Fascinating.
We might be seeing the Bradley Effect here: with O’Donnell being so completely demonized by the MSM, poll respondents are lying to pollsters about their intent to vote for her.
Possible.
What I’d say about polls right now is that we haven’t had a real poll in quite some time.
The last one was 10/14 - it was Rasmussen, and it had Christine at -11. Rasmussen charges for crosstabs and they have some weird system of charging every month so I didn’t want to spend $20 and forget to cancel, just to look at one set of crosstabs.
I suspect that Rasmussen did talk to more Rs than the others, but did they talk to the whole 37% of Rs?
She has come across very reasonable in the debates, not at all a crazy idiot, and I expect that she’ll win, provided we have a similar turnout advantage that we’re seeing in other states.
It’s possible there was a Bradley effect in those old polls with the wide margins.