In terms of the advance polls and the weather and the election-specific effects, we don't know. One can only intuitively assume that in a winter election, the parties were probably telling the voters they had identified as core voters that if the weather was nice, they should get out there to vote, and here are the advance polls. They're pulling their vote in the advance polls because you don't know....
Those are election-specific effects. Is that why we had more voter turnout in advance polls in 2006 than in 2004? Probably, but that's the only answer we can give in the absence of empirical data.
Jon, did you want to talk about it?