A Rothschild-Stiglitz Approach to Bayesian Persuasion
Matthew Gentzkow, Emir Kamenica
American Economic Review · 2016 · Volume 106 · Issue 5
Abstract
Rothschild and Stiglitz (1970) represent random variables as convex functions (integrals of the cumulative distribution function). Combining this representation with Blackwell's Theorem (1953), we characterize distributions of posterior means that can be induced by a signal. This characterization provides a novel way to analyze a class of Bayesian persuasion problems.
Keywords
Search, Learning, Information and Knowledge, Communication, Belief, Unawareness
Enable JavaScript for citation tools and additional paper interactions.