This study investigates experimentally how mutual monitoring affects effort when employees are compensated via rank‐order tournaments. Theory and anecdotal evidence suggest that mutual monitoring may either decrease effort by facilitating collusion or increase effort by stimulating competition. In our first experiment, we find that mutual monitoring increases effort, because participants do not attempt to collude but rather behave competitively. This result leads us to expand our theory and develop hypotheses to predict that the effect of mutual monitoring depends on whether employees have the inclination to collude or compete. Specifically, we predict that mutual monitoring decreases effort when employees are inclined to collude and increases effort when employees are inclined to compete; that is, mutual monitoring will not change the basic inclination created by the workplace setting, but will “turn up the volume” on the effect that such inclination has on effort. Consistent with our predictions, our second experiment finds that mutual monitoring leads to lower effort when participants have a collusive inclination and (eventually) higher effort when they have a competitive inclination. Overall, the results from these two experiments suggest that allowing employees to observe each other's productive effort in tournament incentive settings may have positive or negative consequences for the firm, depending on whether environmental factors predispose employees to collude or compete.

This study investigates how relative performance information (RPI) affects employee performance and allocation of effort across tasks in a multi-task environment. Based on behavioral theories, we predict that the social comparison process inherent in RPI induces both a motivation effect that results in increased effort as well as an effort distortion effect that results in the distortion of effort allocations across tasks away from the firm-preferred allocations. We also predict that both effects are magnified when the RPI is public compared to private. We argue that although the motivation effect will generally benefit performance, the effort distortion effect may be detrimental to performance. We design an experiment that isolates these two effects. Consistent with our predictions, we find that RPI induces both motivation and effort distortion effects and that both effects are magnified when the RPI is public rather than private. Although the motivation effect increases performance, we demonstrate that the effort distortion effect can decrease performance. By isolating the motivation and effort distortion effects, our study provides insights into the costs and benefits of RPI in a multi-task environment. As such, it informs accountants regarding the design of information systems and when tasks should be aggregated or disaggregated across employees. Data Availability: Data are available from the authors upon request.