We provide a new explanation for why central banks have become transparent over the last three decades. We apply recently developed social interaction panel regression models for the observational data, which allow the identification of peer effects. The identification is based on variations in the past monetary policy régime exogenously determined with respect to transparency. Previous literature has argued that domestic factors such as macroeconomic stability were behind the trend toward greater transparency. In contrast, our results indicate that transparency primarily increased because of a favorable global environment and, importantly, because of the peer effects among central bankers. Central bankers thus learned from each other's experiences regarding transparency. To our knowledge, our paper is the first econometric analysis of peer effects among public institutions or in the macroeconomic literature.