By Jake Fisher
http://online.wsj.com/article/SB10001424052748704224004574489773943949070.html
People love to talk about competitive parity in sports, most of the time to say that it doesn’t exist. The Wall Street Journal article, linked above, takes a stand that the salary cap/floor structure of the NFL is hurting the competitive parity in the league, as evidenced by the many blowouts this season.
But I don’t think this claim holds too much weight. In relation to other major leagues, the NFL has one of the strictest salary bounds and most competitive parity.
The NFL has a salary cap and floor. The floor is around 85-90% of the cap figure, as stipulated by the league’s collective bargaining agreement. These bounds are fairly tight, especially when looking at other sports. When comparing the standard deviations in team payrolls as a percentage of the league average in 2007 for MLB, NFL, NHL, and NBA, football is the lowest. This makes sense. In baseball there is no cap or floor and in basketball, the salary cap is “soft” meaning there exists a significant number of exceptions whereby teams can rise above the cap and fall below the floor. The Wall Street Journal article makes it seem like in the NFL the difference in the highest team’s payroll and the lowest team’s payroll is huge (six times Peyton Manning’s salary). Though it sounds big, this difference is less (on a percentage basis) than the difference in other sports. The actual dollar figure is high only because the salary cap and floor are larger in football than in other sports, since the NFL generates the most revenue of the major leagues.
It makes sense that strict payroll restrictions create more competitive parity in the NFL than in other leagues. And from what I’ve seen this plays out in reality. Looking at regressions of payroll against wins (from 2005 to 2007), the NFL R2 appears to be the lowest of the major sports. Though all sports have statistically significant trends of payroll spending to winning, the trend in the NFL seems to be the least pronounced.
