This project propose to test the hypothesis that medical partnerships can effectively respond to an inherent moral hazard threat, or shirking, by monitoring their members. While it has long been believed that partnerships are inefficient due to an inability to control opportunism of members, widespread occurrence of this organizational form belies this belief. This project will examine the methods forms use to address the potential to shirk, the relative effectiveness of these methods, and whether monitoring methods differ systematically across types of groups or markets. A model is developed which describes the decision-making process in medical groups. In order to evaluate the relationship between medical group structure and their choice of analysis will be performed using survey data collected by the Medical Group Management Association's Center for Research in Ambulatory Health Care Administration in their annual Physician Compensation and Production Survey and Cost Survey, with additional data describing market characteristics pulled from the Area Resource Files. Stochastic frontier estimation, similar to improvements attributable to monitoring the behavior of physicians in response to groups' choices of incentives and monitoring methods will be examined by estimating individual physicians' demand equations using Random Effects. Instrumental Variables estimation, as in Gaynor's and Gertler's 1995 article in Rand Journal of Economics. Group choice variables will be instrumental to collect for endogeneity and random effects will account for possible self- selection of physicians into groups on the basis of group culture.
