The primary policy option that has been suggested to states over the y
ears for shifting care for Medicaid beneficiaries away from hospital s
ites and toward office sites has been to raise Medicaid fees to physic
ians on the assumption that this increases the geographic availability
of office-based care, which then naturally attracts Medicaid clients
away from hospital sites. This study uses Medicaid claims data from th
e state of Alabama to assess the role of geographic provider availabil
ity relative to other factors in families' decisions to select a hospi
tal or an office for an illness visit to a primary care or to a specia
list physician. The authors examined the last ambulatory visit for an
illness made by continuously enrolled children under age 8 in the firs
t half of 1991. The authors found that both higher Medicaid office pra
ctice density and the presence of larger Medicaid office practices wer
e associated with choice of an office for ambulatory care, whereas the
presence of larger Medicaid hospital practices were associated with c
hoice of a hospital for care. Controlling for provider availability, h
ospital users were less likely to have sought previous care for illnes
s during the year, and were more likely to be rural residents, to have
traveled away from their home counties for care, and to be eligible f
or Medicaid through Aid to Families with Dependent Children, rather th
an through the program's income expansions. We conclude that increased
office-based provider availability must be coupled with improved acce
ss for new, remote, and very low income families if Medicaid clients a
re to be expected to voluntarily shift their choice of site for ambula
tory care.