Mickey Kaus lists the most common theories, but there are at least three reasons they are not completely convincing.ÂÂ
First, the NH polling average for Sen. Barack Obama was 38.3% and the ultimate result was 36% – within the margin of error and thus not supportive of the Bradley/Wilder effect theory. Whether Obama benefitted from a reverse-Bradley effect in Iowa is an interesting question, but one I think is unanswerable. Obama’s lead in IA may have been inflated by the “second choices” phase of the Democratic caucus, but this could be a bandwagon effect as much as a reverse-Bradley effect.
Second, as Jay Cost points out at RCP, late deciders did not break to Sen. Hillary Clinton, which runs contrary to the the Lazio effect or the Feiler/Skurnik theory Kaus posits. It also discounts Clinton’s debate performance as a factor.
Third, the Independent vote was 44% this year, as opposed to 48% in 2004. This might lend slight weight to the Congestion Alert effect, but it is far from clear that Independents likely to have voted for Obama decided to vote for McCain in the GOP primary instead. I suppose it could be argued that the Underdog voter decided to do this, but it is not as though Obama and McCain match up well on issues presumably important to otherwise Democratic-leaning Independents.
Accordingly, I am inclined to agree with Jay Cost’s conclusion:
Clinton won by mobilizing the traditional Democratic coalition that is demographically inclined to her. You don’t just win elections by persuading people you’re the best candidate. You win elections by getting those people out to the polls. This appears to be what Clinton did. Accordingly – the implication is that the polls were wrong not because of last-minute shifts. They were wrong because they underestimated Clinton’s ability to draw out her base.
As I have noted in prior posts, the media tends to cover polls much more than they do organization, despite the fact that turnout operations have been key to the election results in 2002, 2004 and 2006.
Among political junkies, Clinton’s narrow victory in NH will add to the mystique of Michael Whouley, who joined the Clinton campaign in NH last month.
The pre-primary polls may not have captured his efforts because pollsters likely adjusted their turnout models after supposedly missing Obamamania in IA.
Update: At Time magazine’s site, Clinton’s NH director Nick Clemons is taking credit for the ground game. That may be true also, though Whouley is known for staying in the shadows and hates talking to the media. Either way, it’s another indicator of the importance of organization.
Update x2: At Pollster.com, Charles Franklin reaches similar conclusions to those here, though he did not have the data about the late deciders. He also suspects the pollsters’ likely voter screens were not accurate (as do I, for the reasons stated above).
Update x3: Rasmussen points to the same factors, but again missing the data on the late deciders.
Update x4: Zogby chimes in, claiming that his Sunday night and Monday polling had it as a two-point race, but that the three-day average required him to report a huge lead for Obama. He also suggests problems with the turnout model and the Congestion Alert effect.
I hate nuts and boltsy politics a lot really but I’ve noticed I’ve heard a lot about how “Independents” broke but not so much about how “Undecideds” broke. So in my head I’ve been figuring mostly the pollsters, who are very biasy and really a lot more appreciate what Hillary’s negatives mean, probably overestimated the Iowa bounce in a race where people made up their minds mostly a long time ago. In a three man race of whiny liberals who only talk about teh changeyness there’s not a whole lot of reason to pull up your flag and plant it somewhere else just cause someone else maybe has more better changeyness. But I imagine I guess there’s an actual reason to be found.
[…] Karl at Protein Wisdom says it wasn’t the late deciders who decided it. Follow the link to the Jay Cost piece he […]
bradley…..wilder….effect….theory…..zzzzzz…………
[Shortened so as to not break the site’s css – Karl]
Because all polls are full of crap, it’s just that this one was more so.
This post is interesting to me because I’ve just assumed that the Clinton Machine was getting predictably large turnouts. If the pollsters have been underestimating the power of the Clinton Machine on the Democratic Base, then they have done worse than merely underestimate — they’ve misunderestimated.
Thanks, Eleven, you knucklehead, one long word to screw up the page.
“As I have noted in prior posts, the media tends to cover polls much more than they do organization, despite the fact that turnout operations have been key to the election results in 2002, 2004 and 2006.”
Of course – it isn’t as boring and is much more fun, even though organization is very important.
Historians cover battles more often than they do logistics or planning, because battles are exciting and colorful. Putting together a schedule of trucks to run from rear base to forward depot over which roads carrying what and then back down which roads doesn’t quite have the same pizzazz.
Unless it’s Kaptain Karl Rove doing the planning. Then the media gets to introduce concepts like pure evil, brainwashing, and stolen votes.
Hey, PURE evil is always interesting and draws the eyes. It won’t be the planning, just the Evil!
jdm, install this Javascript to forever eliminate that text overrun problem.
Yup.
