Our research on these attacks — as well as the Orlando shootings by Omar Mateen and the Charleston church shootings by Dylan Roof — shows that news coverage framed these shootings very differently. Only the attacks perpetrated by Muslims were routinely called terrorism.
Even before we did our study, research showed disproportionately high media coverage of terrorism committed by Muslims — even though right-wing extremist groups have committed more attacks than Muslim in the United States since Sept. 11, 2001. Indeed, on the same weekend as the Charlottesville attack, the white nationalist Jerry Varnell was arrested for attempting a Timothy McVeigh-style bombing, but with little media attention.
Our research on the Orlando and Charleston shootings focused not on how much these attacks were covered, but on how they were covered. Our statistical analysis used a tool called topic modeling, which identifies common themes in a collection of articles and clusters them together using an algorithm. Essentially, we identified the relevant frames in thousands of articles from major national and regional U.S. newspapers.
Although the Orlando and Charleston shootings had key similarities — both were committed by individuals, involved firearms and were plausibly hate crimes — they were not covered similarly.
First, the graph below shows that coverage of Mateen used “terrorism,” “terrorist,” and “radical” three to four times as frequently, while Roof’s coverage used “mental health” 3.5 times as frequently.
Even within the coverage that focused on terrorism, there were differences. Articles that discussed Mateen and terrorism focused on Islam and violence. But articles that discussed Roof and terrorism tended to focus on the question of whether his attack constituted terrorism. The coverage of Mateen didn’t really ever ask that question. This was despite weak evidence tying Mateen to the Islamic State, compared to stronger evidence tying Roof to right-wing extremist groups.
The same pattern emerged in coverage of the Charlottesville and Barcelona attacks. We gathered stories published within five days of each attack that included the names of the drivers of the vehicles. The same statistical models showed stark differences in how the coverage was framed:
Once again, coverage of Barcelona referred to terrorism and religion substantially more than did coverage of Charlottesville. Even the Charlottesville coverage that mentioned terrorism did so within the context of debating whether Fields’s attack was terrorism. The same does not appear true for coverage of Younes Abouyaaquob.
We do not believe that these differences in coverage are intentional or nefarious. Moreover, we cannot say how far this bias extends, as we have examined only these cases.
However, the striking differences in these prominent examples are important. Other research has found that framing attacks as terrorism changes public opinion — leading people to be more supportive of the Patriot Act, increased funding for anti-terrorism activities, more prone to authoritarian thinking, and more hawkish on U.S. foreign policy. This coverage may also help to make Americans feel less favorably toward Muslims overall and also strengthen specific stereotypes of Muslims — especially that they are violent — which shape American attitudes on the War on Terror.
The frames of news coverage take on even greater significance at a moment in which President Trump has advocated for restrictions on travel to the U.S. from Muslim-majority countries, even as his administration recently cut funding to fight right-wing extremism. The question that news outlets must confront is whether these differences in news coverage of attacks by Muslims and right-wing extremists are warranted and, if not, how to change them.
Bryan Arva (@barv23) is a PhD student in the Department of Political Science at Pennsylvania State University and a researcher at University of Maryland’s National Consortium for the Study of Terrorism and Responses to Terrorism (START).