The Delusions Of The Poll Denialists
Steven Taylor asked this morning where the poll denialists had gone now that the polls were showing Mitt Romney surging in the wake of last Wednesday’s debate. In that regard, Buzzfeed’s Ruby Cramer talked yesterday with Dean Chambers, the guy who originated the whole “skewed polls” meme just about two weeks ago:
Mitt Romney’s rise in the polls has produced conservative elation and, in some quarters, rather exotic explanations of the sources of Romney’s rise.
Indeed, when the Pew Research Center said Tuesday that Romney was leading President Obama by 4 percentage points among likely voters, conservatives were quick to embrace the new numbers. The national polling unit that would have been branded last month as “skewed,” was evidence this morning of a strong Romney rally, and some in the movement noted the change of mood with amusement.
“What are out formerly disgusting and now beloved polls telling us this morning?”tweeted the author David Limbaugh, Rush’s brother, Wednesday morning.
But Dean Chambers, the founder of the movement to unskew, says this isn’t just conservatives reversing the new numbers. The pollsters are actually becoming less skewed and more honest, he says — and the blogger is taking much of the credit.
“I exposed the polls for being very inaccurate,” Chambers told BuzzFeed. “I showed Romney leading in August and September when everyone thought Obama was ahead. Now pollsters are skewing less.”
On his website, unskewedpolls.com, Chambers did not alter, or re-weight, the new Pew poll at all. The poll is listed in his tracking table as having “0 Skew” — a first for the website.
Chambers attributes Romney’s rise in the polls to less unskewing — to taking a larger Republican sample size — and also to Romney’s solid debate performance. Pew says that it adjusted its Republican sample size because more voters are now identifying as Republicans, but Chambers dismisses that claim.
“I’m not buying that,” he said. “I think they polled a more balanced sample for the sake of being accurate. We’re getting closer to the election, and they’re all gonna claim their reputaiton based on their numbers in October.”
So, you see, they haven’t gone away they just think that they’ve scared the pollsters into creating less “skewed” polls than they have been to date. There’s no real evidence, of course, that there has been any significant change in the D/R/I balances of the major polls in the last week, but Chambers has decided to take an uptick in Romney’s numbers that is clearly tied to the results of last week’s debate and claim that it happened because he scared the pollsters. If it turns out that the polls start moving back in Obama’s favor, you can bet your bottom dollar that these people will be back in full force.
My brain hurts.
Yet another example of how right wing “reality” diverges from the real thing. Makes me want to bang my head up against a wall.
What’s surreal about this whole asinine “debate” is that the left’s chattering classes don’t even know what they don’t know.
Check out for example NBC/WSJ/Marist’s latest survey for Florida. Their likely voter set is 63% white. The only problem with that, of course, is that in ’08 the Florida electorate was 71% white. Then go ahead and compare the white, black, Latino breakdown in their Oct. 3rd poll with the racial breakdown for their Oct. 11th poll. White voters in Florida apparently in one week suddenly disappeared en masse. Also compare the age-based demographics of their Oct. 3rd poll with the age-based demographics of their Oct. 11th sample. Back over to you, Katie.
Look, you don’t need a ouija board to figure out what’s been going on here. All you need are two eyes and for them not to be wired shut.
That’s not true. The Pew poll, which you linked to, has a partisan breakdown of registered voters of: 33.5% R, 32.9% D, 30.3 I. Last month (9/19), the partisan breakdown of the Pew Poll was 35.8% D, 29.5% R, 31.2% I.
I’m not saying that Pew was “scared” into changing its poll sample by the unskewed guy, but it’s a fact that there was a change of +4 Republican and -2.9 Democrat in the Pew samples from Sept to Oct. I would not be surprised if other polling organizations have similar changes in their samples as well.
The polls could not be trusted two weeks ago, and they cannot be trusted today. There is some oversampling due to using 2008 as a model, but this is not the larger problem. The polls are missing a large block of voters, and they are misrepresenting another. This is not intentional, and there is no way to “unskew” these polls.
