Paris School of Economics
Date and Location
Thursday,?April?4th,?2019 – 13:30 – 15:00
Lille: Room?R217 – Visio?
This paper studies the discredit of scientific knowledge into fake news. Scientific knowledge is seen as a collection of models, which are simplified representations of reality. An informed sender is assumed to have a more educated perception of models’ likelihood, but cannot prove one to be true. A receiver is in a situation of model-uncertainty and has ambiguity sensitive preferences. When the sender and the receiver have non-congruent interests and communication is costless, communication is a cheap-talk game over models. I show that when ambiguity aversion grows, it is much harder for the sender to be credible. Yet, under the assumption that the receiver is a maxmin expected utility maximiser, I show that all equilibria of this game can be ranked by informativeness and that the sender is better off playing the most informative one. This is, a credible sender has an incentive to convey as much information as possible to an ambiguity averse receiver. These results shed new light on the possibilities of scientific communication and on the specificity of ambiguity aversion in that context.