Can partisan media contribute to healthy politics?

On Monday at 5 pm, I’m participating in a South by Southwest panel entitled “How Partisan Media Contributes to Healthy Politics.”  I prefer to think of this as a question: can partisan media contribute to healthy politics?  For my contribution, I want to do two things.  The first is report on the available social science to show that partisan media might not be as powerful as is sometimes suggested.  I think that’s an important piece of context for this discussion.  The second is to raise some questions about whether and how partisanship—an often maligned notion—can play a valuable role in democracy.

The audience for partisan news is not as big as you might think.

What percentage of Americans watches cable news for 10 minutes or more per day?  Only about 10-15%, if you simply add up the audiences for Fox News, CNN, and MSNBC.  This is based on calculations by political scientist Markus Prior, drawing on detailed data about what people actually watch and not what they report in a survey.  Survey reports of news consumption are often highly inaccurate.  Consider this comparison of a 2008 Pew survey to data on viewership from the Nielsen Company:

In the survey, almost a third of Americans believe they watch one of the three cable networks “regularly.”  It’s not quite clear what “regularly” means, of course.  This is one of the problems of using survey questions to measure media exposure.  But if we assume that a regular viewer should watch at least an hour per week, then in reality only about 6-7% of Americas meet that description.

And even those numbers may be too high, because they double-count anyone who watches more than one of those channels.  The seemingly inconceivable possibility that someone might watch both Fox and MSNBC leads to the next point.


Most people are news omnivores.

Most people’s “diet” of news isn’t all that skewed by their partisans.  There is actually a lot of overlap viewers of various cable news networks.  Markus Prior reports that people who watch at least 1 minute of Fox News each week devote about 7.5% of their news consumption to Fox but 3.7% to other cable news channels.  The same is true of CNN viewers.  This is consistent with the research of Michael LaCour, who tracked media usage via devices that participants carried with them and that regularly recorded the ambient sounds around them.  It is also consistent with the research of Matthew Gentzkow and Jesse Shapiro, who examined news consumption on-line and found that most consumers read ideologically diverse new outlets.

Unsurprisingly, if you isolate people who watch a lot of Fox News or a lot of MSNBC, their viewing habits reflect more skew.  But this is a small group of people.  The same is true of people who read political blogs: they are anything but omnivores, according to my research with Eric Lawrence and Henry Farrell, but they are also a small fraction of the public.

Prior has an excellent summary of these points:

Automatic tracking of television viewing using two different technologies reveals that most people avoid cable news almost entirely. A large segment watches cable news infrequently and nonselectively, mixing exposure to different cable news channels. In the small slice of heavy cable news viewers, however, partisan selective exposure is not uncommon.


Partisan news may not polarize partisans, but attract polarized partisans.

There is surprisingly little research that attempts to deal a fundamental issue.  Do people who watch partisan news become more polarized, or do people with polarized views simply like to watch partisan news?  In one experiment, political scientist Matthew Levendusky randomly assigned people to watch partisan news that either did or did not share their political outlook, or to a neutral news source.  He found that partisan news that reinforced subjects’ political outlook made their attitudes modestly more extreme.  This effect was stronger among those who said that they preferred to consume news that shared their political outlook—suggesting that even if the people who watch partisan news are already pretty partisan, partisan news will make them more so.

However, other research by Kevin Arceneaux and Martin Johnson arrives at a different conclusion.  They conducted a set of experiments and allowed people to choose whether they watched their side’s partisan news, the other side’s news, or entertainment programming that had no news content.  They found that the news shows had no effects on attitudes as long as people were allowed to choose.  This suggests that, in the real world, partisan news doesn’t polarize.  If anything, it may be that polarization creates an audience for partisan news.

A few experiments isn’t much of an evidentiary base.  Much more needs to be done.  But it’s worth noting that we don’t really know that partisan news is polarizing us, and with more evidence, we may find that it isn’t.


Learning to love partisanship.

As you can tell from the title of the panel, it was deliberately framed as a provocation.  It’s sometimes (often? always?) hard to like partisan news and even partisanship itself.  But here is the trade-off I want to emphasize.  We want politics to involve calm, civil, rational deliberation about the common good.  Partisanship doesn’t necessarily facilitate that goal and can actively detract from it.  But we also want politics to be full of active, eager, and engaged citizens.  Partisanship does a very good job of facilitating engagement.  It’s one reason why voter turnout was so high in the late eighteenth century during the heyday of strong party organizations and a largely partisan press.

