Dynamic Non-Bayesian Persuasion

Abstract

If a sender in a persuasion game can use a sequence of experiments rather than a single experiment, does this change the sender's value? We show, within a large class of non-Bayesian updating rules, divisibility, introduced in Cripps (2018), exactly characterizes the receiver's updating rules under which the sender is indifferent between static and dynamic persuasion in any environment. Consequently, restricting attention to static persuasion is without loss precisely under divisible updating rules.