An Economic Response to Unsolicited Communication Appendices

There exist two ways to increase participant information: the sender can know r, and the recipient can know s. Given the sequence of communication, we first analyze informed senders, then add informed recipients. In the latter, private values become common knowledge. In both models, increasing information increases efficiency relative to a perfect filter. Note that perfect correlation of sender and receiver values (assuming invertibility) is a subset of the common knowledge case. As before, the recipient need not know sender value ex ante but she can infer it ex post . This follows from the fact that knowing either private value and knowing the correlation function is sufficient to compute the other value. In the analysis that follows, we make one additional assumption, that the recipient can credibly commit ex ante to a policy. Since information is known, commitment can be enforced via external reputation systems or a trusted third party. It does not matter how it is enforced as long as the recipient has the proper incentives to execute the policy to which she committed. As in (Rasmusen, 2001) precommitment and reputations improve welfare. We also relax our assumptions on the distribution of communications values. The following mechanisms work regardless of the magnitude of value, the density function