Tuesday’s primary in New Hampshire is not likely to be as dramatic as the Iowa caucuses, with one candidate, Mitt Romney, holding a clear lead in all of the polls.
Still, the state is known for defying pollsters’ and pundits’ expectations. And as in other early-voting states, the performance of the candidates in New Hampshire will affect the tenor of news media coverage and voter sentiment heading into South Carolina and Florida, where Mr. Romney could be more vulnerable.
There is also a close battle for second in New Hampshire and, to the extent it matters, another close battle for fourth place.
As I did before Iowa, I will present each candidate’s FiveThirtyEight forecast for New Hampshire along with what the forecast model deems to be his 90 percent confidence interval. A candidate’s actual result should fall within this confidence interval 90 percent of the time and outside of it 10 percent of the time.
Mitt Romney
538 forecast (most likely outcome): 39 percent
High end of forecast range: 47 percent
Low end of forecast range: 27 percent
First things first: Could Mr. Romney actually lose Tuesday night?
On Monday, after seeing a Suffolk University tracking poll that had Mr. Romney’s support dropping to 33 percent, I wrote that that such a loss had become “thinkable.” Thinkable, however, does not equate to likely, and Mr. Romney rebounded to 37 percent in the Suffolk poll when a fresh set of results were released Tuesday morning.
The FiveThirtyEight forecast model gives Mr. Romney a 98.9 chance of winning New Hampshire. It gives Ron Paul a 0.6 percent chance and Jon M. Huntsman Jr. a 0.4 percent chance.
These numbers ought to check out reasonably well with less complicated, back-of-the-envelope methods of estimating Mr. Romney’s losing chances. His lead over Mr. Paul is now about 20 percentage points in the forecast model. One candidate in our database, Jimmy Carter in the New York primary of 1980, lost despite having a lead of about that size. And Walter Mondale lost New Hampshire in 1984 to Gary Hart despite having a 17-point lead, which falls just shy of that threshold. But perhaps Mr. Hart should be given some extra credit because he beat Mr. Mondale by nearly 10 points.
So Mr. Romney has lost just enough of his lead that actually losing in New Hampshire not be without historical precedent. But it’s just not something that’s going to occur very often. Mr. Romney also has several good defenses; he has the strongest ground game in New Hampshire and his favorability ratings in the state remain high.
Although our forecast model says that Mr. Paul is slightly more likely than Mr. Huntsman to pull off an upset, I would suggest that Mr. Huntsman’s odds are slightly shorter. The reason is that Mr. Huntsman’s votes are more likely to come directly from Mr. Romney’s pile, essentially doubling their net effect.
Even if Mr. Romney wins, the question will be whether the outcome will be close enough that he is deemed to have underachieved expectations and will get negative coverage as a result. This type of coverage often reveals more about the news media than about the candidates or the voters, so I’m not eager to establish an arbitrary threshold that would distinguish a “good” win from a “bad” win. Winning the New Hampshire primary is a big deal, period.
I would suggest, however, that there is a tangible difference between a win that is called early in the night and one that takes several hours to declare. In the first case, voters in South Carolina and Florida will see a night’s worth of favorable coverage for Mr. Romney, but much less so in the second case.
Ron Paul
538 forecast (most likely outcome): 19 percent
High end of forecast range: 27 percent
Low end of forecast range: 11 percent
As was the case in Iowa, I’m not sure that Mr. Paul’s actual performance is as important as that of the other candidates, since his supporters are more or less fully committed to him regardless of the outcome.
Still, that does not mean that there is no uncertainty in his forecast. Mr. Paul has a good field operation in New Hampshire, and his campaign has targeted independent voters there as much as registered Republicans.
Mr. Paul also does well with young voters — it would not be a great exaggeration to say that the median age of voters at a Paul rally is half that of other candidates.
If these groups turn out in great numbers — there is an argument to be made that pollsters are underestimating them — Mr. Paul should finish above 20 percent. If not, he will be somewhere in the teens.
Jon M. Huntsman Jr.
538 forecast (most likely outcome): 17 percent
High end of forecast range: 26 percent
Low end of forecast range: 9 percent
Mr. Huntsman has strong momentum headed into the primary and has nearly caught up to Mr. Paul in our forecasts. The Suffolk tracking poll, moreover, actually had Mr. Huntsman beating Mr. Paul in interviews conducted on Monday night alone, although it was a small sample size.
Although our forecast model is quite aggressive about accounting for a candidate’s momentum, these effects may be especially strong in the early voting states. Mr. Huntsman has also spent more time in New Hampshire than any other candidate, which could lead to voters giving him the benefit of the doubt when trying to decide between him and another alternative. For these reasons, I would probably take the “over” if using our model’s 17 percent forecast as a betting line.
However, Mr. Huntsman will need quite a lot of momentum to have a viable plan once he leaves New Hampshire, since he is far back in the polls in South Carolina and Florida. If he beats Mr. Paul and clears 20 percent of the vote, the aesthetics of his performance will look better in the news media coverage.
Rick Santorum
538 forecast (most likely outcome): 12 percent
High end of forecast range: 20 percent
Low end of forecast range: 6 percent
There is a potentially favorable development for Mr. Romney hidden in the polls: it looks as though Rick Santorum and Newt Gingrich could perform fairly poorly in New Hampshire. Because Mr. Santorum and Mr. Gingrich appear to be the main threats to Mr. Romney in South Carolina, that could give Mr. Romney a buffer even if he also is viewed as underachieving on Tuesday.
Mr. Santorum remains slightly ahead of Mr. Gingrich in most surveys and therefore in our forecasts, but the momentum he achieved after his near-win in the Iowa caucuses stalled out a few days too soon. Our forecast model says that Mr. Santorum could still get as much as 20 percent of the vote, meaning that he retains some chance of finishing in third place and a long-shot chance of taking second, but he could also finish in the single digits.
Mr. Santorum, who has said that his main goal in New Hampshire is to beat Mr. Gingrich, has been seeking to lower expectations, having spent much of Saturday in South Carolina rather than New Hampshire.
Newt Gingrich
538 forecast (most likely outcome): 11 percent
High end of forecast range: 19 percent
Low end of forecast range: 5 percent
Mr. Gingrich, who had once appeared to be the most formidable threat to Mr. Romney in New Hampshire, has been in steady decline in the polls, although his numbers have stabilized a bit over the past several days.
Because a “super PAC” that supports Mr. Gingrich is set to unleash millions of dollars in negative advertising against Mr. Romney in South Carolina, Mr, Gingrich will continue to be a part of the story there regardless of his finish in New Hampshire.
However, just because Mr. Gingrich is a part of the story will not mean that he is a viable candidate; a top-three finish in New Hampshire would increase the chances that he could be something other than a spoiler. Working against Mr. Gingrich is that his ground game is not very strong in the state.
There is also one small factor that could help Mr. Gingrich: his name is listed above any other major Republican on New Hampshire’s 30-candidate ballot, something that has been shown to have a favorable effect in academic research.
Rick Perry
538 forecast (most likely outcome): 1 percent
High end of forecast range: 3 percent
Low end of forecast range: 0 percent
If both Mr. Gingrich and Mr. Santorum perform poorly in New Hampshire, then perhaps Mr. Perry can hold out hope that voters will take one last look at him in South Carolina.
Still, he could be headed for a finish in New Hampshire that might look fairly pathetic for someone who is supposed to be a nationally viable candidate. Most polls have Mr. Perry with just 1 percent of the vote, and there is a chance that he could finish behind obscure candidates like Buddy Roemer who have not been included in the debates.
Mr. Perry also got some bad news this week when a key group of conservative leaders, some of whom had once been favorably disposed toward him, did not give him their backing when they met over the weekend. One member of the group, Gary Bauer, endorsed Mr. Santorum instead.
We will be live-blogging the New Hampshire results beginning at about 5 p.m. tonight.