That Steaming Manure Pile of Political Polling

July 26, 2012

As I sit down to write, I’ve just returned from shoveling manure at the barn.  Which makes me ideally qualified to comment on the latest NBC/WSJ poll, as it’s a stinking pile of horseshit, too.

Last week’s CBS/NYT poll sent a shudder through the obama-cult chattering class.  Showing a drop in Dear Leader’s approval/disapproval rating from an already anemic 42/46 to a ruinous 36/48, the CBS poll was generally perceived as indicative of the backdraft of obama’s decision to go nasty on Romney.

With the writing seemingly on the wall, obama’s sycophants braced for more bad news.  Would the upcoming NBC/WSJ poll confirm the worst?

_

Do You Believe in Miracles?

The progosphere gave a collective sigh of relief when a veritable miracle occurred yesterday.  The PRAVDA  MSNBC poll showed obama with a slim 49/48 approval advantage, and a comfortable 49/43 lead over Romney in the ubiquitous “if the election were held today…” question.

To get these stellar results, all NBC had to do was survey +11 Democrat/Lean over Republican/lean (a D/R/I of 46/35/19). In comparison, the D/R/I in the 2010 mid-terms was 35/35/29.  In 2008, D/R/I was 39/32/29.  Rasmussen currently estimates it at 34/35/31.

Let that sink in for a moment:  +11 Democrats yielded +6 support for obama.

_

Working Overtime at the Fudge Factory

With so much riding on this poll, NBC took no chances that it would turn out unfavorably for its employer, obama.  Across the board, segments preferential to obama were over-sampled:

Group                             NBC     USA

Bachelor’s                      24%    18%

Post-grad                       18%     10%

Union household          20%     17%

2008 BO/McCain          +9        +6

NBC’s population also comprised registered voters (RV), who consistently show greater preference for obama than “likely voters” (LV), which is a technical term used by pollsters when referring to those people who are likely to vote in the upcoming election.

As a hint to how likely NBC’s sample will find their way to a polling booth on November 6th, 12% didn’t vote in 2008.  Another 4% can’t remember who they voted for.

Yet everyone accepts NBC’s results with a straight face.  Real Clear Politics robotically plugs it into its influential RCP averages, alongside results from firms that survey “likely voters” (see above for definition), like Rasmussen and Quinnipiac, also the only two pollsters to weight party identity.

_

Just An Outlier

Ah, but Rasmussen is a “conservative polling outlet”, we are reminded by the completely neutral The Hill.  Its polls are just outliers that can be safely ignored.  Rasmussen has indeed received criticism in the past for missing the mark, particularly in 2010.  Its methodology for determining party identification has also been questioned.  Is Rasmussen fudging its results, or just applying a more sophisticated approach to gauge the current political waters?

obama lackeys, desperate for good news, struggle to interpret auguries contained in the plethora of conflicting polls. As a rule, they pounce on any poll favoring Dear Leader, while viciously lashing out at any negative results, and not just those from Rasmussen.

  • In May, at the height of its push of the “War Against Women” meme, OFA got all in a tizzy after a CBS poll showed Romney leading among women;
  • Kos refuses to believe that Romney is ahead!
  • This minor blogger’s reaction epitomizes proglodyte, head-in-the-sand approach: “The 36% favorability … just is too small to be believed”, she insists, pointing to contrary numbers from earlier (Dem-skewed) polls;
  • Over at the once regal, now tawdry, The New Republic, Nate Cohn assures the Faithful that the 6-pt. approvals drop in CBS’ poll “just wasn’t as bad as it might seem.”  How so? Because CBS “has consistently found Obama with lower favorability ratings.”  Which, of course, can’t be true of our beloved barry, so “those numbers were fairly meaningless”;
  • Little Ezra, WaPo wag and obama butt-monkey, has a sure-fire approach to sleeping at night:  “ignore individual polls” and trust in the RCP average, which shows obama with “a small but persistent lead of between two and six percentage points….”


In April, Mother Jones columnist, David Corn, attempted to wrap his big, ivy league-educated brain around the perplexities of the wildly fluctuating polls.

