Red Media, Blue Media

By Shanto Iyengar and Richard Morin
Wednesday, May 3, 2006; 6:00 PM

Thirty years ago, you had to visit the public library to read something other than your local newspaper. Today, thanks to information technology, newspapers, radio and television networks the world over are no more that a keystroke away. Does this dramatic expansion of available news outlets mean that Americans -- and others worldwide -- will be exposed to a more diverse "marketplace of ideas," gain familiarity with new points of view and become more tolerant? Or will consumers stick with their preferred (politically compatible) news sources while screening out those sites offering unfamiliar or disagreeable information and perspectives?

As yet, there is little evidence to indicate that consumers apply a political litmus test in deciding which news sources to use. We designed this study to find out whether people do in fact prefer news reports from sources that they believe to be sympathetic or compatible with their views. We observed whether attention to the identical news story was increased or decreased when the story was attributed to Fox News, NPR, CNN or the BBC. The results demonstrate considerable polarization in exposure to news. Republicans have distinct preferences for particular news sources (they go to Fox and avoid NPR and CNN). Democrats avoid Fox but divide their attention between CNN and NPR. When the news focuses on controversial issues, partisans are especially likely to screen out sources they consider opposed to their political views. The study design was as follows: Using the MSNBC daily news feed (which includes news reports from a variety of sources), we randomly assigned news stories (for purposes of the study) to one of four sources -- Fox, NPR, CNN, or BBC. Participants were provided a brief headline accompanied by the logo of the news organization and asked to indicate which of the four reports displayed on the screen they would like to read. (They could also click a "can't say" box.) They repeated this task across six different news categories -- American politics, the war in Iraq, "race in America," crime, travel, and sports. We also included a baseline or control condition in which all source logos were deleted; here participants could only choose between the reports based on the text of the headlines.

All other aspects of the presentation were equalized across the different news organizations. For instance, the placement of a particular story or source on the screen was randomized so that no particular news source gained from being the first or last on the screen.

We were able to supplement the "drop-in" sample with a representative national sample of adult Americans made available through the market research firm of Polimetrix, headed by Stanford professor Douglas Rivers. The Polimetrix sampling methodology is available at The use of a second sample permitted us to compare the media choices of Post readers with those of the population at large. This was especially important given the paucity of Republicans (only 11 percent) in the Post's participant pool. The PMX sample, by contrast, was 33 percent Republican.

Our main expectation was that we would find a stronger demand among Republicans and those with conservative political views when reports were assigned to the Fox condition, while readers on the left would be more interested in stories assigned to CNN or NPR. Since the BBC is a foreign news source with a well-deserved reputation for independent journalism, we expected similar preferences for the BBC label among Democrats, Republicans and non-partisans alike. We further expected that the effects of the source manipulation on news story selection would be strong for political topics where partisan divisions are intense, but would have little impact on neutral topics such as travel and sports. As shown in Figure 1 and Figure 2, this is precisely what we found -- the divide between Republicans and Democrats proved considerable when news stories dealt with national politics or the war in Iraq, but was relatively small when the news turned to travel or sports. Even though the partisan divide is greater in the case of hard news, it does not disappear entirely for non-political subjects. Republicans prefer Fox News, even when reading about possible vacation destinations.

Despite the fact that Fox won the Republican "vote" in both samples, there were clear differences in the intensity of preference. While Republicans in the national sample preferred Fox by a landslide margin, Republicans among Post readers were more willing to look elsewhere, resulting in a narrow (although still statistically significant) margin for Fox.

The Democrats did not converge on a single source, but tended to divide their selections between CNN and NPR. Nationwide, Democrats opted for CNN; in the Post sample, they preferred NPR. In one important respect, however, the two groups of Democrats behaved similarly -- they demonstrated an equally strong aversion to Fox. Finally, independents lived up to their designation and revealed no particular source preference, although among those in the Post sample, they were just as averse as the Democrats to offerings from Fox. A more pointed indicator of partisan preference is whether respondents perceive media treatment of President George W. Bush as biased. We asked respondents to indicate whether they thought the news media were either insufficiently or overly critical of the president. As shown in Figure 3, beliefs about the media's treatment of President Bush is a very powerful determinant of the overall preference for Fox news reports. (These results are based only on the hard news categories.) Fox loyalists come disproportionately from those who think the media are too tough on President Bush. On the other side, people who feel that the media should provide more critical coverage of the administration are much more likely to select NPR and CNN over FOX.

As a final indicator of the relationship between partisan preference and news selection, we present results comparing story selection rates with and without source labels (see Figure 4). As part of the design, one-quarter of the study participants were provided with the news reports without source labels. Therefore, we were able to compare the fraction of the study participants who selected the same story when it was either unlabelled, or attributed either to Fox News or CNN and NPR. (The small number of Republicans made it difficult to carry out this story-level analysis in the Post sample, hence we limit this analysis to the national sample.) In this analysis we are comparing hit rates for the same story; any difference in the number of people selecting the story can only be attributed to the presence or absence of the source label.

Figure 4 provides considerable evidence of political selectivity: the very same news story on crime or Iraq or politics or racial issues attracts a different audience when labeled as a Fox or NPR report. The effects of the story label proved most powerful in the case of Fox. On average, adding the Fox label to a story more than tripled the Republican "hit rate" for hard news stories! Consistent with our expectations, the effects of the Fox label were weakened for non-political news. Nonetheless, the effects of the Fox label doubled the selection rate for travel and sports stories among Republicans. While Republicans were drawn to the Fox label, they avoided CNN and NPR. On average, the probability that a Republican would select a CNN or NPR report was around 10 percent.

As for the Democrats, they were just as averse to Fox as the Republicans were to CNN and NPR. But unlike the Republicans, they did not converge on a particular news source. Although the CNN and NPR labels boosted interest among Democrats, the effects were weak. After all, the expected selection rate for any particular source is 20 percent (participants could choose one of the four reports or "can't say"). Thus, the Democratic selection rate for CNN-NPR is just slightly above the "no preference" baseline. Overall, the results suggest that Democrats are exposed to a greater range of news sources than Republicans

Finally, independents lived up to their designation. The effects of story labeling on their selections was miniscule. Most independents had no news preferences; they typically selected the "can't say" option.

No matter how we sliced the data -- either at the level of individuals or news stories -- the results demonstrate that Fox News is the dominant news source for Americans whose political leanings are Republican or conservative (the results presented above are even stronger if we substitute ideology for party identification). Fox's brand advantage among Republicans is especially strong when the news deals with political subjects. The effectiveness with which Fox attracts Republicans suggests that "news with an edge" -- the motto of one popular Fox News show--is no impediment to market success.

Unlike the Republican enthusiasm for Fox, Democrats showed only lukewarm preferences for CNN and NPR. Perhaps the Democrats' brand loyalty is weaker because they find CNN and NPR content insufficiently slanted to their liking. Alternatively, Democrats may be less inclined to seek out one-sided news coverage that confirms their view of the world.

One thing is certain. The importance of source labels to news consumption will only grow as technology diffuses and consumers increasingly customize their online news menus. As this trend progresses, there is the real possibility that news will no longer serve as a "social glue" that connects all Americans; instead, the very same lines that divide voters will also divide news audiences.

Shanto Iyengar is Professor of Communication and director of the Political Communication Lab at Stanford University. Richard Morin is director of Washington Post polling and a staff writer.

© 2006 The Washington Post Company