The Determinants of Physicians’ Location Choice: Understanding the Rural Shortage.

A long-standing challenge in the US health care system is the provision of medical services to rural areas, where 25% of the population live, but only 10% of physicians operate. This paper develops a model of physicians' location choices and uses it to explore the impact of policy changes (loan forgiveness and salary incentives) on the geographical distribution of physicians. I build a structural spatial equilibrium model in which physicians are heterogeneous along their specialty, demographics, and skill. Identification of the parameters of interest is challenged by the possible correlation between unobserved characteristics of location and wages, as offered wages are higher where amenities are fewer. To overcome this issue, I collect micro-level data from physicians' directories on doctors' medical school, residency, and first job choices. This wealth of information and structural methods of demand à la Berry, Levinsohn, and Pakes (1995) allow me to back up the unobserved characteristics and be exactly identified. I allow individuals the preference to remain close to their residency location and let each medical resident's job choice set depend on his or her skill. I find that all residents display a strong retention preference and that primary care physicians in particular are 3.5 times more likely to pick a job within the same state and 4 times more likely to pick a job within the same area as their residency. I then use the model to analyze the performance of current policies targeted at bringing physicians to rural areas. I find that 0.5% more residents have picked rural areas due to loan forgiveness alone. Monetary incentives in the form of bonus payments averaging $7,500 are responsible for a further 0.2% increase. I find that even higher initial monetary bonuses would lead to only 1.5% more physicians choosing rural areas, confirming that physicians are just not that responsive to monetary incentives.