Indeed, if you look at partisans in the public, they look like ideal citizens in many respects.  In a December 2011 YouGov poll, 65% of people who identified as “strong” Democrats or Republicans said they were “very much interested” in politics.  Only 35% of those who identified as independents with no partisan leaning said that.  Partisans are more likely not only to follow politics but to participate in it.  Indeed, it is sort of odd to expect people to care deeply about something but then tell them they’re not allowed to have strong opinions.  It’s like saying, “You should love baseball, but please don’t actually root for a team.”

I’m not suggesting that partisanship is an unalloyed good.  Partisans can be misinformed if they are buying the spin their side is selling—spin that, by the way, they can usually hear in neutral news outlets doing “he said, she said” reporting, not simply in partisan news.  Partisanship militates against other democratic goals, like tolerance for opposing points of view.  Or compromise.

And, in any case, we’ve only got an hour in this panel, so we’re hardly going to resolve this.  I just think it’s worth exploring these tensions in the folk theories we have about politics.

If you’re at South by Southwest, please drop by the panel!  I’m joining Christina Bellantoni, the political editor of the PBS Newhour, and James Kirchick, a writer, blogger, and fellow at the Foundation for Defense of Democracies.

10 Responses to Can partisan media contribute to healthy politics?

  1. John Jay March 10, 2013 at 10:01 pm #

    “If you’re at South by Southwest, please drop by the panel! I’m joining [. . .] James Kirchick, a writer, blogger, and fellow at the Foundation for Defense of Democracies.”

    Good luck with that . . .

  2. Jumper March 11, 2013 at 10:49 am #

    Try comparing all that with newspaper and text website news. And rate text/print perceived reliability in comparison with TV.

    I don’t know how typical I am, but I rate print I disagree with about as high as TV that I do agree with, as far as accuracy and gross spin levels.

  3. Matt Jarvis March 11, 2013 at 6:05 pm #

    Great. Now I have to redo my media lectures for Political Behavior. Thanks a lot!

  4. xpostfactoid March 12, 2013 at 3:36 pm #

    I have a pretty strong appetite for partisan written news but watch next to no TV news. Am I statistically insignificant?

  5. Nameless March 14, 2013 at 6:31 pm #

    This article has a big and seemingly obvious problem which renders most of the conclusions moot.

    The discussion starts with partisan media, but then somehow inexplicably “partisan media” is replaced with “cable news”, losing the majority of partisan viewers in process. (Prior’s article confirms, on page 23: “The large majority of partisans followed mostly local news.”)

    Yes, FNC viewership is low. As per Prior’s article, fewer than 10% of adults watch FNC more than 1 hour/week. The problem is that FNC viewership is the wrong number to look at. I’m pretty sure that many Americans aren’t even aware that FNC exists. For most, Fox is the local broadcast channel that has Simpsons, and Fox News is the evening TV program on that channel. It is hardly less partisan, but definitely more popular.

    Hard data is hard to come by, but there’s a Nielsen-sourced estimate that the cumulative viewership among all local Fox affiliates is ~6M for the 10 pm news, and 2M each in three other popular time slots. The estimate does not come with definitions, but compare that with the viewership of the highest-rated FNC show, The O’Reilly Factor in the 8 pm timeslot: 3M viewers aged 2+ and 500K viewers in the 25-54 age bracket. (In case you are wondering how this discrepancy is possible, the answer is that the median age of the viewer of TORF is supposed to be somewhere in the vicinity of 70.)

    It’s not clear whether the 6M figure for the 10 pm news refers to everyone or to some narrower demographic like 18-49’s or 25-54’s (they are the most “interesting” brackets for advertising and often those are the numbers being cited implicitly.) In either case, local news beat FNC at least by a factor of 2. I would also suspect that FNC’s viewers are more regular (its viewers are more likely to watch the same show every day) and that you need to add up local TV viewers from nearby time slots, at least from 5, 10, and 11 pm (the person who watches local news at 10 isn’t likely to watch it again at 11, and the person who watches at 5 is less likely to tune in at 10 or 11.) On the other hand, the 3M viewers of The O’Reilly Show at 8 pm are likely to overlap a lot with the 2M viewers of Hannity at 9 pm. Both these factors further increase the difference between the total number of different 1+-hour-per-week viewers of local Fox affiliates and FNC.

