited their needs (McDonald, Donahue, Janes, & Cleghorn, Reference McDonald, Donahue, Janes and Cleghorn2006).
Finally, older people who are homeless have unique needs regarding safety. They encounter violence on the streets and in shelters (Cohen et al., Reference Cohen, Onserud and Monaco1992; Lee & Schreck, Reference Lee and Schreck2005; North et al., Reference North, Smith and Spitznagel1994) and are thought to face higher threats to safety than their younger counterparts because poorer health may mean they are seen as easy targets (Dietz & Wright, Reference Dietz and Wright2005). Risks of victimization are especially high for older women and transgendered people, but older homeless men are also at high risk of physical abuse (Cohen, Reference Cohen1999; Dietz & Wright, Reference Dietz and Wright2005; Gonyea et al., Reference Gonyea, Mills-Dick and Bachman2010; Grossman & Lundy, Reference Grossman and Lundy2003; Lee & Schreck, Reference Lee and Schreck2005; North et al., Reference North, Smith and Spitznagel1994; Tully & Jacobson, Reference Tully and Jacobson1994).
Building a Research Agenda to Inform Policy and Practice
A strong knowledge base is necessary to address the needs of diverse groups of older people, develop provincial and national strategies to end homelessness, and design community services for this group. At present, the evidence base is limited by significant gaps. Table 1 outlines our suggested research agenda.
Research agenda for improved responses to homelessness among older people
First, better estimates of the prevalence of older homelessness in Canada, and an identification of profiles of risk, are needed. Understanding how people become homeless for the first time, and which sub-populations are more likely to become or remain homeless in later life is especially urgent. We suggest that researchers make better use of data gathered through the Homeless Individuals and Families Information System (HIFIS) – an administrative tool for collecting detailed information on shelter users – to identify user profiles, patterns of shelter use, and people at extreme risk of long-term homelessness. Understanding homelessness can be enhanced if organizations link administrative data within and between cities and geographic areas. Footnote 19 Carrying out this research on people over age 50 (instead of the standard age, 65), on those who are 40 to 49 and approaching old age (see Walsh, Hewson, Dooley, & Pauls, Reference Walsh, Hewson, Dooley and Pauls2013), and accounting for over-represented sub-populations can provide much-needed projection trends. Research suggests that many Canadians are financially vulnerable or asset poor (Brandolini, Magri, & Smeeding, Reference Brandolini, Magri and Smeeding2010; Rothwell & Haveman, Reference Rothwell and Haveman2013), and that inadequate or limited access to pensions may contribute to poverty in later life, particularly for disadvantaged groups such as immigrants (McDonald et al., Reference McDonald, Dergal and Cleghorn2007) and women (Wakabayashi & Donato, Reference Wakabayashi and Donato2006). Footnote 20 Over-represented sub-populations such as Aboriginal people and persons from the LGBTQ community also experience systemic discrimination and exclusion across the life course that likely place them at high risk of continued poverty and/or homelessness in later life (see Addis et al., Reference Addis, Davies, Greene, MacBride-Stewart and Shepherd2009; Patrick, Reference Patrick2014).
Second, building appropriate strategies for diverse groups of older people who are homeless – or at risk of becoming homeless – requires a better understanding of intersecting needs for affordable housing and care in mid- to later life. Research assessing the supply of market-based and social housing in mid- to late life, both with and without available care, is needed. We suggest a two-pronged approach that focuses on re-housing people over age 50 who are homeless, and ensuring support for those who are precariously housed as they near “old age” and are vulnerable to situations that lead to homelessness (e.g., poverty, job loss, family conflict). Here Canada may look to international studies that consider how complex structural factors, including economic conditions and available housing, shape risks of homelessness in late life (see Byrne, Munley, Fargo, Montgomery, & Culhane [Reference Byrne, Munley, Fargo, Montgomery and Culhane2012] and Culhane et al. [Reference Culhane, Metraux, Byrne, Stino and Bainbridge2013] in the United States; and Crane & Warnes [Reference Crane and Warnes2001] in the United Kingdom). When researchers develop guidelines for housing strategies, they must be mindful of the following four areas: (a) rehousing people with limited financial resources as they transition from shelters or hospitals; (b) ensuring access to safe and affordable community housing, with links to health and social care supports; (c) ensuring an income base, and housing for at risk groups, including groups that are currently over-represented among homeless populations (e.g., Aboriginal people, older immigrants, LGBTQ people); and (d) ensuring access to appropriate services and long-term care when necessary. While implementing such changes relies heavily on political will, the recent literature underlines the importance of long-term, affordable, and secure housing for older people who are homeless or at risk of becoming so.
Third, and closely related, the research, policy, and practice agenda must draw closer attention to identifying intersecting inequalities experienced by over-represented sub-populations. Given histories and current practices of colonialism, homophobia, racism, and sexism, older Aboriginal people, LGBTQ people, immigrants, and men/women may face particular challenges with regards to aging and homelessness. Therefore, research addressing multiple intersecting categories, including gender, class, “race”, ethnicity, health status, and age, in specific geographic contexts, is necessary (see Brotman et al., Reference Brotman, Ferrer, Sussman, Ryan, Richard, Orel and Fruhauf2014; Klodawsky, Reference Klodawsky2009). There are serious gaps in knowledge regarding the impact of sexual orientation on older people’s risks of homelessness, and on the experiences of Aboriginal people who are aging in precarious situations. For members of these over-represented sub-populations, the structural and systemic challenges associated with income security, access to health care, and safe affordable housing can only be expected to continue – if not worsen – as they age. In this regard, a life course perspective that is attuned to the impacts of intersecting and cumulative inequalities could provide a basis for policy development as well as services and support.
Finally, addressing the needs of older homeless people will require greater knowledge on existing service structures and the challenges of implementing current strategies such as housing first. The literature clearly outlines that older people who are homeless have complex, intersecting needs for health and social care, income support, and housing. Access to services, however, is limited by rigid institutional boundaries, few options for affordable housing and care, Footnote 21 and age-based thresholds for programs that exclude those aged 50 to 64. We suggest that researchers and decision-makers assess current and planned resource allocation, and the ways in which these resources could be revised to permit seamless movement across policy and service structures, in order to develop more-comprehensive support.
Conclusion
Homelessness among diverse groups of older people is a significant form of social marginalization that should be a pressing concern for decision-makers, gerontologists, and housing advocates alike. With homelessness among older people expected to rise – and more people experiencing homelessness for the first time in later life or remaining in situations of homelessness as they age – a lack of research knowledge leaves policy-makers and practitioners with few directives from which to address the needs of older homeless people. Our review was intended to set the stage for an inclusive Canadian research, policy, and practice agenda that targets homelessness among older people. We have reviewed the state of knowledge on aging and homelessness; discussed the available statistics and the estimated prevalence of homelessness in Canada; articulated pathways into homelessness in later life; and drawn attention to within group variations and over-represented sub-populations in Canada. We have identified several gaps, including (a) accurate estimates of the older homeless population (including the diversity that exists within this population); (b) the availability of affordable housing; (c) the effects of intersecting inequalities on homelessness in later life; and (d) appropriate resources and supports. On the basis of these shortcomings, we advocate for the development of an agenda that will address gaps in knowledge through focused research, and stimulate a targeted response to homelessness that integrates the needs of diverse groups of older people in Canada.
