Field study evidence of subjectivity, bias mitigation, residual bias, and interdependencies in performance measurement and reward systems

This paper explores the roles, uses and impacts of subjectivity in performance measurement and reward systems in a field study spanning four large diverse organizations. We draw on the informativeness principle to frame our theoretical contribution. Our findings, drawn from data collected from both senior and subordinate-level managers, contribute to the literature in several ways. First, we document subjective practices across performance measurement and reward system (PMRS) processes including subjective measures, ratings and rewards, thus providing unique holistic field insights into subjectivity. We find that subjectivity is pervasive, sometimes hidden in seemingly formulaic processes, frequently used in a confirmatory role, and that it is purposeful rather than a residual solution when objective measures fail. Our research design also enables us to identify interdependencies among subjective practices at different stages of PMR processes. We observe subjectivity being used to improve risk sharing, capitalise on the observability of agent action choices, penalise incongruent actions, enhance employee sorting and task allocation and thus move toward better contracting outcomes and HR decisions. In view of the well-accepted risks of subjectivity including bias, cognitive errors and corruptibility, we also document risk mitigation strategies adopted in our participant firms. The firms in our study invest considerable resources in controlling and validating subjective assessments. Calibration emerges as a critical risk mitigation strategy. However, we identify residual biases remaining despite mitigation, including asymmetric rating adjustments, centrality bias created by practices designed to limit leniency, and excessive weight on potentially weakly informative contributions in calibration processes