Heterogeneity in tournaments with incomplete information: An experimental analysis

We focus on the relationship between a player's effort provision and tournament heterogeneity in a setting where players only know the distribution of their opponents' abilities. By isolating whether increases in heterogeneity influence optimal effort provision in cardinal, ordinal, and piece rate tournaments, we show that a model in which ability and effort are complements can be empirically distinguished from a model in which ability and effort are neither complements nor substitutes. To discriminate between the two models, we conduct a laboratory experiment where subjects participate in a real effort task and are paid based on performance relative to a group of opponents that may be relatively homogeneous or relatively heterogeneous. In these laboratory data, the level of effort provision is independent of tournament heterogeneity, lending support to the model in which ability and effort are neither complements nor substitutes.