On its face, the MediaMatters report on the alleged rightward shift of the guest list on Sunday morning talk shows isn’t anything most of us didn’t already know.
The shows, which often are little more than preening sessions for the usual suspects — members of Congress, administration apologists, presidential hopefuls and non-threatening journalists — have been stale and predictable for some time, but the Washington press corps remains addicted to them, perhaps because they are a chance to watch conventional wisdom congeal in front of their eyes. So every Monday we’re subjected to another round of day-after quarterbacking on who’s up, who’s down, and who looked like they had a spray-on tan.
And since the shows are given such oversized importance, that’s what makes the MediaMatters study such a necessary read. If you haven’t looked through the package yet, the liberal media monitor studied the ideological makeup of the 7,000 guests who appeared on the three major Sunday shows from 1997 through 2005, which encapsulates Bill Clinton’s second term, George W. Bush’s first term, and 2005. It found that “the left has of late found itself outnumbered, in some ways substantially, on the television shows that define the Washington conventional wisdom.”
The report found that during Clinton’s second term, “52 percent of the ideologically identifiable guests were from the right, and 48 percent were from the left. But in Bush’s first term, Republicans/conservatives held a dramatic advantage, outnumbering Democrats/progressives by 58 percent to 42 percent. In 2005, the figures were an identical 58 percent to 42 percent.” The study also reports that during both the Clinton and Bush administrations, “conservative journalists were far more likely to appear on the Sunday shows than were progressive journalists. In Clinton’s second term, 61 percent of the ideologically identifiable journalists were conservative; in Bush’s first term, that figure rose to 69 percent.”
All of which brings us to the subject of methodology. This part of the report gave us pause: “These classifications do not represent what each person actually said when she/he appeared on a show on a given date … we simply classified each guest based on her/his general partisan or ideological orientation.”
Reading further, we find that this rule itself was open to interpretation. “While the vast majority of guests are clearly identifiable by party or ideology (or as having none),” the group writes, “there are a few whose public stances make such classification more difficult.” As a result, “there were certain instances in which it was necessary to ascertain the purpose of a guest’s appearance in order to correctly classify her/him.”
To try and get around this problem, MediaMatters chose to classify all administration officials as representing the party in power. Fair enough. Those who had served under both Clinton and Bush were “neutral,” with the exception of guests like former “terror czar” Richard Clarke, who served in both Republican and Democratic administrations but was obviously a Bush administration critic when he made his appearances on the Sunday shows. Hence, he was labeled “progressive.” Then there is someone like Zell Miller, who as a Democrat (in name only, to be sure) has been highly critical of the Democratic Party, who is coded as a conservative.
We’re concerned that all of these exceptions to the classification system really begin to add up. And what seems to be lost in this analysis — aside from a few prominent exceptions — is the all-important matter of context. As Vaughn Ververs asked last night on CBS’ Public Eye blog, “what real conclusions can be drawn from a study examining labels rather than what was actually said? Are all public officials, commentators and even journalists simple automatons, spitting out the party line without independent thoughts of their own?”
Granted, there’s little doubt that guests like Colin Powell or Donald Rumsfeld spouted the party line during their segments when they were members of the Bush administration — but defining journalists or former administration officials isn’t quite so easy. David Brooks for example — a conservative — would probably be tough to pin down on several issues (and incoherent on others), as would many other daily reporters and columnists.

Isn't this kind of analysis of the shortcomings of the Media Matters piece kind of irrelevant unless you take your own advice and look at a sample of what the speakers on the shows actually said, Mr. McLeary?
Maybe they're wrong. Why not do some reporting and prove them wrong? Or right?
Posted by darrelplant on Wed 15 Feb 2006 at 04:28 PM
As the author of the study discussed in this entry, I have to object to your core characterization, that because the report answers some important questions, but not every important question, it "suffers from one flaw." That's like saying that oranges suffer from the flaw of not being apples.
Is context important? Of course. Could a content analysis make accurate ideological categorizations of the comments of nearly 7,000 guests on Sunday shows? Maybe, if you had about 20 years to do it. We decided to address one question: who gets to appear on the Sunday shows. Of course one could address other questions, but that's hardly a "flaw" in our study.
Furthermore, you pick out from the report's methodology section a couple of EXCEPTIONS to the coding rules, then argue that they "really begin to add up." Well, they don't. Zell Miller made 2 appearances on the Sunday shows during the period covered. Two, out of nearly 7,000. As for the other guests for whom we had to ascertain their purpose for appearing, most of them were attorneys appearing on behalf of clients; we coded them according to the client's ideology or partisanship.
Many of the doubts you raise were also raised by CBS's Public Eye blog, which we addressed at length here. I'd encourage people to read it.
Paul Waldman
Media Matters for America
Posted by Paul Waldman on Wed 15 Feb 2006 at 04:59 PM
Exactly the type of non-reporting CJR Daily is supposed to be fighting against. And this is the second time in a week that I've had to say so. WTF?
Props to Paul Waldman for the reply, and for the good work.
Posted by phaedral on Thu 16 Feb 2006 at 10:49 PM
Good critique. Any report that calls Clarke a "Progressive" needs to examine itself.
Posted by TomMessner on Fri 17 Feb 2006 at 09:18 AM