In some respect, all current major strategies to ameliorate homelessness aim to house people who are now on the streets. Housing first strategies simply put homeless people in housing without major restrictions; supportive and wraparound housing strategies place people in housing with integrated services; services first strategies aim to ameliorate immediate problems and personal deficiencies that proponents believe must be solved before someone can be stable in housing. Each of these strategies has its strong points and may be the right solution for some situations, but all may seem expensive to the government and the public. Burt (2003) reminds us that although housing and serving the homeless is expensive, homelessness is expensive regardless – and the solutions we choose greatly affect the quality of life of those experiencing homelessness. Research overwhelmingly supports housing first and supportive housing over services first approaches, so these options will be explored in more depth. It is also important to realize the diversity of approaches within housing strategies. Some emphasize more centralized plans while others aim to decentralize housing for people experiencing homelessness. As one consequence of approach, Lee, Farrell, and Link (2004) demonstrate that exposure to homelessness leads to public motivation to reduce homelessness and shift policy accordingly. Mixed housing and decentralized strategies emphasizing broader exposure may reduce existing outgroup stereotypes that stigmatize the homeless, increase willingness to support policies to alleviate homelessness, and shift perceived blame from individual to structural causes for those exposed to people experiencing homelessness (Lee, Farrell, & Link, 2004). Centralized strategies – although not exposing the broader community – may create stronger exposure to better achieve the same goals for a smaller population. It is important to weigh these possibilities for sustainable strategies since positive public attitudes about homelessness in Nashville translate into support for policies that alleviate homelessness.
With primary strategies focusing on housing, the issue of efficiency has boiled down to how to most effectively target low-income or subsidized housing to those who are most likely to become homeless without it. Thus, there have been a significant number of studies and analyses done based on the demographic data available about the homeless population in order to determine what the primary determinants of homelessness are in order to create the most effective entry requirements for low-income housing (Allgood & Warren, 2003; Bohanon, 1991; Early, 2002).
Early (2002) created a housing model that incorporated both income-based targeting and household demographic characteristics. The model indicates that the upper income limits currently in place needed to be lowered in order to increase the effectiveness of low-income housing, as a significant number of families living in subsidized housing would not become homeless if their subsidies were removed. Early (2002) argues that this is primarily due to low-income housing officials basing access on policies of first-come, first-serve or hardship and recommends using absolute income as an appropriate measure for differentiating among low-income households to increase effectiveness in targeting (and thus reducing the cost incurred by subsidizing housing for those who have the means to live elsewhere).
Yet income alone overlooks other factors that may put an individual at greater risk for homelessness, such as mental health problems. Allgood & Warren (2003) also found that eligibility criteria for subsidized and public housing favors drug-free, single women with young children, a cluster which is less likely to experience extended homelessness, while single men and those with behavioral problems actually form a larger demographic proportion of the homeless population. While this does not necessarily mean that low-income housing is entirely ineffective, those seeking to lower the costs and incidence of homelessness must recognize that single men and individuals with behavioral problems require different types of housing or services which must use different eligibility requirements to target these clusters than the current system of subsidized or public housing.
Housing First Strategies. Housing first strategies are based on the notion that people are homeless because they do not have a home, not because of personal deficiencies that transcend housing status. As stated previously, 7.4% of people had been homeless at some point in their life and 3.1% of people have been homeless within the previous 5 years (Shinn & Tsemberis, 1998). This means that more people are experiencing homelessness and that homelessness is far more temporary than previously suspected – indicating that homelessness is a temporary state, rather than a permanent trait (Shinn & Tsemberis, 1998). An experiment in New York City demonstrates that housing first programs (i.e., Pathways to Housing) get people off the streets faster, reduce health costs when compared to continuum of care programs and the control group, and generally cost less (Gulcur, Stefancic, Shinn, Tsemberis, and Fischer, 2003). Tsemberis and Eisenberg (2000) showed that residents with psychiatric disorders placed in housing first programs in NYC had a higher housing retention rate (88%) than residential treatment (47%) after 5 years. Overall, housing first has been the most successful option to reduce homelessness for people on the streets and in psychiatric hospitals in New York City(Gulcur, Stefancic, Shinn, Tsemberis, and Fischer, 2003; Stefancic & Tsemberis, 2007).
Burt (2003) suggests housing subsidies as the most straightforward and fast strategy for reducing homelessness, coupled with education and job training for those living in poverty. Shinn and Tsemberis (1998) tentatively claim that in many cases subsidized housing is sufficient to end homelessness for families. Additionally, Shinn (2007) showed that 80% of sheltered families in New York City that received housing subsidies remained stable – versus 18% stability for those who did not receive subsidies.
Even people that fit the stereotype of chronic homelessness can be well served by housing first. Clients with severe psychiatric disabilities can obtain and maintain housing when given the opportunity and necessary supports (Tsemberis & Eisenberg, 2000) – greatly reducing the resources necessary for serving this population. As far as abusers and addicts – the mainstay of the services first argument – Padgett, Gulcur, and Tsemberis (2006) found no significant difference in substance use between treatment first and housing first users. It is even possible that services first approaches are doing some harm, as services without housing can keep people experiencing homelessness from using mainstream health services instead of expensive emergency care (Shinn & Tsemberis, 1998).
Supportive Housing.Permanent supportive housing (PSH) is one of the 4 guiding principles in the Strategic Framework for Ending Chronic Homelessness in Nashville (The Mayor’s Task Force to End Chronic Homelessness, 2004). PSH is defined in the Strategic Framework as “permanent, affordable housing linked to health, mental health, employment and other support services” (p. 12). A client in PSH has direct access to all of their health needs, including transportation, economic support, community support, and healthy social networks. Solid social networks and organizational ties are associated with better health outcomes, higher self-esteem, social support, and social capital (Cattell, 2001). This also supports decentralized, integrated housing, wherein people who are formerly homeless are given access to established social networks. It is likely that any housing strategy should include some facilities for PSH.
Culhane, Metraux, and Hadley (2002) found that placing homeless individuals with severe mental illnesses in supportive housing experienced a marked reduction in shelter use, hospitalizations, length of stay per hospitalization, and time incarcerated. Costs of services used totaled an average of $40,451 per person per year prior to entering supportive housing and $16,281 less after placement in supportive housing compared with a control group not placed in such housing. With the annual cost of placement being $17,277, the net annual cost of providing housing to these individuals was only $995. In addition, upwards of 70% of the individuals placed in supportive housing retained their housing.
Martinez, Burt, and Burt (2006) documented a 56% drop in use of acute health care for people sheltered in supportive housing. Wright (2006) found a 75.2% decrease in physical and mental health care costs after the institution of wraparound services for people with co-occurring mental health and substance use disorders. Proscio (ND) notes that people experiencing homelessness placed in supportive housing reduced residential mental health service use to zero within one year. Comprehensive supportive housing in Nashville currently costs $11,500 per person and already shows similar reductions in service use and costs for clients.