As soon as the results started trickling in the explanations began as to how the polls could have been so wrong. Was it the voters reaction to the media's coverage of a show of emotion from Hillary? Or could it have been those damn Diebold machines again?! On MSNBC, WaPo's Eugene Robinson brought up one explanation that's now being floated around: the Bradley Effect, in which people supposedly lied to pollsters about whether they would vote for a black candidate.

Download Download

NBC's political director Chuck Todd on the Bradley Effect and why it may have been an influence in New Hampshire primary and not in the Iowa caucus:

In Iowa - a very public caucus . In New Hampshire, you have a voting curtain. And the privacy of the vote versus not.

And Howard Fineman explained why he thinks that could be a problem for Obama if true:

It's Barack Obama's burden and blessing that he can't really - I don't think - talk about this a whole lot. His whole message is about bringing people together. He can't start complaining about miscalculations in the polls.

Do note that there seems to be some reason to doubt that the Bradley Effect explains the outcome, especially considering the polls were actually right about Obama, they just didn't account for the surge in votes for Hillary. Ezra wonders if it wasn't the caucuses that did the discriminating.

Don't look to me to postulate one way or another (I will say that Pam's "Tweety Effect" echoed by Rachel Maddow holds some appeal for me though). We'll be hearing lots more about the reasons for the pollster's failure in NH in the days to come. I'm just enjoying the fact the polls lied so the pundits now have to eat crow and it's starting to look like we might have an actual race on our hands where the voters and not the powers that be get to do the deciding. Who'd have thunk?