Income and employment among homeless people: the role of mental health, health and substance abuse

Background: The homeless population is among the poorest of the poor in the United States. Employment and government programs are potential sources of income, but many homeless people face potential barriers to work: many have serious mental and physical disabilities, and many more have alcohol and drug disorders. As a result, most homeless who work do so either for a few hours per day or only some days, which provides little income. General Assistance, a public program of last resort, also provides a low level of income support. More income might be gained through higher levels of work or participation in income support programs for people with disabilities.

Aims of the Study: To investigate the characteristics of homeless people that impede them in the labor market and in government program participation, paying particular attention to their mental and physical health, as well as their alcohol and drug problems.

Data: Data are from a survey of the homeless population in Alameda County, California, conducted from 1991 to 1993. Our sample is 471 homeless adults randomly selected from area shelters and meal providers, who were reinterviewed approximately 6 months later, regardless of domiciliary status. Mental health and substance use problems were assessed using the Diagnostic Interview Schedule, a structured, psychiatric interview that uses criteria based on the American Psychiatric Association's Diagnostic and Statistical Manual of Mental Disorders 3rd edn (revised). Employment between the first and second interview is categorized as none, low level (less than 6 hours a day or fewer than half the days between interviews) or higher level (at least 6 hours a day for at least half the days).

Analytical Procedures: The models of employment status and program participation are recursive in that homelessness at the first wave of the survey is treated as given. Thus we explore whether, given their initial homelessness, persons can gain or maintain access to income between the two interviews, conditional on the sample member's homelessness, health and disability at the first interview. Using maximum-likelihood methods, we estimate a generalized ordered logit model of whether the person works not at all, at a low level or at a higher level. Participation in disability programs and GA are estimated as probit models over the subsamples of potentially eligible participants.

Results: While a surprisingly large number of homeless people work, few homeless persons are able to generate significant earnings from employment alone. Physical health problems that limit work or daily activities, in particular, are barriers to employment. Drug and alcohol abuse and dependence are positively associated with lower work level but are negatively related to higher work level. Program participation is quite low relative to eligibility. Those with physical health problems are substantially more likely than those with mental health problems to be in the more generous disability programs. Substance use disorders are also a barrier to participation in disability programs.

Discussion: Mental health, health and disability play a large role in the employment and program participation of the homeless and persons at risk for homelessness. Physical disabilities are a barrier to employment, and those with substance use disorders are most likely to work at lower levels that provide less income. Rates of participation in government programs are low, and people with major mental disorders have especially low participation rate in disability programs. The low rates of participation, particularly in the disability programs, suggest the need for continued research in improving access to income support programs among eligible homeless populations. © 2000 John Wiley & Sons, Ltd.