World Series time! Enjoy Premium-level access to most features through the end of the Series!
April 15, 2010
Ahead in the Count
Labor Market Discrimination
You see guys like Jermaine Dye without a job. Guy with [27 home runs and 81 RBI] and can’t get a job. Pretty much sums it up right there, no? You’ve got some guys who miss a year who can come back and get $5 million, $6 million, and a guy like Jermaine Dye can’t get a job. A guy like Gary Sheffield, a first-ballot Hall of Famer, can’t get a job…We both know what it is. You’ll get it right. You’ll figure it out. I’m not gonna say it because then I’ll be in [trouble].
The fact is that African-Americans earn less than Caucasians in the United States of America in 2010. Blacks earn less than whites who have the same education and experience, work in the same areas, and work in the same sectors. This is a uniform result in basically all economic research on the topic. This wage gap is real, and its impact is large. It is not something that economists or any other individuals should take lightly, but its causes are less than clear.
There are different kinds of discrimination, and it is important to separate prejudice from discrimination. This is not to belittle discrimination, but to define the difference so that we can talk about this topic seriously. Prejudice is an adverse opinion without reason or logic. Discrimination is treating two groups differently. There does not need to be prejudice for there to be discrimination. Most economists believe that the basis for the wage differential between blacks and whites is not based on prejudice as much as discrimination. The reason is simple. Suppose that car dealerships were all run by white racists who wanted only whites to sell their vehicles. Now say there were plenty of qualified black salespeople who were not paid the same amount of money as whites to do the same job, but who were equally capable. This type of situation is difficult to sustain, because all you need is one greedy person to realize that if he builds a car dealership and offers low enough wages that only blacks will accept, he make a ton of money selling cars without paying much in the form of sales commissions and, thus, run the racist dealers out of business. He doesn’t need to be a holy man or even a nice guy. He doesn’t even need to like blacks. He just needs to like money more than he dislikes blacks, and the competition would drop like flies. The free market can free us from these kinds of racists.
There is another more dangerous type of prejudice than employer prejudice, however. That is customer prejudice. When customers are prejudiced, that can create a much larger wage gap. Car dealers could all be enlightened individuals who see beyond race, but if they want to sell cars and car buyers are racists, a problem is created that economic market forces won’t solve. Of course, that is where legal remedies enter. However, proving discrimination is far from easy as evidenced by blacks being paid less than whites for similar work.
There is a third type of discrimination, one without prejudice. That is statistical discrimination. It is also difficult to fight, because the enemy is subtle. If certain traits are seen as more common in one race and are believed to affect productivity, they can be applied reductively, lumping everyone together because race can be used as a proxy for other variables. The human being is prone to categorize, and this is where we get into trouble. This is also difficult to combat, but remains horribly unfair. A major example of this is the wage differential women deal with, whereby women earn less than men doing equivalent work. Some of this, at least, is due to the fact that women leave the labor force more often than men when they have children. Knowing this, firms adjust their investments in long-term employees, favoring those who they believe are more likely to stay. Again, this is illegal, but proving it in court is difficult, suggesting why this type of wage differential persists. Otherwise, women would earn salaries on par with their male counterparts more often.
Taking all this to baseball, we have to consider which of these conditions are the most appropriate to describe the sport's labor market. One of the most unique things about baseball from an economist’s standpoint is that labor productivity is very measurable, far more than in nearly any other industry. This is why many economists believe that baseball is a perfect natural laboratory to study these kinds of things.
For example, take the major leagues before Jackie Robinson broke the color barrier. I mentioned the example of the more-greedy-than-racist car dealer earlier, and how simple employer discrimination can be remedied by market forces. All we needed was somebody to come in and open a new firm and…
That’s where the problem lay. You can’t just build a new baseball team and put them up against the Red Sox, can you? There were actually people trying to build baseball teams of great black players, many of whom were superior to their white counterparts, but were forced to play in the Negro Leagues because the major leagues banned them. Without the market force of free entry, there is a coordination that can take place between owners. Fortunately, Branch Rickey did his part to correct this, and signed Robinson for the difficult task of breaking the color barrier.
