Nyhan Wonders About a Selig Quote

Posted on May 27, 2008 by


Brendan Nyhan – of Brendan-Nyhan.com – asked me (via e-mail) to look at the following quote from Major League Baseball commissioner, Bud Selig.

“…. I think we have more parity than any [sport].”

The partial quote appears in an article by Tom Verducci (The Bizarro Superman) examining the level of competitive balance in Major League Baseball. 

Here is the full quote from baseball’s commissioner:

“This is one time I can say,” Selig boasts of the parity, “that this is exactly what we tried to do. I think we have more parity than any [sport].”

The complete Selig quote makes two contentions:

1. The policies baseball has put into place in recent years have improved competitive balance.

2. Baseball now has more competitive balance than any other sport.

Nyhan wondered if the data supported the second point.  Before I get to that issue, though, let me examine the first point.

Payroll Parity in Baseball

From 1996 to 2000, the New York Yankees won four World Series titles.  In four of these five seasons the Yankees also led all of baseball in payroll.  These two facts led many to conclude that money was destroying baseball’s competitive balance. The supposed negative impact of money on the game of baseball was articulated by Baseball’s Blue Ribbon Panel (a report that was reviewed in detail in The Wages of Wins).

In 2002 baseball faced another labor strike.  With the players about to walk, though, an agreement was struck.  This agreement instituted a luxury tax, a policy designed to address baseball’s payroll disparity and supposed competitive balance problem.

Here is the basic story the powers-that-be in baseball were telling in 2002:  There is substantial payroll disparity in baseball.  Some teams spend much, others spend relatively little.  And the teams that spend quite a bit have an unfair advantage.  As Bob Costas stated on page 58 of Fair Ball (a quote we mentioned twice in The Wages of Wins)”The fact is, the single biggest indicator of a team’s opportunity for success from one year to the next is whether that team has a payroll among the top few teams in the league. Period.”

Given this view, the luxury tax was needed to restrain the spending of the big-spending teams.  And if this policy was successful, baseball’s competitive balance would be improved.

Before we get to the competitive balance story, let’s look at the disparity of pay in Major League Baseball.  Table One reports – for the American and National League – average payroll, the standard deviation of payroll, and the coefficient of variation from 1988 to 2008. 

Table One: Payroll Dispersion in Baseball from 1988 to 2008

The coefficient of variation – or standard deviation divided by the mean (or average) – is a measure of dispersion used when we wish to compare data sets where the value for the standard deviation is quite different.

For example, in 1993 the National League had an average payroll of $28.5 million with a standard deviation of $10.3 million.  Ten years later average payroll in the National League was $72.9 million while the standard deviation was $22.9 million.  If we only looked at standard deviation, it would appear that the dispersion of pay had increased in the National League.  But if we look at the coefficient of variation, we see values of 0.361 in 1993 and 0.314 in 2003. In other words, if we just look at these two years, spending in the National League was more equal.

Of course, we want to look at more than just two years.  Let’s start in the National League.  From 1996 to 2002, the average coefficient of variation in the National League was 0.346.  From 2003 to 2008, this value fell to 0.334 (despite the fact that coefficient of variation in both 2007 and 2008 surpassed the average value seen from 1996 to 2002). Although we see less payroll disparity in the NL, it’s not a very big drop. 

Turning to the American League, we an average value of 0.408 from 1996 to 2002.  From 2003 to 2008, though, this value increased to 0.511.  Yes, the dispersion of payroll has increased in the American League in the post-2002 era.

If we eliminate the Yankees from the AL sample, the coefficient of variation does drop from 0.397 to 0.380.  Yes, even without the Yankees the AL has more disparity.  But we do see a little bit less after the 2002 agreement.

Of course, the Yankees were not eliminated after the 2002 agreement.  Given that this agreement was designed to limit the spending of this team, we should expect to see spending by New York to change after 2002.

Well, it certainly did change. Just not the way Selig intended.

Here is what the Yankees spent from 1996 to 2002 (both payroll and payroll divided by league average is reported): 

1996: $52.2 million, 1.63

1997: $59.1 million, 1.56

1998: $63.2 million, 1.57

1999: $88.1 million, 1.83

2000: $92.9 million, 1.66

2001: $112.3 million, 1.72

2002: $125.9 million, 1.87

Now here is what this team has spent since the 2002 agreement:

2003: $152.7 million, 2.15

2004: $184.2 million, 2.67

2005: $207.3 million, 2.85

2006: $194.7 million, 2.51

2007: $189.6 million, 2.29

2008: $209.1 million, 2.16

Prior to the 2002 luxury tax the Yankees had a high payroll, but relative to the average in the league, the Yankees payroll was always less than double.  After the 2002 agreement, though, the Yankees have always had a payroll that was twice the average.  This spending has not resulted in another World Series title.  But the spending still continues. 

In sum, the 2002 agreement did not do much to stop the Yankees or reduce payroll disparity across baseball.  Given that this was the intent, Selig’s policy has to be considered a failure.

The Competitive Balance Story in Baseball

Now let’s look at the second half of the story.  Is baseball the most competitive sport?

To answer this question we need a measure of competitive balance.  And although there are many metrics out there, if you want to compare competitive balance across sports (as I did a few days ag0) the Noll-Scully measure is a good choice. Just to review… the Noll-Scully is the ratio of the standard deviation of winning percentage (or standing points in hockey or soccer) to the ideal standard deviation you would see if all teams were equal in strength.  If you look at the major professional sports leagues (baseball, basketball, hockey, American football, and soccer) across time, you see an average ratio of 1.8 (this average is based on more than 600 season observations). 

