This paper examines the effect of Medicaid recipiency on the level and site
of medical service use among homeless single men and women in New York Cit
y. Simple regressions of Medicaid on service use indicate that Medicaid sig
nificantly increases the likelihood that homeless individuals receive servi
ces, especially emergency and inpatient hospital services. In further analy
ses that control for health status, use instrumental variables procedures,
and examine differences between a similar population in 1985 and 1987, we f
ind that Medicaid neither increases nor diminishes access to emergency room
s. We find some evidence suggesting that Medicaid does improve access to no
nhospital medical care.