Objective: This paper compares the prevalence of mental illnesses and alcohol and drug abuse and the residential histories of homeless individuals identified as having a mental illness and individuals who are not so identified. The cohort consisted of single persons applying for shelter over a 12-week period in Westchester County, a suburban county in New York State.
Methods: The sample of 201 persons (89 percent male, with a mean age of 37) represented 77 percent of consecutive single shelter applicants in a single-point-of-entry system over the study period. Information from an intake assessment was augmented by a semistructured interview to reconstruct subjects' residential history for the last five years, including periods of homelessness and time in institutions.
Results: Twenty-one percent of the cohort was classified as having mental illness. Seventy-two percent had a diagnosis of drug abuse or dependence, and 51 percent had alcohol abuse or dependence. For individuals with a mental illness, the use of cocaine and heroin was significantly lower, but alcohol use was somewhat greater, compared with other homeless persons. Persons with mental illness also experienced homelessness of some kind over a significantly longer period (a mean of seven years versus a mean of three years for other subjects), and they spent almost twice as many weeks during the previous five years literally homeless. Institutional time, most of which consisted of time in jail or prison, was equivalent for both groups.
Conclusions: Not only is residential instability heightened among shelter users with mental illness, but over time public institutions play a critical role in their accommodations. For some homeless persons with mental illness, the circuit of shelters, rehabilitation programs, jails, and prisons may function as a makeshift alternative to inpatient care or supportive housing and may reinforce the marginalization of this population.