The Baseball Economist: The Real Game Exposed (17 page)

Whenever an individual takes into account the behavior of other individuals before acting, economists classify the behavior as “strategic.” We see players making strategic decisions frequently on the field. For example, consider the strategic interaction between a base runner and a pitcher. If the base runner sees that the pitcher has a slow delivery, he may attempt to steal. Fearing a steal, the pitcher may speed up his delivery and throw a fastball to the catcher to prevent a steal. Each player is making a strategic response to the other. Similarly, players may make strategic decisions off the field based on the behavior of their colleagues.

Because many choices in life involve strategic behavior, economists have developed models to explain decisions governed by strategic incentives. Consider two players of equal ability in a world in which they are only two baseball players. In the labor market for players, each of these players will command the same salary, equal to half of the total sum of money owners are willing to pay players. However, if one player becomes better than the other, the better player will garner more of that sum, in proportion to the improvement in his performance. Each player’s performance affects the income of his colleague. If they both sit around drinking beer and eating doughnuts, they will earn the same amount of income as if they both worked out every day. The only way to earn more money is to be
relatively
better than the other player. Harvard economist Robert Barro describes why this is the case:

To a considerable extent, a team’s or athlete’s output is measured not so much by absolute skill—how far a ball is hit or how fast a race is run—but by comparisons with the skills of other performers. How much difference would it make if the longest home run went 600 feet or 300 or whether 100 meters could be run in 8 or 10 seconds? These numbers matter mainly in relation to what other athletes can do (now and in the past).
⁵⁵

Given the monetary incentive to be relatively better than other players, both players will work as hard as they can only to find themselves earning the same amount before they exhibited the effort. The incentives encourage the players to bring the best possible quality of baseball to fans.

Assuming that steroids improve performance, introducing steroids as an option for improvement changes the contest. If one player starts using steroids and his performance improves, not only will his income rise, but the income of the other player will fall. In order to keep up— because otherwise the labor market will value his natural services less— the formerly clean player must begin taking steroids too. Using steroids creates an externality because the actions of a player spill over onto other players. However, both players end up back where they started: even in skill and income. Except, if steroids have negative health consequences, then both players will be worse off than in a world without steroids. How is this possible?

The incentives of this strategic game lead both players to make decisions that are sub-optimal—meaning the situation could be improved—and neither party can improve the outcome by unilaterally choosing not to take steroids. This is a curious outcome that results from the fact that if one player decides not to take steroids, it is in the best interest of the opposing player to take them, because doing so will increase his income. Also, if the other player chooses to take steroids, the competing player must also take steroids to avoid losing income. A common tool economists use to predict human behavior reveals why this happens.

It is called
game theory
. Each player in the game makes choices based on payoffs that are affected by the behavior of the other. A matrix of payoffs, such as those illustrated in Figure 9, models the incentives for action. The numbers in the boxes make it easy to see the consequences of a choice given what the other party chooses. The payoffs below represent hypothetical monetary rewards (in millions of dollars) to both players for the four possible outcomes of the steroid game:

This is a symmetric game, which means that both players receive the same payoffs for the same choices. This is a realistic assumption for two players with similar skills: each ought to suffer and gain equally from the other’s steroid use. In a world without steroids, both will receive $2 million on the labor market. If one player decides to use, but the other does not, the using party will gain from superior performance,
while the abstaining player will lose income. If both use steroids, neither player will be better than the other, thus their salaries will be the same as in the world without steroids ($2 million). However, the difference between the Use/Use and Abstain/Abstain outcomes is that the players will suffer health consequences of using the drugs that may include minor discomfort, future medical bills, and shortened life span. For this example, I’m guessing the cost to be $500,000. If there are costs to using steroids, then both players would be worse off in a world with steroids than in one without, even though they end up receiving the same salary from their teams.

The payoff matrix is an easy way to observe the incentives for use based on what the other player chooses. The arrows in the matrix point toward the optimal choice for each player given what the other player chooses. Player A’s choices are represented vertically. Player B’s choices are represented horizontally. If Player B decides to use steroids, Player A must choose between earning $1.5 million (if he uses) or $1 million (if he abstains). Therefore, we expect Player A to use when Player B uses. If Player B decides to abstain, Player A must chose between earning $4 million (if he uses) or $2 million (if he abstains). Again, we expect Player A to use when Player B abstains. Because this is a symmetric game, players face the exact same payoffs to their decisions. The matrix shows that no matter what the other party does, each player will always be better off using steroids.

This particular game theory model is sometimes known as the “prisoner’s dilemma,” because it reflects a game criminal prosecutors sometimes play with suspects. “If you sell your partner up the river, you’ll get a deal. But if you stay quiet, and your partner gives you up, I’ll put you in jail for the maximum.” Both suspects know that whatever the other does, each is better off confessing. The prosecutor benefits from the deal, because he does not have to waste resources taking the case to trial. The prosecutor relies on the fact that some criminals will follow the incentives to rat each other out. The same incentives may induce baseball players to use steroids.

