Stability of Panel of Experts in Persuasion Games

This paper is based on the work of Bhattacharya and Mukherjee (2013). As in their model, I study the persuasion game where the uninformed decision-maker relies on the messages of a panel of experts. An expert’s preferences are biased but state- independent and characterized by his ideal action, called an agenda. This framework allows us to analyze the optimality of the diversity and amount of experts in the panel. Intuitively, you want to hear different points of view as more information is revealed if you talk to experts with various agendas distributed all among the political spectrum. The authors in Bhattacharya and Mukherjee (2013) show that in the world, where you could choose any agent from a continious spectrum, only a panel from extreme agents can be optimal and the new expert is always beneficial for the decision-maker. I amend their model in such a way, that we are now left with choice: to accept new expert in a panel or not. I analyze the conditions, under which acceptance of this expert would be optimal for the decision-maker and come to the result, that the extreme panel of experts are more stable in that sence: there exist conditions on an expert, under which the decision-maker is always better-off. Moreover, under commitment the question of stability of the panel disappears, because a decision-maker could easily guarantee himself a utility that is not less that before.