David Leonhardt, author of the NY Times article on coaching tenure in the NBA, kindly wrote and pointed out an observation in his piece that I overlooked:
The pattern holds in almost any important category of coaches. Winning black coaches have been replaced sooner than winning white coaches on average, and experienced black coaches have served shorter tenures than experienced white coaches. The same is true among losing coaches, among rookie coaches and among coaches who played in the N.B.A. and those who did not.
Clearly, something is going on. This does not rule out the hypothesis advanced in my earlier commentary, but it causes me to consider an alternative. Suppose the league office exerted pressure on teams to hire a minority coach, i.e. someone other than the team would have picked absent the pressure. The match would not be as good as when hiring was unrestricted, and tenure might be shortened as a result.
As an economist, I have an innate preference for this hypothesis over the "mistaken belief" hypothesis advanced earlier. It may be in the league's interest to promote diversity among coaches, but individual teams care little about their contribution to it. The league can put pressure on hiring, but firing is another matter. I would like to see the results of tests between these and other hypotheses on the tenure of coaches in the NBA.