The results of empirical research on whether media attention encourages terrorism differ according to the study period. The present contribution identifies which episodes from 1970 to 2010 are characterized by Granger-causalities. Structural breaks are systematically taken into account, and both domestic and transnational terrorism are considered. Data on terrorism are drawn from the Global Terrorism Database (GTD), while the media attention variable is constructed by the number of New York Times' (NYT) articles that contain the word terrorism. A bidirectional Granger-causality between transnational terrorism and media attention was found only from 1999:09 to 2002:07, indicating a temporary effect of 9/11.
terrorism media Bai-Perron test GTD Toda-Yamamoto procedure NYT