Pith. sign in

REVIEW

Perfect Bayesian Persuasion

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2402.06765 v1 pith:K7QTSUD5 submitted 2024-02-09 econ.TH

Perfect Bayesian Persuasion

classification econ.TH
keywords senderbayesianbeliefequilibriummodelspayoffpersuasionreceiver
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

A sender commits to an experiment to persuade a receiver. Accounting for the sender's experiment-choice incentives, and not presupposing a receiver tie-breaking rule when indifferent, we characterize when the sender's equilibrium payoff is unique and so coincides with her "Bayesian persuasion" value. A sufficient condition in finite models is that every action which is receiver-optimal at some belief is uniquely optimal at some other belief -- a generic property. We similarly show the equilibrium sender payoff is typically unique in ordered models. In an extension, we show uniqueness generates robustness to imperfect sender commitment.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.