There’s something odd going on in the polling of the 2008 Republican presidential field. In the key states of Iowa and New Hampshire, former Massachusetts governor Mitt Romney seems to be pulling away from the pack. But in national polls, Romney is way back, by some measures in fourth or even fifth place, behind not only Rudy Giuliani and John McCain, but also still-undeclared candidate Fred Thompson and, sometimes, Newt Gingrich.
Here are the numbers: A new Des Moines Register poll puts Romney ahead with 30 percent of those surveyed, with McCain at 18 percent and Giuliani at 17 percent. The poll, of likely Republican caucus participants, did not include Thompson or Gingrich. Meanwhile, two New Hampshire polls, one by Zogby and the other by SurveyUSA, have Romney in front with 35 percent and 32 percent of the vote. Those polls stand in stark contrast to the last eight Gallup national polls, which have put Romney at anywhere between three and nine percent, well behind the frontrunners.
What’s going on? “The New Hampshire polls are not surprising,” says Republican pollster David Winston, who is not affiliated with any presidential campaign. “That’s basically Romney’s backyard. What you’re seeing in New Hampshire is that people just know him. He’s been a fixture on the news for years now.” Still, even with that familiarity factor, Romney seems to be moving up in New Hampshire. Most polls taken in the first three months of this year put him behind McCain and Giuliani, so it seems safe to say his efforts in New Hampshire are bearing some fruit.
Iowa is another story. The Des Moines Register’s numbers, which put Romney at 30 percent, with McCain at 18 percent and Giuliani at 17 percent, are at odds with several other polls in the state. For example, a new Research 2000 poll puts Romney at 16 percent, versus McCain’s 18 percent and Giuliani’s 17 percent, and a new Zogby poll puts Romney at 19 percent, with both McCain and Giuliani at 18 percent. All three polls were taken at virtually the same time.
The safest conclusion to draw from that is to be skeptical of the Des Moines Register survey. “My theory about surprising polls is that when a single poll is very surprising, you need to see something else to validate it,” says Winston. “It’s the only one I’ve seen that has the race that way.” Winston suggests the discrepancy could be traced to the definition of likely caucus participants. “Your definition of what a likely Republican caucus goer is can be different from someone else’s,” he says. “That’s where a lot of error comes in.”
But even if the Des Moines Register poll is wrong, there’s no doubt Romney is doing better in the polls in Iowa and New Hampshire than he is nationally. In the past, the conventional wisdom has been to give more weight to the state polls, because that’s where the early momentum will be won or lost. Now, however, that is probably less true. “Those national numbers take on increasing significance because of February 5,” says Winston, referring to the day in which 17 states, including big prizes California, New York, Texas, and Michigan, will hold primaries. (Florida has decided to go even earlier, on January 29.)
About 60 percent of the nation’s population will be represented in the February 5 races. So this time around, the national polls mean much more than they used to. And so far, Romney has been unable to pull out of single digits.