Forty years after the publication of Edward Said’s Covering Islam, the book Covering Muslims positions itself as a “conscious echo” (p. x) of that work. The material, time period, and scientific approach differ, but now as then the state of affairs in news coverage is dismaying. Articles mentioning Muslims or Islam in American newspapers between 1996 and 2016 are strikingly and pervasively negative in tone.

The scale of Covering Muslims is prodigious. In the United States alone, Erik Bleich and A. Maurits van der Veen scrutinize more than a quarter of a million articles mentioning Muslims or Islam, appearing in seventeen national and regional newspapers, plus more than half a million additional articles for comparisons with other groups, not to mention yet another half a million and more newspaper articles in Britain, Canada, and Australia, all stretching over the same time period from 1996 to 2016. This analysis is further supplemented by examination of a selection of newspapers in East Asia, South Asia, and Africa, analyzed over a shorter and more recent time period. Much is done with this mass of material and the methodology is cutting edge. Sentiment analyses register variation in tone; collocation analyses recover variation in content; topic modeling adds inductive exploration to deductive interrogation.

Science is cumulative. Political science tends to be adversarial. Covering Muslims is a model of how political science can and should be done. It is unblemished by the desire to show that previous researchers have got things wrong—just the reverse. Its principal strategy is deductive, with deductive emphasized for a reason. In political science, deductive is synonymous with reasoning from first principles. Bleich and van der Veen introduce a companion usage—deducing predictions from previous empirical claims. Treating previous research as a seed bed of promising ideas, they systematically harvest it for testable predictions about the degree, not merely the direction, of negativity (tone) over time and across countries, types of newspapers, and topics.

What do we learn? Not simply that newspaper coverage of Muslims is negative in tone. How much more negative, when, and compared to whom are everything in the quantitative departments of political science. With a few assumptions of scalability, Bleich and van der Veen put a number on how overwhelmingly negative is the coverage of Muslims. We learn also that the insight that foreign news tends to be more negative goes a good way towards explaining the tone in coverage about Muslims and Islam. Only 9% of articles mentioning Muslims or Islam are set uniquely in a domestic location (p. 53). In addition, the attention to Muslims and Islam in American news in the period examined is strongly connected to coverage of violence. Articles mentioning violence make up 73% of all the pieces in the corpus (p. 53). The focus on terrorism did worsen after 9/11, but instructively, the main change was not in the tone, which was already negative, but in the scale and topic of coverage (chapter 4).

Coverage in tabloids is even more negative, not surprisingly given reliance on sensationalism, but that is not the main point. The main point is that coverage is intensely, relentlessly, negative even in the national newspapers of record, the New York Times and the Washington Post. This is what I would have expected, some might say. But we doubt anyone could say that they expected that ideologically right-leaning newspapers are not distinguished by negative coverage. And the reason, it is fascinating to learn, is that right-leaning newspapers publish more articles about religion—Islam as well as Christianity—and articles on religion generally have a less negative tone. Another unexpected finding: domestic political debates on value conflict are not found to be important independent drivers of negativity in coverage.

All in all, Covering Muslims is a catalogue of demonstrations of just how remarkably negative the coverage of Muslims and Islam is in American newspapers and beyond. Coverage of Muslims is decisively more negative than coverage of other religious groups—Catholics and Jews and, perhaps more surprisingly, Hindus, too. In the smaller subset of articles about Muslims in the domestic context, coverage of Muslims is more negative than that of other minority groups. In other English-language democracies, Canada, Australia, and Britain, coverage of Muslims and Islam is relentlessly negative in the same way as in U.S. coverage.

This raises the question of whether there exists a globally negative news discourse about Muslims. The authors make a highly plausible case that this is unlikely. This is the most tentative part of the analysis, relying on the least amount of material, collected from a more recent period than examined in the rest of the book. Still, the main argument is very likely to be correct. Where Muslims and Islam have a larger place in domestic society and politics, the coverage is more varied, more average, less dominated by articles mentioning violence and extremism. For example, the coverage in an English-language newspaper in a majority Muslim country like Malaysia includes articles on Islamic finance, debates on Islamic jurisprudence, and announcements of burials.

So, newspaper articles mentioning Muslims or Islam are overwhelmingly negative in tone in Anglophone democracies. We suppose the same could be demonstrated in non-Anglophone newspapers in Europe as well had they been analyzed using the same methodology. Taking this insight in Covering Muslims as the starting point, what are the new important questions to be raised for discussion?

One is how we should think about what happened in the United States just after the completion of the empirical part of this study—the election of a Republican president, Donald J. Trump, who in the mold of European far-right leaders publicly constructed Muslims and Islam as an existential threat. How, when coverage is nevertheless relentlessly negative, should we think about the role of far-right political leadership and power? Would the quite provocative conclusion that it does not much matter be warranted based on the argument advanced in Covering Muslims? Why or why not?

A second question is how we should think about the relationship between news coverage and the construction of villains in wars and conflicts in which the United States is involved. One way to think about that question is through a historical counterfactual that, perhaps one day, may open itself to empirical analysis. What would be the result had a similar analysis been done on newspaper coverage during the Cold War, and perhaps especially, in the period of McCarthyism? In that period, Communists and “Soviets” were singled-out as the principal foreign threat to the nation. Today they are no longer in the limelight. Can this example serve as a counterfactual to suggest that it is possible for coverage of Muslims to change and become less relentlessly negative, if and when the constructions of America’s friends and foes in the main wars and conflicts in the international arena change? Or will the additional factors associated with negativity in coverage of Muslims prevent this from ever happening?

For us as public opinion specialists, the larger point that Covering Muslims brings out into the open is the challenge of how to theorize connections between media coverage and public opinion. The headline result of the study is the temporal constancy of outlier levels of negative coverage about Muslims. The puzzle is that, in Western Europe, apart from oscillations in response to specific events (e.g., the 2015 refugee influx following the Syrian war), the trajectory of evaluative attitudes towards Muslims is not negative. And despite the growth of the far-right, who insist on seeing all Muslims as an existential threat, the main tendency across the largest share of the public in a range of Western European countries is to look past the far-right’s tropes and distinguish ordinary Muslims from extremists who label themselves Muslim (The Struggle for Inclusion, chapter 2). As always in public opinion research, measures and measurement are limited and imperfect. But the contrast between the overwhelming negativity in media coverage and the more multifaceted patterns in public opinion suggest that the influences of media coverage on public opinion is a far cry away from direct and uniform.

The deeper significance of media studies like Covering Muslims, then, may be the light they shed, not on mass attitudes, but on the sources of biases in perceptions of mass attitudes. We ourselves are midway through a study of such perceptual biases. The guiding idea is that openings in public opinion to build coalitions favorable to the inclusion of Muslims are systematically not perceived. Covering Muslims prompts a hypothesis about how such perceptual failures can come about. The relentlessly negative coverage of Muslims in news media can be a key factor encouraging overestimation of negative public sentiment towards Muslims. The risk is that politicians on both the left and right, as well as Muslims and non-Muslims in society at large, conclude that most voters in contemporary liberal democracies favor the exclusion of Muslims. This may be a natural mistake to make. In The Struggle for Inclusion, we present evidence why it will be a tragic mistake to make.