On Friday, October 14, 2005, the American Statistical Association sponsored a debate on the meaning of the 2004 exit polls with Warren Mitofsky, Director of the National Election Pool (NEP), and myself. Basically, this was to be a debate on whether the exit polls served as evidence that the 2004 US presidential election was stolen. I had said they were. Mitofsky and everyone else associated with the NEP said no, despite what seemed like almost incontrovertable indications.
From my perspective, the debate went very well. Here is the paper I delivered, "Polling Bias or Corrupted Count? Accepted Improbabilities and Neglected Correlations in the 2004 U.S. Presidential Exit Poll Data." See also the accompanying slide presentation which included a good deal of information, not included in the talk I gave because I had so much to say and so much information to convey in a short time. (Basically the slides were an independent presentation going on at the same time as I was giving a different talk. I gave the slides as a handout, along with other supporting documentation.)
As it turned out, Mitofsky ignored his abstract and even the debate, rather supporting my points by explaining how the poll was done and why it ought to be trusted. Here are my accompanying notes of what transpired at the presentation.
Although the debate was filmed, Warren Mitofsky and the National Election Pool refused to permit publication of the video. Also it was the last such debate ever held. Mr. Mitofsky died suddenly and unexpectedly from an anuerism shortly afterwards.
(and also a note on the unwillingness of US Political Scientists to either challenge or defend the election results)
A similar event was set up by the Swarthmore Political Science Department. The chair of the department , Rick Vallely, was also national chair of the American Political Science Association at the time. He invited every prominent academic election expert -- including all of those who had publicly criticized me or otherwise asserted confidence in the election result. When push came to shove, most declined, and those who orginally agreed to participate withdrew. The last of these, Christopher H. Achen of Princeton University, withdrew a month before the event because, he explained, "something came up."
Steven F. Freeman