Yglesias on Payroll and Wins

Matthew Yglesias has posted a comment on an article by Greg Johnson in the October 4th issue of the LA Times. Johnson’s article, entitled “Seriously, Money Isn’t Everything”, was also noted in this very forum. The argument advanced by Johnson, which builds upon quotes from noted sports economists Andrew Zimbalist and Rod Fort (as well as some economist from California named Berri), is that payroll and wins have a surprisingly weak relationship in baseball.

The weak link between payroll and wins is not only observed in baseball, but as noted in The Wages of Wins, also in basketball and football. All this leads one to ask why it is the case that teams that spend the most in the three major sports don’t generally win the most.

Yglesias offers this argument: “Money differentials do matter, accounting for about 20 percent of the variance in MLB win-totals, but they don’t really matter all that much. And, well, good for A’s fans. Nevertheless, it’s hard to avoid noticing that this implies that baseball teams are, as a rule, managed extremely poorly. Absent a salary cap, the teams with the most money to spend ought to be able to perform much better than the teams that spend much less. Moneyball is all about Billy Beane’s ability to use statistics to identify some player-attributes that were undervalued in the baseball labor market and assemble quality teams at bargain-basement prices.”

“That strategy depends, however, on significant inefficiencies existing in the first place. And the inefficiencies need to be quite large, since the payroll differentials are enormous — the Yankees’ salary total is 2.7 times the median. On top of that, you’re looking at a kind of meta-inefficiency. The owners prepared to shell out the highest amount for players ought to be able to also outbid stingier owners for the services of the best GMs, scouts, managers, etc. In a perfect environment, in other words, differential management-ability to entrench, rather than mitigate, payroll-differentials. That all these inefficiencies exist winds up being good for the fans, but it’s still rather odd.

…Part of the story here has to be the simple fact that team revenue is only very vaguely associated with team success on the field. Even the Nets haven’t been able to simply poach the Knicks fan base merely by fielding a vastly superior team over a period of years and most teams are even worse-positioned to do so. Still, I don’t really see what explains the overall situation . . . owners still do have a clear financial interest in fielding winning teams, but they genuinely don’t seem to be trying very hard. Perhaps the problem is that major pro sports teams essentially never go out of business no matter how bad they get.

To summarize, Yglesias is arguing that payroll differentials indicate that the labor market in sports is plagued both by inefficiency and a weak link between revenue and wins. In other word, teams do not fully understand the link between player action and wins and since a fair amount of revenue is guaranteed, teams have little incentive to fix this problem.

In The Wages of Wins we offer a different explanation. In football, where the payroll-wins link is the weakest, we see a great deal of inconsistency in player performance. This point in made in Chapter Nine of our book, and also previously in this forum, as well as in the New York Times by both Martin Schmidt and myself. Player performance in football depends crucially on the actions of the other players on the field, as well as the coaching staff. Consequently it is very difficult for decision-makers to forecast exactly how a given free agent or draft choice will perform in future seasons.

In baseball Yglesias has focused specifically on the Moneyball hypothesis. Yes, as Jahn Hakes and Raymond Sauer have noted in the current issue of the Journal of Economic Perspectives, there is evidence that on-base percentage was undervalued in baseball’s labor market. But Hakes and Sauer also argue that in 2004 this problem was resolved. Given this work it is not clear that the Moneyball hypothesis explains the weak link between wins and pay.

In The Wages of Wins we argue that the same story we advance for football also applies to baseball. Although baseball players are more consistent than football players, there is still a fair amount of variation in performance. Exhibit A might be the Detroit Tigers in 2006.

When we turn to basketball, though, we see a different story. Relative to their counterparts in football and baseball, basketball players appear to be much more consistent from season to season. Yes, player performance can change in the NBA. It just doesn’t seem to happen as often.

Despite this consistency, though, we still see a weak link between pay and wins. To explain this weak link we turn to the subject of player evaluation. Player salary in the NBA is dominated by how many points a player scores. And it does not seem to matter much if the player scores efficiently, or if he rebounds, generates steals, or avoids turnovers. Basically if a player can score, he will score a major payday. One should note that this same pattern is also seen in the coaches’ voting for the All-Rookie team.

The best example of the inordinate focus on scoring is the New York Knicks. In previous posts [Larry Brown or Isiah Thomas and Isiah Thomas Discovers Role Players] I detailed how Isiah Thomas assembled a collection of expensive scorers last year only to find that scoring alone cannot win basketball games.

So player evaluation may not be perfect in the NBA. Why do these inefficiencies persist? It is important to note that wins definitely matter to decision-makers. In Chapter Five of the book we note that team wins do have a significant impact on gate revenue in the NBA. And although I am too lazy to look up the studies on a Saturday, I know sports economists have shown a similar relationship in baseball.

So teams do have an incentive to win in professional sports. But I still think it is difficult for decision-makers to adopt new information. If you make decisions according to the conventional wisdom and lose, future employers might say you just had “bad luck.” If you go against conventional wisdom and lose, though, then future employers will conclude you are not very good at your job. Consequently, to protect the possibility of continued employment in the future, one might expect decision-makers to be slow to adopt new information.

As Hakes and Sauer found, on-base percentage was undervalued before 2004. The statistic, though, existed for decades. It took the Oakland A’s success to convince others to change their thinking. Such a pattern suggests that although finishing first is important in sports, when it comes to making changes to the decision-making process, too many managers in sports would probably prefer to go second.

– DJ