“[T]here is a fundamental dynamic to the race to consider: Many voters tend to like Obama but are disappointed [in] the economy. And many voters—perhaps some of the same voters—don’t like Mitt Romney … but they fancy the idea of a business-savvy Mr. Fix-It who can turn around the economy. Consequently, voters in the middle … will be tugged in opposite directions over the next five-and-a-half months, as they sort out conflicting impulses. Consequently, poll results will see-saw.”

Eureka! the genius Corn has solved the riddle of the shifting polls — those dullard voters who are not members of Phi Beta Kappa simply change their tiny minds from one day to the next.  Consequently, it has nothing to do with those wide anomalies in party affiliation from one poll to the next.

_

Comparing Apples to Oranges, and Kumquats and Lingonberries.

Such divination is impossible so long as assorted pollsters proffer results like the latest out of Michigan. Rasmussen has obama +6; Mitchell, Romney +1; and PPP, obama by a whopping +14!  In 2008, obama won Michigan 57/41 over McCain (+16).  PPP would have us believe that support for the Hopey-Changey One has remained virtually intact among Wolverines.

What’s behind the disparity?  Mitchell is a local firm about which I admittedly know very little.  Public Policy Polling (PPP) is an openly pro-Democratic shop.  Founded by a prog activist, and frequently working hand-in-hand with Daily Kos and SEIU, PPP readily admits it skews its samples.  It also relies on IVR, generally considered less accurate than live calls.

PPP cleverly varies its methods of distortion from poll to poll.  Usually, its Democrat proportion hovers in the mid-forties. Other times, it will increase union household representation three-fold.  As a favor to Kos, PPP notoriously found Scott Walker and Tom Barrett dead even just before the Wisconsin recall vote.  PPP always makes sure to issue one honest poll just before an election to substantiate its boast of accuracy.

In its recent Michigan survey, PPP had a D/R/I of 32/28/40 — only +4 Dem, in line with that state’s voter registration. But the high indy is odd.  Delve deeper into the crosstabs, and we find 31% of respondents described themselves as “moderate”.  Yet these moderates split 70/20 obama v. Romney.  Has PPP detected a silent groundswell of obama support not reflected in any other poll in the country, or is something fishy going on?

Many other pollsters release skewed results without bothering to weight the samples.  A May national poll by Reuters with a D/R/I of  47/38/15 (D +9) had obama up by +7.

Again:  +9 Dem yielded +7 obama.

On the whole, big media outlets polls consistently skew democratic, including, for reasons unknown, FOX.

Even without all the intentional distortion, it’s foolish to equate polls of RV with LV.  In the past, RV polls have averaged +2 to +4 support for Dems vs. LV.  RCP blends 24 Michigan polls for its latest average, which has obama +4.2:

  • The 15 surveys of 10,768 LV, yielded +1 obama;
  • The 9 surveys of  7,082 RV yielded +10 obama.

Of the nine RV polls, six were by PPP and two by NBC.

_

Why Serve Fudge?

Wouldn’t it be in the best interest of polling firms to produce the most accurate results possible?  It would if their only motivation was their reputation.  Hence, the results from the established firms all fall within a narrow range, and show a very close race with low approvals for both candidates.  But as we saw, places like PPP are in the bag, while others hire out to the highest bidder; MSNBC is nothing less than obama’s ministry of propaganda.

Falsely inflating obama’s support helps his campaign in a number of ways.  If approvals for Romney can be shown as weak, it bolsters OFA’s attempt to portray its opponent as unlikable.  And a recent Pew poll shows many voters still forming an impression of Romney.

That same Pew poll, however, revealed that 90% of voters feel they already know enough about obama to make up their mind  — and most find him unpalatable.  From these figures, analysts have extrapolated that “the president’s approval rating among undecided voters is languishing in the 20s.”

Which makes preventing despondency from infecting obama’s own base paramount.  Enthusiasm among Democrats is down 22% from 2008, while GOP enthusiasm is up 16%. (Yet another poll, this one by Gallup.)  Until news outlets like RCP get more discriminating, a smattering of outrageous polls will suffice to mask the crumbling of obama’s support.

_

Call It What It Is

We now live in a fact-free society, where everyone feels they have the right to say whatever and believe whatever.  Enough is enough.  Next time you read or hear someone cite one of these ridiculous polls, speak out and call it what it is: ‘a load of horseshit!’


(c) 2012 by True Liberal Nexus.  All Rights Reserved.

Advertisements