The pissed-off voters from 2010 are still pissed-off, but they have mostly tuned out. They are not voting for President Obama, and they do not care if anybody knows it. They are either not answering the phone or they are hanging up. If they were counted, they would tilt the polls more towards Mitt Romney.
The under-decided voters are not counted as undecided, but they are up for grabs. These are mostly Clinton Democrats, and they do not like either candidate. They are counted as supporting one candidate or the other, but this support is tepid. When Bill Clinton said that even he could not have fixed the economy in four years, he pulled them to the President. This was the vaunted Obama surge and landslide.
Mitt Romney’s debate performance and President Obama’s non performance overrode Bill Clinton’s OK. These voters are now tepid Romney supporters, and this is the Romney surge.
Partisans do not understand these voters, and these voters can be spooked easily. Once spooked, they stampede to the other side. Democrat partisans thought these voters were Progressives. Republican partisans now think these voters are Conservative. Everything any partisans does will be attributed to their candidate, and no amount of denial will remove it.
Clinton voters are more fiscally conservative than Progressives, and they are more socially progressive than Conservatives. If Republicans begin to bring up abortion, gay marriage, etc., there will be a stampede to Obama, and it looks like Republicans are starting their nonsense. Whoever spooks these voters last will lose.
To all those who want to classify me as something, go F*CK yourself. Keep your stereotypes to yourself.
Look, why don’t we just pass a law that all polls have to have the exact same partisan breakdown always? 33.3% Republicans, 33.3% Democrats, and 33.3% Independents.
Sure, polls that would have to confine themselves to this no matter what state they were polling would be worthless at best, but at least they would make us feel better.
Doesn’t Chambers just drip sticky arrogance with the boast about working the refs? I must resist the urge to engage the troll.
@TastyBits:
Why were the polls correct in 2006, 2008 and 2010 but not today?
@David M:
Why were the polls correct in 2006, 2008 and 2010 but not today?
Secret knowledge. Shhhhh.
@David M:
The 2010 polls were missing them, and they shunned the exit pollsters. This was the reason everybody thought it would be a long night. They have tuned out politics, and since 2010, they have been unresponsive to polls. These pissed-off voters are not Republicans.
I should have added that the pissed-off voters are not impressed with the House or Senate Republicans. I am not sure how they will vote in these races, but they are not necessarily going to vote Republican.
@TastyBits:
You’re ignoring the actual question: Why were the polls correct in 2010 but not now? The results of the 2010 elections were what the polls predicted, so I’m not sure what you’re trying to say about them either.
@David M:
After 2010, they are not answering the phone. If a model can account for them outside the actual response numbers, then it may be able to account for them, but I doubt it. We will know soon enough.
@TastyBits:
Any actual real-world evidence of this?
@David M:
Nope.
And this has been another episode of wishful thinking by Tasty Bits.
@TastyBits:
After 2010, they are not answering the phone.
Is there some Tea Party directive I don’t know about?
@mantis:
There are a number of these voters who overlap with the Tea Party, but many of them were Reagan Democrats. They are to the right of Clinton Democrats on social issues, but they do not endorse the Republican platform.
I do not see them being captured in the polls. I did not trust the polls two weeks ago, and I do not trust them now. In another two weeks, Romney supporters will be “unskewing” the polls again.
@TastyBits:
You have no evidence for your claims.
I do not see them being captured in the polls.
What does that mean?
@mantis:
It is an analysis of the non-partisan voters and how they fit into the polls. If my analysis is wrong, we will know shortly.
@TastyBits:
It is an analysis of the non-partisan voters and how they fit into the polls.
Assertions are not analysis.
If my
analysisguess is wrong, we will know shortlyFTFY
@mantis:
I stand corrected. They do not exist.
@TastyBits: “The 2010 polls were missing them, and they shunned the exit pollsters. This was the reason everybody thought it would be a long night. They have tuned out politics, and since 2010, they have been unresponsive to polls. These pissed-off voters are not Republicans.”
Do you have any evidence about the ‘going Galt group’?
@mantis:
Tasty Bits: “After 2010, they are not answering the phone.”
Mantis: “Is there some Tea Party directive I don’t know about? ”
You know who else used telephones?