Abstract:
This paper presents an empirical test for the hypothesis that US news coverage of al-Qaeda causes al-Qaeda attacks. To isolate causality, disaster deaths worldwide provide an instrumental variable crowding out al-Qaeda coverage. Studying daily al-Qaeda coverage by CNN, NBC, CBS, and Fox News, as well as the NYT and the WaPo, results consistently produce a positive and statistically significant effect. At its mean, al-Qaeda coverage is suggested to cause 1.2-2.3 al-Qaeda attacks (equivalent to 5.8-10.9 deaths) in the upcoming week. Results are remarkably consistent across media outlets and it appears unlikely that attacks are simply delayed when coverage is low.