The campaigns are pouring millions of dollars into North Carolina and the polls show a tight race, but Nate Silver doesn’t think that the state is worth the investment. While he is certainly right that North Carolina is unlikely to prove decisive, it’s easy to envision how the Tar Heel state could play a pivotal role in 2012. The state’s demographics, a small number of undecided voters, and Obama’s elevated standing effectively ensure a tight race—whether Obama wins or loses.
There’s plenty written here on the fundamentals in North Carolina, so let’s start by slightly reformulating Silver’s maxim: The most important states are not those that are closest in an absolute sense, but rather those that are closest to the national average in a close election. This slight reformulation is important, because while North Carolina isn’t the closest state to the national average right now, it could be pretty close to the national average in a close national election. Silver doesn’t think that’s likely because he believes that the current sequence of the states is likely to hold through November: If Obama is doing better in Colorado than North Carolina today, then it is very unlikely that Obama could win North Carolina without winning Colorado.
That’s more probable than the reverse, but I don’t think that’s quite an ironclad rule. The sequence of the states in July or August doesn’t necessarily resemble the sequence in November, as this Nate Silver article from four years ago illustrates. This is especially true if there are big changes in the race, as there were four years ago. Obama surged ahead in Michigan in October 2008, but barely advanced at all in the interior South. So the question is: if Romney surges and fights to a tie or even a win, where will Romney make the biggest gains?
One way to judge might be the number of undecided voters in a given state, and by that metric, we wouldn’t expect Romney to suddenly jump to a big lead in the Tar Heel state. According to the RCP average, just 4.4 percent of North Carolina voters are undecided, while most battlegrounds are north of 8 percent. So even though Romney has a 1 point lead in the RCP average today, it’s tough to envision him gaining a comfortable lead, even if Obama’s current 3-point national lead vanishes as Romney consolidates undecided voters. The opposite is true in a state like Iowa, where nearly 10 percent of voters are undecided. If Romney sweeps undecided voters, perhaps Iowa would become a lean Romney state. Or perhaps Michigan would become a true toss-up.
North Carolina’s unique demographics make it even easier to envision a tight race in the event of a close national election. Most undecided voters are white, working class, independents, so a Romney comeback would entail big gains with that group. But Obama needs very few white working class independent voters in North Carolina—fewer than any other battleground state. So as Romney surges back into a tie nationally with big advances among independents, he might not make many gains in North Carolina.
There’s also reason to accept a great deal of uncertainty about the composition of the electorate, and that's a problem in a state like North Carolina, which is ultimately about turnout, turnout, and turnout. Yeah, there are probably some persuadable voters outside of Charlotte or something, but at risk of oversimplifying, this is a state of conservative white evangelicals, African Americans, and post-industrial northern transplants. In a state like North Carolina, there are issues about likely voter models three months before the election, since if the non-white share of the electorate is a point or two higher or lower than the current polls assume, Obama moves into the lead or falls out of contention.
With turnout likely to decide the outcome of North Carolina, it’s hard to confidently say that Romney has an overwhelming edge based on a one-point lead in the polling averages. In 2008, for instance, SurveyUSA’s early polls found McCain leading and whites representing between 74 and 76 percent of the electorate. SurveyUSA’s October polls found that number reduced to 72 or 73 percent, and Obama ultimately won with the exit polls showing whites at 72 percent of the electorate. The bottom line is that everyone knows that the Obama campaign’s chances in North Carolina are unusually dependent on their ability to register and turnout non-white and young voters, and everyone also knows that polls just can't gauge these efforts well at this stage.
Given the challenge of estimating the composition of the electorate three months out, the small number of undecided voters, and the substantial investments made by both campaigns in the state, North Carolina could easily remain close in a tight national election, even as Obama suffers losses elsewhere. Now broadly speaking, Nate Silver is right about North Carolina: The state is not as vital to the electoral math as its competiveness would suggest. While Silver argues this is because Obama’s down by 2 points in North Carolina while up 3 nationally, I’d place the blame on Virginia: It’s tough to envision Obama winning North Carolina without winning Virginia, and a victory in the Commonwealth is generally sufficient for Obama to win reelection.
Even so, it is possible to envision scenarios where North Carolina can play a critical role. Here's one example:
This might seem outrageous at first, but it is internally consistent: Romney surges ahead with gains among undecided voters, but Obama’s enduring strength among postgraduates and minorities, combined with an effective GOTV and voter registration effort, squeezes out a win in North Carolina. It’s not wholly inconsistent with the current data, either. In fact, if you look at Obama’s share of the vote in each of the battlegrounds, you’ll note that Ohio is really the only state deviating from this possibility.
% Obama: RCP Avg
Of course, this highly unlikely scenario doesn't need to occur to justify substantial investments in North Carolina. Any close election would probably involve a tight race in the Tar Heel state, as well as several states that currently seem to tilt Obama's way, including Colorado and Ohio. Heading into election day, the Obama campaign would want to hedge its bets against undecided voters by having an alternative route to victory built on turning out Obama’s base groups. Given the close race in North Carolina, Romney can't take any chances.
So I don’t see the campaigns making a big mistake in North Carolina. The state’s demographics all but ensure a close race—whether Obama wins or loses—and that’s confirmed by polls showing a tight race with a small number of undecided voters. Virginia makes it difficult for North Carolina to be as important as the other big East Coast swing states, but it would be unwise for the campaigns to abandon a state that could be close in a tight national race, especially this far out from November.