![]()
1
THE MOST POWERFUL NAME IN NEWS
Over the course of the last 20 years, the American news media has undergone rapid and entirely transformative change, and itâs only now that science can start to describe the effects this change has had on American politics and society. The clearest example of the change in how news is presented and consumed in America comes from Fox News, a channel which has found enormous ratings success while presenting a consistent ideological viewpoint. While there has been a great deal of hand-wringing about what effect Fox News may be having on the American political system â a recent bestseller about the channel notes in the title that Fox has âdivided a countryâ â there has been relatively little scientific analysis of the effects that Fox has had on actual Americans, their attitudes, and their behavior in real elections. This book moves beyond past work, to show how what is said on Fox News directly impacts Presidential elections, the Republican Presidential primary, campaign contributions, the political knowledge of the American public, and the social views and behaviors of Americans.
Unlike most of the books written about Fox News, this is a work of social science. Whether the effects Fox News has on politics and society are good or bad is beyond the scope of this work. Since 1998, Fox News has called itself âthe most powerful name in news,â and the purpose of this book is to show exactly how this power has played out. As the later chapters show, the effects of Fox coverage on American politics are enormous. One day of positive coverage of the President, for instance, can increase his approval rating by as much as a full percentage point. Fox coverage is also somewhat responsible for the radical swings in support between a large number of candidates in the 2012 Republican primary, and is one of the major factors driving campaign contributions, especially to lower-tier candidates. For these lower-tier candidates, a single positive mention on Fox was worth more than $20,000 in increased contributions over the following few days, with most of it coming from small donors. Media coverage shapes the general election, too: while campaign events such as the debates are often thought to have a huge effect on the race, the results here show that itâs not the events that matter, but the coverage of those events in the media that pushes voters to one side or the other. On the individual level, watching Fox leads many viewers to concentrate their attention on a small subset of issues that arenât frequently discussed in other outlets, leading them to do worse in answering political knowledge questions than they would if they werenât watching any news at all. Those viewers are also much more likely to think that President Obama is not a citizen, that the US found weapons of mass destruction (WMD) in Iraq, and that global warming is a hoax. These effects go beyond political attitudes and behaviors as well: discussions of gun control on Fox increase the number of gun sales in the United States, with effect sizes into the hundreds of thousands per month.
The Rise of Ideological Media
Before widespread access to the internet, Americans got their news from newspapers, or evening news broadcasts, or the single cable news channel, or the radio. This was broadcasting: unless you were part of the small minority of Americans listening to a political talk radio show, subscribing to a newsletter or firing up a modem to read and post comments on a message board, you were getting news that was designed to appeal to the widest possible audience. This news was, by and large, non-partisan, and, as Prior (2007) points out, since it was on all of the channels at the same time, anyone who wanted to watch TV during the news had no choice but to watch it. If you lived in a medium-sized city, you might well have had the choice between two newspapers, one with an opinion page that leaned a little left, and one with an opinion page that leaned a little right. Shows like Crossfire or Sunday morning talking head programs like The McLaughlin Report had plenty of opinions, but they were designed to be balanced between conservatives and liberals, Democrats and Republicans, and even so, they were the subject of intense mockery from Saturday Night Live and similar programs for their hyperbolic arguments. News outlets werenât for Republicans, werenât for Democrats, but were trying to capture the widest possible audience, which meant staying above the partisan fray.
But if thereâs one thing thatâs true of any American institution of the twentieth century, itâs that advances in telecommunication technology changed everything. Newspapers, which had already been struggling, had to start sharing ad revenue with websites, and lost the profits from the exceptionally lucrative classified ad business entirely. To keep up with the times, many, if not most, began giving away their content online for free, in the hopes that ad dollars would follow. There arenât too many American cities with two newspapers any more.
People with any interest in politics could now find news sources that catered directly to their opinions and interests: itâs easy to forget that Drudge Report was one of the first breakout websites of the internet. People of any political stripe, from the mainstream to the obscure, could find an online community that reported news from their perspective, and didnât force them to listen to opposing points of view in the way that the political debate shows had. These sites often werenât run by professional journalists, but rather by activists and advocates, who saw their job as being the advancement of a point of view or a cause, rather than the advancement of knowledge. The line between news and opinion, which had previously been sharp and clearly delineated, became a little blurrier (Sunstein 2007).
Perhaps the most important shift, however, came from the huge increase in the number of cable channels available to Americans. According to Nielsen data for 2013, the number of channels available to the average American had risen from 41.1 in 1995 to 189.1 (though the number of channels they actually watched hadnât gone up since 2008, even as the number of channels available increased by more than 50 percent). Nearly every cable package now includes at least three 24-hour cable news channels â CNN, Fox News and MSNBC â and many include several less popular ones such as BBC World News, not counting the financial news channels that generally have little political content.
