This month’s edition of The Journal of Sports Economics has a study on the Rooney Rule and the hiring and firing of African-American coaches in the NFL over the last 21 years, entitled “Has the NFL’s Rooney Rule Efforts “Leveled the Field” For African American Head Coach Candidates?” (h/t to Phil Birnbaum of Sabermetric Research). The study makes a couple of different findings. First, that African-American coaches prior to the institution of the Rooney Rule were more successful than their white counterparts, in terms of wins, playoff appearances, etc., but were fired . Second, that there are no racial differences in head coach performance since the Rooney Rule was put into place.
I’ll discuss each of those things, plus my thoughts on the Rooney Rule at the end. Phil Birnbaum does a good job of addressing the first issues, the evidence of discrimination based on coaching performance pre-Rooney Rule. Phil is a bright guy and generally does a good job breaking down these studies. Football subject matter knowledge is not his forte, but he is on the right path with his criticism of the findings here.
The findings are based on 29 seasons and 5 total African-American head coaches (Art Shell, Dennis Green, Ray Rhodes, Tony Dungy, and Herm Edwards). Those 5 coaches were at 7 total jobs, as Rhodes moved directly to Green Bay after being fired in Philadelphia, and Dungy moved from Tampa to Indianapolis in 2002.
As Birnbaum points out, the authors found significance between the African-American coaches when they looked at average win totals and first year win totals compared to white counterparts. The African-American coaches won 9.1 games per year (8.0 for white coaches) and 9.6 in their first season (7.0 for white coaches). However, the statistical significance disappeared when they used the Sports Illustrated preseason rankings as a proxy for team quality, and they also noted that African-American coaches took over teams that had a better record the year before, than white coaches.
This is true, but with this few coaches, I think we can dig deeper. For example, I wrote about the 1991 Minnesota Vikings last year due to how much talent they had on offense and how poorly they played in Jerry Burns’ final season. When Green took over for the underperforming Burns, he had a team that already had two Hall of Fame offensive linemen in their primes in Zimmerman and McDaniel, Terry Allen at running back, along with receivers Cris Carter, Anthony Carter and Steve Jordan at tight end. It ranks as one of the best offensive groups of the last thirty years not named the 49ers, the Manning era Colts, Air Coryell, or the Elway/Terrell Broncos. And since we don’t have a large sample size, that’s one case of a really talented underperforming team where an African-American coach took over a favorable situation.
I looked at the average simple rating system for each of the seven teams in question. The average result was that these teams were +2.4 points per game above average over the previous 3 seasons prior to the African-American coach being hired. The average result in the first season was +1.9 points per game above average. Basically identical. Tony Dungy to Tampa Bay was the only one from the above group that actually took over a below average team. Even when the coach took over a team with a losing record, it was because of an unlucky season, such as when Dungy took over in Indianapolis in 2002 after a 6-10 season that involved Edgerrin James tearing his knee, and with Peyton Manning already in place.
If the authors had accounted for team quality beforehand, they would have seen that African-American coaches had better records because they took over better teams. It’s entirely possible that the better organizations were more willing to give chances to African-American coaches at a time when discrimination was still present in many organizations. But that is correlation, and not causation of better African-American coaches on average leading to more wins.
As for the finding about firing, there were only five firings. I’m sure Ray Rhodes being fired in Green Bay after one season at 8-8 skews the findings with such a small sample size, when a regression includes variables like wins and years with the team. In reality, he was a questionable choice to replace Holmgren, he was just coming off a 3-13 season at Philadelphia that got him fired, and when he went 8-8 with an in-his-prime Favre, a team that had been to the playoffs for six straight years, and had been in Super Bowls, the leash was short. I suspect that if the New England Patriots hypothetically replaced Bill Belichek with John Fox in 2011 and the team failed to make the playoffs and regressed badly, Fox would be out in short order.
I point all of this out to question the methodology and conclusions pre-Rooney that African-American coaches were better, and thus this was the sign of discrimination. I do think there was discrimination, bias, whatever you want to call it. I would just simply point to this: there were only five African-American head coaches, total, before 2003, and two at the time the rule was implemented.
The NFL had legitimate reason to be concerned about its hiring practices when the rule was instituted. I know it’s not always a popular rule, but I support the spirit behind it. It’s also, I should note, a decision by a private business and not an edict of government. All those in favor of private business having the discretion to make business decisions should support the Rooney Rule. It’s not affirmative action from the government. The owners looked at themselves, their constituency and players, and decided they weren’t doing a good enough job of hiring qualified African-American candidates for head coaching jobs based on the nature of their business.
The average coaching performance has been roughly equal since the Rooney Rule, with African-American coaches winning 8.1 and other coaches 8.2 per year per the study (I’m wondering if there is an error somewhere, average should be 8.0, unless they are excluding partial seasons where a coach was fired). 30% of coordinators from 2002-2008 were African-American; 16% of coaching hires among the coordinator positions have been African-American. Madden and Ruther found no significant differences between performance of coordinators based on race.
I think there is an interesting question going forward of what standards the NFL wants to see to discontinue the Rooney Rule. After it was implemented, the number of African-American head coaches rose quickly to 5 in 2005, and now sits at 7 as we enter the 2011 season. At some point, the benefits of the rule will be outweighed by the negatives–the perception of certain “sham” candidates by teams that have no interest, the difference where African-American interim coaches like Leslie Frazier can be hired directly, while white ones like Jason Garrett cannot without further interviews of minority candidates. In my opinion, these negatives have been outweighed by the fact that guys like Mike Tomlin and Raheem Morris have received opportunities earlier than they may otherwise have gotten, and the hiring practices have moved to more adequately represent the assistant coaching ranks.
Still, there will be a point very soon that the league will need to revisit the rule. The momentum created by the success of Lovie Smith and Tony Dungy meeting in a Super Bowl and Tomlin winning another will do more to break barriers than continuing to require mandatory interviews. As important in demonstrating how far the league has come is the case of Marvin Lewis. It is a rare case that a coach gets a ninth year when he has never won a playoff game, regardless of any other characteristic of that coach. The fact that, with the highs of Super Bowls and the lows of bad seasons, the African-American coaches over the last nine years are in line with other coaches is a positive.
I hope the league has some milestones in mind, whether they be hiring rates among coordinators or some other measure. The ultimate success would be the termination of the Rooney Rule in the near future, as a sign of how far the league has come in twenty-five years.
[photo via Getty]