What Didn't Happen In Ohio

For those of you who best understand the exit poll issues, what is your reaction to the following article?

http://www.tompaine.com/20050505/articles/what_didnt_happen_in_ohio.php

What Didn't Happen In Ohio

Russ Baker

May 05, 2005

Russ Baker —an investigative reporter and essayist—is a longtime TomPaine.com contributor. He is involved in the development of a new not-for-profit organization dedicated to revitalizing investigative journalism in America. To read more about the problems in the 2004 presidential election and proposals for reforming our electoral system, see Best Of TomPaine: Election Irregularities In 2004.

Back in January, I wrote a piece for TomPaine.com questioning widely circulated claims that the election in Ohio had been stolen. I had done some poking around, anticipating that at least some of the frightening anecdotes filling our mail boxes and raging on talk radio would be borne out. They weren’t. In spot checks on a few popular fraud anecdotes, I found credible alternative explanations such as incompetence, structural problems, politicization of decision-making and other failings— but no evidence of deliberate fraud designed to hand the election to Bush.

I looked especially closely at the theory that fraud is the only way to explain the large gap between the early exit polls, which showed Kerry doing very well, and the final result giving Ohio’s key electoral votes to Bush. According to this theory, there was no way the actual tally could vary so greatly from the exit polls. The proponents of this view essentially accuse the legendary exit pollster Warren Mitofsky, and a media consortium, the National Election Pool (NEP), of some kind of complicity— or at least willful denial. I found no evidence whatever of either.

For casting doubt on the conspiracy theory, TomPaine.com and I received virtual barrels of e-mail, most from angry anti-Bush activists who could not believe that their hard work had been for naught. I also heard from Steven Freeman, a University of Pennsylvania professor and author of a widely cited study that served as the primary basis for the pro-theft-theory folks, The Unexplained Exit Poll Discrepancy . His remarks, and my response to them, appeared on TomPaine.com.

Privately, I heard from many Democratic officials, election reform advocates and analysts from inside Ohio and elsewhere, who believed my reporting to be accurate, and who were more than a little perturbed by the frenzy, which they found a counterproductive distraction from the serious ongoing effort to reform election practices. Since the debate refuses to die, this seems a good time to trumpet the arrival of not just one, but two, new technical analyses that cast further doubt upon the conspiracy theories out there. The author of the first is an earnest young fellow in San Diego named Rick Brady.

“Brady's paper is a must-read for those still genuinely weighing the arguments on the exit poll controversy,” writes Mark Blumenthal, a longtime Democratic pollster on whose website blog, MysteryPollster.com (“Demystifying the Science and Art of Political Polling”) Brady sometimes posts.

Brady’s point-by-point refutation of the Stolen Election thesis, in which he exposes fallacies, misuses of data and other technical sloppiness, can be found here. These range from an inapt comparison with German exit polls to reckless application of out-of-date margin-of-error statistics.

Meanwhile, a growing chorus of voices is raising doubts about the methodology and conclusions of a loose-knit coalition of academics called U.S. Count Votes (USCV) which has been at the forefront of the Ohio Fraud movement. As Warren Mitofsky told me privately back in January (he’s now gone public with this) —and demonstrated to me in some detail why—he finds the fraud theory highly implausible. Recently, WashingtonPost.com columnist Terry M. Neal interviewed Mitofsky about the findings of USCV. Mitofsky said :

"The trouble is they make their case very passionately and not very scholarly. I don't get the impression that any of these people have conducted surveys on a large scale."

Although many of the USCV people have degrees in statistics and math, those are general skills that constitute only a part of the toolkit needed to design and deconstruct complex polls. That’s not to say they don’t have some legitimate points, just that they don’t have the chops for such a powerful conclusion.

Like the USCV folks, Rick Brady—author of the new study— is no polling expert. He has been deeply involved with graduate-level statistics primarily while earning a master's degree in public planning, but appears to have approached the exit poll mystery with the best qualifications—an agile and open mind.

The other study comes from Elizabeth Liddle, a U.K.-based former USCV contributor and Ph.D. candidate in psychology/cognitive neuroscience who published her own independent study, which demonstrates fundamental problems with the fraudniks’ conclusions.

She begins by acknowledging her own concerns with the situation in Ohio. “I believe your election was inexcusably riggable and may well have been rigged,” writes Liddle. “It was also inexcusably unauditable. I am convinced that there was real and massive voter suppression in Ohio, and that it was probably deliberate. I think the recount in Ohio was a sham, and the subversion of the recount is in itself suggestive of coverup of fraud. I think Kenneth Blackwell should be jailed. However (and I'll come clean now in case you want to read no further) I don't believe the exit polls in themselves are evidence for fraud. I don't think they are inconsistent with fraud, but I don't think they support it either.”

Specifically, Liddle asserts that the exit polls were not just wrong in so-called battleground states, as the fraudniks assert, but everywhere. “My analysis shows that the swing states were not in fact more wrong than the safe states,” writes Liddle. “This evidence shows that the greatest bias was [actually] in the safest blue states... Moreover, the pattern of polling bias is the same as in the nearest comparable election, 1988, another two-horse race where there was also a large significant over-estimate of the Democratic vote and another losing Democratic candidate (Dukakis).”

Liddle explained to me that, since 1988 at least, voter sampling has consistently over-polled Democrats. I’ve heard a variety of explanations for this, but in general, it’s not hard to imagine that Democrats might be at least marginally more inclined to explain their political decisions to exit pollsters, who, after all, are representatives of the often-reviled “liberal” media.

In fact, it seems that Republican voters are overall slightly less likely to accurately express their preferences to in-person interviewers, even in precincts where they constitute a sizable majority. For fairly complex reasons, a slight undersampling of Bush voters produces a larger gap between exit polls and final results in (A) evenly split precincts than in highly partisan precincts, and in (B) highly Republican precincts than in highly Democratic precincts. Not knowing this, says Liddle, one could look at certain precincts and immediately, if incorrectly, smell something foul.

So, absent the emergence of true polling methodology experts screaming theft, we may reasonably conclude that no evil genius rigged the results. Instead, what we experienced was probably an amalgam of system failings, miscalculations, incompetence, and, in some cases, the variably successful exertions of biased election officials. These are, at worst, symptoms of gaming the system, a deplorable practice hardly limited to this election or, historically, to one party. The anomalies being cited, including by Christopher Hitchens—apparently without any notable independent verification—in a widely cited Vanity Fair piece, may prove to be invalid, or attributable as well to other factors. Perhaps fraud occurred on an isolated basis, but no one has come forward with careful documentation—as opposed to hysterical—unscientific allegation.

Until the public becomes confident in the underlying integrity of the electoral apparatus in this country, none of the urgently needed improvements to that system can take place. That’s why the conspiracy-mongering must cease. Can we instead please turn now to the many substantive proposals already being proffered to make things better—including pending legislation? Let’s keep our eye on the real ball that’s in our court.