Abstract
Misinformed citizens are deemed both widespread and threatening the efficiency of democratic and economic systems. This paper focuses on the role of politicians as originators of misperceptions as a response to “bad news”. I exploit the random timing of mass shootings in the U.S., arguing that these events update voters beliefs on the state of the world, i.e. the causes of mass shootings and in turn policy support for gun-control. Applying quantitative text-analysis and machine learning techniques to twitter-data I demonstrate that politicians running on a pro-gun platform as opposed to anti-gun politicians (i) seek to distract, i.e. sway public attention away from the mass shooting and (ii) provide misleading information about the causes of mass shootings. The results are consistent with a conceptual framework in which politicians that face bad news communicate strategically in order to (i) prevent belief updating or (ii) distort the belief formation process of voters towards supporting policies other than gun-control.