Date of Award

Fall 2009

Degree Type


Degree Name

Doctor of Philosophy (PhD)

Graduate Group

Managerial Science and Applied Economics

First Advisor

Mark V Pauly


This study seeks to determine the effects of the local uninsurance rate on the quality of care delivered to Medicare beneficiaries. While a great deal of research exists on the negative consequences of uninsurance for the uninsured, few studies have attempted to address the spillover effects of this population to the health care system in a local market. Theory suggests that a high local uninsurance rate in a market has the potential to cause a decrease in the shared quality of care provided to Medicare beneficiaries. It also suggests however that a higher uninsurance rate may result in increases in access to care by Medicare beneficiaries as well as improvements in the unique quality provided to these individuals. The implications of a high uninsurance rate on outcomes are therefore ambiguous. These concepts are tested using data on hospitals and Medicare beneficiaries in 100 large MSAs. The effects on outcomes of care are examined by exploring the relationship between local uninsurance rates and mortality from a variety of specific conditions. The results are mixed. In smaller markets, the spillover effects of the uninsured are more likely to be negative. In larger markets however, the effects are more likely to be positive. The mechanisms behind these results are examined by looking at the effects on shared and unique quality. The effects on shared quality are tested by examining the availability of specialized hospital services in markets with varying levels of uninsurance. The results show that hospitals in markets with high levels of uninsurance provide fewer unprofitable services. The effects on unique quality and access are tested by examining the relationship between local uninsurance rates and utilization by Medicare beneficiaries. The results exhibit weak evidence that Medicare beneficiaries in areas with higher uninsurance rates use more care. The study concludes that while local uninsurance rates do not appear to have strong negative consequences for Medicare beneficiaries, spillover effects at the market level do exist and warrant further exploration. In controlling for Medicaid rates in the models, stronger relationships are evident between Medicaid and Medicare which may be an important avenue for future research.