Team Incentives and Reference-Dependent Preferences

This paper examines a multi-agent moral hazard model in which agents have expectation-based reference-dependent preferences `a la K˝oszegi and Rabin (2006, 2007). The agents’ utilities depend not only on their realized outcomes but also on the comparisons of their realized outcomes with their reference outcomes. Due to loss aversion, the agents have a first-order aversion to wage uncertainty. Thus, reducing their expected losses by partially compensating for their failure may be beneficial for the principal. When the agent is loss averse and the project is hard to achieve, the optimal contract is based on team incentives which exhibit either joint performance evaluation or relative performance evaluation. Our results provide a new insight: team incentives serve as a loss-sharing device among agents. This model can explain the empirical puzzle of why firms often pay a bonus to low-performance employees as well as high-performance employees.