"Selective disclosure"

Abstract : This paper investigates theoretically and experimentally the common situation in which an interested party can voluntarily disclose informative but noisy evidence to a decision maker. Distinguishing whether the decision maker is fully rational or affected by an "all you see is what there is" updating bias, we investigate the non-obvious effects of the amount of disclosable evidence in terms of : i) the total number of pieces of evidence the interested party observes ; ii) the total number of pieces of evidence she can disclose among those she observes. Subjects in the role of the interested party disclose selectively, even if only imperfectly (i.e., some favorable evidence is withheld and some unfavorable evidence is disclosed), which limits the gains from deception and the costs of selection neglect. Subjects in the role of decision makers are systematically deceived only when the interested party observes more evidence than she can disclose. Consistent with the theory, a higher amount of disclosable evidence does not necessarily lead to higher information transmission, even though the welfare effects generally differ for rational and boundedly rational decision makers.