Hidden Testing and Selective Disclosure of Evidence

Abstract

This paper contrasts a decision maker's payoff under public and private information acquisition by a biased advisor. Both players agree on the optimal choice under certainty, but differ in how they trade off the loss from errors. The advisor can sequentially acquire informative test outcomes. If acquisition is private he decides in the final period which realizations to verifiably disclose. If players' preferences are sufficiently misaligned, the decision maker is weakly better off under private rather than public information acquisition. The effect on the advisor's payoff depends on the direction of his bias.