Skip to main content
ABC News
In Michigan, Momentum for Romney?

A new Public Policy Polling survey of Michigan ahead of the Republican primary there shows Rick Santorum with 37 percent to Mitt Romney’s 33 percent, essentially tied, since the poll had a four-percentage-point margin of error.

This poll is much better for Mr. Romney than the one the same polling firm put out last week, when it had Mr. Santorum ahead by 15 percentage points.

As the headline suggests, I am of mixed minds about this poll. Surely it qualifies as good news for Mr. Romney, but it may overstate his momentum.

The original Public Policy Polling survey had been something of an outlier. The 15-percentage-point lead it had given to Mr. Santorum was larger than any other poll of the state, which had him ahead by margins ranging from 3 to 10 percentage points.

When a poll is an outlier, you should expect its results to revert to the mean to some degree. Indeed, there is some evidence of this in the Public Policy Polling survey. Their new survey projects that 41 percent of the voters in Michigan will be evangelical Christians, and that 31 percent will identify themselves as “very conservative,” figures that are fairly realistic as compared with how Michigan has voted in the past. (In the 2008 Republican primary, 39 percent of Michigan voters identified themselves as evangelical.) By contrast, the previous survey had 48 percent of the electorate as evangelical and 38 percent as very conservative — unlikely figures in a state where independents and Democrats can easily vote by changing their registration at the polling site.

The demographic shifts appear to account for perhaps 4 or 5 points of the 11-percentage-point swing toward Mr. Romney in the poll. Still, that leaves room for forward movement toward Mr. Romney.

Our forecast model, which looks at polling and polling alone, now gives Mr. Santorum about a five-percentage-point lead in Michigan. That translates into a 72 percent chance of victory, according to the model.

I am simplifying a bit, but essentially what this means is that in past primaries and caucuses, candidates who held a five-percentage-point polling lead with a week or so to go until the voting wound up winning about 72 percent of the time.

There is reason, apart from the polling, to think that Mr. Santorum’s chances are not that strong, however, and that his lead is more tenuous than normal. Specifically, Mr. Romney has a substantial advertising advantage in Michigan, something that could continue to pay dividends for him.

In fact, the betting market Intrade now gives Mr. Santorum just a 35 percent chance of winning Michigan.

Maybe some smart person has developed a really good way to model the effect of advertising in a state. But intuitively, that feels like a bit too much of an adjustment relative to what the polls say. In the Public Policy Polling survey, there was no deterioration in Mr. Santorum’s favorability numbers, in contrast to the sharp declines Newt Gingrich experienced in various states after advertising barrages against him.

Mr. Romney’s favorability numbers improved in the poll, however, suggesting that his base voters might be coming home to him. Meanwhile, the number of undecided voters in the poll decreased significantly, suggesting that voters who might have had doubts about Mr. Romney after his losses in Colorado and Minnesota were returning to his camp.

It would certainly not be out of line to characterize Michigan as a tossup. As my colleague Michael Shear notes, the Republican debate on Wednesday could prove decisive in Michigan if either Mr. Romney or Mr. Santorum emerges with a clear win.

The same could hold in Arizona, where the debate is being held. Our model now gives Mr. Romney an 81 percent chance of winning in Arizona. But that does not account for a new Public Policy Polling survey, to be published on Monday night, which is likely to show a tighter race there.

Mr. Santorum has continued to gain ground in national polls, particularly the Gallup national tracking survey, which now has him with an eight-percentage-point lead. National polls can be somewhat lagging indicators because they do not reflect voters who have been exposed to the campaign, as they are in Michigan and Arizona. At the same time, this weighs against the idea that Mr. Santorum’s numbers are in some state of collapse.

Mr. Santorum holds polling leads in Super Tuesday states including Ohio and Oklahoma, but the results are likely to be greatly affected by what happens in Michigan and Arizona.

Nate Silver founded and was the editor in chief of FiveThirtyEight.