There are still more votes to be counted, but as of now the tipping point state was...Colorado, same as in 2008. They're still counting the votes in Colorado, but as of now Barack Obama's lead is an impressive 4.7 percentage points, larger than his leads in Florida, Ohio, and Virginia. A uniform 4.6 point shift away from Obama would have given Mitt Romney those three states, while leaving Obama with Colorado and a winning 272 Electoral Votes.
Meanwhile, as I look at the results so far, Obama's lead in the national vote is up to 2.3 points, which makes his electoral college advantage just north of 2 percentage points. It's possible that the national lead will still grow a bit, which will mean that the apparent EC advantage will fade, at least unless the Obama lead in Colorado grows as well. It seems fairly safe at this point to say that it will be close to 2 percentage points.
First, that answers Sean Trende's (sincere and plausible) pre-election question about the difference between the state and national polls. As it turned out, the polling averages underplayed Obama's lead in both the states and nationally, but the gap between them was more or less accurate.
That gap is up a bit from the 1.7 point EC advantage that Obama had in 2008, and the small edge that John Kerry had in 2004 (when he came fairly close to winning Ohio while losing the national vote). Of course, everyone knows that the gap was in the other direction in 2000.
In other words, for three cycles now Democrats have been using their votes more efficiently in presidential elections.
We don't exactly know why, but it seems to me that it's a big question. It might have been a totally random effect of the ebb and flow of votes. I might have been an Obama-specific effect, having something to do with the particular appeal of Obama to some groups, or perhaps the particular antipathy towards him from others.
Or it could be structural, with (for now) Democrats just having a better distribution of supporters in presidential elections than Republicans.
It is rare in recent American political history for one party to have a persistent Electoral College advantage (compared to the national vote). Through 2004, there didn't seem to be any partisan advantage. Again, that certainly could still be true. It may disappear next time around.
But if it does persist, it's pretty enormous -- it would mean that Republicans begin presidential elections two points in the hole. Remember, this is apart from any advantage or disadvantage in the total number of votes. It's purely about how votes translate into wins or losses under the EC system. If, for example, the recent Democratic national vote victories (and remember, that's five of the last six elections) are a result of a mild national tilt towards them -- something that I'm certainly not claiming, but it's possible -- then this EC advantage would be above and beyond it. If, again, it's a long-term structural phenomenon.
It's going to be very interesting to see whether this distortion survives the Obama elections, and also whether Republicans will soon react to it by flipping in favor of reforming or eliminating the Electoral College -- and whether Democrats, who have been the anti-EC party at least since 2000, flip to support of that procedure.