This article studies how politicians react to feedback from citizens on social media. We use a reinforcement-learning framework to model how politicians respond to citizens’ positive feedback by increasing attention to ...