The Influence of Social Identity on the Performance Effects of Relative Performance Evaluation

Abstract:     
This paper examines the influence of social identity on the performance of auditors compensated using a relative performance evaluation (RPE) plan. We posit that when RPE is used in a peer group with high social identity, group members are likely to perceive the effect of RPE in reducing common error. On the other hand, when RPE is used in a group which has low social identity, group members are not only unlikely to perceive the effect of RPE on the common error, but they are also likely to perceive RPE as adding to the total error due to increased salience of the idiosyncratic error. We use proprietary and survey data on performance measures and outcomes for 145 junior and senior auditors from a large internal audit organization to explore the effect of social identity on the performance effects of RPE. This organization introduced a pay-for-performance incentive plan in which supervisors engaged in RPE to varying degrees. We develop arguments for the presence of higher social identity for the junior auditors and lower levels of social identity for the senior auditors. Results indicate that, consistent with our predictions, the performance effect of using RPE in the presence of high levels of uncertainty is positive for junior auditors, but negative for senior auditors. Our results demonstrate the importance of understanding the influence of workplace social characteristics on the effectiveness of RPE.

Number of Pages in PDF File: 43

Keywords: Relative Performance Evaluation, Social Identity, Incentive Contracting

working papers series