With just six weeks to go, it is now evident that there is a clear-cut way in which the 2012 election cycle is unique from its predecessors. And not, as it happens, in a particularly good way.
Polls late in the election season have always invigorated those out in front, and to some extent dispirited those running behind. As a result, supporters of the trailing candidate or party have always beseeched supporters to either discount or ignore bad polling, and press on as if the race was a true dead heat.
But, in my two decades of following elections, never have I seen such a broad effort, from the most wacky corners of the blogosphere to people in positions of high authority in the campaign in question, to attempt to smear and discredit pessimistic data. That Republicans have decried media conspiracies in the past is old hat to anyone that has seriously followed campaigns. But it is a bit mournful to see the GOP's war on science extend now to simple mathematics.
More on the rantings of the haters skeptics after the jump. But first, the numbers:
PRESIDENTIAL POLLING:
NATIONAL (Gallup Tracking): Obama 48, Romney 45
NATIONAL (Ipsos/Reuters Tracking): Obama 49, Romney 42
NATIONAL (PPP for Daily Kos/SEIU): Obama 50, Romney 45
NATIONAL (Rasmussen Tracking): Obama 47, Romney 47
ARKANSAS (TalkBusiness/Hendrix College): Romney 56, Obama 35
FLORIDA (Washington Post): Obama 51, Romney 47 (LV); Obama 52, Romney 43 (RV)
NEVADA (Public Opinion Strategies--R): Obama 46, Romney 46
NEVADA (PPP): Obama 52, Romney 43
NEW JERSEY (Monmouth University): Obama 52, Romney 37
OHIO (Gravis Marketing--R): Obama 45, Romney 44; Obama 45, Romney 38, Johnson 11
OHIO (Washington Post): Obama 52, Romney 44 (LV); Obama 52, Romney 41 (RV)
DOWNBALLOT POLLING:
FL-SEN (PPP): Sen. Bill Nelson (D) 46, Connie Mack IV (R) 37
FL-SEN (Washington Post): Sen. Bill Nelson (D) 54, Connie Mack IV (R) 40 (LV); Sen. Bill Nelson (D) 55, Connie Mack IV (R) 37 (RV)
MA-SEN (Rasmussen): Sen. Scott Brown (R) 48, Elizabeth Warren (D) 48
MI-SEN (Rasmussen): Sen. Debbie Stabenow (D) 53, Pete Hoekstra (R) 37
NV-SEN (Public Opinion Strategies--R): Sen. Dean Heller (R) 44, Shelley Berkley (D) 39, David VanderBeek (IAP) 4
NV-SEN (PPP for the League of Conservation Voters): Shelley Berkley (D) 48, Sen. Dean Heller (R) 44
NJ-SEN (Monmouth University): Sen. Bob Menendez (D) 49, Joe Kyrillos (R) 34
OH-SEN (Gravis Marketing--R): Sen. Sherrod Brown (D) 44, Josh Mandel (R) 43
OH-SEN (Washington Post): Sen. Sherrod Brown (D) 53, Josh Mandel (R) 41 (LV); Sen. Sherrod Brown (D) 51, Josh Mandel (R) 39 (RV)
PA-SEN (Susquehanna Research--R): Sen. Bob Casey (D) 46, Tom Smith (R) 41
IA-03 (Benenson Strategies for the DCCC): Rep. Leonard Boswell (D) 45, Rep. Tom Latham (R) 45
IL-08 (PPP for the League of Conservation Voters): Tammy Duckworth (D) 52, Rep. Joe Walsh (R) 38
MA-06 (Stinson Strategies for the NRCC): Richard Tisei (R) 47, Rep. John Tierney (D) 35
NJ-03 (McLaughlin and Associates for the Runyan campaign): Rep. Jon Runyan (R) 51, Shelley Adler (D) 34
NY-11 (Siena College): Rep. Michael Grimm (R) 48, Mark Murphy (D) 38, Hank Bardel (G) 6
PA-04 (Tarrance Group for the Perry campaign): Scott Perry (R) 52, Harry Perkinson (D) 30, Others 8
A few thoughts, as always, await you just past the jump...
Though he has gotten an unthinkable amount of attention this week, we here at the Daily Kos Elections Polling Wrap were very early to the party in making fun of noticing the contributions of one Dean Chambers.
Chambers, a right-wing blogger, has found his 15 minutes of fame by (in defiance of all political science, math, and logic) "unskewing" polls by applying the most optimistic possible weighting for Republicans to every national presidential poll. By this method, he has concocted an alternate political "reality" where a 3-6 point Obama lead in the preponderance of polling is, instead, a lead for Mitt Romney in the high single-digits.
When Chambers came across my radar a few weeks ago (and I alluded to him last week on the Wrap), I regarded him as a somewhat amusing sideshow to the election. Alas, as has become abundantly clear this week, when one looks at the Republican Party, a sad fact has become self-evident.
They are all Dean Chambers now.
