The goal of this research project is to investigate the role of public hospitals in our current health care system. It is known that public hospitals provide a disproportionate share of medical care to the poor and uninsured, but the direct effect of public hospitals on access to medical care for the uninsured in their community is not well studied. Public hospitals may replace care that would have been provided by private institutions or increase the overall care provided. Previous work in this field has measured uncompensated care provided by hospitals. This study will instead directly. study the impact of public hospitals upon the behavior of individuals living in the surrounding communities by using a population-based survey containing outpatient and inpatient utilization information. An economic model of the effect of public hospitals upon individual consumer behavior will be developed to address the following research questions: (1) are individuals less likely to buy private insurance if a public hospital is available (2) are the uninsured and Medicaid enrollees more likely to use physician and hospital services if a public hospital is available? Individual-level data from the 1987 National Medical Expenditure Survey (NMES) will be linked to county-level characteristics such as the presence of a public hospital, hospital beds per capita and practicing physicians per capita as well as state-level characteristics such as Medicaid generosity. The empirical analysis will begin with a multinomial logistic regression to predict insurance coverage. This estimation will yield predicted values for insurance coverage which will then be included as regressors in generalized Tobit estimations of physician and hospital utilization. This methodology will allow a more precise analysis of public hospitals than has been previously possible.
