10-21-2026
Managers often try to compare like with like. But that intuition can break down when workers compete repeatedly with the same colleagues. New research in the Review of Accounting Studies from Purdue University’s Mitch Daniels School of Business explains why. In “Relative Performance Evaluation and the Level Playing Field,” Assistant Professor Lin Qiu and coauthors Tim Baldenius and Jing Li examine how organizations can design performance-based compensation when employees repeatedly interact and can influence one another’s behavior. Their model suggests that differences among employees — rather than always creating a disadvantage — can sometimes make relative performance evaluation more effective.
The findings have important implications for leaders designing sales contests, bonuses, performance rankings and other incentive systems.
Relative performance evaluation (RPE) links an employee’s compensation to how well that employee performs compared with a peer. A sales contest, for example, might reward an employee when their sales exceed those of a coworker.
RPE can be attractive because it helps filter out common factors affecting everyone’s performance, such as market conditions or other shared shocks. But the researchers identify a potential problem when employees repeatedly compete against the same colleagues: collusion.
Because one employee’s success can reduce another employee’s reward under RPE, coworkers may have an incentive to coordinate by reducing their effort. The researchers’ model shows that when employees interact repeatedly, differences in productivity or performance measurement can make such coordination more difficult.
That produces a counterintuitive result: some degree of workforce heterogeneity can actually make RPE more attractive.
The effect is especially relevant in jobs where additional effort produces relatively small increases in performance. The model predicts a stronger role for this mechanism in lower-skill industries such as retail, hospitality and transit than in higher-skill industries.
The research also challenges the assumption that employees who are similar should necessarily receive identical incentives.
Under some conditions, the most cost-effective way to discourage employees from coordinating on lower effort is to give one employee a stronger incentive than another. The stronger bonus gives that employee a greater reason to keep working, which can destabilize an informal agreement among coworkers to ease off.
Importantly, the model predicts that this form of differential incentive design is more likely to be useful in lower-skill industries than in highly skilled fields such as financial services, scientific research and software development.
For leaders, the broader implication is that incentive design should focus on how compensation shapes employees’ strategic responses to one another. Under some conditions, differentiated incentive structures can help sustain effort and make the overall compensation system more effective.
The researchers compare three approaches:
Their model indicates that, under some conditions, JPE or IPE can be more attractive when employees are relatively similar. As differences among employees increase, however, RPE can become more effective because those differences make collusion harder to sustain. The research also shows that worker skill, productivity differences and team stability can influence which system is most effective.
The broader lesson is that effective performance management is not simply about creating a level playing field. It is about designing a system that produces the behaviors an organization needs. In workplaces where employees compete repeatedly, the composition of teams — and even carefully designed differences in incentives — can affect whether peer competition sustains effort or creates opportunities for employees to coordinate on doing less.