This is where Robinson had to face a new kind of problem mentioned earlier, which is customer discrimination. Angry white racists were not happy to see Robinson in the National League. As a result, bringing in Robinson cost owners a chance to make money from those fans who boycotted attending games as a result of his signing. Even if adding Robinson to the roster also added to their win column, it was a risk for their bottom line, but the Dodgers were willing to take it. This is where Robinson’s role was so crucial: because both Rickey and Robinson were aware that people would stereotype “black ballplayers” based on the small sample size of one, it was essential that Robinson show the strength not to respond, to just play. He ignored critics and refused to fight. Racists fans, umpires, opposing players—he had to rise above all of it, and he did just that, something few people in the world could have done. This combated customer discrimination that could have kept blacks out of baseball longer. Remember that Robinson debuted in 1947, 17 years before the Civil Rights Act was passed.
Part of the beauty of baseball, as I mentioned earlier, is that productivity is far more measurable than in other labor markets. It was easy to see Robinson him win the batting title in 1949 with a .342 average, and to count his 37 stolen bases that year. That was his production, and it was easily isolated and valued. It made it much harder to discriminate against him in baseball as it might in other industries.
The question is whether there is customer discrimination against Dye, because employer discrimination just is not going to cut it as an explanation. It is pretty clear that it would be next to impossible to coordinate an effort on the part of all 30 teams to consciously discriminate against black players, especially with some many teams looking for a competitive edge. However, if customers prefer white players to black players enough that it hurts the owners at the gate, then that would be another story. Fortunately, baseball profit is driven heavily by making the playoffs. As the playoffs expanded after the 1993 season to allow for more home post-seasons games, the value of advancing to the postseason grew even larger. Any racism among fans would have a hard time outweighing the flush of cash that would come with a playoff appearance. However, it does not need to outweigh this effect. It just needs to have an effect to change salaries in an unfair way.
The other possibility is statistical discrimination, as mentioned earlier. If being black is perceived to be correlated with other traits that affect productivity, then it risks being used a proxy for those traits if they are difficult to measure. This is where the Sheffield allusion by Hudson is particularly interesting. It seems as if the media perception of Sheffield is that he is a bad guy. Right or wrong, that is the common attitude towards him. Sky Andrecheck has studied the extent that teams do pay attention to perceived clubhouse effects, and treat them as though they affect value by about 1.5 wins at the extremes. If black players are more likely to be identified with this negative clubhouse effect than white players, even if it is perceived rather than real, that is where African-Americans could be getting treated unfairly in baseball’s labor market.
At this point, most studies have shown little evidence of discrimination. J.C. Bradbury highlighted a number of these on his website yesterday. Of course, there could be a lot of information missing that leads to a false negative in these tests. For instance, if race is correlated with position and certain models overvalue or undervalue the contribution at that position, there could be a lot of error that creeps into that analysis.
That is why the question should be a theoretical one as well as an empirical one: both approaches remain important.
To me, the real question comes down to whether black players’ clubhouse effects are treated differently than those of white players. If Brett Myers gets a mulligan for domestic violence, while Elijah Dukes does not, that is where racism could enter into the equation. Clubhouse effects are the murky area of baseball where productivity is harder to measure, and it becomes more plausible that general managers might not notice their subconscious biases. If black players are underpaid relative to their contribution, I doubt that such a tendency would be true across the board because of all the reasons listed above, but it is certainly plausible that white malcontents and black malcontents are treated differently.
Consider the July 2009 Forbes’ article listing the “The 10 Most Disliked People in Sports”, in which the top eight were not white. Looking through the list, it’s not really all that surprising, but that’s not the point. I don’t really see any problem with people taking issue with Michael Vick, but my question is whether there are white people getting away with bad behavior who are not disliked in the same way. That would not only be related to statistical discrimination but to customer discrimination, black malcontents could be forced to take lower salaries if fans disliked them in advance while white malcontents could get salaries commensurate with their on-field production.
As far as Dye is concerned, he is not a player with a reported track record of “poisoning clubhouses,” and so in his case, I have a hard time believing that racism is playing a role. In his case, I suspect that the fact that UZR reports his fielding has been 61 runs below average over the last three years, while the Baseball Info Solutions’ Defensive Runs Saved for Dye in that time frame is negative 46 runs is working against him. Dye is best employed as a designated hitter at this stage of his career, and with PECOTA projecting a .262 TAv for 2010, it is likely he not getting contract offers to his liking because he is essentially not far above replacement level.
However, the question Hudson raised does not need to be limited to looking at one ballplayer, and it is an important one. Looking at baseball can teach us all kinds of things about economics because productivity is so well documented. The wage differential could be yet another economic lesson learned from the sport.