Just as widespread access to the internet allowed Americans to choose news content that was in line with their predispositions (Sunstein 2007), the rise of multiple news channels allowed Americans to pick the television news that would tell them mostly what they wanted to hear, and politically active Americans have largely taken advantage of the opportunity. It took some time for the channels to achieve their present form: MSNBC began broadcasting in July 1996, with Fox News following a few months later in October, but neither was an ideological force in American politics until later. Foxâs political tone was fairly evident from the beginning, but wasnât available in most of the country until after the 2000 Presidential election, when cable providers noted the ratings it had achieved during the coverage of the election and its aftermath. MSNBC was started as a combination of repackaged stories from Dateline NBC, delayed re-broadcasts of The Today Show, viewer emails and links to stories on the channelsâ website. There was some opinion commentary, though this was reasonably balanced: conservative firebrand Ann Coulter was one of the hosts on the first day of programming. It wasnât until the channelâs Countdown: Iraq dropped subtitles and became Countdown with Keith Olbermann in 2003 that the liberal tone of the channel started to be established (Sherman 2014).
The rise of these channels is important because of their reach and their audience. While there are plenty of individual websites with strong political viewpoints, the largest of them â TheBlaze, Drudge Report, Newsmax, Infowars â are viewed by about 20 million people per month (the most popular liberally oriented site, Daily Kos, gets about 6.5 million views per month). In contrast, Fox News can expect more than 2 million viewers on a slow news night during primetime, vastly outpacing online conservative sources, and MSNBC gets about 800,000 viewers on a slow news night (Pew Project for Excellence in Journalism 2010). For both, big news events can dramatically increase those numbers: Fox has garnered as many as 11.5 million viewers on a single night (the final Presidential debate of the 2012 election).
Relative to the network evening news broadcasts, these arenât a lot of viewers: the three evening news broadcasts on ABC, NBC and CBS typically get 6â8 million viewers a night each. But while the evening news broadcasts get far more viewers than their cable competitors, their audiences arenât nearly as politically polarized, moderating their political impact. Using data from the 2010 American National Election Study (ANES), it is possible to estimate that ideological composition of forty-four different news sources, across television, radio and print media (full results are in Table 1.1). The most polarized audiences are all for conservative radio shows: the audience for The Rush Limbaugh Show is 83 percent conservative, just 4 percent liberal, and he barely makes the top five. The most polarized audiences for television shows are all for programs on Fox â Hannity, Fox & Friends and The OâReilly Factor â with MSNBCâs Rachel Maddow Show following with the most polarized liberal audience. The most politically polarized network program isnât even a news show, but The Tonight Show (then with Jay Leno), with an audience that leans conservative, mostly because of their age. Of the evening news broadcasts, the most polarized audience is for CBS Evening News, coming in at twenty-three out of the forty-four sources, with an audience about as polarized as that of The Ellen DeGeneres Show. ABC and NBCâs broadcasts are a little less polarized than that, both with audiences that lean a little conservative (again, largely due to age), and the venerable PBS News Hour manages to have an audience that isnât polarized at all: exactly 38.1 percent liberal, 38.1 percent conservative, and 23.8 percent moderate in the ANES data (a closer examination of these figures can be found in Chapter 5 and methodology note 5.3).
Even though MSNBC and Fox News donât get nearly the viewership of their network rivals, theyâre doing a good job of speaking to one particular ideological audience. Theyâre important not just because of the number of viewers they get, but because those viewers are liberal or conservative, giving them the opportunity for outsize influence over the political and social views of those segments of the population.
TABLE 1.1 Ideological Composition of Most Polarized Media Sources, 2010 ANES
These changes in the media landscape are often spoken of as unprecedented, but, in reality, America is simply moving back to the old way of presenting news. In the eighteenth and nineteenth centuries, American political parties were closely affiliated with news outlets, and many newspapers and books were produced with either direct or indirect party funding. The debate over the ratification of the Constitution was played out in partisan newspapers, between dueling Federalist and Anti-Federalist opinion pieces, which were never separated out from the ânewsâ portion of the papers, to the extent that objective journalism even existed. When the first political parties in post-Constitution America began to take form, Thomas Jefferson and Alexander Hamilton immediately began funding newspapers to promote their political movements (Leonard 1986). These proved successful enough that the Federalists in control of Congress passed the Alien and Sedition Acts partially to shut down opposition newspapers. The law allowed Federalists to jail writers who were critical of them (it even included a sunset provision, phasing the law out after the next election to ensure that it couldnât be used against them if they lost control of Congress or the Presidency, as they did), and deport foreign-born writers for doing so (Rosenfeld 1997). These newspapers traded in partisan sensationalism â claims that John Adams wanted to make himself king, that Hamilton was giving bribes to a member of Congress, that Jefferson was having an affair with his dead wifeâs half-sister who happened to be his slave (though at least some of these claims were true), and the state of journalism didnât get much better over the next century, with the rise of yellow journalism in the nineteenth and early twentieth century (Miller and Dershowitz 1951). It wasnât until the early to mid-twentieth century that journalism started to become professionalized and aspire to objective reporting of facts, clearly separating out opinion from news. This push was embraced by television news broadcasts, both as a matter of ethics and for profit (Muhlmann 2008, Iyengar and Kinder 2010). In an era when three evening news broadcasts were competing for a national audience, none of them could afford to have a political slant that might alienate a large slice of the potential viewing audience. This conformity to a non-political tone was further enforced by the Cold War, and the notion that politics â and, to some extent, debate â stopped at the waterâs edge. In sum, the recent move back to a fragmented, partisan media without a clear separation between news and opinion is less a historical anomaly than a move back to traditional media practices in the US. The anomaly was the type of news Americans were used to for most of the twentieth century â but this doesnât make the move back any less jarring.
