Romney, Obama, and Media Bias

by John Sides on May 31, 2012 · 7 comments

in Campaigns and elections,Media

So this Politico story by Jim VandeHei and Mike Allen is making the rounds.  How about we move beyond the claims and counter-claims of Democrats and Republicans to some actual data?  The Project for Excellence in Journalism has been tracking the tone of news coverage since the summer of 2011.  More about their data and methodology is here.

This is the tone of coverage for Obama and Romney through the first week of May (click to enlarge):

What this shows is that every point in the past 10 months, Obama has received more negative coverage than positive coverage.  The tone of Romney coverage has shifted depending on primary campaign events, but, as of the end of April, positive coverage still outweighed negative coverage.  At that point, Romney received about as much positive coverage as Obama received negative coverage.


Now, these data don’t speak to the specific subject matter of the Politico story, which focuses on a couple of stories about Romney that aired recently in the New York Times and Washington Post.  But, in the news media as a whole, the PEJ data suggest that there hasn’t been “blatant bias” toward Obama.

{ 7 comments }

Adam Schiffer May 31, 2012 at 1:11 pm

This nicely highlights some of the most common pathologies of bias controversies. First, the Politico piece is just silly, falling into the age-old trap of claiming systematic bias based on cherry-picked anecdotes. It would take a liberal columnist about 5 minutes to assemble a mirror-image story in which the press beats up on Obama on an issue while giving Romney a free pass (say, Rev. Wright vs. Romney’s role in the Mormon church). But where does that leave us?

So I appreciate the introduction of systematic data. I’m a big fan of PEJ’s analysis and have used it in my work. However, it does raise a more subtle but equally important caution about bias claims. Though you don’t do it here, many bloggers have used that data as evidence that the press is biased *against* Obama. While a quick glance at the chart lends face plausibility to that notion, it’s highly misleading without taking a step back and understanding the composition of the coverage. The pos/neg tally includes ALL statements printed or broadcast about the candidates, regardless of their origin. Thus, during the height of primary season, a high proportion of statements about Obama came from Republican presidential candidates. Meanwhile, he wasn’t actively campaigning, so there was no countervailing stream of positive remarks to offset the Republican attacks.

“Balance” can serve as a useful baseline for ascertaining bias only when reality is balanced, so to speak. Once the general election swings into full gear, PEJ’s data will be far more instructive as to whether the coverage contains a systematic slant.

matt w May 31, 2012 at 3:34 pm

In some ways I find this kind of analysis unilluminating, because it’s impossible to set the baseline of the analysis that the candidates deserve. To pick a hopefully uncontroversial example, if in the 1991 Louisiana gubernatorial runoff, the tone of coverage of Duke was more negative than Edwards, would that reflect bias, or the objective reality that Duke was a worse candidate? Or would it be bias against white supremacists? But then that raises the question of whether bias is always a bad thing.

matt w May 31, 2012 at 3:34 pm

I shouldn’t say it’s unilluminating, just unilluminating on the specific question of media bias.

Steven Miller June 1, 2012 at 5:40 pm

I’m not convinced this model makes sense. Just looking at positive vs. negative mentions in aggregate in the media space has little to do with the actual impact. If this were weighted by the actual influence and reach of individual media properties would the result look anything like this?

I think it’s quite doubtful it would.

John Sides June 1, 2012 at 8:06 pm

Steve: How would it look different? Do more widely read/viewed media outlets have different views of Obama or Romney than less widely read/viewed outlets? I don’t really have a strong reason to suspect that they do.

J.W. Hamner June 1, 2012 at 8:47 pm

But what is the baseline media coverage of an incumbent vs. a challenger? Would we really be surprised if a sitting president generates more negative press than fluff pieces 3 years into their term? That a challenger would get a lot lighter coverage as the press was still getting to know him/her?

Obviously we can’t know that since the data set started in 2011, but I’d hesitate to draw strong conclusions from it even though /especially because it confirms my own biases about media coverage.

StewartIII June 2, 2012 at 12:13 pm

NewsBusters| Pew, That Stinks: Liberals Use Rotten Charts to Claim ‘You Can’t Sanely Argue’ Pro-Obama Bias
http://newsbusters.org/blogs/tim-graham/2012/06/02/pew-stinks-liberals-use-rotten-charts-claim-you-cant-sanely-argue-pro-ob

Comments on this entry are closed.

Previous post:

Next post: