Believe Me, You are (Not) that Bad

Abstract

This paper studies the effect of incentive schemes incorporating status classes on workers' performance. I focus on performance comparisons between similarly skilled workers that belong to different status classes. A theoretical framework predicts that, under certain conditions, low ability workers attain high performance when they are assigned to a high rather than a low status class, and that high ability workers achieve high performance irrespective of the received status. These predictions are tested in a laboratory setting, where subjects are randomly assigned to a high status or a low status condition and constant performance feedback is provided. The experimental data support both predictions: low ability subjects assigned to the high status condition outperform their low status counterparts by 0.53 standard deviations in a cognitively challenging task, and high ability subjects display high performance outcomes in both status classes. Moreover, I explore the subjects' beliefs about performance as a mechanism to explain these results. I find that low ability subjects assigned to the high status exhibit performance targets that were as high as those elicited by high ability participants. This suggests that these workers used status to believe that they were good performers, and performed accordingly.JEL Classification : D03, C91, D84, M54, Z13