Will 2000 Happen Again?

In the last week or so, conventional wisdom has begun to settle on the possibility of an Electoral College/popular vote split. The situation is straightforward: Thanks to a persistent lead in Ohio, Obama ekes out a victory in the Electoral College, but Romney wins a bare majority of the popular vote.

Setting aside the question of politics—i.e., how the parties and public would react to the second such split in just over a decade—just how likely is this scenario? If you go by current polling, it looks like a solid possibility; just this morning, the pollsters at Quinnipiac University found Obama with a five-point lead in Ohio. If you take an average of the three major polling averages—Talking Points Memo, Real Clear Politics, and Pollster—Obama has a 2.4-percent lead over Mitt Romney in Ohio. By contrast, that same average of averages shows a slight 0.2 percent lead for Romney nationally. If this holds, we could have the split.

Of course, in addition to accepting this is as a serious possibility, it’s worth considering the caveats. The first is that, as calculated by FiveThirtyEight’s Nate Silver, the odds of such a split are still low. According to his forecast, there is a 5.4 percent chance that Obama wins the Electoral College but loses the popular vote, and when you consider the relationship between state and national performance, this makes sense. Rarely are the two divergent—a candidate who is winning in the national polls will see that advantage translate to each individual state. If Obama were to lose two points in the national polls over the next week, you would see a similar swing in most states, and vice versa.

Someone has to get to 50 percent, and if that’s Mitt Romney, the expectation should be that he brings enough states with him to win 270 electoral votes. To wit, if Romney wins 60 percent of white voters—his “magic number” for the presidency—he’ll almost certainly take Ohio, and he might even peel off states like Wisconsin and Iowa. And if Obama wins key battleground states—like Virginia and Colorado—odds are good that he’ll have a small popular vote victory to go along with his win in the Electoral College.

There is one thing that complicates the view that a split is not likely to happen—the South. The latest Gallup polls are probably outliers—no other poll shows Romney with a six- or seven-point lead—but they do contain some useful information. Last week, Gallup gave a regional breakdown for its numbers, and election observers noticed something remarkable: Obama was winning by a solid margin in every region but the South, where he was losing in a huge way:

  Obama Romney Margin
East 52 48 O+4
Midwest 52 48 O+4
South 39 61 R+22
West 53 47 O+6

If we see a similar regional gap on Election Day—where Romney loses in every region but the South, where he racks up a huge score—then I would bet on a split between the Electoral College and the popular vote, even if Obama prevails in most contested states.

It’s hard to say how the country—and particularly, the conservative movement—would react to that outcome, but it wouldn’t be good.