Essays on Health Care and Inequality

Abstract

This dissertation consists of three chapters analyzing the role of institutional factors in observed health inequities. In the first chapter, I measure how differences in neighborhood environments contribute to Black-white disparities in primary care outcomes. I study this question in the context of preventive health care take-up among Medicare enrollees, a setting with full insurance and yet persistent racial gaps in recommended care. Leveraging variation from Medicare enrollees who move across areas, I show that local environments have sizable and immediate impacts on the probability of receiving recommended care and that Black enrollees on average live in areas with lower causal effects on outcomes. Within areas, Black enrollees receive worse care than white enrollees when residential racial segregation is greater. Using this insight, I estimate that residential sorting across Primary Care Service Areas (PCSAs) accounts for 28% of the Black-white gap in having an annual primary care visit, while heterogeneous effects by race within PCSAs account for 16% of the gap. Although individual-level differences such as preferences for health care utilization contribute to racial disparities in outcomes considered, local area effects play a substantial role as well. In the second chapter, I study whether heterogeneous hospital quality generates different local health care environments by race in the context of emergency care. I test whether there are differences in the ordered quality ranking of local hospitals for Black and white emergency patients. To control for underlying factors that affect hospital choice, I exploit ambulance company preferences as an instrument for hospital characteristics and estimate the impact of being treated at a hospital with a higher share of patients who are Black. Using mortality following hospitalization as the primary outcome, I find substantial race-specific effects, with Black patients experiencing better outcomes at hospitals with greater Black patient shares in the medium to long run. The results are consistent with a hospital choice model that exhibits institutional comparative advantage and positive Roy selection by patient race, pointing to the limitations of broad-based hospital quality measures. In the third chapter, I examine trends in the trajectories of Black medical school applicants over the years 1979-2020. I find that Black applicants have grown to 10% of the applicant pool in recent years, but acceptance rates remain below those of white and Asian applicants. For Black students who do matriculate to medical school, graduation rates lag behind those of white and Asian medical students. Among the cohorts of graduated MDs, medical schools affiliated with historically Black colleges and universities (HBCUs) continue to play an important role in graduating Black physicians, accounting for 14.9% of Black physicians in 1984-1999 and 14.7% in 2000-2015. Taken together, the evidence highlights continuing gaps for Black students in the physician pipeline, with need for targeted actions.