Not to toot my own horn, but I was on this Unskewed fella pretty damned early.
Indeed, before he inexplicably became the most-quoted polling analyst in the GOP, the subject of serious profiles in places like Slate and The Atlantic, and even merited mention in a column by Eugene Robinson, readers of our Daily Kos Elections Polling Wrap were introduced to one Dean Chambers.
The purpose, in our case? Gentle mockery:
Speaking of national polling: did you know that Mitt Romney is up by eleven points? He really is! According to the most unintentionally funny attempt at "unbiased polling" you ever will read. Go ahead and read the write-up. While you are at it, check out our man's archives, where you will learn that if you just take the left-wing bias out of polling, Mitt Romney is clearly favored to win 350+ electoral votes. Just...read the whole thing. Oh, and one more thing: You're welcome.But, in our amused chuckling, something strange and awful happened. People on the right began to mimic the rantings of this small-bore Examiner polling blogger. And the issue of "skewed polls" suddenly, and inexplicably, became among the most discussed items in the electoral and political conversation for the past week.
What is mournful about that is not that Republicans in general, and Chambers in particular, are skeptics about polls showing Mitt Romney at risk of getting routed in November. It is human nature to be doubtful of pessimistic outcomes, especially when your heart is wholly invested in said outcome. Democrats were, to be sure, devout polling skeptics in both 2004 and 2010, while in the interim, it was the Republicans who were certain that what they were seeing on paper was not what they were going to see on Election Day.
My beef with Chambers, and the wave of Republicans that have followed his lead, isn't that it is analysis of polling that runs counter to my own.
My beef is that is simply poor analysis. It would barely qualify as acceptable polling analysis for a middle school student, and even then only because it would be mildly impressive for a middle schooler to be analyzing political polls.
That anyone in the Republican Party, or the political press, is taking him seriously is the biggest indictment of all in this whole sordid episode.
(Continue reading below the fold.)
For the uninitiated, this is a fairly typical bit of Mr. Unskewed's analysis:
Given what the current polls are showing, even a 54 percent to 40 percent margin among independents for Romney in that heavily skewed CNN/ORC poll, it is reasonable to predict that Romney could very well win among independent voters by a margin of 55 percent to 45 percent. Democrats could support President Obama by a 85 percent to 15 percent margin while Republicans show support for Romney by a margin of 90 percent to 10 percent. With Romney closing the sale for his candidacy with strong performances in the debates, the enthusiasm factor among the Republican based, and the lesser of such a factor among the Democratic base, leads to Democrats voting 10 percent less and Republicans voting 10 percent more. These factors, and the electorate resembling the partisan trends measured by Rasmussen Reports in interviewing tens of thousands voters, which shows the electorate this year made up of 37.6 percent Republican voters, 33.3 percent Democratic voters and 29.2 percent independent voters.There are no shortage of assumptions here, and they run from mildly errant to totally absurd. And, for the moment, let's leave aside the fact that Chambers appears at the end to be predicting the most lopsided defeat of a presidential incumbent since Herbert Hoover got dropped by a 57-40 margin in 1932.
Calculations using those numbers and parameters described above suggest a popular vote projection of Romney winning 58.3 percent to Obama's 41.7 percent. That is comparable to the popular vote majority received by Ronald Reagan in 1984.
But the most basic problem here, as several responsible journalists have managed to point out during Chambers' 15 minutes of fame, is that his method of "unskewing polls" is to simply take a party weighting that favors his political ideology, and apply it to every poll. Voila!
In Chambers' case, he uses a recent survey conducted by the Republican polling firm Rasmussen Reports which claimed that Republicans make up 38 percent of the current electorate, while Democrats make up 33 percent and independents 29 percent.
There are a number of inherent problems in this kind of "analysis."
For one thing, as has often been noted in the past two weeks, party ID is easily the most fluid of the demographic subcategories often associated with polling. People change their self-identified preferred political party with some frequency. And it goes without saying that they will change that more often than they will change ... say ... their gender or ethnicity. Even ideology, which is a state of mind as much as anything, is more permanent than one's party ID, which can flit from Democrat-to-independent-and-back-to-Democrat based on the news of the day with some voters.
