Designing Incentives for Teams

Description

Most organizations treat an employee’s compensation contract as private information, and they do not share the details of that contract with other employees. Most employees do not know the incentive packages offered to their peers, nor do they know the compensation budget offered to their boss. What remains unanswered is whether this real-world feature about contracts makes teamwork more difficult to incentivize as employees care about how much effort each of them will put in, and consequently, about the incentive pay each will receive. They need to be sure that they each have strong incentives to be convinced to put in a high effort themselves.Our proposed research begins with a standard model of moral hazard in teams based on the theory in Buffa, Liu and White (2022). Within our theoretical framework, we derive each agent’s equilibrium effort level under public and private contracts. Moreover, under private contracts, we consider three types of beliefs that each agent forms about the other agent’s effort. Our model implies that it is more difficult to incentivize agents with private contracting and the principal chooses to delegate pay-setting when agents are sufficiently skilled.Based on the theoretical predictions of our model, our proposed research aims to address the following questions empirically:(1) Are effort levels from employees when contracts are private different from those when contracts are public?(2) What effort levels does each employee anticipate other employees will exert when contracts are private?(3) Is there a theoretical rationale to explain the prevalence of private contracts in practice, such as employees envying others’ contracts?(4) Is there a better way for the employer to incentivize a team when contracts are private, such as designating a team-leader with authority to decide his teammate’s contracts?Having recognized the difficulties of obtaining clean data on employee compensation, we resort to an experimental approach. Experiments run in a lab allows us to implement exogeneous variations in the observability of contracts and assess how experimental subjects react to those changes. We categorize subjects into two groups: “principal-subject” and “agentsubject.” Each principal-subject owns a project and recruits two agent-subjects to implement the project. Each agent subject is associated with a skill level. The principal-subject has discretion over the agent-subjects’ compensation, and each agent-subject responds to the principal-subject’s offer with an effort level. We compare the agent-subjects’ effort choices under different contracts observability and ask them to report their beliefs about others’ private contracts. When contracts are public, we also ask the agent-subjects whether they believe the contracts offered to the others are fair, and the principal-subjects whether they would like to make the contracts private. When contracts are private, through varying the agent-subjects’ skill levels, we observe whether a “team leader” emerges endogenously among the two agentsubjects.