To listen to or read much of the coverage of the New Hampshire Democratic primary tonight, you'd think the winner wasn't Hillary Clinton, but voters determined to defy polls and pundits. Indeed, the Clinton campaign itself, which spent much of the last 48 hours lowering expectations for NH, seemed as surprised as anyone else.
But HRC did win, and we're now into a contest whose outcome simply cannot be predicted.
So how did HRC beat Obama after losing in IA? Well, on one level, it's obvious that Clinton did better in a two-and-a-half candidate race than in a three-candidate race. She wasn't stuck with a capped and eroding share of the vote after all.
Moreover, the exit polls indicate that the Democratic primary vote broke down along the lines everyone expected before the Iowa results. HRC won women, people with family income under $50,000, union members, and registered Democrats, while Obama won men, upscale voters, non-union voters, and independents. Obama won big among the youngest voters, as did Clinton among the oldest (though Obama failed to do nearly as well among thirty-somethings as he did in Iowa).
It wasn't much about ideology: all three leading candidates performed almost exactly at their statewide percentages among every liberal, moderate and conservative category.
And it wasn't much about turnout patterns, either: levels of both Democratic and independent participation were up sharply from 2004, though Democratic turnout was up a bit higher. (Oddly, Bill Schneider of CNN suggested that McCain beat Obama by attracting higher-than-expected indie particicpation to the GOP primary. But since registered indies represented 42% of Democratic primary voters, and 34% on the GOP side, while total Dem turnout appears to have been nearly 25% higher, it's hard to credit that theory, particularly since McCain didn't exactly crush the field among indies).
So: were all the polls just wrong, or did something happen in the last day or two?
I'm sure I don't know, but I'm more likely to think the polls were wrong than believe HRC's tearing up or anger at the polls themselves moved thousands of voters. One theory we're going to hear about is the "Wilder factor" (named after former Gov. Doug Wilder of VA): African-American candidates tend to underperform their poll numbers because people are more likely to indulge their racial prejudices in the privacy of the polling booth (a privacy that doesn't, of course, exist in the Iowa Caucuses).
One thing I haven't heard a thing about tonight is early/absentee voting. It's possible that a lot of Granite Staters voted one way a couple of weeks ago and then reported their preferences to pollsters another way.
In any event, tomorrow is early enough for speculation about the impact of NH on the contest as a whole.