What Happened To The Polls? The ZbornakEffect?

Ah, theories abound.

1. There was a conspiracy, somehow, because pre-election polls are just so much more valid than actual vote counts.

2. The Bradley/Wilder effect -- voters were afraid to tell pollsters they didn't want to vote for a black person, so a certain percentage of them lied about their preferences. But wait -- the pre-election polls did NOT overstate Barack Obama's support. He averaged 36.7%, according to Mark Blumenthal's compilations.

If anything, they understated Hillary Clinton's support by nine points. Let's name this phenomenon the "Dorothy Zbornak" effect -- for some reason, older women voters refused to disclose their preferences to pollsters, or refused to admit that they favored Hillary Clinton.

Mickey Kaus postulates an application of the reverse Bradley effect -- that Iowa Democrats somehow felt social pressure to stand up in front of their peers and cast a vote for a viable black candidate.

3. The exit polls were wrong. -- No -- by the third wave, they were basically correct. In Iowa and New Hampshire. So something is working for Edison-Mitofsky Research.

4. Ballot placement helped Clinton. Probably true, but it does not account for the discrepancy.

5. Obama actually won New Hampshire because both he and Clinton were allocated the same number of delegates (true) and he has more New Hampshire superdelegates than Clinton (true). Cute. But by that score, Clinton has a national lead because she has more pledged superdelegates nationwide than Obama.

6. Feiler's Faster Thesis with the Skurnik appendage. That is -- voters process information rapidly -- and they process through the information even more quickly. And uninformed late deciders usually rapidly assimilate and process late-breaking news. Again, from Mickey Kaus. Combine this with some evidence that working women are unusually late deciders...

Comments

I think there are a few reasons that should all be taken together.

First, there were still a significant amount of undecideds in all of these polls. Most of them apparently broke towards Hillary.

Second, I'm sure several polls, learning from Iowa, overestimated how many Independents would vote in the Democratic primary rather than the Republican. They didn't all go to McCain, possibly, but also Ron Paul and Rudy Giuliani. With Obama "assured" of a victory, perhaps these Independents who were on the fence felt they could have their cake and eat it to by voting Republican.

Third, women were clearly undersampled as Hillary turned out a lot more than expected and won most of them.

Is that enough to account for the discrepancy between the polls and the results? Maybe for some of them. The others have explaining to do.

Posted by VA Blogger | January 9, 2008 8:46 AM

Is there actually any doubt that some tactical tears swung women back to her and helped encourage the defection of Edwards voters?

Posted by Seth | January 9, 2008 8:50 AM

The Wilder Effect is overstated.

White men who were polled did exactly what they said they were going to do.

Posted by nctodc | January 9, 2008 8:51 AM

I would add a couple more possible explanations.

7) The debate "ganging-up," BIll Clinton saying he can't make HIllary into a man, the crying episode, the sexist hecklers all conspired to give Hillary a late surge among female voters who say her as being unfairly treated because of her gender. Basically, the gender card was palyed masterfully, and worked. The evidence for this is the wide gender gap not seen in Iowa.

8) Too many thought Obama was a lock on polling and many independents decided to go play in the republican primary instead. Under this theory, the polls actually hurt Obama because people thought he was going to win and they were less sure McCain would beat Romney. The evidence for this is that the dem turnout advantage in Nh was not nearly as high in IA, suggesting that McCain took many Obama independents.

9) Michael Whouley did his legendary magic in a state that is in his backyard.

Posted by RKA | January 9, 2008 8:54 AM

It was the staged crying scene that did it. I'm amazed that wasn't number one on your list.

Posted by paleo | January 9, 2008 9:10 AM

My take: And before we pile on polls, consider that (1) polls were mostly on the field over the week-end when the bounce was at its highest, when bounces are notoriously volatile and it's a very hard polling period; and (2) independents apparently massively switched their allegiance at the last minute from the Dem to the GOP race. We knew Obama and McCain couldn't both do well, and that's what happened. And (3) Rasmussen, ARG and Suffolk all suggested that the race had stabilized and that Clinton was slightly coming back. Remember, ARG said last night that Clinton had regained her lead among female voters. Frankly there were signs in the polls that Clinton was very much alive. The media is trying to push the responsibility of their misguided coverage on polls, when frankly it's the way they portrayed them.

