Jon Ralston: The consequences of flawed public polling

By Jon Ralston

Friday, Oct. 8, 2010 | 2:02 a.m.

Harry Reid is dead (again), one in an occasional series:

With the release of three polls this week showing Sharron Angle inching ahead of Reid (although within the margin of error), I am reminded of what happens when I ask audiences these days to raise their hands if they will vote against Harry Reid, no matter what — say, Angle resorting to a Second Amendment remedy on an especially obnoxious mainstream media member.

Without fail about half the hands in the room go up (and I bet some are afraid to do so). And at some events, such as one Thursday in Northern Nevada, nearly two-thirds of the attendees raised their hands.

I bring this up because Reid has such an imbued Anybody But Harry vote that he has had to employ a long-predetermined strategy of scorching the earth with Angle’s own incendiary words so that she would be burned beyond any recognition and unelectable.

That mission either has been accomplished — or simply cannot be. She either has been so marginalized that she actually is a walking political corpse. Or even Angle could not reanimate the ambulatory dead man known as Harry Reid.

Which brings me to the subject of today’s analysis — polls and what they really tell us. Polling is arcane to most laymen — and, alas, to most media folks. But public polling, which is as flawed this cycle as any I have covered, can create and drive a narrative.

If you think I am alone in my observation about public polls, consider what the inestimable pundit Charlie Cook wrote a couple of weeks ago:

“More than in any previous cycle that we’ve witnessed, perceptions of the ebb and flow of races are being driven by state- and district-level polling. This does not mean that there is better polling, just more … Probably 90 percent of the public polling in statewide and district races is mediocre at best, and much of it is very close to worthless.”

Exactly. And why? Because it is done on the cheap and the internals of the poll often expose, as Cook and others (including yours truly) have pointed out, why they are worthless.

Cook made another key point — that polling done for news organizations often is less than reliable:

“I should echo an argument made several weeks ago by my good friend and competitor Stu Rothenberg. He scoffed at those who mistakenly believed that polls conducted independently from the candidates and parties were inherently better or more reliable than campaign polling. My view is that most academic polling, as well as the polling sponsored by local television stations and newspapers, is dime-store junk.”

Mason-Dixon, driving the Las Vegas Review-Journal narrative in this contest, has used an array of samples and ballot tests, without much consistency. But it’s not just local news organizations that are doing a disservice and driving a narrative — let’s look at the polls released this week:

• CNN: 42-40, Angle. The survey did not poll anyone under 35 — some of those young ’uns do vote — and even worse, it gave independent voters (only 15 percent of the electorate here) comparable weight as Democrats and Republicans, who have twice the registration of nonpartisans.

• Fox: 49-46, Angle. These are push-button polls by an offshoot of Rasmussen Reports. This one surveyed more men than women (no one thinks that will be the actual composition) and more Republicans than Democrats (even Republican strategists don’t think the enthusiasm gap will make up for a 5 percentage point deficit in registration).

• Rasmussen: 50-46, Angle. Rasmussen, another auto-dialer company, has long been thought to skew GOP, although it has had some success in calling races. But the ballot test here is off by not having a “none of the above” option and not listing the other candidates.

Some blind Anglophiles or Reidophobics will insist that this is some kind of partisan analysis. It’s not. It’s science. Those sampling errors could invalidate those polls — they at least render them questionable.

As Cook put it: “The far more sophisticated polling is done by top-notch professional polling firms for campaigns, parties and major business and labor organizations. These polls are considerably more expensive and the methodology is more rigorous.

“Most of these surveys are not made public, but insiders can be made aware of them.”

So, as Cook points out, there are two separate conversations, the public and private. And the private one is that Reid is slightly ahead, not Angle. I would bet a lot of money both campaigns think he is slightly ahead.

And yet: Raise your hands if you won’t vote for Harry Reid no matter what happens.