Friday, November 12, 2004

The Exit Poll Issue

Steven Freeman of the University of Pennsylvania has posted a paper arguing that the exit poll results that favored Kerry should be taken seriously. This is not an accusation of fraud, merely a call for someone to provide a rigorous explanation of the discrepancy between the exit polls and the precinct returns.

While the internets have been all over this issue, the big media outlets have dismissed it as conspiracy mongering. The problem is that no one has given us a reason to dismiss this issue. We are just being assured, over and over, that the exit polls must have been wrong, because they were merely exit polls.

Freeman’s paper begins a serious investigation of this issue by showing that the likelihood that the discrepancy between the exit polls and the tallied results is due to the margin of error of the polls is vanishingly small. The odds that random variation could have caused the differences we saw are 250 million to one. If the exit polls were wrong, something must have been systematically biasing them. If so, we are entitled to know what it is.

Prof. Freeman’s paper is not peer reviewed. While it would be nice if it were peer reviewed, waiting for peer review would mean that the legitimacy of this election will only be determined by future historians. We need an explanation now.

Link via echidne

Note added:

This NYT article is the sort of big media coverage I was talking about. It raises the issue of the discrepancy, but then simply changes the subject, pointing out that there is no direct evidence of vote fraud. This is all well and good. But it still doesn't explain where the discrepancy came from.

No comments: