Withholding Verifiable Information

Abstract: I study a class of verifiable disclosure games where the sender's payoff is state independent and the receiver's optimal action only depends on the expected state. The sender's messages are verifiable in the sense that they can be vague but can never be wrong. What is the sender's preferred equilibrium? When does the sender gain nothing from having commitment power? I identify conditions for an information design outcome to be an equilibrium outcome of the verifiable disclosure game, and give simple sufficient conditions under which the sender does not benefit from commitment power. These results help in characterizing the sender's preferred equilibria and her equilibrium payoff set in a class of verifiable disclosure games. I apply these insights to study influencing voters and selling with quality disclosure.