Spinsanity: Countering rhetoric with reason
Home | Columns | Posts | Topics | Email list | About | Search

Media Research Center's network bias claim in the Condit affair

By Bryan Keefer (bryan@spinsanity.org)
July 23, 2001
[Email this to a friend]

[Sidebar to Spinning Chandra: Making the Condit Affair Partisan]

The Media Research Center report claiming to show network news bias in the Rep. Gary Condit/Chandra Levy scandal is misleading.

The report claims that in 92 percent of the stories about Condit, the networks have failed to include his party ID. "Normally, a "Republican" or "Democrat" label is presented nearly every time a member of Congress is cited, as in "Rep. Gary Condit (D-CA)."," says the study. "But since May, the three broadcast networks have practically erased the "D" from Condit's political identity, detaching the scandal-plagued politician from the rest of his party."

The first and most glaring problem is that there is no baseline to compare this number to. To be meaningful, the study would need to establish a pattern of labeling that is out of the ordinary. For example, if might take into account how often Senate Majority Leader Tom Daschle is identified as a Democrat, how often House Speaker Dennis Hastert is identified as a Republican, and perhaps even how often President Bush is labeled a Republican. The latter would be especially interesting, given that MRC has claimed bias against Bush. Regardless, the 92 percent number is meaningless without a statistical baseline to compare it to.

In television appearances, Bozell and others citing the study have claimed anecdotally that Senator Packwood was consistently labeled a Republican in stories about his sexual peccadilloes, as was former Speaker of the House Newt Gingrich. Here's Bozell on "Hannity and Colmes" on July 13:

BRENT BOZELL, MEDIA RESEARCH CENTER: Well, sure. I mean, when only 8 percent of the stories on this congressman mentioned his party affiliation, you have to scratch your head and wonder, especially when you compare it to the way those very same networks cover the Republicans.
Just think of the way they covered Newt Gingrich during the book signing scandal or whatever that was. I mean...
HANNITY: "Republican Congressman Newt Gingrich." Did you go back...
BOZELL: No.
HANNITY: ... and look at those numbers, Brent?
BOZELL: We didn't look at the numbers, but I just remember every single mention. I can remember every single mention that Dan Rather had of Ken Starr. It was always "the Republican special prosecutor Ken Starr." What about Katherine Harris in Florida? "The Republican secretary of state." They make it a point to mention that they're Republican, which is fine by me, but when it comes to a Democrat, no.

Argument by anecdote, especially given the source in this case, is no substitute for actual data.

There is a second major problem with the study: it takes only verbal labels into account. Working from transcripts, MRC may not have access to the captions attached to pictures of Condit. This is a flaw that seriously compromises MRC's findings, yet the report makes no acknowledgement of it.

Finally, the MRC study is compromised by its incompleteness. It takes only ABC, NBC and CBS into account; CNN, Fox News and MSNBC are all excluded. In a study billed as thoroughly assessing media bias, this is an inexcusable flaw.

Subscribe to our email list and you'll always know who's trying to spin you.

Home | Columns | Posts | Topics | Email list | About | Search

This website is copyright (c) 2001-2002 by Ben Fritz, Bryan Keefer and Brendan Nyhan. Please send letters to the editor for publication to letters@spinsanity.org and private questions or comments to feedback@spinsanity.org.