Tournaments as coordination devices: Theory and experimental evidence

Thomas C. Schelling proposed that individuals can coordinate more successfully than conventional models of coordination predict if they use payoff-irrelevant features to label strategies. However, recent experimental evidence illustrates that salient labels may lose much of their effectiveness when the payoffs are asymmetric. In this paper, we consider an asymmetric coordination game with two inequity-averse players. The model features a preplay tournament in which the two players compete to justify their future dominance. We argue that, after observing the result of the tournament, both players are more willing to select the label-salient strategy because the inequality in equilibrium is justified by the performance gap in the tournament. To test this prediction, we design an experiment that allows us to adjust the justifiability of inequality. The experimental results demonstrate that a preplay tournament may constitute an effective focal point that increases the likelihood of coordination.