ABSTRACT
In a setting involving a firm using multiple performance measures, we experimentally examine the effect of an evaluator's perspective on the relative performance evaluations of two managers. The performance outcomes of the two managers are linearly equivalent, but one manager's performance includes an equivalent number of above target and below target outcomes (e.g., the mixed manager), while the other manager's performance only includes above target outcomes (e.g., the positive manager). In this setting, we provide new evidence on the negativity bias and whether the bias is moderated based on evaluators' role and the importance of the measures with negative outcomes. Participants are assigned to the role of the supervisor, the mixed manager, or the positive manager and asked to evaluate each manager's performance. We predict and find that participants in all three roles exhibit the negativity bias. In addition, we predict that the strength of the negativity bias exhibited by mixed or positive managers relative to the supervisor's negativity bias depends on whether the measures with negative outcomes are more or less important. As expected, we find that when negative outcomes involve less important measures, the relative performance evaluations of mixed (positive) managers are similar to (differ from) those of supervisors. In contrast, when negative outcomes involve more important measures, the relative performance evaluations of positive (mixed) managers are similar to (differ from) supervisors. Understanding whether and when managers' relative performance evaluations differ from their supervisors is important, in part, because conflicts and potentially dysfunctional behavior are likely to arise when their relative performance evaluations differ.
