Abstract
We develop a general theoretical model of the effect of wage dispersion on team performance which nests two possibilities: wage inequality may have either negative or positive effects on team performance. A parameter which captures the marginal cost of effort, which we estimate using game-level data from Major League Baseball, determines whether wage dispersion and team performance are negatively or positively related. We find low marginal cost of effort; consequently, wage disparity is negatively related to team performance. Game and season-level regressions also indicate a negative relationship between inequality and performance. We discuss a variety of interpretations of our results.
Acknowledgements
The authors thank Deborah Cobb-Clark and José Rodrigues-Neto and seminar participants from universities around Australia for their comments. The newly formed Australian Baseball League, the Canberra Cavalry and Sarge kept us going on this topic through the hot Australian summer.
Notes
1 Lazear (Citation1989) adapts the model of Lazear and Rosen (Citation1981) to allow workers to increase their probability of winning in two dimensions by introducing sabotage.
2 Scully (Citation1974) argues similarly.
3 Avrutin and Sommers (Citation2007) are an exception to the general rule. They find that inequality is not a significant predictor of team winning percentage using the Gini coefficient but analyse a very short data window, 2001–2005.
4 with constant k across players.
5 Montreal Expos (1969–2004) and Washington Nationals (2005–present) are treated as the same team, as are Angels of California, Anaheim and Los Angeles.
6 These details and the data are available from the authors. Link to be provided on acceptance.
7 (20) where
is the salary of player j and w is the team wage bill.
8 Full results of these regressions are available from the authors.
9 Richards and Guell (Citation1998) look at whether teams attempt to maximize attendance rather than winning, but attendance is only one small part of team revenues.