The Optimality of Simple Contracts: Moral Hazard and Loss Aversion

This paper extends the standard principal-agent model with moral hazard to allow for agents having reference- dependent preferences according to Köszegi and Rabin (2006, 2007). The main finding is that loss aversion leads to fairly simple contracts. In particular, when shifting the focus from standard risk aversion to loss aversion, the optimal contract is a simple bonus contract, i.e. when the agent's performance exceeds a certain threshold he receives a fixed bonus payment. Moreover, if the agent is sufficiently loss averse, it is shown that the first-order approach is not necessarily valid. If this is the case the principal may be unable to fine-tune incentives. Strategic ignorance of information by the principal, however, allows to overcome these problems and may even reduce the cost of implementation.