As one team after another gets eliminated in the baseball playoffs, no doubt many fans are grousing that certain strikeout- or error-prone players are getting paid too much. New research suggests one likely reason for this disconnect between an athlete's paycheck and performance.
In a paper just published in the Journal of Sports Economics, economist Andrew Healy of Loyola Marymount University in Los Angeles uses Major League Baseball as “an ideal laboratory for testing theories about compensation” — specifically, the theory that salary offers are warped by “memory-based biases” on the part of employers. Analyzing salary and performance data for all MLB hitters who signed free-agent contracts from 1985 to 2004, he concludes salaries are disproportionately affected by how well a player has performed in recent months.
“Teams reward players for performing well in the immediate past, ignoring other evidence of a player’s quality from his earlier performance,” he writes. “When choosing salary offers, teams have short memories.”
Healy’s crunching of the numbers finds that looking at a player’s performance from two and three years ago is a significantly better predictor of future performance than how he did in the most recent season. Salaries, however, tend to reflect that more recent history.
“Teams are not equally prone to underweighing earlier performance relative to recent performance,” he adds. “Controlling for total payroll, the teams that win the most games use past performance data most effectively.”