Media reach to minimize Bush's poor poll results

››› ››› PAUL WALDMAN & KURT DONALDSON

A recent poll conducted by CBS News that placed President Bush's approval rating at 34 percent has become the target of misleading and uninformed attacks by conservatives in the media.

A recent poll conducted by CBS News that placed President Bush's approval rating at 34 percent has become the target of misleading and uninformed attacks by conservatives in the media.

Among polls taken in the past few weeks testing opinions of the president, the CBS poll result for the president's approval is at the low end; recent polls compiled on the Polling Report website put Bush's approval at an average of approximately 40 percent. The lower approval rating reported by the CBS poll, apparently, is what set off most of the discontent among conservative media figures.

The bulk of the attacks by conservatives concerned the fact that CBS' sample included a higher percentage of Democrats than they contend accurately reflects the general population. But in fact, according to Republican strategist Richard A. Galen in a column on the conservative Cybercast News Service, the adjusted percentages in the poll -- 37 percent Democrats and 28 percent Republicans -- do in fact reflect the proportion of Democrats and Republicans in the general population.

Moreover, pollster Mark Blumenthal has noted on the Mystery Pollster weblog that even if the sample were weighted to make it more closely reflect previous CBS poll samples, the results would likely be similar because of Bush's low approval among independents.

As Media Matters for America noted, during the February 28 edition of The Big Story with John Gibson, Fox News correspondent James Rosen minimized the results of the CBS poll by contrasting it with the February poll showing the highest ratings for Bush, the Diageo/Hotline poll (subscription required), which reported 45 percent approval for Bush, and adding an incorrectly calculated margin of error to show a "theoretically" high rating for Bush.

Later in the same program, host John Gibson asserted that the sampling of the CBS poll is questionable. "Of course, it's weighted with more Democrats, so you've got to take that into account," Gibson said. Additionally, on the February 28 edition of Fox News' Special Report, host Brit Hume, like Rosen, cited a more favorable poll for Bush by Rasmussen Reports, implied a problem with the sample of the CBS poll, and asserted, "Yes, there's good reason to be skeptical of this CBS poll. It's wildly oversampled Democrats, it appears, anyway." And Sean Hannity, co-host of the Fox News' Hannity & Colmes, offered a similar argument, falsely stating, "The first thing we find out that nearly two to one they polled Democrats."

Further, on the February 28 broadcast of his nationally syndicated radio program, Rush Limbaugh stated, "The sample that they used in this poll, 28 percent Republican, 37 percent Democrat, after they weighted the Republicans up ... I mean, you got a sample of 40 to 27, Democrat over Republicans. ... This is not representative of the -- of the population of the country in any way, shape, manner, or form. Nor is the fact that Bush has 34 percent. Nor is the fact that Cheney has 18. You just know that's not possible. It simply isn't possible."

But Galen's Cybercast News column suggested that conservatives' concerns about the sample were misplaced:

* CBS had a sample of 1,018 respondents which they weighted to reflect 28 percent Republicans; 37% Democrats; and 34% Independents. Not likely voters, but adults in the American population.

* In the general population, those who claim to be Democrats outweigh those who claim to be Republicans by 7 to 9 percentage points.

Moreover, on his Mystery Pollster blog, Blumenthal explained that a sample with a lower percentage of Democrats would have had little or no effect on the results, since the CBS sample differs from others not in its percentage of Republicans, but in the fact that it contains slightly more Democrats and slightly fewer independents:

Some will no doubt seize on the fact that the latest CBS News sample is a few points more Democratic on party ID (37%) than on their last three surveys (34% in late January, 33% in early January and 32% in December), although the Republican percentage (28%) is about the same as the last three surveys (27%, 29% and 28% respectively). However, the difference in the party results does not explain the drop in the Bush job rating, which occurs across all three categories.

In fact, even when MP [Mystery Pollster ] recalculates the CBS job approval results for the most recent survey using the average party composition reported on their last three surveys (33% Democrat, 28% Republican, 39% independent or other), the Bush approval percentage still rounds to 34%. The reason is that my recalculation just increases the number of independents at the expense of Democrats. However, Bush's rating is now so low among both subgroups as measured by CBS that the adjustment makes little difference.

To sum up, Bush's ratings are so low among independents that a sample with fewer Democrats would yield virtually the same overall approval rating as the most recent sample used by CBS.

The sample was not the only issue about which conservatives criticized the CBS poll. Some (for instance, Bob Moran of the Republican polling firm Fabrizio, McLaughlin & Associates, in a letter posted on the National Review website) have taken issue with the fact that the CBS poll surveyed adults and not registered voters. Moran wrote: "CBS and most news orgs poll adults instead of reg. voters for two reasons (a) it costs less and (b) it makes the data more Democratic." It does indeed cost slightly less to poll adults instead of registered voters, since fewer potential respondents are rejected and therefore fewer calls must be made. But Moran's conspiracy theorizing aside, the reason that news organizations (unlike partisan polling firms) poll adults is that they are looking to find out what the public thinks. The public includes people who are not registered to vote. When they do election polls to determine how people will vote, they poll voters. When they want to know the state of public opinion, they poll adults.

