Competition in physician private practices

This dissertation is comprised of three separate essays that share a similar theme: physician private practices. The first essay revisits prior research that has proposed a general framework for physician cost function estimation. The econometric specification is a multi-product cost function, and my model addresses and deals with many measurement problems and data complications in physician practice surveys. In particular, I expand upon existing theory to include those practices that yield zero outputs. This modification allows allow me to utilize a more representative and larger sample. The results of my analysis confirm that previous estimates have yielded a downward bias in reported marginal cost estimates, a point often speculated in the literature but not formally addressed empirically. The second essay uses the econometric specification from the first essay to develop a means for testing market power in physician private practices. Using the 1998 American Medical Association Physician Socioeconomic Monitoring Survey, I test the hypothesis of physician collusion in an oligopolistic setting for five fields of specialty. The collusion hypothesis is tested in a traditional conjectural-variation framework that reveals physicians are not highly collusive. However, a Lerner index shows that physicians do possess a high degree of monopoly power, consistent with many of the claims brought forward by the Department of Justice, primarily due to consumer price insensitivity. The third essay uses the marginal cost estimates of the second paper to revisit the Harberger partial equilibrium approach for computing welfare loss. Deadweight loss estimates are computed by physician, practice, and office visit. The reported losses are indicative of a loss in consumer surplus from the monopoly power that physicians possessed in 1998. The third essay concludes by showing that significant price markups led to substantial losses in welfare by the consumer and third-party payer