Learning and signalling by advisor selection

In this paper we consider a model where a policy maker uses advice in order to (1) obtain information about the consequences of an innovation (information motive) and (2) to support political legitimacy of her decision (persuasion motive). We conduct our analysis in the context of a cheap-talk game with three players: (1) a policy maker, (2) the median voter in parliament or of the electorate and (3) an advisor. The advisor has private information about the consequences of policy. Communication between an advisor and a recipient improves as their preferences are closer aligned. If the preferences of the policy maker and the median voter are different the policy maker faces a trade-off. On the one hand, she wants to gain information to judge whether the innovation is worthwhile. On the other hand, she needs to convince the median voter whether the innovation is desirable.