Crime news effects and democratic accountability: Experimental evidence from repeated exposure in a multiweek online panel

Abstract

Publics hold chief executives uniquely responsible for national well-being, and they learn about national conditions through news. But when news disproportionately covers problems, what happens to democratic accountability? Here, we experimentally test how leader approval changes when crime loses prominence in news for a sustained period. We create an online news environment coding real news in real time, then experimentally filter news for nationally diverse U.S. panelists over 1 week. We find causal evidence that reducing crime news raises presidential approval and depresses problem importance evaluations for crime. No other leaders are credited, and reducing all problems produces no further gains. These effects persist well after exposure but dissipate within a week. We conclude with broad implications for journalism and democratic judgment.