This paper investigates reporting honesty when managers have monetary incentives to overstate their performance. We argue that managers who report about their performance will take into account how their report affects their peers (i.e., other managers at the same hierarchical level). This effect depends on the design of the organization's control system, in particular, on the reward structure and the information policy regarding individual performance reports. The reward structure determines if peers’ monetary payoff is increased or decreased when managers claim a higher level of performance. The information policy determines if managers will be able to link individual peers to their reports and affects the nonmonetary costs of breaking social norms. We present the results of a laboratory experiment. As predicted, we find that participants are more likely to overstate their performance if this increases the monetary payoff of others than if their reported performance decreases others’ monetary gains. In addition, overstatements are lower under an open information policy, where each individual's reported performance is made public, compared to a closed information policy, where participants only learn the average performance of the other participants. Our findings have several important implications for management accounting research and practice.

This paper investigates managerial discretion in compensation decisions in a team setting, in which a measure of the team's aggregate performance is readily available from the accounting system. Specifically, we examine the willingness of managers to obtain additional, costly information that would supplement this measure and allow the managers to more accurately assess individual contributions to team output. Using theory from behavioral economics that incorporates social preferences (i.e., fairness and trust reciprocity) into the managers' utility function, we predict and demonstrate experimentally that managers' willingness to obtain the costly information increases as the team's aggregate performance becomes a more noisy measure of individual performance. Further, we predict and demonstrate that managers' willingness will be greater for relatively high versus relatively low levels of aggregate performance. The study contributes to the literature on subjective performance evaluation by identifying how social preferences influence managers' use of discretion in evaluation processes. Data Availability: The data from this study and the set of instructions for the experimental task are available from the researchers upon request.