Non-Cooperativity in Bayesian Social Learning

TL;DR

This paper models Bayesian social learning in networks, showing that observation levels optimize learning but individuals tend to defect from societal best outcomes, highlighting conflicts in information sharing.

Contribution

It introduces a Bayesian social learning model on directed networks and compares it to complete graph scenarios, revealing insights into observation incentives and societal outcomes.

Findings

Observation levels maximize learning in both models

Individuals have incentives to defect from societal optimum

Competition over information affects collective learning outcomes

Abstract

We describe a Bayesian model for social learning of a random variable in which agents might observe each other over a directed network. The outcomes produced are compared to those from a model in which observations occur randomly over a complete graph. In both cases we observe a nontrivial level of observation which maximizes learning, though individuals have strong incentive to defect from the societal optimum. The implications of such competition over information commons are discussed.