The Theory of Incentive Mechanisms and the Samuelson Critique of a Contractarian Approach to Public-Good Provision

The assessment that the implementation of efficient outcomes by means of decentralized spontaneous solutions is to be expected for private goods, but not for public goods has become part of the conventional wisdom of our discipline. This paper uses a mechanism design approach to clarify under which conditions this proposition is indeed correct. The main result is that the following has to be satisfied: There is a large number of individuals and only mechanisms which are robust, in the sense that they do not exploit assumptions about individuals' probabilistic beliefs, are considered.