Strategic gradual learning and information transmission

Abstract

Prior to advising a decision maker, the expert needs to gather some relevant information. This often takes time, and therefore, even if the expert's learning process is unobservable, the timing of advice can be informative in itself. If, in addition, the expert can choose which experiments to perform, the timing of her advice may reveal not only the amount but also the type of information at her disposal. This paper studies a dynamic information transmission problem where a biased expert acquires information by performing a sequence of experiments, which cannot be observed by the decision maker. It is shown that, even in the absence of an objective reason to expedite information transmission, artificial time pressure can increase the amount of information transmitted and be beneficial to both players.