Feasible Joint Posterior Beliefs

Document pages: 51 pages

Abstract: We study the set of possible joint posterior belief distributions of a groupof agents who share a common prior regarding a binary state, and who observesome information structure. For two agents we introduce a quantitative versionof Aumann s Agreement Theorem, and show that it is equivalent to acharacterization of feasible distributions due to Dawid et al. (1995). For anynumber of agents, we characterize feasible distributions in terms of a "no-trade " condition. We use these characterizations to study informationstructures with independent posteriors. We also study persuasion problems withmultiple receivers, exploring the extreme feasible distributions.