Appearance Alert!
MRC's Brent Bozell on FNC's The Kelly File, Friday 9:40pm ET/PT

TV Tries To Sink Bush With Weighted Poll

Liberal Networks Show Their Bias By Pouncing On CBS Poll Based On A Skewed Anti-Bush Sample

No one would claim right now that President Bush fits the adjective "popular." But the hubbub this week over the "record low" approval ratings CBS News discovered for President Bush raises an important question: which media-outlet polls are trusted and which are not?

If the pollsters for Fox News Channel are ignored, then should the pollsters for CBS, the network where Dan Rather had to resign his anchor chair for sloppy Bush-bashing journalism, be treated as trustworthy and nonpartisan?

A glance at a roundup of presidential approval rating polls shows that the grand canyon between Bush's approval and disapproval ratings of 25 percent (34 to 59) is far outside the usual margin. An early February Fox News-Opinion Dynamics poll had a gap of only three percent (44 percent approve, 47 percent disapprove). No one noticed. The other media-elite polls placed the gap in the mid-teens: ABC (14), NBC (15), CNN (17), AP (17), and Time (14). FNC's brand-new poll has the gap up to 15 points, 39 to 54.

If these approval-rating polls were like Olympic judges, wouldn't the media throw out the high mark and the low mark? Not if you're the "mainstream" media. Instead, they championed the "record low" CBS number as the definitive current Bush approval rating. CBS naturally hyped the poll on its evening and morning shows. ABC promoted it on Tuesday's Good Morning America. NBC's Today promoted it twice Tuesday in news reports and once in a Matt Lauer interview with Chris Matthews on Wednesday. (Katie Couric also mentioned it without citing CBS by name.) None of those reports came close to questioning the poll's internals, that featured more Democrats than Republicans.

As the blogosphere quickly discovered from CBS's online report, CBS "weighted" its sample to reflect an ideal cross-section of American adults. They adjusted the number of self-described Republicans up to 28 percent and Democrats down to 37 percent, and independents with the rest. That's hardly the exit-poll breakdown the networks found on Election Day 2004 (37 percent GOP, 37 percent Democrat, 26 percent independent).

The question for these networks: did the last ten CBS polls get any fraction of the publicity from competitors as this one drew?

The cable networks also publicized the CBS poll. Unsurprisingly, on MSNBC, both Chris Matthews on Hardball and Keith Olbermann on Countdown led off their evening shows with the bad news for Bush.

CNN liked the poll so much they began mentioning it at 9 PM on Monday night, when Larry King asked comic Jon Stewart for his reaction. On Tuesday, CNN promoted CBS's Bush poll in these PM hours: 3, 4, 5, 6, 7, 8, and 10. OnThe Situation Room at 4 PM, blogs reporter Jacki Schechner offered one mention that conservatives complained the poll's internals were skewed. That didn't stop the CBS poll from leading Paula Zahn Now at 8.

Fox News discussed the poll's biased sample both on Tuesday's Special Report with Brit Hume and Wednesday's Fox and Friends.

Pollster Bob Moran wondered on "The Corner" blog of National Review Online: "Why is the sample so Democrat? One reason may be because almost every question bangs the President and I would guess that the hang ups they get are vastly more Republican than Democrat. Think about it. Why would a Republican sit on the phone and answer loaded anti-Bush questions for 15 minutes?"

Is CBS really the gold standard for polling data? One critic on the CBS Public Eye blog suggested, "Let's see CBS conduct a poll of CBS's approval rating. That should give a good indication of the validity of any of CBS's other polls, no matter how they are weighted." - Tim Graham