Here a Poll, There a Poll, Everywhere a Poll

Comedy Central’s Jon Stewart joked Monday about the “latest Reuters/MSNBC/Zogby/Harris/bathroom poll,” with “a margin of error of completely wrong.” At least one news organization is beginning to think the comedian may be onto something.

CNN, one of the pioneers in 1988 in the use of news media-commissioned voter tracking polls -- snapshots of how voter attitudes are changing -- is thinking of giving them up this fall because there are just too many of them.

As the number of competitive news organizations has increased, so has their desire for proprietary information to report what Michael Traugott, professor of communications studies at the University of Michigan, calls “good, cheap news.”

The landscape got notably more crowded in the lead-up to the New Hampshire primary, “where there were six, seven, eight or nine polls coming out each morning,” including CNN’s own joint venture with USA Today and Gallup, said Tom Hannon, CNN’s political director. “With that many tracking polls out there, I’m wondering, should I save my money?” He added that the polls were of “marginal utility” and that he would prefer to come up with “other things that are a better tool and more unique.”

The polls released Monday, the day before the New Hampshire primary, ranged widely in their results, showing Sen. John F. Kerry of Massachusetts with a lead over former Vermont Gov. Howard Dean that ranged from 21 percentage points to 3. By Tuesday, the six polls that were released had largely converged, with several showing Kerry winning by 10 to 13 points. He eventually won by 13 percentage points.

Compared to four years ago in New Hampshire, when some pre-primary polls ended up wrong, or even two weeks ago before the Iowa caucuses, where polls wrongly predicted a close race, the New Hampshire polls “as a group were marvelous this time out,” said Gary Langer, director of polling for ABC News.

CNN’s misgivings notwithstanding, the polls are not likely to go away. Without their own polls, news organizations said they would have to rely on those generated by the campaigns. The day before October’s California recall vote, Langer said, Gov. Gray Davis’ campaign was telling journalists that it had a poll showing the race was dead even; Arnold Schwarzenegger eventually won by a wide margin.

“It is true that the zone has basically been flooded,” said Elizabeth Wilner, NBC News’ political director. But if NBC didn’t have its own Reuters/MSNBC/Zogby International poll, she said, “you’d be forced to rely on” those from the campaigns themselves. “It’s just one more tool you can use to form a picture.”

As the polls have proliferated, they have taken up an increasing amount of air time. Hannon, worried about what he called an “increasing susceptibility to journalism by polling,” a couple weeks ago challenged two CNN reporters to get through the day without referring to poll numbers, which he called “journalistic crutches.”

For Traugott, the new glut of polls “is a bad trend, because it feeds the worst instincts of the press,” to report on the horse race. “The primary focus of the story shouldn’t be on who is ahead or behind, it should be on the dynamics of the electorate and why there is this shift back and forth between Kerry and Dean,” he said. Moreover, he said, in some polls “the samples are small, so the shifts are statistically meaningless, but they are being reported as real.”

However, viewers “like to know who is going to win,” said Marty Ryan, executive producer of political programming for the Fox News Channel. “It’s the American way of looking at things.”

Viewers “want to know who is up and who is down,” said CNN anchor Wolf Blitzer, although he acknowledged that the polls can “be misleading from time to time, yeah.” Still, “based on my experience, they do show some trends,” he said, adding that “we’re in the business of providing as much information as possible to our viewers.”

ABC News hasn’t commissioned its own tracking polls this campaign season, but Langer calls them a valuable tool when done responsibly and used judiciously. Polls, he said, “are results of interviews with real people, not the blowhards and the spinmeisters. It’s you and me and the lady across the counter and anybody with a phone.” However, he said the polls need to be vetted first to make sure they are done reliably, and he noted that some in the news media are too “data hungry and math-phobic.”

It’s not just tracking polls that are shaping up as an issue this campaign year. Many network executives were dismayed Tuesday afternoon when the first wave of data from their specially commissioned exit polls of New Hampshire voters showed up almost immediately on the Internet. Not only was the first round of information not indicative of how the primary would eventually play out, they said, but the networks themselves had agreed not to report the results until the polls had closed, so as not to influence voters.

The networks also are being cautious with their exit polling this year, fearful of a repeat of the 2000 presidential election, when bad data resulted in wrong on-air projections.

Likewise, results of the first wave of data from a separate Los Angeles Times exit poll, which showed Dean leading Kerry at one point, were also being widely quoted Tuesday afternoon in newsrooms around Manchester, leading some journalists to mischaracterize the closeness of the race before the final results were in.

“An exit poll is not complete until it is complete,” said Susan Pinkus, director of the Los Angeles Times Poll. “You can’t look at first-wave data and say this is how it’s going to be.” She added that the data were not supposed to be widely disseminated.