Dissertation Title： "Measuring Health and Assessing the Impact of Health System Change"This thesis explores each of the steps in the process linking health system change to output: changes in access, changes in health, and the valuation of health.

The first chapter derives an economic framework for measuring and valuing health and presents estimates of the change in population health over the past 40 years. We define "health capital" as the discounted value of the current and future utility associated with individual health. We measure health capital empirically using data on mortality and morbidity from chronic disease. Our estimates suggest that health capital at birth increased by about $100,000 between 1970 and 1990, and health capital at age 65 increased by about $169,000. The increase in health capital at birth and age 65 are both greater than the increase in average medical spending over that time period.

Chapter 2 examines the relationship between managed care penetration growth and change in the number of hospitals and emergency rooms. I find that growth in managed care penetration can explain between 19 and 37 percent of the decrease in hospitals and between 0 and 56 percent of the decline in emergency rooms in low income areas between 1984 and 1994. None of the decline in hospitals or emergency rooms in higher income areas can be attributed to managed care. I also find that distance to the nearest hospital in California cities increased by 22 percent for the poor and 16 percent for the non-poor over this period. I estimate that this would lead to a one percent drop in inpatient utilization by the poor and a 0.8 percent decline for people in higher income areas.

The third chapter examines the impact of hospital closings on health outcomes. I look at the effect of hospital closures on rates of admission for avoidable hospitalizations. I also examine differences in mortality and length of stay for two urgent conditions. I find weak evidence that people experience adverse health outcomes when an urban hospital closes, and that this effect increases with poverty status. Overall, the results are inconclusive but suggest that further study is necessary.