With this average in mind, let’s look at the history of competitive balance in the American League and National League. Specifically, let’s look at the average level of competitive balance – as measured by the Noll-Scully – for each decade since the start of the 20th century.

1901-09: 2.347 (AL), 3.022 (NL)

1910-19: 2.506 (AL), 2.180 (NL)

1920-29: 2.243 (AL), 2.163 (NL)

1930-39: 2.615 (AL), 2.214 (NL)

1940-49: 2.146 (AL), 2.343 (NL)

1950-59: 2.351 (AL), 1.934 (NL)

1960-69: 1.909 (AL), 2.064 (NL)

1970-79: 1.873 (AL), 1.757 (NL)

1980-89: 1.656 (AL), 1.581 (NL)

1990-99: 1.617 (AL), 1.668 (NL)

2000-07: 2.083 (AL), 1.677 (NL)

When we look over this data we see that competitive balance in baseball generally improved overtime.  We also see that in recent years, competitive balance has worsened a bit in the AL.  This trend can be seen if we look at competitive balance for each of the past 12 seasons.

1996: 1.699 (AL), 1.370 (NL)

1997: 1.526 (AL), 1.424 (NL)

1998: 1.971 (AL), 2.170 (NL)

1999: 1.877 (AL), 1.935 (NL)

2000: 1.310 (AL), 1.723 (NL)

2001: 2.479 (AL), 1.658 (NL)

2002: 2.696 (AL), 2.052 (NL)

2003: 2.507 (AL), 1.768(NL)

2004: 2.106 (AL), 2.203 (NL)

2005: 2.036 (AL), 1.398 (NL)

2006: 1.840 (AL), 1.291 (NL)

2007: 1.694 (AL), 1.326 (NL)

After the games of May 25, the Noll-Scully in the American League this season stood at 1.791 (a mark around average).  In the National League the mark was 1.949 (a mark a bit worse than average).

Okay, let’s put all these numbers in perspective.

In the first half of the 20th century, the actual standard deviation tended to be more than twice the ideal.  In the latter half of the 20th century this ratio tended to be below two.  And then in the first few years of the 21st century we again see marks in excess of two.  In recent years, though, the Noll-Scully in both leagues is back below two.

So what’s going on?  The general trend seen across the 20th century we have discussed before.  In The Wages of Wins (and an article published in Economic Inquiry) we note that competitive balance in baseball is primarily influenced by the size of the underlying population of talent (an argument based on the work of Stephen Jay Gould and also found in The Baseball Economist).  As baseball integrated and looked beyond the United States for talent, competitive balance improved.  Following this reasoning, expansion will make competitive balance worse (an argument also echoed in The Baseball Economist). 

The focus on the underlying population of talent suggests that league policy – such as the reserve clause, the amateur draft, and/or free agency – will not have much of an impact on competitive balance.  In fact, this is what we report in our

Economic Inquiry paper (an article that empirically examined all these issues).

Does this argument explain what we see in the AL from 2001 to 2005? In these years the Yankees and Oakland A’s were very good (only one of these teams spent a great deal of money) while the Detroit Tigers and Tampa Bay Devil Rays were very bad. What explains the performance of these teams?  It’s hard to say that money is the only factor that matters. 

Both Detroit and New York spent more than Oakland. The Yankees, despite spending more than three times what Oakland spent, only averaged three more wins per season.  The Tigers, again despite spending more, averaged more than thirty fewer wins per season.

A sample of five years is quite small and difficult to explain.  But one suspects the difference in the results achieved by these teams is not strictly about money, but really about managerial skill (i.e ability to choose productive players).

Competitive Balance in the Major North American Sports

Selig argues that his policies have caused balance to improve.  When we look at the data – both on payroll and competitive balance in baseball – it’s hard to share his perspective.

Of course Selig was comparing baseball to other sports.  What do we see if we look at recent data from the NHL, the NFL, and the NBA?  As I noted a few days ago, competitive balance in the NBA remains poor.  From 1999-2000 to 2007-08, the Noll-Scully was in excess of two every season, and last year it was in excess of three. 

Across this same time period in the NHL the Noll-Scully was always less than two with an average mark of 1.547 (last year it was 1.037). The NFL has a very similar level of competitive balance.  From 200o to 2007 we see an average of 1.587, with the highest mark only reaching 1.694 (in 2005).

When we compare baseball to these other sports we see that Selig is right if other sports only means the NBA.  But if we look at the NHL and NFL, it’s hard to see that baseball has reached the level of parity we see in those sports.  It’s very close, but I don’t think baseball has surpassed hockey or football. And remember, that’s what Selig argued.  The commissioner argued that baseball had more parity than other sports.  The data doesn’t seem to support that particular argument.

Let me summarize what we have learned.  The luxury tax, designed to reduce payroll disparity, has not really been successful.  Of course, it also doesn’t matter.  It does not appear that baseball had much of a competitive balance problem in 2002.  And it doesn’t appear that it has much of a problem today (and no, you can’t thank Selig for that particular outcome).

What we do see today is that the Yankees – at least at the moment – aren’t winning as much.  Yes, the team is still spending.  But the wins are not happening as often in 2008.  Ultimately, I think, seeing the Yankees fail to win is what makes Selig most happy.  And on that point Selig and I agree.  After all, aren’t all non-Yankee fans happy to see New York out of first place?

– DJ