The blah-blah effect…and the blah-blah effect…and the blah-blah effect…
Sounds like something that goes between an instrument and an amplifier – that has no user’s manual. The more complicated the pollsters make it, the less reliable polling is. Sorta like climate models. One (or is it two?) big joke(s).
Face it, pollsters.
You are trying to pull a rabbit out of a hat. Most people with half a brain who are polled hang up on these idiots, because the questions are always black and white, and leave no room for any shade of gray. I have been called twice, and after agreeing to answer the questions the first time, I made it about halfway through the “poll”, and then excused myself, and hung up. I have never seen a poll question that leaves any room for shading, or room for a “Yes, but…”
The only question that matters is: “Who are you going to vote for?” And by now, a lot of the “pollees” are gonna say anything, just in retaliation for interrupting their dinner. I know that the last time I was called, I gave them a polite earful of baloney.
Because, as far as I am concerned, that what polls are.
KISS! Blah-blahs are subjective. “Nuff said.
Thanks, Jeff, interesting little script. I think I like Karl’s approach better tho’.
I’m an idiot – I see your link does address the undecided or I guess late deciders is the right word. I wish I had some more zest for the politicsy part of politics… I guess that’s the political science part but I guess cause it’s so close to what I do for a living I just filter that stuff out cause it feels like a lot of work to really get a grasp on all the variables and stuff enough to get the analysis part started.
You know but maybe I can get more in the spirit as things go along.
You have a flair for it though. I bet if you had been doing this stuff at like 25 your life would have been very different and probably I wouldn’t even know you except maybe I would recognize your name in the news.
Bah. Polls are an attempt to mechanize the process — to generate a simple metric for politicians to follow. Wandering around out in the hinterlands talking to grubby real people, or buttonholing passers-by in a city environment, is sooooo tedious. (Dangerous, too, nowadays.) Numbers are much simpler. X% in favor, Y% against, if X .gt. Y go to re-election… it doesn’t work. It can’t work.
Ask Nancy and Harry. They took office in a glow of poll results that said X was much greater than Y, and set their sails accordingly. What they found out was that the pollsters hadn’t asked the next question. Why should they? — they hadn’t been paid to. But X turned out to be the sum of the absolute values of positive and negative numbers. Nobody’d asked what the sign was, and -X’ + Y turned out to be bigger than X’.
Frustrating. But compared to that, a few percentage points in an election is bang-on accurate.
Regards,
Ric
eh, I notice you left out the real reason this couldnt be a bradley/wilder effect. This is DEMOCRATS who are genetically unable to be racists. Only Republicans can suffer from this effect. At least that’s what the democrats keep telling me.
THE DARK VEIL OF CHANGE
Change a). verb
Change 1.To make or become different.
2007 Webster’s dictionary
Political “change” in 2008
Will the candidate please state the obvious, over and over again
Change is the course in 2008
There is no incumbent president running as a candidate,
Change is the only course of the event,
Change shall happen no matter who runs, who wins or who loses.
Change is the only possible outcome,
Change is not an option in 2008
Change is the event.
This 2008 political “change” transmutation frenzy is seized upon like a world stopping innovation that has just been discovered.
Invented only by those who embrace “change” as their political logo.
Change is all shinny and new, oh my God this candidate has the “change”,
did you here what they said; that candidate has the “change”
What is political “change”?
Does anyone really know?
Yes — Yes I do
Political “change” is an uncertain obscure intangible indefinite promise broken before it is heard or spoken.
It is an insult when a candidate declares that he/she is “change”
but has yet to never inform America of the substance of these “changesâ€Â
It is equally insulting that these “change candidates” believe America is so unable to recognize “change”, they must use the word “change” at the beginning and end of every sentence.
It is equally insulting when a candidate declares this election is about
“you” as in “you the people” and “giving back†to “you” the people,
America when will you demand,
“giving what back�
The intangible utterances of a promise?
The unattainable utopian political dream world
declared so many times for so many years.
The “giving back” that America has never actually
had in their collective possession ever.
Political “change†is a formula used as a shield that the politician hides behind to conceal and obfuscate and obscure all political truth.
Political “change†is a dark veil, a verbal hand trick, subterfuge practiced by masters of auditory illusions.
Political “change†is cunning with a deceitful scheme that is executed with machine like precision to excite the masses into blind hysteria, a “change” promise which the masses manifest into a fruitful and prosperous future frenzy of underlying false hope
Beware of the candidate that speaks only of “changeâ€Â
For “change†is the master’s auditory illusion,
That in the end
“changeâ€Â
is always more of the same.
[…] the pre-election polls did not greatly overstate the share of the vote Obama was expected to win. As noted here yesterday: …the NH polling average for Sen. Barack Obama was 38.3% and the ultimate result was 36% […]