How Does Team Composition Affect Effort in Contests? A Theoretical and Experimental Analysis

Abstract

When companies organize salespeople into teams to compete with one another, how does the ability composition of team members affect effort? And how may the effect of team composition on effort depend on the type of contest the teams are competing in? Using a game-theoretic model, we derive the very sharp predictions that when winning the contest depends on the average team sales, the ability composition of a team has no effect on team effort, and the stronger and weaker members will expend identical effort levels. However, when the contest winner is determined by the minimum or maximum output contribution within the team, heterogeneity in team composition exerts a deleterious effect on team effort, with the stronger and weaker members expending unequal effort. Our model also shows that in a contest between two symmetric teams, when team members are homogeneous, all three contest metrics yield identical team effort; however, when team members are heterogeneous, the average metric elicits the highest team effort. We also discuss cases when the minimum and maximum metrics are optimal. We test the model's empirical validity using two incentive-aligned experiments in which participants make effort decisions strategically. The experimental results exhibit broad support for the theoretical predictions.