Referenced Information and Belief Update: Evidence from India

When an individual receives a piece of information, he decides if and how to incorporate it into his set of beliefs. This is particularly important for referenced information, often used by policymakers as a tool to achieve some form of behavior. In this paper, we ask how individuals update their beliefs about Covid-19 mitigation measures when they receive referenced information about vaccine efficacy and the ineffectiveness of treatments like hydroxychloroquine. If individuals follow some form of Bayesian updating process, belief polarization should either decrease or remain unchanged after individuals are presented to the same piece of information. We test this prediction with an online experiment in which participants are randomly assigned to a treatment group that receives referenced information. Our results show that polarization increased in the treatment group, suggesting that belief update does not follow a pure Bayesian process. This finding calls into question the effectiveness of traditional information dissemination in polarized environments.