How'd Clinton Pull it Off

There are a few theories going around this morning as to how all the polling got it SO wrong. Election morning polls showed Obama with anywhere from a 4-12 point lead, the average was around 8. The results came in last night and we see a Clinton squeaker. So how'd she do it?

Over at the Rothenberg Report it's suggested that it may because of John Edwards dip, citing his drop from support in Iowa to New Hampshire he suggests that those that broke from Edwards, who we would think would break to Obama, instead went to Clinton. The problem is this treats the results out of two different states as if they were from the same set of folks. I don't think this explains the Clinton surprise as John Edwards finished at 17% which is consistently where he's been in polls out of New Hampshire since Iowa. In fact John Edwards is the one thing the polls got right, so I don't see how we can say that there was a "collapse" of Edwards explains Clinton's success, in NH there was no break from Edwards he finished right where he always was. I think Edwards may have played a role, but I'll get back to that at the end.

The next explanation I've heard is the contrarian nature of NH, and their desire to stick it to the media. I suppose this could have played a role but don't really know a way to test it other than anecdotally and I've not found anyone doing so. It's an interesting thought and I think it may be part though not all of the answer.

The other explanation I've heard was from Chris Matthews this morning who thinks that what happened was "The Bradley Effect" in which white survey respondents tell a pollster that they're voting for an African American candidate as they don't want to appear racist but when they are alone in the voting booth don't do so. It's interesting because it was while watching MSNBC that I saw this dismissed prior to NH, pundits citing Harold Ford's race in Tennessee as an example where an African-American got a vote percentage beating his poll numbers. It was suggested that this wasn't a factor in Iowa, as the Caucuses involve voting in public.

I think though, looking at the exit results vs. the poll results it's hard to make the Bradley Effect argument. Among white males we find that they broke for Obama. These are the people that I think are most likely to have lied to a pollster, for a white male it seems to be more progressive to be voting either for a minority or a woman. Had we seen this number break more in Edwards direction I'd be ready to call foul, but it didn't.

This leaves us with a final explanation, one that holds up with the exit data. Where the real surprise break for Clinton occurred was with Women. I'm not inclined to believe that women would not be proud enough voting for a fellow woman that they would lie to a pollster. Also I'm not quite ready to call white women liars, I mean, my mother is one. What this indicates to me is that there was a large block of women who changed their mind late in the game. Was it the "cry"? Was it President Clinton? Was it the fear-mongering by the Clintons? Was it John Edwards attacks at the debates? Was it Obama's muttering "you're likeable enough"? Or just the general media pile-on? I think the answer is probably yes to all of these, combined they made Senator Clinton seem like she was surrounded on all sides, vulnerable, but still out there fighting. That's something that must have gotten to a lot of women in the last couple days, and whether it just turned out more or changed their minds it doesn't matter. Through the Clintons' political skill and missteps of the other campaigns women sided with Clinton in force and she was able to pull off the upset.