Local differences in racial preferences

R.E. McCormick, R.D. Tollison J. of Economic Behavior Org. 44 2001 201–219 207 the league’s desire to control the property rights to team movement to venues where other NBA franchises are located. 15 The basic customer discrimination hypothesis does not stand up in these results. Of course, we have not shown that white fans do not prefer white players or that black fans do not prefer black players or the reverse. Until we can observe the racial composition of those buying tickets, this issue remains open. 16

4. Local differences in racial preferences

To investigate whether there are regional or local differences in racial preferences, we parse the data into racial groups based on the black percentage of population in the under- lying SMSA. The quartiles of this variable in our sample are 5.7 percent black at the lower end and 18.2 percent at the top end. We define an SMSA as White if it falls in the lower quartile; Black if occurs in the top quarter; and Mixed if it lands in the middle half of the distribution. Table 3 reports the breakdown of the individual SMSAs. We reestimated the attendance equation allowing for separate coefficients on racial com- position in the three-racial classifications. The results on the racial composition variable, white to black player ratio, are reported in Table 4. We suppress the printing of the other coefficients. They are nearly the same as those reported in Table 2. This technique reveals that there is a pattern in the relation between team racial com- position and home attendance that seems to depend on the underlying racial composition of the community, a result that was not apparent using the simple black percentage of the population as a control variable. There is no statistically credible relation between the racial composition of the team and attendance in racially mixed or white SMSAs, but there is a relation in the predominately black SMSAs. In the last category, a whiter team is associated with lower attendance. By contrast to the earlier pooled results, we find that there is an attendance relation with team racial composition; the F statistic on the interaction class is 2.72, which is significant at the 5 percent level. 17 These results suggest that black fans have a preference for black players in largely black areas, while there is weaker evidence 15 The Clippers moved to Los Angeles without the formal approval of the NBA. A litigation ensued which was ultimately settled out of court. 16 There is an endogeniety issue that clouds this result. If the marginal cost of a minute of given player quality is lower for blacks than whites, relative playing minutes may affect ticket prices, so an OLS regression of demand does not work to estimate the attendance effects of racial composition. To address this question, we estimated the attendance equation jointly with a ticket price equation. The structural ticket price equation has price depending on stadium capacity, year, population, income per capita, and the ratio of white to black players on the team. In the structural estimates, ticket price is positively related to the player ratio variable. However, two-stage estimates of the attendance equation do not alter the previous conclusion about racial composition and attendance. The racial composition variables remain insignificant. The two-stage estimate of the composition coefficient is 8886, and the t ratio is 0.348. The insignificant relation between racial composition and attendance is not blurred by an endogeniety problem. 17 This result holds over alternative specifications where we omit the black population variable, and where we substitute the minutes played ratio variable in lieu of the player ratio variable to measure the racial composition of the team. In the latter case, the coefficient in the predominately white cities borders on statistical significance at the 10 percent level, and is positive. 208 R.E. McCormick, R.D. Tollison J. of Economic Behavior Org. 44 2001 201–219 Table 3 The distribution of black population across NBA cities, 1990 SMSA Racial classification Black percentage of total population in 1990 Atlanta Black 26.0 Boston Mixed 5.7 Chicago Black 19.2 Cleveland Mixed 16.0 Dallas Mixed 14.3 Denver White 5.3 Detroit Black 20.9 Golden state Mixed 8.6 Houston Mixed 17.9 Indiana Mixed 13.8 Kansas city Mixed 12.8 Los angeles Mixed 8.5 Milwaukee Mixed 13.3 New york Mixed 18.2 Philadelphia Black 18.7 Phoenix White 3.5 Portland White 2.8 Sacramento Mixed 6.9 San antonio Mixed 6.8 San diego Mixed 9.9 Seattle White 4.8 Utah White 1.0 Washington Black 26.6 Table 4 Parameter Estimates on Racial Composition in Attendance Equation Across NBA SMSAs a Sample size: 157; F ratio: 16.29; R 2 : 0.717 SMSA classification Parameter estimate t Ratio Prob |t| Blackest quartile − 206074.6 − 2.43 0.0165 Mixed half 12874.6 0.52 0.6062 Whitest quartile 39733.4 1.20 0.2333 a These parameter estimates are the interaction of the SMSA classification and the racial composition variable, total number of white players divided by total number of black players on a team. The F statistic on the interaction class is 2.72 which is significant at the 5 percent level. that white fans likewise prefer white players in predominately white areas, all else equal. In the racially mixed areas, no such relation is revealed. 18 These results do not provide an explanation for a salary wedge between white and black players. In general, we find no relation between race and attendance. In more detailed examination of NBA venues, we find that black players are sometimes preferred over whites, 18 In pairwise tests the coefficient on team racial composition in the blackest quartile is significantly different from the other two coefficients. The coefficients on the racial composition variable in the mixed and white SMSAs are not statistically different from each other. R.E. McCormick, R.D. Tollison J. of Economic Behavior Org. 44 2001 201–219 209 and white players are sometimes viewed as better by some fans. Thus, the source of the wage differential identified by Kahn and Sherer remains at issue. Indeed, according to the model of racial discrimination, black players ought to make more money in certain NBA markets than comparable white players.

5. Playing time results