Abstract
Prior econometric studies of physician fee determination report that fees are positively related to the proportion of the market area that is black and negatively related to the proportion of the market area that is white, but the studies provide only sketchy explanations for these results. This article presents a price discrimination model which explains the empirical results and provides the specific prediction that low income self-pay consumers in the black community constitute the group that pays higher prices for physician services. The study then replicates prior econometric results on a more recent national database, but finds that the results are sensitive to specification. When geographic differences are controlled for in the empirical model, the results fade and a statistical test indicates that the expanded specification is superior to the specification which replicated the price discrimination result. The conclusion is that there is no compelling evidence of price discrimination in physician services markets based on race.