ABC News
Romney’s Early-State Obstacle Course

We’re getting to the point where it’s worth paying a fair amount of attention to the early-state polls. At the same time, with all the volatility in the Republican field, it’s sometimes hard to tell whether a strong poll in Iowa or a weak one in Florida reflects something unique to that state or a national trend instead.

So here’s a way to help us sort through the polls: something I’m calling State Strength Scores.

The basic idea is to take a candidate’s polls in each of the four key early-voting states — Iowa, New Hampshire, South Carolina and Florida — and compare them to his overall polling trendline from all national and state surveys at the time the poll was conducted. So, for instance, if Gov. Rick Perry had polled at 15 percent in a South Carolina poll during a point in the campaign that he was polling at 10 percent nationally, that would indicate his numbers were 50 percent stronger in South Carolina than they were elsewhere.

Got it? The actual method is a little more complicated than that (to get very technical, it involves a regression of a candidate’s state polls against his national polling trendline at the time, omitting the constant term) but operates on the same basic principle. I’ve rounded the scores to the nearest 10 percent just to make them a little more legible.

Here’s how to read these numbers. Jon M. Huntsman Jr.’s score in New Hampshire is 110 percent, which happens to be the highest number in the table. That means we’d expect Mr. Huntsman to get a 110 percent bonus in New Hampshire relative to his national polls at the time. Right now, with Mr. Huntsman polling at only 2 percent nationally, that bonus is only worth another 2 or 3 percentage points on top of it, meaning that he’s still just polling in roughly the 5 percent range in New Hampshire. If, however, Mr. Huntsman were to rise to 10 percent in the national polls, a 110 percent bonus would be worth 11 additional points in New Hampshire — meaning he’d poll at about 21 percent there and could be quite dangerous.

Obviously, this method involves some degree of extrapolation, especially for candidates like Mr. Huntsman who never have polled all that well nationally. Still, it should be a nice complement to looking at the polls on an ad-hoc basis.

We see, for instance, that Rick Santorum is in something of the same situation in Iowa that Mr. Huntsman is in New Hampshire. Right now, Mr. Santorum’s campaign isn’t going anywhere. But if he could get to the high single digits in the national polls, we’d expect him to be solidly in the double digits in Iowa, making him a factor in the state.

Mitt Romney, on the other hand, gets about a 20 percent penalty in Iowa. That implies that if he’s polling at 22 percent nationally, he’d poll at about 18 percent in the state (which looks to be about right from the latest polls). This puts Mr. Romney in a tricky position. He’s certainly not a lost cause in Iowa and got a lot of votes there in 2008. But if he were roughly tied in the national polls with a conservative candidate like Newt Gingrich, that penalty is enough to imply that he’d be a slight underdog there.

Mr. Romney, however, has a firewall in New Hampshire, where he is outperforming his national polling numbers by about 50 percent. One way to look at this is that things would have to go really, really bad for Mr. Romney for him to lose New Hampshire. The other way to look at this is if he does lose it, he’s in deep, deep trouble.

One candidate whom we haven’t discussed very much but who could play a role here is Representative Ron Paul. He’s the only Republican candidate to have a positive State Strength Score in both Iowa and New Hampshire. Mr. Romney might prefer that Mr. Paul won Iowa rather than someone like Rick Perry who could resurrect more of a national campaign with a victory there. But the wild card of Mr. Paul winning Iowa, which is certainly not impossible, could open up the prospect of an upset in the New Hampshire as well.

South Carolina might not seem to be a great state for Mr. Romney, but his polling has been all right there, whereas the two candidates who had seemed to have the most inherent strength in the state, Herman Cain and Mr. Perry, are on a downward trajectory in the national polls.

Mr. Romney has been overperforming his national polling slightly in Florida, a state his campaign now seems to see as exceptionally important (I agree). The only issue is that it looks to be an above-average state for Mr. Gingrich as well, who has targeted older voters.

The bottom line is that almost any winner of Iowa apart from Mr. Romney himself would present some kind of threat to him in one of the next three states. (The exceptions are probably Michele Bachmann and Mr. Santorum.) Mr. Romney certainly does not need to win Iowa to win the nomination. But unless he builds up more of a cushion in the national polls before the voting there, a loss for his campaign in the caucuses would at least make for an exciting January.

Nate Silver is the founder and editor in chief of FiveThirtyEight.