X hits on this document





14 / 16



10. Looking at Nielsen ratings for local broadcasts of NBA games during the 1996-1997 season, Kanazawa and Funk (2001) found that these ratings were positively related to the number of White play- ers represented on the two teams. A qualification is that the Chicago Bulls, shown in our Table 1 to be the whitest NBA team during the 1990-1999 period, may have boosted the Nielsen ratings not only through the Bulls’racial profile but also by featuring the returning superstar Michael Jordan. Thus, the findings of Kanazawa and Funk for the 1996-1997 season may to some extent conflate a fan response to greater White representation with a “Michael Jordan effect.” That is, the importance of superstar prowess may well have transcended race in this case.

11. We interact team White with time rather than using a random-effects model because of the degrees of freedom limitations. The restriction appears to be reasonable given the systematic change (approximately linear) in team composition apparent from Figure 1.

  • 12.

    Again, results are not sensitive to inclusion of the two expansion teams, Toronto and Vancouver.

  • 13.

    Similar results are obtained if we enter the ratio between TWHITE and POPWHITE in place of

the interaction. The insignificance of either of these matching variables when not interacted with time is consistent with the negative results noted by Berri, Schmidt, and Brook (2004) in their gate-revenue regressions based on data from the 1992-1993 through 1995-1996 NBA seasons.

14. As noted, extant studies of NBA player salaries show that these five variables (points scored, total rebounds, assists, blocked shots, and sometimes shooting efficiency) are the statistically significant determinants of salaries. Other productivity factors (e.g., free-throw percentage, steals, and offensive rebounds) are not statistically significant in salary regressions. The studies, however, do not yield clear guidance as to the relative weights of these variables. See Berri (2003), who reported on evidence from 11 salary studies.

15. We do not run this regression for starters and bench players separately due to the small number of White player observations.

16. Recent evidence on salary distribution in the NBA shows that, even though the overall Black-White wage differential appears to have diminished, the degree of dispersion in NBA wages nev- ertheless increased during much of the 1990s. Hill and Groothuis (2001) showed that whereas the mean NBA salary rose approximately 78.5% between the 1993-1994 and 1997-1998 seasons, the median sal- ary increased by just 31.3%. Hamilton (1997), examining salary data in the mid-1990s, found evidence that at the upper end of the salary distribution (75th and 90th percentiles), Whites earn more than their Black counterparts, although Hamilton found no statistical difference in salary at lower percentiles.

17. Meanwhile, teams’abilities to bid freely for the players they want do not seem to have been con- strained in any meaningful way by the post-1983 “salary cap,” because there are numerous exceptions to this limit—including the “Bird Exception,” which allows a team to acquire players from others teams up to the level of the salary cap and then freely bid above the salary cap to keep any of its own players who have become free agents (Hill & Groothuis, 2001, p. 133). Wage setting in the NBA does seem to have been significantly altered by the subsequent 1998 collective-bargaining agreement, with Hill and Groothuis (2001) documenting a reduction in wage dispersion after this new policy took effect. This could have no more than a minimal impact on our empirical work, however, because we end our sample in 1999.

18. Analysis of team draft picks could offer a means of more definitively separating team preferences from player preferences.


Becker, G. S. (1971). The economics of discrimination (2nd ed.). Chicago: University of Chicago Press. Berri, D. J. (2003). Economics and the National Basketball Association: Surveying the literature at the tip-off. Mimeo. Bakersfield: California State University.

Document info
Document views27
Page views27
Page last viewedSat Oct 22 18:38:31 UTC 2016