Understanding the Influence of Government Controlled Media: Evidence from Air Pollution in China

This project establishes several empirical facts to better understand the influence of government-controlled media in autocratic regimes. We document that government control and bias on the controversial topic of air pollution are salient; that individuals do not discount repeated information, even when they know the information comes from one biased government source; that individuals have difficulty interpreting conflicting pieces of information; and that the population is heterogeneous in the degree to which they update. These patterns are difficult to reconcile with Bayesian updating and frameworks that assume homogeneous populations.