Posted by dan | January 9, 2008 9:11 AM

Just because Obama's numbers were spot on doesn't immediately kill the Bradley hypothesis. We don't even fully understand it.

And the most recent example of it, the 2006 Maryland Senatorial race (http://en.wikipedia.org/wiki/Bradley_effect#Diminishing_effect.3F) showed a similar effect - the black candidate's numbers were spot on, while the white candidate surged ahead with independents. Add in Clinton's added appeal to female voters, and it's not surprising that undecided women broke heavily for an "experienced underdog" white female over the "less-qualified frontrunner" black male.

Remember, this was the first secret ballot of the primary season. We'll have better data as more races come in.

Posted by ly_yng | January 9, 2008 9:14 AM

My take: And before we pile on polls, consider that (1) polls were mostly on the field over the week-end when the bounce was at its highest, when bounces are notoriously volatile and it's a very hard polling period; and (2) independents apparently massively switched their allegiance at the last minute from the Dem to the GOP race. We knew Obama and McCain couldn't both do well, and that's what happened. And (3) Rasmussen, ARG and Suffolk all suggested that the race had stabilized and that Clinton was slightly coming back. Remember, ARG said last night that Clinton had regained her lead among female voters. Frankly there were signs in the polls that Clinton was very much alive. The media is trying to push the responsibility of their misguided coverage on polls, when frankly it's the way they portrayed them.

Posted by dan | January 9, 2008 9:15 AM

The comments need an edit button - UNDECIDEDS broke late for Hillary, not independents, which Obama won pretty handily.

My bad.

Posted by ly_yng | January 9, 2008 9:18 AM

I think it was her subterranean (and false) direct mail. I heard about these pieces directly from female undecided voters in NH and had to go look them up to see what they were referencing.

False claims about Obama's pro-choice record, debunked earlier in the campaign:

http://tpmelectioncentral.com/2008/01/hillary_mailer_hits_obama_on_abortion.php

Rove-style politics of fear:

http://www.politico.com/blogs/bensmith/0108/ProHillary_mail_Be_very_afraid.html

DOVER, N.H. – Facing the prospect of defeat in tomorrow’s primary, Hillary Clinton just made her strongest suggestion yet that the next president may face a terrorist attack – and that she would be the best person to handle it.

She pointed out that the day after Gordon Brown took office as the British prime minister, there was a failed attempt at a double bombing in London and Glasgow.

“I don’t think it was by accident that Al Qaeda decided to test the new prime minister,” she said. “They watch our elections as closely as we do, maybe more closely than some of our fellows citizens do…. Let’s not forget you’re hiring a president not just to do what a candidate says during the election, you want a president to be there when the chips are down.”

http://www.boston.com/news/politics/politicalintelligence/2008/01/clinton_heighte.html

Posted by bob | January 9, 2008 9:43 AM

I would suggest we think about New Hampshire in terms of the last couple of months, to see where the candidates really came from, and what they achieved. Remember that two months ago, Hillary was inevitable, with a massive lead in New Hampshire - and on the day she won by about 3 per cent, after a weekend of blatant manipulation (tears), scaremongering and hatemail. This suggests to me that the longer trends are still running in Obama's favour, and that it would reasonable to see this as a Hillary "blip" in favourable northeastern territory, among mostly white workers. You can run the story as Hillary's comeback, but the longer trend is more a story of Hillary decline from supremacy to a serious fight. Given that New Hampshire is likely to be a state where the demographics line up more in her favour than many others, I think that longer analysis should have Hillary feeling extremely nervous at this point. The Independents may have gone to McCain, as others suggest (which would fit with their known loyalty in New Hampshire), and enough women may have been swayed at the last minute by tears/sentiment/Gloria Steinem etc. Either way, when you consider the lead Hillary once had in the polls, this is not a great night for her, far less a major comeback. If Obama wins handily in South Carolina and Nevada, I would guess that once again the fall of the House of Clinton will seem like the real story - and that is what the longer-term trends suggest at this point.