From the February 28 edition of Fox News' The Big Story with John Gibson:

ROSEN: John, good evening.

No doubt the poll numbers are not where President Bush would want them to be, but most major surveys put him above 34 percent. The Hotline, for example, has him at 45 percent, with a margin of error that could go up to seven points, meaning theoretically the president could be at 52 percent approval ratings, by the Hotline's guide.

An average of all the major surveys has the president at 40 percent.

[...]

GIBSON: President Bush is on his way to India, leaving behind some of his troubles at home. A new CBS News poll puts the president's job approval number at an all-time low, 34 percent. Of course, it's weighted with more Democrats, so you've got to take that into account.

From the February 28 edition of Fox News' Special Report with Brit Hume, with Nina Easton, Washington bureau chief of The Boston Globe:

HUME: The president's approval rating has fallen to an all-time low of 34 percent and has dropped eight points since last month, that according to a new CBS poll in which just 30 percent approved of how President Bush is handling the war in Iraq. And for the first time, most Americans say in this poll that the president doesn't care about people like them.

The latest Rasmussen Reports poll also shows support for the president dropping over the past week, but his approval rating in that survey is 43 percent. Rasmussen all but perfectly forecast the outcome of the '04 presidential election, so is there a reason the CBS poll numbers are so much lower?

It turns out, even after weighting their sample, 37 percent of respondents in the CBS poll called themselves Democrats, compared to just 28 percent who said they were Republicans.

[...]

EASTON: Well, the good news is it's not as bad as the old man's polls in 1992. Former President Bush dropped to 29 percent the summer of 1992. So he could take a little bit of heart in that.

But the polls -- whatever numbers you come out to, and I think probably the average is probably the best --

HUME: Yes, there's good reason to be skeptical of this CBS poll. It's wildly oversampled Democrats, it appears, anyway.

EASTON: Even so, it's not great news. And it's not unlike Bill Clinton's approval numbers in 1994. And there's a lot of temptation, I think, to compare it to 1994 when, of course, the Republicans swept into Congress, Democrats lost control of Congress for the first time in 40 years.

From the February 28 edition of Fox News' Hannity & Colmes, with Fox News host Oliver North:

HANNITY: Look at, for example, you have these polls that came out today in CBS. And if you look at the headline, it says, "Bush lowest number in his presidency." The first thing we find out that nearly two to one they polled Democrats.

NORTH: Sure.

HANNITY: Number two, the questioning was flawed. They ask about, for example, wiretapping, but they don't point out, oh, it's an Al Qaeda member that's a suspect and it came from outside the country. They never give the specifics. It's like -- you almost believe in conspiracies, only I'm not a conspiracy theorist.

NORTH: No, and I don't think it's a conspiracy. I think it's more of what you would call the so-called mainstream media doing what they do best, that is telling only part of the story. You had to read down about five column inches in it to find out what was the basis for this poll.

HANNITY: Coming from the very same network that, 55 days before an election, presidential election, forged documents, what can we expect? Why do they even have any credibility?

NORTH: Well, they have credibility, only because there is a two-party system in this country. There are some people who hold their noses and vote. Look at -- Joe Lieberman is a responsible Democrat.

From the February 28 broadcast of The Rush Limbaugh Show:

LIMBAUGH: CBS poll -- there is something about this company, this news division and George W. Bush that -- I mean -- it goes back a whit -- but was something is festering in this -- this poll is as irresponsible as any poll that they have ever done. This poll that shows Bush with a 34-percent approval rating. By now, some of you have gotten the -- the details. But let me give them to you if you do not know.

This is a survey, first of all, of just adults, Americans. This is not a survey of voters. It is not a survey of likely voters. It is not a survey of registered voters. It's just a random sample of "Will you answer the phone today, and I got some questions for ya." And the questions are absurd. And the questions are misleading. And the questions lie. And this poll was produced to make news. Just as they all are these days. This whole -- this poll was nothing more than a way for CBS to get its editorial position out there as a so-called news story. That's all this poll was.

The -- the sample that they used in this poll, 28 percent Republican, 37 percent Democrat, after they weighted the Republicans up. In other words, the sample was even more Democrat when it came back. Democrats were 40 percent of the sample. Republicans were only 27 percent, and the independents were given a huge weighting. Well, not even a weighting. They were just given a huge percentage of the respondent poll, or pool.

Some -- someone -- I mean, you got a sample of 40 to 27, Democrat over Republicans, I mean that's -- that's a dead giveaway what the intent here is. And they haven't even -- the people at CBS didn't even bother changing this to put both the numbers in the 30s to have a relatively equal. This is not representative of the -- of the population of the country in any way, shape, manner, or form. Nor is the fact that Bush has 34 percent. Nor is the fact that Cheney has 18. You just know that's not possible. It simply isn't possible.

We've changed our commenting system to Disqus.
Instructions for signing up and claiming your comment history are located here.
Updated rules for commenting are here.