Likeability in subjective performance evaluations: does it bias managers’ weighting of performance measures?

Abstract

In this paper, we investigate how subordinate likeability induces bias in managers’ subjective performance evaluations. Based on the affect-consistency heuristic, we expect managers who use multiple performance measures to subjectively evaluate their subordinates’ performance to place greater weight on likeability-consistent performance measures than on likeability-inconsistent measures. Hence, we predict that likeability and performance information interact in affecting managers’ performance evaluations. The results of our experiment support this prediction. In line with prior research, we find evidence of likeability bias in subjective performance evaluations: likeable subordinates receive more favorable evaluations than dislikeable ones. We further find that participants adjust their performance evaluations in the presence of likeability-consistent performance information to a greater extent than in the presence of likeability-inconsistent performance information. Thus, in accordance with the affect-consistency heuristic, our results indicate that likeability bias occurs due to a differential, biased weighting of performance measures. Additionally, we find that perceived likeability is also affected by subordinates’ performance, which in turn partially mediates the effect of subordinate performance on evaluations: good performers are more likeable than poor performers. Hence, this can exacerbate likeability bias. We discuss the implications of our findings for the design of performance evaluation systems in practice.