Heterogeneity and Performance in Tournaments: A Test for Incentive Effects using Professional Tennis Data

Abstract

This paper provides an approach to test whether greater heterogeneity of contestants leads to lower effort exertion in elimination tournaments, as predicted by conventional tournament models. This prediction is difficult to test with real world data because effort is difficult to measure. Based on a simple behavioral model, testable implications are derived and an identification strategy is suggested that allows to test for an incentive effect of heterogeneity even when effort is unobservable. The application uses data from professional tennis tournaments and provides evidence that heterogeneity affects the incentives to exert effort.