Posted by maximus | January 9, 2008 9:59 AM

Matt's strenghts are on analyzing substance, not polls. The "Obama's number proved right hypothesis" is a transparent fallacy, because in opinion polls, the interviewees are allowed to say "undecided," and the idea that 100% of undecideds break one way is unheard of. Often, undecideds do break in favor of the winner in a higher proportion than the winner's margin (e.g., winner wins 52-48, undecideds break 65-35), but it's unheard of for every single undecided to break for one candidaet. Here, on the very last day, the crying story completely dominated, so there may be some tiny element of truth in Matt's theory, but his theory is clearly insufficient to explain what happened.

Posted by RaymondA | January 9, 2008 10:22 AM

Remember that a woman running for president and having a chance to win is so new a situation in this country where women have had the vote for only eighty seven years means that Hillary never was inevitable and we cannot ever call her that.
Those who insist that she was all summer were always overstating her level of preparation and the incredible tightness her campaign had until then achieved.
Remember too that in Iowa perhaps Hillary never had a chance with that evangelical electorate: does an easterner ever do well there? The same thing that kept dodd or biden or Bradley from ever getting any chance there might have kept hillary and Rudy from doing well there.
Remember too that Obama was too confident coming out of Iowa, telling NH it was their job to finish this for him, talking of History and all that: in this way too he is a neophyte high on his own fumes.
Hillary decrying in the debates that being the first female president was no little change: it would be seismic shift in possibility.
Remember that if Hillary's alledged ineviotability was a turn-off then so must have been Obama's. A great deal of this turn-around must reflect badly on Obama's ability to close a race.
Hillary didn't give up: that more than a lot of things must be seen as appealing to alot of people.
Hillary quite possibly benefitted from Dodd and Biden leaving the race: thier message was steady experience matters and thier few percentage points may have shifted to the other candidate talking experience.

Posted by Michael | January 9, 2008 10:40 AM

@ Michael

First of all, the Hillary campaign spent quite a long time running from the stance of inevitability. You may not have agreed, but that is how she ran for a long time, and how it was presented. As for the evangelical electorate, there is no reason to think that they would take to Obama any better than Hillary. Obama certainly did not take New Hampshire for granted, and your remarks about fumes are offensive rather than informative. As for the seismic idea of a female president, and women's voting rights, perhaps you would care to remind us how long African Americans have had full civil rights, and why a woman is more "seismic" than a non-white president? As for the Hillary did not give up idea - well, nor did Gravel or Kucinich, and they hardly made much of an impact. Dodd and Biden are not likely to have been factors in this game. It is more plausible to see Hillary squeaking out a win in a state she should have won handily (judging by the vast majority of polls over the last 2 months), and achieving a victory that will probably help Obama more in the long run. Obama voters will know they have to turn out and fight to the end, and this will motivate them very strongly. The Clinton campaign has begun going negative very early, and this may well turn off uncommitted voters. Finally, the demographics in NH are more in Hillary's favour than those in Iowa, let alone SC or Nevada. Overall, this is a better night for Obama than it looks, and Hillary has used a lot of ammunition to produce a relatively mediocre result. The comeback narrative sounds good, but is really the impact of poorly sampled polls in a short period of time, coupled with a high rate of indecision.

Posted by mahant | January 9, 2008 10:49 AM

I blame patronage, which actually seems to be working for John McCain. Depending on how you define it, of course. Maybe that didn't work for Barack Obama?

Posted by Horsey | January 9, 2008 10:54 AM

I support Obama, but as a woman I would have voted for Hillary last night if I lived in NH. And not because of the tears which I actually thought were phony. I have been sick to death of the sexist BS coming from the media regarding Hillary and things just came to a head Sunday and Monday with MSNBC having Scarbourgh, Matthews, Schuster etc... excoriate Clinton, with the NYPOst and Boston Globe covers, with the Drudge gloating, with Dana Millbank's "Fired Up, Ready to Bore" nastiness the day before...

