| In Chapter 1 I examine the effect of managed care on the access to and quality of hospital services provided to the uninsured and its effect on their health status. There are three main conclusions to my paper. First, managed care penetration disproportionately reduces the revenues of safety net hospitals, increasing their probability of closing and of terminating services generally used by the uninsured---emergency rooms, obstetrics and alcohol and drug treatment centers. Second, with the expansion of managed care, uninsured patients have shifted toward government hospitals, whose quality---as measured by the proportion of patients that die of heart attack---has deteriorated. Finally, I also find a negative impact of managed care on the health of the uninsured.;In Chapter 2, with Janice Seinfeld, we examine the relationship between managed care penetration and hospital's adoption of technologies during the period 1982--1995. First, we find that managed care has a negative effect on hospitals' technology acquisition for each of the thirteen medical technologies we consider. This effect is stronger for those technologies diffusing during the 1990s, when managed care sector is at its largest. Second, we rank the technologies according to their cost-benefit ratios (CBR). We find that managed care enrollment has a disproportionately negative effect on the adoption of technologies with higher CBR. Finally, the empirical evidence suggests that there may be long-term reductions in medical cost growth resulting from increased managed care penetration.;In the final chapter, joint with David Cutler, we compare the US health system to the one in Canada, Spain and the United Kingdom. We find that, although the US has significantly higher health expenditures than the other countries, it has not better health outcomes. The US outperforms other countries in the treatment of heart disease and strokes, but it does worse than the other countries for the treatment of diabetes. The discrepancy is not attributable to differing performance by income. Nor does it appear to be attributable to measurement issues in determining a person's quality of life.;We conclude that the US may do better in the treatment of acute, technology intensive conditions and worse in the treatment of chronic conditions that require constant monitoring. |
