The Big News Out Of Iowa

The pollsters blew it big time.

But a great deal of the "news" in the coming week will be the poll results from New Hampshire.

As recently as Sunday, all the pollsters said it was too close to call: three or four candidates in the Iowa Democratic caucuses were ranked within the margin of error between, roughly, 18% and 25%, depending on which poll and which candidate.

As I write, with 98% of the caucuses reporting, Kerry got 38% and Edwards got 32%.

No poll of which I am aware ranked anybody higher than 25%. Here's the skinny from Zogby's vaunted three-day tracking poll for the week ending last Saturday:

Zogby had Kerry and Dean in a dead heat: Kerry actually got more than twice as many votes — or twice as much support, however you have to look at it — as Dean (38% of delegates to 18% of delegates). And Zogby had Gephardt and Edwards in a dead heat: Edwards actually got nearly three times as many votes — or three times as much support — as Gephardt (32% to 11%).

And don't miss this sage observation from the same webpage: "There are limits to pre-caucus polls. But can anyone doubt that this race has tightened among four serious challengers?"

If pollsters can be so wrong the day before the event, when are they ever right except by coincidence? And why do we depend on them so much?

This isn't a story you'll read about in the newspapers or see on TV or hear on the radio because it's a story about how wrong the media can be — the media who spend so very, very much time telling us all about polls — and they don't want to have to engage in the kind of self-examination that would be required by an honest appraisal of why they're so wrong the very day beforehand.

Okay. Okay. A caucus isn't like a primary election where an individual goes in privately and casts a vote. Still... let's see how the pollsters will have done the day before the New Hampshire primary.

P.S. See Lead and Gold.

[Follow-up: "Poll Tax".]

Lane Core Jr. CIW P — Tue. 01/20/04 06:59:07 AM
Categorized as Media & Political.