Physician Workforce Effect on Health

Abstract

Cities attract both more physicians and healthier people, but whether these two facts are causally related is yet to be determined, as many variables are correlated with both the physician concentration and health outcomes. This paper uses unidentifiable claims data from New Hampshire and treatment-effects analysis to address this question and finds that access to an additional physician per 10,000 residents leads to 4.5 saved lives per 100,000 residents. Using aggregate data and an instrumental-variable approach where I use the procedures carried out across areas joint with the policy-set reimbursement fees to instrument for the number of care providers, I show that these results generalize to the US as a whole. The results are robust to many specifications, to variations in the type of care providers considered, and to variations in how the instrument is constructed.