MSNBC has just called New Hampshire for Hillary Clinton. Thinking out loud on why the polls were so, so wrong. Ideas welcome in the comments.
(1) Independents, who could vote in either the Democratic or Republican race, assumed that Obama had it wrapped up and turned to McCain in order to push him over the top.
(2) The voters in New Hampshire resented the picture the media was painting: Obama is king and New Hampshire is declaring the Clinton hegemony over American politics finished. Wait just a second, said the voters. Let’s keep this debate going.
(3) Clinton cried. Edwards slammed her for it. The media questioned if she showed too much weakness, intimating that a woman couldn’t cry and be taken seriously for high office. Women, who turned out hugely for Clinton in this race, turned to Clinton in the last few days. I actually think Obama got the same percentage of women as he did in Iowa, meaning a large number of women voters who went for Edwards in Iowa turned to Clinton.
(4) Edwards and Obama teamed up on her in the Democratic debate Saturday night. Voters, particularly those women who I just mentioned, didn’t like that. Motivations in (3) and (4) are tied together, obviously.
(4) The strategy of answering questions showed voters the depth of her knowledge.
(5) There is a well-known effect that hits black politicians. They tend to do better in polling than they do when voters actually head into private polling booths. You can guess why. This effect doesn’t occur in a caucus, because participating in a caucus requires voters to stand up for who they want in a public setting. There is social pressure. (I can’t for the life of me remember the name of this effect. Anyone want to remind me in the comments?)
(6) All of the above.
I’m going with (6). And by the way, all this was incredibly premature.
A big victory for Clinton tonight. Nevada is up next. The political powerhouse in Nevada, UNITE-HERE Culinary Workers Union Local 226, was set to endorse Obama tomorrow, which many believed would basically hand him the state’s primary on January 19th. Now we’ll have to wait to see what happens.
Update: Thanks to our readers who IDed (6) as the Wilder/Bradley effect. Here’s evidence that was not in play.