« Back to Results

Empirical Studies on Physician Quality and Treatment Choice

Paper Session

Friday, Jan. 3, 2020 2:30 PM - 4:30 PM (PDT)

Marriott Marquis, Grand Ballroom 10
Hosted By: American Economic Association
  • Chair: W. Bentley MacLeod, Columbia University

Understanding Physician Decision Making: The Case of Depression

Janet Currie
Princeton University
W. Bentley MacLeod
Columbia University


Faulty physician decision making has been blamed for everything from medical errors to excessive procedure use and wasteful spending. Yet medical treatment is often complex, requiring a sequence of decisions that may involve tradeoffs between selecting the choice with the highest expected value or selecting a choice with higher possible payoffs. We show that the best choice depends on a physician’s diagnostic skill so that the optimal treatment can vary even for identical patients. Bringing the model to patient claims data for depression, we show that skilled doctors who experiment more with drug choice achieve better patient outcomes, except when physician decisions violate professional guidelines for drug choice.

Measuring Physician Quality: Evidence from Physician Availability

Joseph Doyle
Massachusetts Institute of Technology


Measuring physician quality is fundamental to understanding healthcare productivity yet attempts to estimate the types of physicians that improve survival can be confounded due to patient sorting. This paper aims to overcome this endogeneity problem by exploiting plausibly exogenous variation in the mix of physicians available to treat the patient on the date of inpatient admission. One innovation is the use of 100% Medicare claims data to characterize the mix of physicians available including specialty training, medical school quality rankings, patient volume, sex, and years of experience. When heart failure patients enter the hospital when more cardiologists are available, patients receive more intensive treatments and are more likely to survive at one year. The results speak to the debate over the value of treatment intensity and specialists in particular.

Physicians Treating Physicians: The Relational Advantage in Treatment Choice

Stacey H. Chen
National Graduate Institute for Policy Studies (GRIPS)
Hongwei Chuang
International University of Japan
Tzu-Hsin Lin
National Taiwan University Hospital


This study examines the agency problems by estimating the informational and relational effects of physician-patients on their invasive treatment. To address potential issues of patient selection, we compare treatment intensity between physician- versus nonphysician-patients attended by the same doctor in the same hospital, and control for patients' previous choices of doctors. To identify the relational effect, we further compare the impacts of physician-patients specializing in the same area as their attending doctors versus those in different areas. We test whether physician-patients receive more care than comparable nonphysician-patients and whether the increased volume results from physician-patients’ relational advantages or their information advantages.

Physicians' Responses to Medical Subsidy Programs: Evidence from Japan

Hiroshi Aiura
Nanzan University
Reo Takaku
Hitotsubashi University


Previous studies find that an expansion of health insurance enrollment encourages physicians to work for an underserved area where many uninsured lived. However, in the countries have had universal health coverage, a further expansion of the generosity of health insurance may not have such an equalization effect. Instead, it may secure sufficient profit to operate in the urban area for physicians, generating a concentration of physicians into cities. To test this hypothesis, we examine how a vast expansion of medical subsidy program changes the behavior of primary care physicians. In Japan, the local subsidization programs which significantly reduce out-of-pocket expense for children's health care utilization, named Medical Subsidy for Children and Infants (MSCI), have spread rapidly since 2000, while there were significant regional differences in the eligibility criteria. By using the census of clinics from 1999 to 2011, matched with municipality-level eligibility criteria of MSCI, we implement difference-in-differences-in-difference analysis. The results show that MSCI increases a monthly number of visits per clinic with the similar impacts indicated by the RAND Health Insurance Experiment. Also, clinics choose to be in a densely populated area under generous MSCI system, suggesting the expansion of the generosity of health insurance system may accelerate the concentration of physicians into an urban area. Finally, we find a significant reduction in the number of consultation days under generous MSCI.
JEL Classifications
  • I1 - Health
  • D8 - Information, Knowledge, and Uncertainty