A Production Function for Physician Services

Abstract

IN recent years, the American public has become greatly alarmed over an apparent shortage of physicians, a shortage which is expected to reach crisis proportions within the next decade. The predicted shortage stems, in part, from the nation's current efforts to make health care accessible to all income classes and from past neglect of medical-school facilities. However, there is also some evidence that the shortage is further aggravated by widespread inefficiency in the use of medical manpower. It is a matter which is increasingly attracting the attention of health-manpower specialists, but one on which opinions still differ widely. Somne students of the health-care sector have argued, as early as a decade ago, that it would be techniically imnpossible to add to the already hea\y patient load carried by American physicians.' In contrast, Rashi Fein's recent book The Doctor Shortage (1967) concludes with the assertion that physicians could increase their productivity still further, either by joining large-scale group practices or by delegating more of their workload to auxiliary personnel. Fein admits, however, that in the absence of more research, one cannot be certain whether the gains from group-practice formation and from increased reliance on paramedical aides will be large or more modest. (p. 144) His argument therefore rests on essentially intuitive grounds. Fein's call for more research on the production of physicians' services is entirely justified. For, in spite of the crucial role played by physicians in the production and delivery of health services, there exists very little enmpirical information on the technical and economic determinants of the physician's productivity. This paper is an attempt to contribute to our understanding of this important parameter. Using a simnple model of physician behavior and a nationwide cross-section sample of physicians, the study seeks to identify the effects of auxiliary persoi)nel and of the mode of practice (solo or group) on the physician's rate of output. Our analysis leads to the conclusion that the average American physician could profitably employ roughly twice the number of aides he currently employs and thus increase his hourly rate of output by about 25 per cent above its current level. This figure takes on added meaning when it is recalled that a mere increase of 4 per cent in average physician productivity in the United States would add more to the aggregate supply of medical services than would the entire current graduating class froin American medical schools. Looked at in another way, our results suggest that, in choosing his combination of practice inputs, the average physician in our sample appears to have priced out his own time at a value much below that implicit in his medical fees. On either interpretation, the results therefore support the thesis that American physicians tend to be wasteful in the use of their scarcest and most expensive resource.