What pundits don't understand is that it is possible to like more than one candidate. I love Obama, but I felt like Hillary deserved better than to have Rush Limbaugh and company singing, "Ding, dong the witch is dead" this morning.

Now I can happily vote for Obama in the SC primary.

Posted by Teresa | January 9, 2008 11:48 AM

Are ground operations stupid?

Remember that all the NH bigwigs like Jean Shaheen to Kathy Sullivan were actively working for Hillary. And they all delivered big time with turning out the vote. Compared to that Obama had no established politician doing the same thing for him. In Iowa Obama had a great ground game.

One other thing, Obama needs to start talking about economy with SPECIFICS more often including in his victory/concession speeches. Economy is the number one issue for voters. As someone said it on NPR today, Obama is like poetry while Hillary is text with specifics.

Posted by my 2 cents | January 9, 2008 12:08 PM

Re the "Sbornak Effect": it's actually *Zbornak*, per Google, after the Bea Arthur character on The Golden Girls.

Posted by Scott Lahti | January 9, 2008 12:41 PM

You realize it's "Dorothy Zbornak", right?

http://en.wikipedia.org/wiki/Dorothy_Zbornak

Posted by Sophia Petrillo | January 9, 2008 12:44 PM

1. After Iowa, polls all started to use models that a) over-estimated the % of ind voters that voted in the Dem primary, b) also over-estimated the % of voters 25 or younger, and c) under-estimated the turnout of women who had not voted in primaries before.

Attribute this to a) McCain b) false belief that younger votes would become as engaged nationally as they were in Iowa c) Clinton's voter ID and mobilization program which worked much better with the younger first-primary women in NH than the older first-caucus women in IA

The fake tears, planted "iron my shirts guy," and talk of media sexism all helped with c, but that all only worked because they had laid the groundwork to mobilize women for months.

2. Lower-income, more economically oriented voters gave up on Edwards; for a few days after Iowa he was gaining in Zogby tracking and then started to drop on Sunday. Instead of undecided lower-income households breaking for Edwards, they broke for Clinton. Credit Bill.

This is why those who say Edwards hurt Obama are all wrong; O's problem is that he didn't appeal to white blue collar voters.

Posted by Gus | January 9, 2008 12:59 PM

Really great post, Marc. Especially coining a pretty awesome new phrase, "The Zbornak Effect."

And maximus, excellent comment... that's the way I've been thinking, too. The other problem Clinton has is that her surrogates are generally abhorent (see: Terry McAuliffe, Sen. Bob Kerrey), and prone to foot-in-mouth disease. Chuck Todd on Morning Joe today told a fascinating story about how major-league Clinton staffers and supporters were stabbing each other in the back, on background of course, convinced they were getting trounced. Nice morale over there at the Clinton camp.

Marc, thanks again for your excellent election night coverage... for IA and NH, you were definitely getting reloaded more than anyone else. :-)

Posted by JenJen | January 9, 2008 2:05 PM

Frank Luntz suggests that the turnout model was broken. Also, I think the Gloria Steinem editorial was more influential than we realize (was it reprinted in the Boston Globe?).

Posted by Ed Unneland | January 9, 2008 3:03 PM

The corporate powers that be have gotten really good at stealing elections and snowing us with exit polling cover-ups. Third time will be another charm. The massive amount of evidence of irregularities has been mostly ignored by our corporate leaders and corporate media. Them that gots the gold makes the rules and the machines that elect the rulers! When they steal this next one they will say "That Hillary was just to polarizing". Are we to really believe black folks arent going to vote for Obama? Thats right. The media is trying to convince us THEY are voting for HillaryD!! Please.

Posted by Jacqueline | January 9, 2008 7:53 PM

Hillary found something that she could do that Obama and the other candidates could not. She cried and got her way.

If any of the other candidates in either party had done what Hillary did in that New Hampshire diner, their campaign would be over. It would have been their Howard Dean "eeehaw" moment.

Posted by Jim | January 9, 2008 9:10 PM