The following was lifted from Hawkonomics, the blog maintained by Stacey Brook (co-author of The Wages of Wins). Although everyone is encouraged to read and enjoy Hawkonomics, Stacey’s blog is primarily written for his students at the University of Iowa.
In today’s USA Today, there is an article stating that a salary cap leads to greater competitive balance. In fact the article quotes David Poile, GM of the Nashville Predators as saying, “[t]he salary cap has brought competitive balance to the league.”
We will tackle this issue formally later in the Sports Economics course, but for now let’s test this hypothesis informally. Since we only have four years of results with a salary cap, the results must be taken with a grain of salt – in other words I cannot make a conclusive statement either way, but will offer some conjectures below. So has the salary cap (payroll cap) brought competitive balance to the NHL?
Let’s see what is happening to competitive balance in the NHL from a few different angles. One way of looking at this is to compare the level of competitive balance – as measured by the Noll-Scully metric – for the four years before the payroll cap and the four years after the payroll cap. Here are the results:
|
Season |
Noll-Scully |
| 2000-01 |
1.858 |
| 2001-02 |
1.581 |
| 2002-03 |
1.592 |
| 2003-04 |
1.633 |
| 2005-06 |
1.637 |
| 2006-07 |
1.600 |
| 2007-08 |
1.037 |
| 2008-09 |
1.369 |
As evident in the table above, the Noll-Scully metric of competitive balance decreased during the four years before the payroll cap, and decreased during the four years after the payroll cap (in bold). Remember that as the Noll-Scully gets closer to zero, competitive balance is improving in the league; so smaller numbers are increases in competitive balance.
But this is really not anything new. Since the 1970’s, the average measure of the Noll-Scully metric has been declining; as shown in the table below. (Note the 2000’s doesn’t include the 2009-10 season, since it hasn’t occurred yet).
|
Decade |
Average Noll-Scully |
|
1970’s |
2.557 |
|
1980’s |
1.969 |
|
1990’s |
1.796 |
|
2000’s |
1.538 |
Hence, only focusing on the decline in Noll-Scully measure of competitive balance in the four years since the introduction of the payroll cap ignores the decades of decline in the Noll-Scully measure of competitive balance WITHOUT a payroll cap.
As we argue in The Wages of Wins (and as detailed in our paper – The Short Supply of Tall People) competitive balance is declining not because of changes in league institutional rules – such as payroll caps – but rather due to the increasing pool of talent to play sports, such as hockey. So I am skeptical that the payroll cap really has anything to do with the observed improvements in competitive balance.
– Stacey
The WoW Journal Comments Policy
Wages of Wins Journal 
TRad
September 30, 2009
Dave, suppose NFL will have an uncapped season. How conclusive would be inferences from such season?
dberri
September 30, 2009
TRad,
In terms of competitive balance, I don’t think an uncapped season makes any difference. Competitive balance in the NFL did not really improve with the salary cap. One of the problems in football is that performance is very hard to predict. Cap or no cap, that will still be the case.
A.S.
September 30, 2009
I don’t find this persuasive at all.
Yes, N-S was decreasing before the cap, and it is decreasing after the cap. But what is the rate of decrease before and after? What other factors could affect the rate of decrease (you mention one – increasing availability of talent). Why isn’t there any attempt to isolate the effects of the cap from the effects of the other factors?
It could be the case that absent the cap, the decrease in N-S would be less (or it would even increase) as a result of the other factors.
Here is what I would like to see done (and I don’t remember any discussion of it in WoW, although my copy is at home so I can’t check):
1. Find out whether a salary cap decreases the disparity between the the big spending and small spending teams. Perhaps through the use of a gini coefficient, or the like. Has the NHL cap decreased the salary inequalities?
2. Find out how many wins are explained by team salary, through regression or some other mathematical technique. That is, if every baseball team spent exactly the same salary, we would expect them all to average 81 wins. But if one team spent twice the average salary and one team half the average salary, how many wins do we expect each team to win?
3. With the combination of (1) and (2) above, we should be able to estimate the improvement (if any) in competitive balance resulting from a salary cap.
