Vaughn:

You take issue with the fact that our study focused on the simple question of who gets on the Sunday shows, while ignoring “the intra-party dynamic” -- the fact that a few prominent moderates like Sens. John McCain (R-AZ) and Joe Lieberman (D-CT) make frequent appearances. While this may be an interesting topic to explore, it has no bearing on our fundamental findings: that Republicans outnumber Democrats, and that conservative pundits outnumber progressive pundits.

You single out the fact that we coded former Sen. Zell Miller (D-GA) as a “conservative,” then imply that this is somehow inconsistent with the fact that McCain is coded as a “Republican.” We discussed Miller in the report's methodology section as an unusual case that required an exception to the general rule being employed, where politicians are coded by their party affiliation. Miller made two appearances, both during the 2004 campaign, when he was actively campaigning for President George W. Bush. The other person we mentioned with regard to this issue was former presidential counterterrorism adviser Richard Clarke, who under the regular coding rules would have been neutral (since he served in both Republican and Democratic administrations), but appeared on the Sunday shows only after he became critical of the Bush administration, so he was coded as progressive. Similarly, Miller appeared only after his switch to the Republican Party in all but name.

Your comparison of Miller to McCain can only be described as laughable. When McCain endorses a Democratic presidential candidate, delivers a blistering attack on the Republican Party from the podium of the Democratic National Convention, and writes a book-length polemic attacking the Republican Party (followed by a polemic attacking conservatism), then we can talk. Until then, he's a Republican.

You then take issue with our categorization of pundits. “I think you'd get some real arguments from Republicans by classifying David Broder as a 'centrist,' ” you write. Here, you have inadvertently revealed just the kind of bias we worked to avoid. You might indeed “get some real arguments” from some Republicans on categorizing Broder as a centrist, but those arguments would be ridiculous. Can you point to a serious person who would actually claim that Broder is anything but a centrist? You might also “get some real arguments” from some Republicans claiming that George Will is a communist stooge, but that doesn't mean those arguments should be taken seriously. By the same token, some people on the left would call Broder a conservative, but the standard we used was not what some people on the extremes might believe, but where these pundits are generally understood to be situated on the ideological spectrum.

You then raise the issue that not all pundits are down-the-line party loyalists on every issue. “Even commentators like David Brooks, the 'conservative' columnist for The New York Times, don't always march in ideological lockstep with any particular party,” you write. Since you see this as a methodological problem, I'm guessing you have never attempted to perform a content analysis that seeks objective, reliable ideological categorizations. I am not suggesting that you should be expected to have such experience, but if you're going to criticize a study's methodology, you should understand something about how that methodology works. Try to imagine for a moment what it would take to come up with a set of coding rules based on this objection that would achieve what is known as “reliability” -- different coders making the same judgments irrespective of their own biases. As you yourself ask, “Is David Brooks 90% conservative, 10% independent?” It is simply impossible make that kind of judgment in a way that would be persuasive to readers from different ideological perspectives.

The alternative, then, is either not to do the study at all, or to do it in a way that answers some questions as comprehensively and objectively as possible with the understanding that it does not answer every question that could possibly be asked about the larger topic. We chose to do the latter.

You also repeat the argument that Republicans are in charge, and therefore it isn't surprising that their voices dominate. First, this ignores the fact that the degree to which they have dominated is far greater than the degree to which Democrats had an advantage during the Clinton years, as the report documents in detail. Second, let's imagine for a moment that Democrats take back the White House in 2008. If Republicans start arguing at that point that Democrats should rightfully dominate the Sunday shows, I'll eat my hat.

Finally, your argument says nothing about the consistent domination of conservative pundits: In some years, they outnumbered progressive pundits by two, three, or even four times, no matter who was in office. Even if one or two of those conservatives are, like David Brooks, reliably conservative but less than fire-breathing, what does that change? Is it your position that the progressive pundits who are invited to appear on the shows -- the likes of E.J. Dionne, Joe Klein, and Eugene Robinson -- are a group of wacky leftists who are farther from the mainstream than William Kristol, Paul Gigot, or Robert Novak?

We would not claim that our study is the final word in any analysis of the Sunday shows -- there are interesting questions that could still be asked, and more analysis that could still be done. But that doesn't make anything about our study “troubling.”

Paul Waldman
Senior Fellow
Media Matters for America