By Kathy Frankovic, CBS News director of surveys
Is there an "Iowa effect?"
In the last few weeks, the Iowa caucuses have received the saturation coverage usually reserved for the last weeks of a general election. But the race won't end with the Iowa caucuses - not by a long shot. It is, after all, a contest for delegates - and on the Republican side, none are going to be allocated on Thursday, no matter who ostensibly "wins." Delegate estimates can be made for the Democrats, but neither party will actually name their delegates to Minneapolis or Denver until the Iowa state conventions are held in June.
So, will what happens in Iowa determine the nominee? Does it even affect the next event in the process - the New Hampshire primary?
Historically, the answers are mixed. And in some circumstances a caucus runner-up benefits more than the winner.
Sometimes Iowa matters. In 1984, Gary Hart finished a distant second in Iowa, getting only about a third the number of votes as front-runner Walter Mondale. But Hart led the rest of the field, and there was a change in New Hampshire polls soon afterward. Prior to the 1984 caucuses, an ABC News tracking poll had given Hart just 13 percent support in New Hampshire, putting him behind both Mondale and John Glenn. In interviewing a week later, conducted over the three days just before the primary, Hart and Mondale were even at 30 percent - and in the interviews done on the very last night, Hart led Mondale 35 percent to 27 percent. Hart beat Mondale in New Hampshire by nine points on February 28; though Mondale became the eventual nominee.
Dick Gephardt's Iowa victory in 1988 was no surprise, and neither was Paul Simon's second place finish. They were Iowa neighbors: Gephardt from Missouri and Simon from Illinois. Massachusetts Governor Mike Dukakis finished third. Dukakis was already the New Hampshire front-runner (another case of being nice to your neighbors), a position he held after the caucuses, too. While Gephardt gained about 10 points in New Hampshire, he still ended up in second place there.
The Iowa bump can also be an Iowa dip. In 2004, John Kerry and John Edwards both moved up in New Hampshire tracking polls after their (respective) first and second place Iowa finishes. Howard Dean had been in first place in New Hampshire polls before Iowa, but his third-place finish there cost him support; he lost the New Hampshire primary and with it his hopes for the nomination. (Dean had actually started losing support in New Hampshire before the Iowa caucuses, but the drop continued afterward.)
On the Republican side, polls taken in New Hampshire in 1988, after the Iowa caucuses, showed Kansas Senator Bob Dole (another Iowa neighbor) gaining on George H.W. Bush, who had finished third in Iowa - behind not just Dole, but also televangelist Pat Robertson! A CBS News Poll showed Bush's New Hampshire lead disappear temporarily: he had led Dole by 24 points before Iowa, but three days after the caucuses, Dole led by three points . However, Bush bounced back to a four point lead in the CBS News Poll right before New Hampshire's primary day. It was a tough campaign, but Bush did win in New Hampshire - and by nearly 10 points.
In 2000, what happened in Iowa didn't seem to directly affect the New Hampshire primary. Al Gore defeated Bill Bradley in Iowa by nearly two to one, but eight days later, he barely eked out a victory in New Hampshire. And even thoughfinished fourth in the Iowa caucuses (behind George W. Bush, businessman Steve Forbes, and even Alan Keyes), he went on to beat Bush decisively in New Hampshire.
So will what happens in Iowa make a difference in New Hampshire? For the last two months, polls in Iowa have shown a tight three-way Democratic contest, with, and fighting for the lead. The race could end that way, especially with the two-part Democratic process, in which supporters of candidates with low initial support reallocate themselves to more viable contenders. That's not a typical outcome for Iowa. And without a clear winner, would anyone benefit? As for the Republican race, we've seen 's lead shrink as moved up, but recent polls suggest Huckabee slipping. Would a Romney win now help him defeat McCain in New Hampshire, or will McCain repeat his 2000 success there even without a strong finish in Iowa?
There's no clear precedent - but at least 2008 will give us a new case study or two about the "Iowa effect!"
P.S. Unfortunately, the people of Iowa and New Hampshire have not benefited from what should have been a week of peace on earth (). While CBS News abstained, eight other polling organizations released Iowa polls that included interviews conducted between December 26 and 31. However, only two organizations have done that in New Hampshire. And voters in the rest of the country, at least, have had a poll holiday.