petardier

The swing between NH polls before Iowa and after was way too big to be believed. We in NH are not that fickle. The polls before Iowa were correct. The polls after Iowa were wrong. All of them. While that defies belief on the surfact, if there is a common reason why they were all wrong, then it is in fact the simplest of all possible explanations, ergo Occam's Razor says it is the best explanation. And the underlying reason is that Clinton supporters were much less likely to be talking to pollsters in the final week. They were older and more likely to be aggravated by the incessant phone calls, ergo less likely to answer. They were less committed to a "cause", ergo less eager to talk to anyone. And their candidate was under seige, ergo they just wanted to ignore the whole thing until the vote.

I've read all the other theories, and none of them fit what I know about New Hamsphire voters better than this.