When should managers use discretion?

Subjectivity in compensation contracting is defined as judgment based on personal impressions and feelings that cannot be verified by a third party. These typically appear in incentive plans through subjective performance measures, flexible weighting of objective data, or ex post discretionary adjustments. Firms utilize subjectivity to achieve five primary benefits: mitigating incentive distortions by capturing hard-to-measure contributions, reducing the risk premium demanded by employees by filtering out uncontrollable external noise, inducing adaptive behavior to unforeseen circumstances, limiting the gaming of objective systems, and restoring perceived fairness to maintain morale. However, these benefits are balanced against substantial costs, including the risk of supervisors reneging on unverifiable promises, rating inaccuracies caused by favoritism or cognitive biases like the halo effect, and the deadweight loss of influence activities where employees focus on ingratiation rather than productive work. Additionally, subjectivity can create uncertainty about measurement criteria and, in environments with weak governance, may serve as a vehicle for rent extraction by powerful managers who seek to mask excessive pay behind an illusion of performance-based compensation.

Managers use discretionary adjustments to objectively determined compensation to account for uncontrollable events, balancing the need to reward current effort with the intent to signal future expectations. When the likelihood of future negative events is high, managers are less likely to adjust current bonuses because they wish to motivate adaptive effort, signaling that employees should take proactive steps to minimize the impact of such events in the future. In contrast, when future event likelihood is low, adjustments are more common as they serve to reward conventional effort and improve fairness perceptions. This strategic use of discretion is significantly curtailed by high compensation interdependence, such as shared bonus pools, where managers fear that adjusting one employee’s pay will demotivate unaffected peers due to the zero-sum nature of the rewards. Consequently, managers are often least likely to use their endowed discretion in the very environments, high-risk or highly interdependent, where performance measures are the most noisy and susceptible to uncontrollable factors. Furthermore, individual managerial attitudes, such as a strong belief in the controllability principle or a preference for purely objective systems, act as significant fixed effects that influence the final decision to adjust pay.

The study demonstrates that supervisors' subjective performance evaluations are significantly influenced by separate objective measures through two primary mechanisms: cognitive distortion and asymmetric controllability adjustments. Cognitive distortion causes supervisors to unintentionally bias their subjective assessments to match the directional level of a known objective measure, even when the tasks being evaluated are unrelated. Furthermore, the researchers identified an asymmetric uncontrollability effect where supervisors use their discretion to compensate for bad luck, specifically by increasing subjective scores when uncontrollable factors negatively impact an objective measure, but they do not similarly reduce scores to punish subordinates for good luck or favorable uncontrollable events. This behavior, characterized as borrowed discretion, suggests that supervisors prioritize maintaining employee fairness and motivation over the strict accuracy of evaluating a specific, separate task. Ultimately, these findings indicate that the exposure to objective data can compromise the intended independence and informativeness of subjective evaluations within a performance measurement system.