This is why responsible pollsters tend not to weight for party ID (Rasmussen and Voter/Consumer Research are two noteworthy exceptions). Simply put, to weight for party identification is to substitute where the electorate actually stands with where you think (or, more appropriately in this case, hope) the electorate stands.
Because party identification is such a fluid statistic, most pollsters go where the data takes them, only weighting for more static factors like ethnicity, age and gender. Adding party ID into the mix means that you are making a guess at what the composition of the electorate will be, from a partisan perspective. Ironically, given the most favored word in Chambers' vocabulary, to re-weight polls in the manner in which he does is the most certain way to ... well ... skew the data.
Especially if you factor in political history, which Chambers and his acolytes apparently have little appetite to do.
Which brings us to the next inherent problem with Mr. Unskewed.
Chambers' analysis presumes an R+5 electorate. How often has that happened in a presidential election in the modern era? The short answer, of course, is never. Looking at the exit polls for every presidential election dating back to 1976, a total of nine presidential elections, Republicans have never comprised a plurality of the electorate. The best they have ever achieved, in 2004, was parity with the Democrats. Over the last four presidential elections, the average has been a Democratic self-identification edge of exactly four points. Which means, in short, that Chambers is assuming that the 2012 electorate will be nine points more Republican, on the margin, than the recent average, and a dozen points more Republican, on the margin, than the most recent election.
Therein lies one of the delicious ironies of the whole "unskewing" phenomenon. Conservative Republicans have decried that the raft of polls showing Obama with big leads by pointing out that the samples have party ID spreads that are as wide, or even wider, than the 2008 exit polls. In some cases, they scoff that polls would assume an electorate that is similar to 2008. But their intellectual polling godfather (Chambers) is assuming an electorate that has not existed since, roughly, the years when Calvin Coolidge was still the occupant at 1600 Pennsylvania Avenue.
But even in that uproariously silly analysis, they are still more credible than Chambers. At least they are comparing 2012 polls to 2008 exit polls, rather than comparing them to an arbitrary set of partisan figures.
But even there, conservative poll critics have nary a leg to stand on. The reason is that there are any number of plausible ways in which the Democratic-Republican spread in voter identification could be the same, or even wider, than 2008.
Consider two theories. First, Eugene Robinson:
That’s why the conspiracy theory is so dangerous for the GOP. If pollsters look at a demographically representative sample of registered or likely voters and find fewer Republicans than might be expected, it could be that voters who once might have called themselves Republicans no longer feel comfortable with the label.Even if Republicans care to deny that the last four years have been damaging for their brand name, there is another, equally plausible, explanation. Political Science professor David Karpf, writing for HuffPo, explains:
There is ample polling data to suggest why this might be the case. Whoever wins in November, it is clear that, while the past four years have been rough on the president’s image, the impact on the Republican brand has been nothing less than brutal.
Voters blame the GOP more than they blame Obama and the Democrats for the gridlock and brinkmanship that have characterized much of the president’s time in office. The Republican Party has taken stands on issues such as abortion and immigration that big segments of the electorate find extreme and unacceptable. Moderate Republicans, as a political species, are all but extinct.
If a polling sample shows Democrats outnumbering Republicans by, say, 32 percent to 24 percent (with most of the rest calling themselves independents), GOP partisans shouldn’t worry about a conspiracy. They should worry that this is a snapshot of how Americans feel about the two major parties.
Pretend we have a state (let's call it Bohio) with only 100 voters in it. (It's a beautiful state, where math problems can be solved on the back of an envelope.) In the 2008 election, 52 of these voters went for Obama and 47 went for McCain. The one remaining citizen voted for, I dunno, Kodos. Exit pollsters asked the Bohioans what party they belonged to, and 39 called themselves Democrats, while 31 were self-styled Republicans, and 30 were Independents.Karpf's fun tongue-in-cheek tone aside, he hits on a legitimate point. In all the fixation on the D-R gap, one simple detail has often been missed. Usually, those polls aren't finding dramatically more Democrats than in previous cycles, they are finding fewer Republicans and more independents. Teabaggers could be part of the explanation for that, as could the continued exodus of moderate Republicans from the fold, choosing to re-identify as independents rather than scoot all the way over to the dark side.