Tball
September 30, 2009
I’d agree, the time frame is too small to reach a useful conclusion regarding competitive balance, but I do believe that anything that puts all teams on similar financial footing with regards to their ability to capture talent works toward competitive balance.
Major league sports also has various vehicles in place to counter the competitive balance fostered by a salary cap. For instance, in baseball, Tampa Bay had one of the smallest payrolls in baseball last year. They succeeded because draft rules limit players they drafted to sign with them or hold out a year. And, once signed, MLB teams get to keep exclusive rights to players for period of years. After drafting first for a number of years, and making some good picks, the Rays were able to sequester enough cheap talent to field one of the most talented teams in baseball despite the lack of a salary cap. Other sports feature similar drafts, although the length of the servitude varies. The NFL has franchise tags. The NBA gives more leeway to teams to sign their own free agents than other teams. These mechanisms counteract the goal of competitive balance through a cap. The more a free market is restricted, the less equal financial footing is productive.
To TRad – it would be one season, so I don’t think we’ll learn anything that is conclusive (not that you were asking me/other readers). I would point out, though, that the NFL will be featuring other free market restrictions (as I understand it) that will counteract the uncapped season and restrict player mobility. Similar to baseball, I believe the restricted free agent period is being extended (I believe players that became free agents within 4 years of signing previously and that period is being extended to 6 years) and that franchise tags and transition tags will grow in number (maximum allowed, not just usage patterns). How many players/team are greatly productive after their 6th NFL season? How many of those players are free agents each year? If a team can tag a couple of them (and tagged players don’t get moved without a trade in the NFL), how much player movement can we expect under ‘free’ agency?
I want to say a read about these new free agency restrictions in Peter King’s MMQB during the preseason, but I don’t have any other sources on this and I don’t recall what I had for lunch on Monday, so take it all with a few grains of salt.
Here’s the real question, when the roof doesn’t cave in, does the next CBA have a salary cap? Salary floor? Will the media impute the fact the roof doesn’t cave in to mean that competitive balance does not require a cap? Interesting times.
Italian Stallion
September 30, 2009
I don’t know much about either football or hockey, but I find it hard to believe that a salary cap doesn’t contribute to balance either sport.
I understand the relevance of pro football performance being hard to predict. But over the long haul there must be some correlation between past performance and future performance that is then correlated to salary.
If not, then paying up for almost anyone would be a economic mistake. You might as well build a team of the cheapest players possible.
Now of course there may be WAY MORE mistakes made in football (and other unpredictable sports) than in predictable sports like basketball. And perhaps because of the resultant volatility we need a longer time frame or different techniques to capture the contribution of a salary cap.
I intuitively suspect that if you gave me an unlimited bankroll I would have a much higher probability of putting together a great football or hockey team than if you tied my hands.
Italian Stallion
September 30, 2009
Tball,
Excellent insights.
Tony S
September 11, 2010
I have a general question on the causes of competitive balance/imbalance.
Do the relative balance found in different sports come from the number of “trials” (possessions, at-bats, etc) in a game? That is, baseball and basketball have a lot of trials and are (in W.O.W. pg61) rated as the least competitive. With a lot of trials per game, the better team is more likely to turn its’ superiority into a win, and I think any way you count it, there are more trials in basketball. The NFL is more balanced than either, and NFL teams get around 12 possessions per game.
The Noll-Scully scores are higher for ice hockey than soccer. These 2 sports can be modeled with the poisson distribution applied to the number of goals scored in a given game, and with the higher scoring in hockey, a greater disparity in winning percentage is not surprising.
dberri
September 11, 2010
Tony S,
The problem with any explanation that rests on the structure of the game is that you can’t explain why competitive balance changes over time. And it certainly has changed. Baseball is much more competitive today than it was in the past.
Tony S
September 11, 2010
Certainly the nature of each sport does not explain changes in a given sport over time. Yes, each league changes balance level, due to revenue or knowledge advantages or talent pool changes or what have you.
However, even with MLB getting more competitive and major European soccer leagues getting more imbalanced (http://www.sportsnexus.co.uk/resources/12028/assets/football/CompetitiveBalance.pdf) the soccer leagues are, by Noll-Scully anyway, much more balanced.