Contests for Ranks: Experimental Evidence

We use experiments to analyze multiple dimensions of the relationship between rank incentives and individual performance. In our experiment (i) rank is defined as subjects' relative position in their group based on their performance in a real effort task and (ii) subjects' earnings are independent of their performance. We find that any rank incentive improves mean performance than no rank incentive, and this result is independent of the group size. In the large group, the mean performance increases strictly in all except at the highest rank incentive, but in the small group the mean performance increases weakly in rank incentives. Finally, the mean performance is significantly higher in the large than in the small group because of a higher prestige effect. In additional treatments in which we do not reveal the identity of the status-prize winners, we find that average performance is identical to that in the baseline treatment without any status prizes. The last result signifies the important role that public revelation plays to enhance the strength of status. The results are important for managerial practices.