Ordinarily, a candidate should benefit from having the support of more definite voters — and most polling firms give them the bulk of the weight in their turnout models, which is why Mr. Santorum leads the poll over all.
The universe of indefinite voters is broader. But those votes don’t count for anything unless the candidate can get the voters to the polls.
That’s something Mr. Romney has had trouble doing so far. In states and counties that would appear to be strong for him, turnout is generally running below its 2008 pace. But in his weaker areas — say, most of the state of South Carolina — it has been steady or has improved some.
(...) So far, Mr. Romney’s campaign has relied more on the “air war” — television advertising. In Florida, where he came from behind to beat Newt Gingrich, he ran far more advertisements than the other candidates.
Almost all of these ads were negative, however, and almost all of them targeted Mr. Gingrich. There are two risks to Mr. Romney if he adopts a similar strategy in Michigan.
First, Mr. Santorum might prove more resilient to the attacks. Mr. Romney will need to develop a more creative line than the one he has used so far, calling Mr. Santorum a “career politician.” That line may produce diminishing returns, since Mr. Romney has leveled a similar charge against opponents like Newt Gingrich and Rick Perry in the past, and since it is slightly discordant coming from someone who has spent most of the past six years running for president.
The bigger problem, however, is that negative ads are no way to increase your base; instead, they may alienate key constituencies within the party.
That could be especially important to Mr. Romney because the window of voting through Super Tuesday might represent the last chance for him to win the nomination by consensus, with voters coalescing around him. To date, they have rejected several opportunities to do so, and the evidence probably weighs against their doing so now, but a decisive set of victories in different parts of the country would allow Mr. Romney to make a credible pitch to voters and party elites.
If, however, Mr. Romney gets only narrow wins in states like Michigan and Arizona after a negatively charged campaign, he would probably be the favorite to win the nomination, but a lot of those winning scenarios would be ugly ones, including an outside chance at a deadlocked convention.
Moreover, such a strategy could make Mr. Romney less resilient if he did tolerably well on Feb. 28 and on Super Tuesday but stumbled later.
Already, Mr. Romney’s unfavorable ratings are reaching dangerously high levels among both Republican and general election voters. If Mr. Romney is viewed unfavorably by the majority of Republicans, the party could be harmed in November even if he is the nominee — through diminished turnout or (less likely) the rise of a conservative challenger who would run as a third party.
A more positively themed campaign would obviously not guarantee a win in Michigan. (The FiveThirtyEight forecast model, following the polls, instead sees Mr. Santorum as the favorite there, although it has sometimes overreacted to temporary shifts in momentum.) But such a strategy is likely to be more sustainable for Mr. Romney, win or lose. It would be best if supplemented by a strong ground game; if it is too late for Mr. Romney to build one on the scale Mr. Obama or Mrs. Clinton had in 2008, he could at least increase the number of personal appearances he makes in each state.
Mr. Romney may also benefit if Mr. Santorum’s numbers fall back down to earth on their own, something that has happened to several candidates after previous bursts of momentum.
However, it is not necessarily obvious that the equilibrium state of the Republican race is one in which Mr. Romney wins, or at least not one in which he wins easily. The only time that he has has a pronounced lead in national polls in either 2011 or 2012 is after surges of his own following his wins in New Hampshire and Florida. But those, too, proved to be short-lived. If Mr. Romney is unable to find his base voters in Michigan — and get them to the polls — it becomes harder to see how he finds the building blocks for a national majority.