Sherry Glied, Ashwin G Prabhu, Norman Edelman
August 26, 2009
Research Objective: This study offers a novel approach to workforce planning in the physician market. Rather than projecting the future demand for physician services, a human capital model is used to estimate the societal cost of producing a physician service. The socially optimal workforce is one at which (at optimal practice scale), the societal cost of producing a physician service is equal to the societal benefit obtained from the service.Study Design: Physician human capital consists of two components: the underlying human capital (productivity) of those who become physicians and the job-specific investments (physician training) added to this underlying capital. The value of physicians' underlying human capital is estimated using a regression analysis of the National Longitudinal Sample of Youth (NLSY). For those in the survey who did not go on to become doctors, income over time is modeled as a function of a rich set of variables measured in youth, including family background, educational attainment and a range of high-school level performance tests. This equation is then used to forecast an age-earnings profile for doctors based on the characteristics in youth of those NLSY cohort participants who subsequently became doctors. Next, published estimates are used to measure the total cost (wherever paid) of investments in physician training. Combining these estimates, the social cost per primary care physician provided visit and Medicare relative value unit (RVU) is determined.Principal Findings: Physicians are drawn from the highest performing group of high school students. The earnings of comparable students who do not become doctors and the predicted earnings of would be doctors are substantially above the population mean. The opportunity cost of physician human capital is thus very high. The estimated societal cost per primary care physician visit is substantially higher than the average co-payment. The societal cost per primary care physician provided RVU is generally higher than the current Medicare compensation rate per RVU. The private return to primary care physician training is relatively low, in the range of 7-9%.Conclusions: At current levels of supply, the marginal social costs of primary care visits appear to be equal to or greater than marginal social benefits of many primary care services. In considering expansions of primary care capacity, it may be efficient to increase the use of complementary, lower-skilled practitioners.