In this game, using steroids is a dominant strategy, meaning that no matter what the other player chooses to do, using steroids is the best strategy. This strategy is also a stable outcome, or equilibrium, from which no player has a reason to deviate. This particular state is a
Nash equilibrium
—named for the Nobel Prize–winning mathematician/ economist John Nash, who is the subject of the book and movie
A Beautiful Mind
—which occurs if no party has any reason to deviate from a strategy as long as other parties do not deviate. It just so happens that both of the parties in the game would prefer the no-steroid outcome to the steroid outcome, because of the health consequences of the latter choice.

Expanding the model to include many players worsens the situation. If better players earn more income and fame than worse players, then all players have an incentive to take steroids. Even the very best players may have to use steroids to prevent the next-best group of players from outcompeting them. If there is a fixed pot of fame and income to go around, this is a suboptimal system. Marginal players can become good, good players can become stars, and stars can become superstars; yet, relative to one another, players are still the same and will be rewarded according to their relative abilities. Players suffer the negative health consequences of steroids and receive nothing in return. All players have an incentive to sacrifice their health to improve performance, but players are not rewarded for their improvements. Even players who abstain from steroids lose, because they suffer lower incomes as steroid-aided players with less talent improve.

Another interesting result of this model is that players have an incentive to violate informal agreements not to cheat. If two players realize the harm of this game, then the players may reach an “I’ll quit if you quit” deal. However, the incentives of the game are such that if one player agrees to abstain, then the other player has an incentive to cheat. The abstaining player is then left with what economists often call “sucker’s payoff.” By getting your competitor to think you are cooperating, you can win big. The big gains from cheating, and the fear that the other party may hurt you by cheating, cause informal agreements to cooperate to break down. The more parties involved in the agreement—over a thousand different players play in the major leagues every year—the more likely it is that parties will cheat. This is why an externally enforced ban on steroids that includes testing and punishment seems to be the only solution.

If banning steroids is a good thing for the players, why has the MLB Players Association fought so hard to limit testing? From the analysis above, it is clear that players ought to want to stop the use of steroids. And strict testing with sanctions seems to be the only practical way to police their use. However, in the media, the players have ended up bearing most of the blame for the whole steroids mess. I think the case against the players is a little overblown. First, the owners have an incentive to keep steroids in the game, which gives them a reason not to push hard for testing. Second, players may be resisting testing for reasons other than wanting to keep steroids in the game.

Though the relative salary differences between players ought to be roughly the same—with superior players making more than inferior players—the overall play of the game can improve with the ratcheting up of individual performances. When the athletic abilities of all players improve, there will be more spectacular plays in the game: diving catches, hard-hit balls, faster pitches, and close plays on the base paths. If the overall level of the game rises, then this ought to create more fan interest, which generates financial gain to owners.

It is possible that owners, who plead that it’s the players who have fought off testing programs, have been covertly hesitant to institute an effective steroid testing program. There is a strong positive correlation between winning and revenue. An owner who can convince his players to take steroids can reap the returns from winning extra games. For most of baseball history, owners have been in a great position. Steroids have been prohibited in baseball, but with toothless punishment and weak detection. An owner certainly could not openly encourage his players to take steroids, but he could set a precedent for players who visit the right “doctors” by giving implicit bonuses, such as avoiding salary arbitration—the process by which owners and players agree on salaries for the upcoming season—more frequently. Marvin Miller, the famed MLBPA leader who opposes testing, indicates that this was the owners’ attitude toward amphetamines in the 1970s:

In most locker rooms, most clubhouses, amphetamines, red ones, green ones, etc., were lying out there in the open, in a bowl, as if they were jellybeans. . . . They were not put there by the players, so of course there was no pressure to test. They were being distributed by ownership. I can’t remember ever having a proposal from the owners, that we’re going to have random testing or testing of any kind.
⁵⁶

But this leads to the obvious question of why the owners seem to be for testing and the players against it. Are owners looking out for the integrity of the game? Yeah right, the same guys who threaten contraction and want to put Spider-Man on the base paths. (I actually find nothing objectionable in either of these proposals, but I am clearly in the minority on this.) Are the owners concerned about the health of the players? Especially considering that the positive effects of steroids occur in the present, and the negative effects occur in the long run?

The owners ought to bear a larger share of the public blame than they have received for the lack of testing, but the player resistance is not just owner propaganda. The MLBPA has publicly opposed drug testing for most of its existence. Maybe the players are principled civil libertarians taking a stand, but I doubt it.

One possible reason for players’ reluctance to allow drug testing has gone largely undiscussed in the midst of the steroid debate. A urine specimen given over to test for steroids can be used for other purposes. Urine contains not just information on the use of steroids, but also other health-related information. Players would prefer to keep some of this information to themselves. I suspect there is a higher than average incidence of THC—the intoxicating chemical in marijuana—in the bodies of major-league players. This is a cohort of men in their twenties and thirties who spend large amounts of time on the road trying to kill time where there is a lot of partying going on. It may not be quite rock star level—these guys do have to play baseball—but I bet it’s way higher than the average American’s. Players most certainly like to keep their partying levels a secret from owners. Pot is not performance-enhancing, but it probably makes being on the Devil Rays at least funny. With the knowledge from tests for performance-enhancers, owners can learn about the bad stuff too.