First-best, second-best and principal-agent problems

Resum

In some pure moral hazard situations the principal can implement a first-best allocation using an incentive contract constructed on the basis of a first-best payment scheme. Such a contract relies on the possibility of discriminate actions according to the outcome by imposing a penalty whenever the observed outcome is lower than the admissible ones. The elimination of inefficient behavior depends basically on the outcome function, and we find that the fine is finite in the more interesting cases. The implementation of the first-best solution does not depend on the principal's risk neutrality. Nevertheless, when the principal is risk neutral, the ef f icient contract is dichotomous. Moreover, we prove that the efficient allocation can be reached through such a dichotomous payment scheme if and only if the principal is risk neutral for a certain range of returns.