Does Media Matter?
By no means is the question of media influence in America a new one, but most social science of the twentieth century found the effects to be subtle, at best. Some of the earliest work on public opinion and politics began with the question of whether the media really matters. This wasnât an idle question: researchers like the Columbia team led by Bernard Berelson in the 1940s were motivated by the propaganda of Nazi Germany (Lazarsfeld et al. 1948). It was thought that a government with control over the media had effectively brainwashed the population of Germany, leading to a fear that the same could happen in the US. The good news was that Berelson and his colleagues found that the mediaâs influence was almost entirely indirect: it mattered only to the extent that it influenced local opinion leaders, what they called the two-step flow of communication (Katz and Lazarsfeld 1955, Katz 1957).
These findings, and others like them, set the stage for whatâs been called the minimal effects model. To simplify, the minimal effects model holds that the media doesnât actually change minds. The media may set the agenda, it may change the issues on which people assess candidates, it may even give people ammunition to defend their opinions â but it doesnât actually change how they feel about a politician or an issue (Finkel 1993, Bennett and Iyengar 2008). The strongest effects of the media were found to be in the area of framing: in a classic experiment, researchers at Ohio State and Purdue showed volunteers one of two news clips about a Ku Klux Klan (KKK) rally in a small Midwestern city. They used the same footage of the rally and counterdemonstrations, and reported the same facts: but half of the group saw a version in which the rally was discussed as a freedom of speech issue, and the other half saw it discussed as an issue of public order. Those that saw the right of the KKK to rally as a free speech issue subsequently expressed much greater support for the rally than those that saw it as an issue of public order. The idea is that the media wasnât going to change anyoneâs mind about free speech or civil disorder or the KKK â but it could change how the concepts were related in the minds of the volunteers (Nelson et al. 1997). This concept of indirect effects has taken hold throughout the study of politics and the media. For instance, the consensus in political science isnât that campaign attack ads make people dislike the candidate theyâre directed at: instead, the debate in political science is about whether or not voters find the ads so distasteful that they decide to just not vote at all (Ansolabehere et al. 1994, Ansolabehere and Iyengar 1995, Ansolabehere et al. 1999).
Some researchers were able to uncover direct effects of media â actual attitude changes, based on the information that had been presented â but these findings were mostly limited to lab studies, and have come under a great deal of criticism. For one, lab studies often make use of student sample, and the students who are recruited into these studies are different from the general public in important ways: theyâre whiter, richer, have more education and are much younger than most Americans. Maybe campaign ads and news coverage impacted the participants in these studies in ways that they wouldnât impact others. Similarly, lab studies generally force their participants to watch the political content, or at least other shows into which the political content is embedded, and thatâs not quite realistic either. These participants might normally have skipped over political ads with a digital video recorder, or not watched the sorts of programming in which political content normally comes up. Even if political information can potentially have an impact on these groups, it doesnât really matter if members of that group are never actually exposed to it. Indeed, as media choice has increased, those individuals who are most likely to opt in to political content are those who have the strongest attitudes to begin with (Prior 2007, Sunstein 2007).
The best lab-based research on the effect of the media on political attitudes comes from researchers at Temple and University of California, Riverside (Arceneaux and Johnson 2013, Arceneaux et al. 2012). In many of their studies, participants were given a choice of what channels to watch, allowing the researchers to avoid the forced exposure issues that have been problematic in other studies. They find that individuals with strong initial political beliefs choose media that matches those beliefs, and when they do happen upon media that goes against their predispositions, their beliefs only become hardened. This leads Arceneaux and Johnson (2013) back to something close to minimal effects: ideological media isnât changing the direction of attitudes, though it may be changing their magnitude. These sorts of experiments can reveal a great deal about the cognitive processes underlying attitude formation about candidates and issues, but, like all experiments, they face problems of external validity. Even the best experiments make use of a relatively small and non-representative sample of the American public, who may or may not act differently because they know that theyâre part of an experiment. In addition, the relatively small sample size of experiments (even the best have sample sizes in the hundreds) means that they may be unable to pick up small shifts in opinion, especially small shifts among subgroups of the population, that may be evident from large-scale samples.
Researchers who have looked for large-scale effects of media content on voting outcomes and polls have generally come up empty-handed (Bartels 1993, Finkel and Geer 1998, Wattenberg and Brians 1999); the conventional wis...
