New Hampshire

Here's an observation about New Hampshire (and, yes, I was dead wrong on Iowa. My analytical ability is highly suspect. And yet I persist).

Compare the polls from Iowa and New Hampshire at this page. You're note that the aggregate spread in Iowa at the very end was Obama by plus 1.6. Obviously, Obama blew the roof off, and the polling prediction wasn't ultimately very accurate (though, of course, it was more accurate than my prediction, as I've noted).

Why was the Iowa poll off the mark? Because Obama's turnout was significantly higher than any predicted by any of the polls.

Similarly, check out the spread on Obama in New Hampshire. He's plus 3.4 (as of Sunday, it may change later in the week). What this doesn't reveal is what the turnout assumptions are. But, assuming they're relatively standard, they may underpredict any Obama success if his turnout is like it was in Iowa.

Will the turnout be higher? I don't know. But this story gives one a reason to think it might be. As does this one.