Researchers’ explanation of the cause of homelessness, whether explicit or implicit, strongly affects the approach taken to studying the costs of homelessness and how resources are targeted, especially with respect to preventing homelessness. The two main explanations identified in the research literature on the costs of homelessness are failures in the housing market, seen primarily in the economic literature (see Allgood & Warren 2003; Bohanon 1991), and mental or physical health issues, prevalent in the psychological literature (see Levine, Perkins, & Perkins 2005; Rosenheck & Seibyl 1998; Shinn & Tsemberis 1998). The economics approach tends to neglect psychological, health, or social problems that could result in people becoming homeless even with a sufficient supply of affordably priced housing. While the psychological literature gives more attention to the social problems that result in people becoming homeless, it often fails to provide solid cost-benefit analyses that would allow for a better comparison of policy alternatives.
Most past studies on the costs of homelessness in the have either a) studied characteristics of those who become homeless or b) studied the traits that mediate the duration of homelessness. The first approach conducts demographic analyses with the intention of improving the efficiency of existing resource use by identifying who is most likely to become homeless and targeting preventative resources at these key at-risk groups. Early (2002) took this approach by studying demographic characteristics, constructing an economic model for homelessness, and using the model to estimate what percentage of people currently living in subsidized housing would become homeless if the subsidy were removed. On the other hand, some researchers study determinants of the duration of homelessness, assuming that those who remain homeless for longer spells use larger amounts of services (Allgood & Warren 2003). These studies examine the characteristics of people who successfully exit homelessness against the characteristics of those who are “chronically homeless.” While a few cost studies of service provision have been completed (see Clasen 2006; Lewin Group 2004), there is a general lack of studies that attempt to comprehensively estimate the average daily cost of an individual being homeless. Without a comprehensive survey of costs, it is difficult to accurately estimate evaluate the cost-effectiveness of policy alternatives to reduce homelessness.
However, the lack of comprehensive empirical studies on the costs of homelessness may be due to two major challenges in conducting this type of research. First, due to the homeless population’s high degree of mobility and unreliable access to communication devices, any attempt to track a specific group of homeless people for an extended period of time runs into severe data collection problems and faces the higher costs associated with long-term studies. Second, in estimating usage of services by contacting service providers, one runs into the problem of repeat clients as some agencies track encounters but not the unique number of individuals that use their services, another variation of the tracking problem. Given these challenges, it has been very difficult for researchers to assess whether costs of homelessness are being driven by a smaller group that use large quantities of services or of larger numbers of users that use fewer quantities of resources.
Thus, from an economic cost perspective, two conclusions can be drawn about issues surrounding the effectiveness of the attempts to improve the efficiency of services provided to the homeless through better targeting. First, the studies that use this approach expose that a strict focus on homelessness as a housing shortage frequently fail to take into account the demographics of those who use the housing and those who are excluded from it. Second, the concept of using income alone to target who should receive housing is a double-edged sword: if income is used prior to demographic analyses as the primary criterion for entry, inefficiency is created by supplying housing to lower risk low-income earners who take spots that could be used to house those at greater risk of becoming homeless without subsidized housing. However, using income to prioritize who receives housing within demographically at-risk groups can increase efficiency by replacing the typical first-come, first-serve or hardship approaches.