Evaluating the Effect of a Policy Change to Hospital Productivity: 80 Hours Work Restriction on Medical Residents

This paper uses a two-year panel dataset on hospitals from the American Hospital Association (AHA) to evaluate the effect a policy change has on the marginal product of medical residents. A weighted 2SLS approach is used to estimate a semi-parametric production function. A policy restricting medical residents to work no more than 80 hours a week is found to result in a net loss of 14 inpatient days per resident annually, which is not statistically different from zero. In addition, the model presented in this paper performs better than past models when estimating first-order effects of inputs in the hospital production function.