Differences in Religious Framing of Muslims and Islam in American Media Coverage Before and After the Trump Campaign

ABSTRACT

Over the last 20 years Muslims have received a disproportionate amount of news media attention. In 2017 they were again thrust into the spotlight when the Trump administration issued a travel ban blocking entry from seven predominately-Muslim countries. With our hand coding of over 900 newspaper articles and mixed modeling techniques, we examine the newspaper rhetoric surrounding Muslims before and during the Trump administration and across US states. We find that after President Trump’s election, fewer newspapers mentioned religion or referenced Muslim claimsmakers. We also reveal that higher state-level religious salience was associated with more newspapers referencing religious claimsmakers in Islam-related discussions. Other state-level political and religious characteristics had no effect on the use of religious elements or Muslims claimsmakers. These results provide insight into the position of Muslims and Islam in the United States, as well as the impact of political rhetoric on media portrayals.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Notes

¹ 5 East North Central states, 1 Mid-Atlantic state, 5 Mountain states, 3 New England states, 4 Pacific states, 6 South Atlantic states, 4 West North Central states, and 3 West South Central states. We did not have access to any newspapers published in the East South Central region of the country.

² Although our key terms of Muslim and Islam are inherently religious, they could be used in a variety of difference topics other than religion, like discussions of politics, immigration, geopolitical violence, or demographics.

³ In a separate analysis, we also considered trying to explain variation at the county-level where the newspaper headquarters were located, but to our knowledge there is no measure of county-level religious salience, like we have for states, which limited our analysis. In our study we had a specific and small group of counties for which we needed religious salience measures and neither the General Social Survey (GSS) and American National Election Survey (ANES) were able to provide even a partial number of the county estimates. Additionally, we were unable to explain any variation at the county-level when we used measures of proportion voting Republican, and Muslim and Evangelical rate per 1,000 people. Likewise, newspaper coverage tends to stretch further than a single county where it is produced and we were more theoretically interested in understanding and explaining-state level variation, which received so much attention during the Trump presidency. For all of these reasons, our three-level mixed models have articles clustered within newspapers that are then clustered within states.