Now it's four years later, and Bohio pollsters want to know the state of the race. Along with asking who they'll vote for (Obama, by approximately the same margin), pollsters ask about Party ID. 38 of these voters now call themselves a Democrat, while 28 say Republican and 34 say Independent. The "party ID" gap has grown from 8 to 10. Does this strike you as odd?
Before you answer yes, take a second to notice how the gap has grown. A couple of our fictitious citizens have stopped calling themselves Republicans and started calling themselves Independents. There's an easy explanation for that move -- since 2008, a couple of Bohioans probably got involved with the Tea Party movement. Though Tea Partiers are deeply conservative, they also self-identify as Independents.
But, if anyone needs any further evidence of how half-crocked Chambers and his whole movement is, feast your eyes on this.
Our guy actually got into the polling game himself 10 days ago.
Leaving aside a stunning lack of knowledge about how polling works (see his explanation of "double-weighting," and his defiance of basic math in that somehow over 2,000 respondents yielded a margin of error of 3 percent), the topline result tells you everything you need to know. It was: Romney 55, Obama 44. For those scoring at home, the last time we saw a Romney lead of that size in this election was ... well ... never. It has never happened. Indeed, Chambers' foray into the polling game almost doubles up the highest Romney lead EVER recorded (a six-point Mitt lead by—who else?—Rasmussen).
This, of course, betrays the most simple rule of polling, one that requires a degree in neither political science nor statistics. If you have one set of data, and everyone else has something entirely different, it is extraordinarily unlikely that it is everyone else that is in error.
This is the horse that Republican partisans have hitched their analytical wagons to. Honestly, it is hard to know whether to laugh or cry.
What is easy to know, of course, is how this guy became the patron saint of Republican polling analysis. Because rather than presenting the 2012 electorate as it is, or as it plausibly might be, he presents instead the electorate as Republicans would dearly like it to be. If reading and watching conservative media over the past four years has taught us anything, it is that they loathe Barack Obama, and refuse to entertain even the possibility that the rest of America don't feel the same level of revulsion.
Thus, when a deluge of polls actually show the president to be (gasp!) a favorite for reelection, they simply cannot conceive of it. Which is where Dean Chambers comes in. He mollifies the conservative masses by offering them an appetizing explanation for what they are seeing. It is not that the public, if not amenable to a second Obama term, finds him more amenable than the alternative. It is that the polls have been doctored to create the illusion that the public feels that way. Which polls? Damned near all of them!
And a startling number of Republicans are willing to buy in. GOP pollster John McLaughlin went so far as to accuse media outlets of producing biased polls to try to discourage Republican turnout. The NRCC actually has developed a website to play the "unskew" game themselves. And, late this week, even Romney spokesman Eric Fehrnstrom got in on the action.
All this obsession over partisan ID and "unskewing" belies one critical truth: The reputations of pollsters depend on being tethered to the planet in terms of their data. As Marist's Lee Miringoff noted after McLaughlin's broadside on his own profession: "Why would pollsters want to look inaccurate?"
That, in the final analysis, is the essential question. And in ignoring that question, Republicans have taken the natural (and admittedly human) impulse to question information that runs counter to their hopes and aspirations to a new and disturbing level.
Their willingness to deny basic elements of political science and math, and more darkly, to ascribe nefarious motives to the entire polling community, would be somewhere between amusing and pathetic were it not for the fact that it is going to lead any number of their adherents to assume, even in a landslide defeat, that the outcome was somehow a "stolen" one.
In this sense, "unskewing" is an odd ideological cousin to birtherism: a willingness to temporarily suspend logic in order to offer a minimally acceptable rationale for undermining the legitimacy of the president, the legitimacy of which far too many on the right never accepted to begin with.