An Experiment on Inequality within Groups in Contest

We study Tullock contests allowing heterogeneity of both rewards and abilities within the competing groups. Our main concern is whether higher degrees of inequality in a group can improve its performance, namely group effort and probability of winning. First, we show that the answer to this question is positive under plausible conditions on players’ cost function. Second, we test these predictions in the lab. Unlike theory predicts, inequality in abilities does not help a team win. Inequality in rewards does help but moderately. The efficient combination of both inequalities, which assigns high rewards to high ability players, substantially increases a team’s performance. Finally, through the analysis of subjects’ beliefs, we provide empirical evidence that overbidding is more severe than we expected