Karl Rove says polls have problems

From the Wall Street Journal Opinion by Karl Rove

There has been an explosion of polls this presidential election. Through yesterday, there have been 728 national polls with head-to-head matchups of the candidates, 215 in October alone. In 2004, there were just 239 matchup polls, with 67 of those in October. At this rate, there may be almost as many national polls in October of 2008 as there were during the entire year in 2004.

Some polls are sponsored by reputable news organizations, others by publicity-eager universities or polling firms on the make. None have the scientific precision we imagine.

For example, academics gathered by the American Political Science Association at the Marriott Wardman Park Hotel in Washington on Aug. 31, 2000, to make forecasts declared that Al Gore would be the winner. Their models told them so. Mr. Gore would receive between 53% and 60% of the two-party vote; Gov. George W. Bush would get between just 40% and 47%. Impersonal demographic and economic forces had settled the contest, they said. They were wrong.

Right now, all the polls show Barack Obama ahead of John McCain, but the margins vary widely (in part because some polls use an "expanded" definition of a likely voter, while others use a "traditional" polling model, which assumes turnout will mirror historical trends but with a higher turnout among African-Americans and young voters).

On Monday, there were seven nationwide polls, with the candidates as close as three points in the Investors Business Daily/TIPP poll and as far apart as 10 points in Gallup's "expanded" model. On Tuesday, the Gallup "traditional" model poll had the candidates separated by two points and the Pew poll had them separated by 15. On Wednesday, Battleground, Rasmussen and Gallup "traditional" model polls had the candidates separated by three points while Diageo/Hotline and Gallup "expanded" model polls had the spread at seven points.

Polls can reveal underlying or emerging trends and help campaigns decide where to focus. The danger is that commentators use them to declare a race over before the votes are in. This can demoralize the underdog's supporters, depressing turnout. I know that from experience.

On election night in 2000 Al Hunt -- then a columnist for this newspaper and a commentator on CNN -- was the first TV talking head to erroneously declare that Florida's polls had closed, when those in the Panhandle were open for another hour. Shortly before 8 p.m. Eastern Standard Time, Judy Woodruff said: "A big call to make. CNN announces that we call Florida in the Al Gore column."

Mr. Hunt and Ms. Woodruff were not only wrong. What they did was harmful. We know, for example, that turnout in 2000 compared to 1996 improved more in states whose polls had closed by the time Ms. Woodruff all but declared the contest over. The data suggests that as many as 500,000 people in the Midwest and West didn't bother to vote after the networks indicated Florida cinched the race for Mr. Gore.

I recall, too, the media's screwup in 2004, when exit-polling data leaked in the afternoon. It showed President Bush losing Pennsylvania by 17 points, New Hampshire by 18, behind among white males in Florida, and projected South Carolina and Colorado too close to call. It looked like the GOP would be wiped out.

Bob Shrum famously became the first to congratulate Sen. John Kerry by addressing him as "President Kerry." Commentators let the exit polls color their coverage for hours until their certainty was undone by actual vote tallies.

Polls have proliferated this year in part because it is much easier for journalists to devote the limited space in their papers or on TV to the horse-race aspect of the election rather than its substance. And I admit, I've aided and abetted this process.

In the campaign's final week, though, the candidates can offer little new substance, so attention turns to the political landscape, and there's no question Mr. McCain is in a difficult place.

The last national poll that showed Mr. McCain ahead came out Sept. 25 and the 232 polls since then have all shown Mr. Obama leading. Only one time in the past 14 presidential elections has a candidate won the popular vote and the Electoral College after trailing in the Gallup Poll the week before the election: Ronald Reagan in 1980.

But the question that matters is the margin. If Mr. McCain is down by 3%, his task is doable, if difficult. If he's down by 9%, his task is essentially impossible. In truth, however, no one knows for sure what kind of polling deficit is insurmountable or even which poll is correct. All of us should act with the proper understanding that nothing is yet decided.

As for me, I've already cast my absentee ballot in Kerr County, Texas -- joyfully, enthusiastically marking the straight Republican column. I would like to have joined the line Tuesday outside the polling place in Ingram, where I've been registered the past few years. But I will be in New York, part of the vast horde analyzing exit polls, dissecting returns, and pontificating on consequences. I'll thoroughly enjoy myself that night, and probably feel guilty the next morning. But this year's 728 national polls and the thousands of state polls made me do it.

Mr. Rove is a former senior adviser and deputy chief of staff to President George W. Bush.