There’s a real chance, as you’ve probably heard, that there’ll be a split popular vote/Electoral College decision on November 6.
As of this writing, the Real Clear Politics polling average has Mitt Romney leading by 0.7 points in the national horserace, while HuffPost Pollster puts the GOP nominee up by 0.4. The electoral map is more decisive, though, at least for now, with Obama ahead in all of his must-win states, and well-positioned to pick off a couple of Romney’s. If the election were held this minute, the odds are very good that Obama would win the Electoral College, and decent that he’d lose the popular vote.
This would officially be the fifth time in history this has happened, after 1824, 1876, 1888 and 2000 (although Sean Trende made a strong case last week that it also happened in 1960). The question is where the disconnect is coming from – why is Obama apparently performing better in swing states than he is elsewhere?
Many reflexively argue that this is a product of Southern and Appalachian antipathy toward the president. He’s faring so terribly in these areas, the logic goes, that it’s dragging down his standing in the national polls (and elevating Romney’s) by a point or two.
On the surface, this makes sense. In a wide swath of territory extending from Oklahoma up through West Virginia, Obama actually found himself losing dozens of rural counties to fringe challengers in this year’s Democratic primaries – something that didn’t happen outside that region. So if he’s facing that much Southern/Appalachian trouble within his own party, then imagine how bad his plight is with all voters there. Obama’s campaign itself made this argument a few months ago, as John Heilemann relayed back in the spring:
[Obama campaign manager Jim Messina] doesn’t give a whit about national polling, in which Obama’s numbers are dragged down by his horrific performance in the Deep South and Appalachia – but is obsessed with the president’s standing in battleground states.
This view was reinforced a few weeks ago when Markos Moulitsas flagged regional subsample data from Gallup’s weekly poll which showed Obama running 4-6 points ahead in the East, Midwest and West – and 22 points behind in the South. So that settles it, right?
First, as Nate Cohn has documented, the Gallup regional numbers are at odds with what most other pollsters are finding. In six other national polls, Obama trailed in the South on average by 7.5 points, which is actually a tick better than the nine-point loss he suffered in the region to John McCain four years ago. Of course, take all of this with a grain of salt; the polls that this data was drawn from are national in nature and aren’t necessarily designed to show the state of the race region by region.
Still, when you think about it, it makes sense that the South/Appalachia wouldn’t really be Obama’s trouble spot, because it’s an area that never liked him to begin with. Don’t forget, this is the part of the country where in many counties Obama actually fared worse in 2008 than John Kerry did in 2004. Even before his popularity began dropping with voters elsewhere, voters in this area had already turned on him. So while he’s not doing well in the South, he’s probably not doing that much worse than he did in ’08.
Patrick Murray, who runs the Monmouth University poll, suggests that a more helpful way of understanding why Obama’s national support is lagging is to divide the country into three types of states – red, blue, and competitive – and to compare Obama’s showing in these states to how he’s doing now. Using the ’08 results and Monmouth’s latest numbers, here’s what Murray came up with:
Obama ’08: +7
Obama ’12: -5
Obama ’08: -15
Obama ’12: -15
Obama ’08: + 24
Obama ’12: +11
It’s not surprising that Obama’s standing has declined in competitive states, but what jumps out is that he’s not doing any worse in red states now than he was in ’08 – but that he has suffered a huge drop in blue state America.
This gets us closer to understanding the potential popular vote/Electoral College split. Obama is still going to win all of the blue states he won in 2008 –New Jersey, California, Massachusetts and so on – but he could do so by significantly reduced margins. This wouldn’t change the electoral math, but it certainly would affect the national popular vote. I asked Murray why Obama would be falling off so much in blue states, and why the damage isn’t worse in battleground areas. His theory is that it’s an enthusiasm issue, and he offers some more data from his poll to support this – the percentage of voters from each party who make it through his likely voter screen:
As you can see, enthusiasm is highest for both parties in the battleground states – not surprising, given how much attention they receive – and lowest in their respective safe zones. But the drop-off is more severe for Democrats in blue states, where they’re almost ten points behind Republicans on turnout likelihood. Here’s what Murray told me:
I really think it’s an enthusiasm gap. It’s offset in competitive states by the fact that there is an active campaign. My theory is that Blue state Dems are a bit disappointed over how the first term has turned out. They don’t want Romney to win but feel that Obama can win their state’s EVs without their support.
This certainly seems plausible. Democrats in, say, Ohio and Massachusetts probably both feel disappointed in how the last four years have gone. But in Ohio, they feel an obvious urgency to vote anyway and to keep the other party from winning, and they’re bombarded daily with paid and free media messages pounding home this point. In Massachusetts, though, there’s no campaign, except from the ads on Boston TV stations that are aimed at New Hampshire voters. Democrats there can shrug and say, “I’d rather Obama wins, but it’s not like me voting here will do anything for him.”
I’d say this might be an argument for ending the Electoral College, if the non-battleground status of most of the country actually is affecting participation levels. But at the very least, it seems like a pretty good explanation for why Obama may lose the national horserace but still win the Electoral College.
Update: I've received a number of emails and tweets pointing out that the Monmouth poll produced a funny result among Latino voters:
— Nick Gourevitch (@nickgourevitch) October 25, 2012
This is true. Overall, Monmouth's poll has Romney leading Obama 48-45 percent, but among Latinos, Obama's edge is only 48-42 -- even though Obama won Latinos by 36 points in 2008 and is performing as well (or better) in most other national polls. I asked Murray about this, and he replied:
I can break that down. Obama is winning Latino voters in Blue states by a more than 2 to 1 margin in my poll (60 to 27) – pretty much where we would expect him to be in those states. He is losing the Latino vote in Red states (53-38 on an EXTREMELY small sample size – so no real impact there) and losing it narrowly in competitive states (48-44).
In other words, my Blue State Latino vote matches expectations. The one place where there might be some sampling error is in the competitive states - which could affect Obama’s margin there by a couple of points (keeping in mind, of course, that we are talking about a small sub-sample with a large margin of error).
So this still seems consistent with the idea that blue states explain the potential popular vote/Electoral College disconnect.