Empirical Studies on the Organization of Health Care

Public Deposited

This thesis contains four essays on the organization (both internal and external) of health care in the US. The first essay examines a mechanism through which individual workers acquire (or maintain) competence, namely that of experience. Specifically, I analyze whether cardiac surgeons who perform more procedures experience an improvement in performance. As my identification strategy, I consider exogenous shocks to the procedure volume of CABG surgeons in Florida caused by the exit of other surgeons from the same hospital. Using this instrument, I find evidence indicating a strong learning-by-doing effect for cardiac surgeons: an additional procedure a year leads to a reduction in the probability of patient mortality by .05 percentage points. Further, I find this improvement in surgeon performance to be completely transferable across different hospital settings, and find evidence of some economies of scope among the different surgical procedures performed by a cardiac surgeon. The second and third essays also deal with physician behavior - focusing on their medical practice style, which is defined as the propensity of the physician to prescribe a particular treatment to a given patient. In the second essay (joint with David Dranove and Hayagreeva Rao), we propose several new statistical methods for assessing the presence and sources of small area variations in physician practice. In the third essay (joint with David Dranove), we focus on "credence good" markets and attempt to answer the question: what prevents agents from always recommending more costly services? We examine whether the market punishes agents who are deemed to be too "aggressive". Focusing on Ob/Gyn physicians, we find that maternity patients prefer not to visit physicians who overprescribe cesarean sections, i.e. who are more aggressive. The final essay (joint with Leemore Dafny and Mark Duggan) looks at the extent to which rising premiums in the private health insurance industry can be explained by the increasing consolidation among insurance carriers. We make use of a large, private database of insurance contracts in order to establish this link. By looking at within-market variation, we find that a one standard-deviation change in concentration leads to a premium increase of 2.5-4%

Last modified
  • 08/14/2018
Date created
Resource type
Rights statement