Let’s try an election without pollsters

The young prince of Democratic politics glided into New Hampshire on a magic carpet of soaring poll numbers, great expectations, adoring crowds and a swooning press corps. The town halls and college auditoriums couldn't accommodate the thousands who turned out to hear his message of hope and change. Even before the first votes were cast in the nation's first presidential primary, the pollsters, pundits and seasoned political reporters who should know better rhapsodized about how Barack Obama, fresh from a big victory in the Iowa caucuses, was on the verge of cinching the nomination and running Hillary Rodham Clinton off the road in New Hampshire.

Obama was more than a candidate, the story line went. His candidacy was a movement, a transformational force in American politics.

It was almost enough to cause Bill Clinton to gag.

"The whole thing is one of the biggest fairy tales I've ever seen," he fumed. And it was enough to move Hillary Clinton close to tears when a woman asked her how she kept going. Then New Hampshire voters rewrote the script. They gave Clinton a narrow victory and a political resurrection. And the pollsters and pundits were left to wonder what had happened, why the Obama fairy tale had not come true, at least in the Granite State, as the poll numbers had predicted.

There were the usual excuses and theories, some more plausible than others, and the usual self-flagellation among pundits and cable gasbags. Maybe the problem wasn't the poll numbers but the way they were interpreted and used. Remember, these same pollsters got the Republican contest right. And even the Clinton campaign's own tracking numbers suggested the former first lady, who finished third in Iowa, was about to go down for the count in New Hampshire.

To me, the real question is not why the pollsters got it wrong. It is, should news organizations even be in the business of conducting horse-race polls and using the results to make virtual predictions about election outcomes before the first votes are cast? Imagine a presidential election where there is no horse-race polling, at least for broadcast or publication. Instead of telling voters which candidates are climbing or sinking, pollsters could measure public attitudes toward candidates and the issues. Of course, I'm not naive. It's not going to happen. Horse-race polling is the crack cocaine of political journalism. And many newspaper readers and television viewers share the addiction.

Consider what Public Agenda, a nonpartisan opinion research organization, said last week in a posting on its Web site: "Predicting elections is one of the most difficult challenges in survey work. Generally speaking, election surveys actually do work fairly well; it's worth remembering that the polling on the Republican side in New Hampshire was pretty accurate. But in the end, even when the election surveys are right, does it matter? Sure, it's fun and even newsworthy to predict elections and try and handicap races, just like it's fun to predict the NFL playoffs. But does it help the voters make good decisions? Does it give leaders the information they need to craft policy? Does it even help the media to cover elections well? Not really. For all their flaws, polls can make a real contribution to democracy if they are used wisely. However, when they are used the way they were in New Hampshire, they can cheapen our politics and erode the news media's credibility.

"The worst part of all this debate is that it makes people skeptical of the truly useful information surveys can provide, by probing the public's values, finding out what worries them and asking them to set priorities for the nation," Public Agenda said. "The danger is that the cynicism engendered by mistakes in election polls will taint all surveys." At least Tom Brokaw, the former NBC news anchor, understands the problem. In an exchange with MSNBC gasbag Chris Matthews, Brokaw suggested the news media should "wait for the voters to make their judgment" before predicting outcomes.

Matthews: "Well, we must then stay home, I guess."

Brokaw: "No, but we don't have to get in the business of making judgments before the polls have closed, and trying to stampede, in effect, the process."

Maybe Brokaw could lead an effort to persuade major news organizations to try an experiment this fall. The television networks and leading newspapers would agree to forgo horse-race polling in the general election, or at least to not report the results, to see if it makes a difference in how the campaigns are conducted and covered.

Meanwhile, voters in the rest of the country should be grateful that contrarian New Hampshire did not turn its primary into a coronation for Obama. Democrats now have an exciting choice to make — to decide whether their party's presidential nominee will be an African-American or a woman. Either way, history is about to be made.

(Philip Gailey is editor of editorials for the St. Petersburg Times. E-mail gailey(at)sptimes.)