    • John Sides March 14, 2013 at 7:57 pm #

      Nameless: I am certain that a lot of people watch local Fox affiliates, and some watch local news on those Fox affiliates. But you haven’t shown any evidence that local news on Fox is a partisan medium. In my anecdotal experience, it is like most local news: mostly crime, weather, sports, and human interest stories.

      • Nameless March 14, 2013 at 9:48 pm #

        Strictly speaking, neither did you or Prior provide any evidence that FNC is a partisan medium.

        But if you want evidence, we can start here. It references a poll where respondents were asked about the network that is their primary source of news. Options were ABC, CBS, NBC, CNN, Fox News, PBS, and NPR. In case of Fox, they did not discriminate between the cable version and the local TV version, but we can presume, based on viewership data above, that local TV viewers dominated the set of those who picked Fox. Respondents were then polled on their knowledge of facts related to the Iraq war (e.g. did we find WMD in Iraq?) and Fox viewers were found to be the most misinformed of all groups. This is a poll from back in 2003 when FNC wasn’t even available in half or so households.

        The same thing is seen in a Stanford study here, respondents were asked simply, “During the last 30 days, on about how many days did you watch Fox News on television?” (again without emphasis on the cable version), and responses were correlated with the degree of misinformation.

        More generally, since almost everyone has a local Fox affiliate and the viewership of FNC is so low, any study of Fox viewers that does not emphasize that they are talking about the cable channel and does not go to great lengths to enhance the signal (and to avoid leaving the set of Fox viewers below the margin of error) will inevitably pick up large numbers of local-only viewers.

        Absent strong evidence to the contrary, we have to assume that local affiliates present the same partisan bias when it comes to reporting on national/global topics (which is done most of the time by broadcasting content from the mother company).

        • John Sides March 14, 2013 at 10:22 pm #

          So your theory is that people over-report their viewership of Fox News because they confuse Fox News with their local Fox affiliates? That’s possible. But that only makes Prior’s point more important: don’t trust what surveys tell you about people’s media habits. The corollary then becomes: don’t trust correlations between survey measures of media habits and things like factual knowledge. My view is I don’t have to presume a partisan bias among Fox affiliates because, absent other evidence, I don’t believe either the survey reports or the correlations people generate from them.

        • Nameless March 14, 2013 at 10:24 pm #

          On further thought, thinking about relative viewership of local affiliates vs. the cable channel will, if anything, underestimate the bias of local affiliates.

          Whenever the person is asked anything along the lines of “where do you get your news” or “what is your primary news source” (as in the 2003 Iraq poll above), most of the time that person will think about the evening news on a local TV channel. Its name notwithstanding, Fox News Channel is, strictly speaking, not news. (Or, at least, it’s less “newsy” than, say, CNN.) Of the highest rated shows on FNC, like TORF or Hannity, are not newsreels, they are political commentaries / talk shows. Whether or not the person is a frequent viewer of TORF or Hannity, he/she will report her primary source of news as Fox iff he/she means “evening news of a local Fox affiliate”.

          A similar type of confusion could also explain the paradoxical finding that 18% of Pew respondents claim to be regular viewers of FNC and that whopping 6% claim to be regular viewers of both FNC and MSNBC:
          * 18% of respondents regularly watch FNC or a local Fox affiliate.
          * 6% of respondents regularly watch shows on a local Fox affiliate and on a local NBC affiliate, and don’t realize the distinction between NBC and MSNBC.

  6. Bryan Gervais March 28, 2013 at 9:13 pm #

    One important question: are “omnivorous” viewers just sampling daytime cable news, which more resembles standard news broadcasts and in which partisanship is more subdued, or are they sampling the more high-octane, pundit-oriented prime-time cable news programs as well? I think its problematic that many media studies (and I recognize this is not by choice of the researchers) do not make a distinction between program types on the various networks. There’s a big difference between watching standard news on CNN during the day and watching Hannity at night, and flipping back and forth between the prime time lineups of MSNBC and Fox News. Unless we are prepared to say that there is no difference between tuning into Shepard Smith and tuning into Sean Hannity, it makes little sense to count them both equally when it comes to partisan news exposure.