Behold: last week, one of the most well-known (and, until this possibly, well respected) GOP pollsters makes the stunning claim that pollsters are intentionally oversampling Democrats, with the stated goal of depressing Republicans to the point that they are dissuaded from turning out:
John McLaughlin, a Republican pollster and consultant to GOP candidates, told the conservative National Review last week that Democrats are lobbying media pollsters "to weight their surveys to emulate the 2008 Democrat-heavy models."
"The intended effect is to suppress Republican turnout through media polling bias," McLaughlin said.
Lee Miringoff, director of the Marist College Institute for Public Opinion in Poughkeepsie, New York, is unconvinced. "Why would pollsters want to look inaccurate?" Miringoff asked rhetorically in a phone interview.
That's a fair question if ever I have heard one.
And then, today, inexplicably, team Romney joined the chorus:
Asked about a new Washington Post poll showing Romney down eight points in Ohio, political director Rich Beeson said their internal polls are telling a different story.
"The public polls are what the public polls are," Beeson said. "I kind of hope the Obama campaign is basing their campaign decisions on the public polls... I have great faith in our data."
This would be exceptionally easy to dismiss as Republicans uniformly shooting the messenger because their presidential candidate (to say nothing of many of their downballot candidates) are on the road to defeat. Except that, in many ways, this cuts a bit deeper than that. McLaughlin's criticism, in particular, is somewhat grotesque, in that it implies that the pollsters are lying in order to influence electoral outcomes.
It is one thing for some random right-wing blogger (a la Chambers) to make that charge. But a member of the community of pollsters? One with a long tenure in the business? It is quite an accusation for anyone outside of the partisan fringe to make. For someone inside the industry, it is an extraordinary charge.
And, it is one that is, by and large, unsupported. College professor David Karpf, writing for HuffPo, absolutely nails it (and, really, go read the whole thing):
As we get closer to election day, you're going to hear more and more rumblings about "partisan voter screens" and faulty "party ID gaps." Dick Morris is already writing about it, so you know the rest of conservative fantasyland isn't too far behind. Ignore those arguments. Simple math and simple history provide a much more compelling explanation: the electorate today hasn't changed much since 4 years ago, but people (particularly tea partiers) have taken to labeling themselves differently.
The idea that a D/R spread that has changed two points, or even four points, from 2008 exit polling is somehow hugely aberrant is a little bit of a joke. Fixating on partisan ID in any event, as I have noted here before, is a tempting (and common) trap. It is one I caught myself in back in 2010, when I noted in several bits of poll analysis that it seemed inconceivable that the partisan breakdowns of polls would be just like 1994, given that the GOP's favorabilities were as bad (or worse) than the Democrats.
The problem? That is precisely what happened. If the years of analyzing polling data (here and elsewhere) have taught me anything, it is that, taken as a critical mass, the polls tend to be depressingly (or blissfully) right when all is said and done.
In other polling news...
- A lot of the fuel for the right-wing polling skeptics has come from the comparatively closer results that have been common with the two national tracking polls coming from Gallup and Rasmussen. The travails of the House of Ras have been well documented, of course. But, today, the Guardian's Harry Enten makes a compelling, and interesting, case that Gallup is every bit the GOP-leaning outfit as Rasmussen. Indeed, Enten invokes a big word in the polling lexicon: outlier.
- In response to a number of queries, both in the comments and on Twitter: yes, I am aware that there is a new daily tracking survey being conducted by the RAND corporation. That survey (in which Obama currently leads 50-43) uses a unique methodology that makes it, properly understood, not a real poll. It employs a panel, and asks them, on a scale from 0-100, what their probability is for supporting a candidate. My issue, as a matter of personal opinion: I have a hard time lumping this poll in with other polls because this is not a genuine trial heat. We don't cast seven-tenths of a vote for Romney and three-tenths of a vote for Obama. That's not how it works. You get off the fence, and you choose. If their methodology is borne out in the results, then RAND will merit inclusion in future cycles, for sure.
- Our daily recipient of the "totally whack-ass poll of the day" includes a pretty whack-ass writeup of the poll, as well. I am referring, of course, to the GOP-sponsored poll showing Richard Tisei ahead in MA-06, even after a public poll there went double-digits in the direction of the Democratic incumbent, John Tierney. What was so off-putting about both the poll, and the reporting of said poll? I will let my esteemed colleague, David Nir, take it from here:
Given that Obama won the district 57-41 in 2008, and Gore and Kerry did about the same, who could possibly believe these toplines?
Well, two people. One is the pollster, Jeffrey Stinson, who happens to also be the former political director of —wait for it—Richard Tisei for Congress. The other is The Hill reporter Alexandra Jaffe, who didn't seem note any possible conflict of interest with this situation. In fairness, Jaffe did point out that the presidential toplines strain credulity... but failing to identify Stinson's former role on the Tisei campaign is really egregious.
It is one thing to cite a poll as emanating from a partisan source. This level of synergy, however, between pollster and campaign ought to wave red flags. Especially when the topline result is so contrary to previous data in the race.