http://www.baseballprospectus.com/article.php?articleid=72


Who benefits from baseball's arbitration process?

by Mark Armour
March 11, 1998



 Each year, players in their third year of service wait anxiously after the
season. Their hope is that they will land in the top 1/6 of players in their
year of service for time spent on the major league roster. If they do, they
are rewarded with arbitration rights. Indeed, this past winter Mark
Grudzielanek, of the Expos, bemoaned his fate of just missing the line, after
a one-day pout last spring over having his contract renewed.

 As most readers are aware, the market for players is biased heavily towards
the team in the first two years. It used to be like this for all players
under the reserve clause. The player sold his talents in what economists call
a "monopsony", where there is only one buyer. In free agency, in contrast,
there is a virtual monopoly held by the best players -- there's only one
Clemens or Bonds, and all teams would love to have that superstar.

 Arbitration lies in a middle ground between these extremes. Each side has to
present a final offer, and an arbitrator or arbitrators (this year they used
three-person panels) must decide which offer is taken. Nearly all contracts
are decided before the arbitrator picks a number. But clearly, players in the
fourth year of service get some benefit, so much so that owners offered to
allow players to become free agents two years sooner if the players' union
agreed to give up arbitration rights. The players refused.

 Another attribute of less monopsonistic markets for players is that they are
supposed to lead to less discrimination. An important paper by Donald Cymrot
about ten years ago showed that minority players who were eligible for free
agency rights received the same return per point of offensive average (or OA;
total bases plus walks plus steals over at bats plus walks) as did white
ballplayers, but whites earned almost 70% more per point of OA than minority
players when neither had free agency. (footnote 1)

 A colleague here at St. Cloud State, Orn Bodvarsson, and I have spent some
time studying whether Cymrot's results could be extended to arbitration.

 Our latest paper (footnote 2) provides the most comprehensive evidence we
can find on whether discrimination is lessened by salary arbitration. And the
answer is relatively shocking: If there was discrimination during the Cymrot
study, we cannot find it any more. We no longer find any discrimination for
minority hitters. If anything, non-white hitters make more per unit of
production (we use slugging average, walks and steals separately rather OA)
than do white hitters. For pitchers, there's a suggestion of discrimination
against non-white pitchers, but the probability of this is about 88%. In
other words, there's a better than 10% chance we found discrimination where
there really isn't any. So the best I can say is that there
kinda-maybe-is-a-little-bit of discrimination against minority pitchers, and
wait for someone with more and better data to do a better study.

 Here, however, are some more results that we can issue with some certainty:

We measured the degree to which a player was underpaid or overpaid in some
year compared to his peers in that same year. This would be the difference
between the predicted salary from an equation relating salary to performance,
and the player's actual salary. We guessed that players who had predicted
salary above the actual would be more likely to file. This turned out not to
be true. However, we did find that non-white pitchers filed more often than
white pitchers.

The threat of filing (i.e., our estimate of the probability that a player
would file) was correlated with salary the player received to play the
following season. The higher the player's threat to file, the higher salary
the player received.

There was some premium received if you actually did file. Atop the payoff for
a threat to file, there was a 20-30% premium paid to those players who
actually did file. This likely accounts for the high number of players who
did file, even though they subsequently settled with the team without going
before the arbitrator.

Combining those points, even though we could not find directly that
discrimination occurred against any players, minority pitchers tended to file
more, and filing more made it more likely those players would receive larger
raises in their arbitration years.
 That players get a big raise in their arbitration year is not surprising. By
the time the third year ends, we know pretty well if the player will have
some value. Those that do not are not going to be given a chance to visit the
arbitrator -- they become the flotsam and jetsam that occupy the waiver wires
or get non-roster invitations each spring. The huge gains many report for the
"average" fourth-year player is bogus in that a big part of the "average" is
no longer playing.
 It's also not surprising that players get some return on actually filing for
arbitration. The process is not fun. There are lawyers and expert consultants
to pay and the psychic cost of having your employer rip you in the hearing.
Some general managers prefer to pay a little extra to a player as a good will
measure to keep them from the process. It probably also helps come free
agency.

 So what are the value of arbitration rights? Remember, the players were
offered two years of additional free agency years (years 5 and 6 of service)
in return for scrapping arbitration in the last protracted contract
agreement. This was the one point the players were adamant in refusing. Since
my sorting argument two paragraphs ago would occur both in arbitration and in
free agency, the sorting argument cannot be the reason why players prefer
arbitration. My guess is that the way arbitration settlements are made is
quite favorable to ballplayers, and it will take some very large concessions
from the owners to rid themselves of this creation.

Footnotes:

1. Donald J. Cymrot, "Does Competition Less Discrimination? Some Evidence,"
Journal of Human Resources 20, Fall 1985, pp. 605-12.
2. "The Value of Arbitration Rights in Major League Baseball: Implications
For Salaries and Discrimination," Quarterly Journal of Business and
Economics, forthcoming Winter 1998. See also Banaian and Bodvarsson, "Reverse
Regression and Salary Discrimination in Major League Baseball: The Case for
Discrimination," in Alvin Hall, ed., Cooperstown: Symposium on Baseball and
the American Culture. Westport, CT: Meckler Publishing, 1990.