Consumer Racial Discrimination: A Reassessment of the Market for Baseball Cards* JOSEPH McGARRITY University of Central Arkansas, Conway, AR 72032 HARVEY D. PALMER University of Rochester, Rochester, NY 14627 MARC POITRAS The University of Queensland, Brisbane 4072 Australia Research on the presence of consumer racial discrimination in the baseball labor and memorabilia markets has produced contradictory empirical results. While studies of baseball salaries find no evidence of discrimination, Nardinelli and Simon (1990) and Andersen and La Croix (1991) use data from the baseball card market to show that the price that consumers pay for a card depends on the player's race. In this paper, we reconsider the evidence of consumer discrimination in the baseball card market. Our study improves on previous research by applying more appropriate econometric meth- ods and using a data set in which card supply is constant and incentives for specula- tive demand are weaker. In contradiction to the aforementioned studies, we find little evidence of racial discrimination. This result proves robust across variable specifica- tions and econometric models. I. Introduction The effect of racial discrimination on wages depends on the source of discrimination. Becket (1957) identified three sources of discrimination: consumers, employers, and co- workers. Competitive forces tend to eliminate wage differentials based on employer and co-worker discrimination (Becket, 1957; Arrow, 1973). Kahn (1991 ) has shown, however, that consumer discrimination can cause long-run wage differentials in com- petitive equilibrium. In order to test for the presence of consumer racial discrimina- tion, empirical research must control for nonracial components of value, such as performance and ability. Considerable research has focused on the labor and memorabilia markets in the professional baseball industry where detailed statistics of player performance are com- piled that measure individual productivity relatively accurately. Controlling for differ- ences in productivity allows estimation of the economic effect of discrimination. In contrast, labor market studies of non-sport industries face the difficulty of accounting JOURNAL OF LABOR RESEARCH Volume XX, Number 2 Spring 1999