On the Exit Poll Debacle

Powerline Blog indulges a reader's conspiracy theory, and then knocks it down to some extent.

Here's what I have heard and what I think.

1. The polling company itself isn't necessarily dirty. Even assuming that there was some underhanded activity going on here, the polling company itself isn't necessarily the bad guy. It only takes one rogue employee to leak something-- remember Bush's debate-prep tape?

The polling company defends itself by noting that they didn't release the data themselves, and furthermore, that the data was actually correct, in the sense that those were the numbers they had-- the numbers, it just turned out, were utterly unrepresentative of the actual situation.

2. But there is good reason to suspect skullduggery. Dick Morris has of course already suggested that the numbers were "juiced" by someone, somewhere. How? The theory he offered that makes the most sense is that someone in the polling outfit leaked to the Kerry campaign which precincts would be subject to exit interviews, and then the Kerry camp (or someone in it) flooded those polling places with plants who would just walk in to the polls, then walk out, then fill out the card claiming they'd voted for Kerry (which they had not, that not being their actual polling place).

Now, there may be some doubts about Dick Morris' reliability, but there are few doubts about Michael Barrone's, and yesterday on Special Report with Brit Hume he suggested the same theory-- and while he didn't, I think, say he suspected an attempt to defraud the poll, he didn't seem to think it was something that could be dismissed out of hand. Apparently this has been done before-- he had a bit of polling jargon to describe the practice: "slamming" the exit polls.

Morris is very insistent that exit polls are the most accurate of all polls, and that it's very unlikely they could be this far off absent conscious design by someone, somewhere.

3. The exit polls have been shown to be utter rubbish, and yet the Democrats and their liberal media Spirit Squad are still quoting from them. The numbers were simply bad-- they showed a coming landslide for Kerry, which was just not what happened. So if the numbers were off on the head-to-head horserace, why are liberals continuing to cite the erroneous polls for the non-horserace data?

How can they keep claiming that based on the polls, at least, the American people are disatisfied with the war in Iraq or the economy? The polls are wrong, guys; you've admitted it. So you can't use them for any purpose. They were unrepresentative and possibly tampered with for all questions, not just Kerry versus Bush.

So knock it off.

Other polls indicate the American people have doubts about Iraq and the economy. Fine-- cite those. But stop pretending that the exit polls can be relied upon to answer these questions.

4. The liberal media is using this fiasco as an attack on blogs. And to some extent, they've got a point-- a bit of one, at least. It was the blogworld that actually reported the data to the public. I don't know who leaked, but it's likely that Kerry partisans leaked to bloggers in order to depress Republican turnout.

To defend some bloggers -- including myself -- conservative bloggers immediately and forcefully reported that the data was likely erroneous when we learned that the male/female split was 41-59. So, yes, we got it wrong, but then we corrected.

At least some of us did. Which brings me to the real villain of this story.

5. William Saletan, of the amateur leftist webzine Slate, published the faulty data, but then never reported that the data was likely erronoeous. Why?

Drudge posted the faulty data; but then he splashed cold water on the numbers by reporting the hinky male/female split. He also reported how badly off the 2000 exit polls had been.

So did NRO, both in the Corner and at KerrySpot.

So did I, for that matter.

William Saletan never did. At least, not that I saw, and not until all the polls in the country had pretty much closed, and the media was ready to start calling states for Bush.

Not nearly good enough, Saletan.

Curious. After making a big statement about how the public needed to be informed as fully as possible, he then withheld information that he obviously had (it's ludicrous that no one in the RNC emailed him, or that he never checked Drudge). Withheld it from the public he claimed he felt such a responsibility to inform fully.

Why?

He will claim that he offered caveats about exit polls earlier in the day, and said that such data do not necessarily forecast a winner. But a general caveat is generally unheeded. When he had specific information that the data was almost certainly wrong, why did he offer no specific caveats as he gleefully continued reporting a Kerry sweep?

Having put wrong information out there, didn't he feel a particular need to debunk his own false reportage?

I don't know why he never saw fit to enlighten his readers. I could guess that, like his comrades in the MSM, perhaps he was so depressed he couldn't bring himself to write the hateful words Kerry might not win; you can't trust these numbers. Which says a lot about media reliability, right there. They want to believe what they want to believe, and, worse yet, they want you to also believe what they want to believe.

But it might be worse still. I can't help suspecting that Saletan wasn't especially upset by the thought that the numbers he was publishing -- numbers he had to know by 5:00 were probably bad -- were helping to depress Republican turnout. Whether those numbers were right or wrong.

Jack Shafer offers some uninteresting musings on the subject, but never even hints at why Slate's exit-poll go-to guy was so blithe about warning his readers that there was specific reason to doubt the numbers he had been reporting all day long.

Until Saletan explains himself, I will continue to have suspicions.

I should repeat that fairly late at night, Saletan did express his belief that Kerry might lose. I can't be more specific than "fairly late," because that part of Saletan's blogging no longer is part of the article. The updates stop at four o'clock or five or so now. Certainly I recall nothing about wildly unrepresentative sampling.

If I recall correctly, however, he began his post expressing the possibility of a Kerry loss thus:

I hate writing this, but...

I'll say, William.