Transcript
BROOKE GLADSTONE: We're back with On the Media. I'm Brooke Gladstone.
BOB GARFIELD: And I'm Bob Garfield. In two recent polls results of note, USA Today declared that General Wesley Clark would defeat President Bush in a head to head presidential matchup. Also, in an online survey by the website Mister Poll, 26 percent said Britney Spears would be the most likely among the list of female celebrities to keep a man as a slave. What was notable about those results is that neither is true -- for reasons I'll get to shortly. For the moment, let's just agree that we as a society are awash in polls, giving our leaders -- and us as citizens -- an unprecedented amount of data about our thoughts, our beliefs, our aspirations. And plenty of those data, as reported by the media, are simply wrong. [CLIP PLAYS]
DAN RATHER: We would rather be last in reporting returns than to be wrong...If we say somebody has carried a state, you can pretty much take it to the bank. [LAUGHTER]
BOB GARFIELD: The congressmen listening to that replay of Dan Rather's infamous Election 2000 boast about exit poll accuracy found the memory very funny, but it didn't seem so hilarious at the time -- especially to the network-anointed winner, Al Gore. Still, Correction 2000 hasn't slowed the polling industry down any. Gallup, Roper, Zogby, Lou Harris, ABC, NBC, CBS, the New York Times, Los Angeles Times, Washington Post, USA Today -- they're all out there asking questions about what seems to be on America's mind. This can include everything from our approval of the president's performance to our least favorite food (liver), to any number of questions America couldn't possibly know the answer to. [CLIP PLAYS]
WOLF BLITZER: Suzanne Malveaux at the White House. Suzanne, thanks very much for that report, and to our viewers, her--here is your turn to weigh in on this story. Our web question of the day is this: Is Saddam Hussein still in Iraq or in another country? You can vote right now. Go to cnn.com/wolf. I'm anxious to hear what you have to say. We'll have the results later in this broadcast.
BOB GARFIELD: I'd tell you how the vote came out, but what would be the point? For one thing, as with the Mister Poll speculation about the bondage proclivities of female celebrities, who would have any earthly idea where Saddam Hussein is. And because the sample of on line respondents is self-selecting, not random, the tally tells us zero about what America at large thinks about the subject. Yet every time you turn on the TV, it seems, there's Wolf Blitzer or Bill O'Reilly or your local anchor asking you to register your vote on line. Betsy Martin is president of the American Association for Public Opinion Research.
BETSY MARTIN: They're really completely meaningless. I think they're really marketing gimmicks. I mean it's entertainment, but it doesn't say anything about public opinion.
FRANK NEWPORT: Now a lot of media outlets will say this is non-scientific. We, of course, say well why put it on if it's non-scientific.
BOB GARFIELD: Frank Newport is editor in chief of the Gallup Poll.
FRANK NEWPORT: Well the reason they put it on is that consultants tell them that that gets viewer involvement; increases ratings. But we don't like it, because it confuses the public. That's bad.
BOB GARFIELD: In the case of on line polls, the media parlor game constitutes journalistic fraud, but through inattention, or worse, the media can also be complicit in misleading the public about poll results obtained with statistical rigor but which, nonetheless, are dangerously misleading.
FRANK LUNTZ: Not only can polls be used to spin an idea. They are quite often used to spin an idea, to spin a proposition, and sometimes even created it to spin a proposition.
BOB GARFIELD: Former Republican pollster Frank Luntz is now host of MSNBC's America's Voices.
FRANK LUNTZ: There are certain pollsters on both the Republican and Democratic side who are known to be in bed with their parties, who are known to twist and turn data to make it show whatever their political parties want it to show, because they're being paid by their political parties to make it show what they want it to show. And the amazing thing is that the news media still reports it.
BOB GARFIELD: Luntz should know. He was censured five years ago by the American Association for Public Opinion Research for spinning his polls on behalf of the GOP. And such behavior, according to Jim Norman, polling editor at USA Today, is both commonplace and easy to accomplish with leading language, cunning choice of background information or just an inflammatory word or phrase.
FRANK LUNTZ: "Tax and spend liberal," a phrase like that, can automatically cause skewing of a question. Proponents of causes know this and use this to get the kind of poll results they want.
BOB GARFIELD: But there needn't be mischief afoot for the public to be misled by a poll result. Consider that USA Today story about the showdown between Wesley Clark and the president. The same story that declared Clark ahead by two points also stated the margin of error as 4 points, which means, statistically speaking, that nobody is ahead. It's a dead heat -- too close to call. This isn't just a statistical technicality. It's the heart of the meaning of poll numbers. Yet it's routinely tossed off as boilerplate. John Zogby is the founder and president of Zogby Polling.
JOHN ZOGBY: I don't think that a lot of the reporters out there understand what the meaning of "margin of sampling error" is, and I think it's perfunctorily reported, and it's the sort of thing that you devote the last second and a half to. [IMITATING] "The margin of sampling error is plus or minus 3." I think that was a second and a half.
BOB GARFIELD: But basic ignorance of or inattention to the underlying science leads to bad reporting. For instance, Zogby's own organization recently was rebuked by the National Council of Public Polls for misrepresenting as scientific a poll of 600 Iraqis, the results of which suggested long term optimism among 70 percent of the Iraqi population. The rebuke, however, came after the story in the Washington Times which relied on the poll data and was headlined: Iraq -- Better than You Think. There's yet one more issue to consider. How much do we even want our politicians to track our opinion and govern around it? The columnist Walter Lippmann considered routine polling to be a cynical opportunistic abrogation of political responsibility -- a perversion of representative democracy that he said was causing the, quote, "catastrophic decline of Western society." And Lippmann never even heard of Britney Spears.
FRANK NEWPORT: Some people argue that we elect representatives and those representatives should there--go off and be brilliant.
BOB GARFIELD: Frank Newport of the Gallup organization.
FRANK NEWPORT: You know, in a democracy, the people are the source of wisdom and ultimate power, and I would like my elected representative to poll all the time, trying to understand that wisdom of the people.
BOB GARFIELD: Certainly politicians ignore public opinion at their own peril. It has swept the likes of Mussolini and Robespierre into power, only later to send then to the executioner. Public opinion about slavery inflamed the North even as public opinion about state's rights emboldened the South. Public opinion killed Nicolai Ceaucescu and spared Captain Dreyfuss and Barabbas. The very public outrage that lured us into the Spanish-American War extricated us ultimately from Vietnam. While there's always the risk of political power being too heavily invested in the wisdom of the people -- think California -- clearly elected officials must keep their finger on the pulse of public opinion. But they cannot do so without the honest brokership of the media who owe it to us to present an actual pulse point and not just the soothing vibration of some scoundrel or fool running off at the mouth.