Pundits always say the race is tightening

By Frank Moraes

Yesterday, I read the following campaign intel from James Pethokoukis, "I just talked to one of my best Team Romney sources who told me that heading into today all the key battleground polls were moving hard and fast in their direction. The source, hardly a perma-optimist, thinks it will be a long night, but that Romney is going to win."

Actually, that's not what it said. If that's what it said, I wouldn't be quoting it here. This is because such information is everywhere right now. Steve Schmidt was promoting this same bullshit line over at MSNBC the other night. George Will predicted that Romney would win with 321 electoral votes. As best I can tell, he means that Romney will win Ohio, Pennsylvania, Michigan, and Virginia. And Minnesota. And Oregon, Nevada, and New Mexico. This would be a surprise, to say the least. But the point of all these confident proclamations is that the election is somehow moving in Romney's direction.

That's why the quote above is so telling. It was written on 4 November 2008 about John McCain. I just changed "McCain" to "Romney." Here's the actual quote:

I just talked to one of my best Team McCain sources who told me that heading into today all the key battleground polls were moving hard and fast in their direction. The source, hardly a perma-optimist, thinks it will be a long night, but that McCain is going to win.

You may recall, that even though Pethokoukis's source may not be a perma-optimist, he was a that-day-optimist. McCain lost big. Obama got 365 electoral votes to McCain's 173.

I think it is interesting that all the liberals I read are predicting modest Obama victories. David Atkins over at Hullabaloo predicts an Obama win 294-244. But he does note that Obama could win Colorado and bring the total up to 303-235. Ezra Klein is predicting only 290 electoral votes for Obama (which means he also thinks that Obama will lose New Hampshire). This contrasts starkly with the Republican landslide fantasies. Romney may well win, but he will eke out a narrow victory if he does.

My only thought is that the only way that Republicans can argue that Romney will win is to assume that the polls are systematically wrong or that undecideds will break heavily for Romney. One can't assume voters are going to break hard for Romney in Ohio but not in Oregon. Thus you end up with something like the Unskewed Polls prediction of a Romney 359-179 victory. The liberal approach is to assume the polls are basically right, the undecideds will break the way they always do, and that Republican voter suppression (and just bad luck because liberals are pessimists) will lead Obama to lose the close ones: Florida, Colorado, and New Hampshire.

My best guess is still that Obama will win 332-206. However, that does not mean I think that is the most likely outcome. That is just the most likely single outcome. I think it is slightly more likely than 303-235. But I would lay even odds that Obama will win 294, 303, or 332 electoral votes. Nate Silver gives the odds of that happening at roughly 35%. But he also gives the odds of a 345 vote Obama victory an 11% chance. That would require Obama winning North Carolina. I just don't see this happening, but maybe 11% is about right.

The main thing is that it is pretty hard to see how Romney wins this race unless the polls are completely wrong. Check out this great interactive battleground state tool over at the LA Times. Start by giving Obama Ohio. If that happens, it is almost impossible for Romney to win. So tomorrow night, it really does all come down to Ohio. If Florida goes for Obama, it will be a blowout. Otherwise, it will be relatively close regardless of who wins. 

(Cross-posted at Frankly Curious.)