Thursday, January 10, 2008

No race effect

Here's the clearest evidence I've seen on the subject (from Matt Yglesias):The polls got Obama's and Edwards' shares of the vote just right. They underestimated Clinton's support by a mile. Why? We'll have to work on that one.

3 comments:

lidzville said...

I don't understand the math. If they got two of the top three contenders close to accurate, where did the votes for Hillary come from? Richardson? Independents? It still all has to add up to 100percent, and if the percentages for Obama and Edwards were close, who comprises the percentage representing Clinton's margin of victory? Did she truck in people from Vermont?

Derek Holmgren said...

So, having spent way too much time reading analysis of the New Hampshire primaries, I thought I might weigh in with my expert (read: completely amateurish and far too lengthy) take on it.

Assuming that the largely accurate poll numbers for the Dems except Clinton are particularly meaningful…

1. The whole White-people-support-Obama-in-polls-but-not-in-the-ballot-box thing might not be true. I mean, if his numbers in the polls pretty much reflect where he ended up, then this probably not a good explanation. Not that the two cases necessarily speak to each other, or that I am remembering correctly, but weren’t the polls showing that Corker would have a slight victory over Harold Ford Jr. in Tennessee in 2006? Wouldn’t these results suggest that either pollsters have unknowingly stumbled upon a way to solve the Bradley effect, or that even in highly racialized campaigns like the one in Tennessee, we have reached the point where those voters still who really care about race are the ones who would be more up front about it anyway?

2. The explanation based upon independents breaking for McCain because they thought Obama would win also seems off-base. Again, it seems like polls had about the right amount of support for McCain (no major bump for him, right?), and besides, there would have been a real disparity in the Obama poll numbers and his results if his supporters flocked to McCain on election day when they thought Obama was safe.

3. Not to make too much out of the poll/results parallels for Obama, but it also seems to show that lazy, unreliable student-voters also pulled their weight on election day. If they had stayed home to play on their brand-new Wiis that they got for Christmas, then there would be greater disparity between the polls and his results. Or perhaps because no one could get a hold of a Wii for Christmas, they had nothing better to do than vote…Could this spell doom for Obama as supply begins to catch up with demand?


If none of those three really played a role, Lidzville's question still isn't resolved: where did Clinton’s jump of upwards of 10% came from? One interesting supposition is that maybe the polls really weren’t that wrong and the undecideds and Biden/Dodd supporters swung hard in her favor on election day. Of course, that leaves the question of why she would garner essentially all of their support.

1. Maybe Biden/Dodd supporters really cared about Clinton’s experience and it was natural for them to pretty much all swing to her – is there any firm info on how this small group ended up voting?

2. The undecideds all went to Hillary because they pitied her. The piling on her seemed pretty bad, and while I was initially happy about it (disclaimer: I would vote for Edwards or Obama or Richardson before her) something nagged at me saying, “that last thing that Obama or Edwards need is for Clinton to become sympathetic.”

3. The proudly independent voters of New Hampshire (or so says the CW), who still had not managed to make up their minds, were so incensed about being told that Obama had the whole thing sewn up that they wanted to stick it to national media and provide other states with a chance to have meaningful input in the process. If you can’t tell, I very much doubt this one, but the behavior might also be just irrational enough to be true.

4. Hillary did well in the debate, and by further sharpening her message and retooling her campaign strategy, she made significant inroads with the undecided voters.

5. Voter fraud – isn’t this always what you are supposed to scream when your candidate doesn’t win? Then you get bonus points for launching into a tirade about those “damnable activist judges”…


All of this leaves me wondering what is going to happen now. Seems like a general blogosphere consensus is that Obama will pick up support has he moves into states with higher African-American populations. But most of his support comes from younger and/or wealthier white people (which is interesting in itself). Is there anything new to suggest that he has decisively reversed Hillary’s early lead among African-Americans?

More bad news for Obama might be that he apparently doesn’t do as well with Dems as Hillary. Perhaps a bit like McCain in 2000, does he run into real trouble in states that are less welcoming to unaffiliated voters wanting to vote in primaries? For instance, in the knock-out, drag-down fight for Colorado’s totally significant (read: not particularly significant) number of delegates, he might suffer because you have to declare for a party in advance in order to vote in the primary. Unless I am completely wrong about those rules, but that is how I remember it and the principle applies to other states. California might be difficult for him because it seems like they have pretty murky rules about all of this.

Will John Kerry’s losing ways somehow infect the Obama campaign now that he has endorsed Obama? As long as I am being facetious, will Ed Perlmutter's endorsement create the wave on which Obama rides to a clear and decisive victory in the primaries and the general election thereby ushering in an unprecedented Golden Age of American prosperity?

At the end of the day, has New Hampshire really changed anything? It seems that Clinton and Obama both have the organizations and fund-raising capacity that pretty much no matter how NH turned out, it was going to be a drawn-out slugfest for delegates anyway. And finally, how long does Edwards stay in the race, and if he leaves, where do his supporters go?

Derek Holmgren said...

Having re-read my post in light of Lidzville's post, I should also point out that when I joke about voter fraud, it was in the same spirit as Lidzville's comment about all of the voters not accounted for in the polls and it was not done to belittle Lidville's way of posing the question.