Research

An agent who receives a stream of messages that contain information about an unknown state of the world might ignore some of them. This might be because of cognitive overload, but it might also be strategically: information that contradicts one’s belief increases cognitive dissonance. A decision maker who experiences disutility in this way wants to be sure to make the right decision—rather than make the right decision. This means that her utility is belief-based.

In our model we consider a Markovian agent who observes a stream of noisy messages that stops in each round with some positive probability. Her anticipatory utility function is defined by a functional equation that takes all possible future beliefs after a message into account. We show that she will always update her belief after a confirming message, but will ignore contradicting messages, if her belief is too strong. The relevant threshold solely depends on the messages’ precision or—in case the precision is unknown and she has maxmin preferences—on the lowest precision she deems possible. 

Such an agent has strictly higher anticipatory utility than an agent who uses every message to update. However, the latter has a higher chance to choose the correct outcome in the end. In a population of strategic agents, who only differ in their initial beliefs, polarization is inevitable.

Our working paper is publicly available here. A more recent version can be found below.