Politics in Washington has never been for the faint-of-heart, but in recent years we Americans who are watching have had to expand our sense of outrage time and again. The President justifies torture and indefinite imprisonment without trial, in defiance of the Constitution. A massive program of internal surveillance is created, and journalists who write about it are accused of treason. Three times the administration has borrowed from our great grandchildren to finance tax breaks for the wealthiest tenth of a percent of Americans.
But they wouldn't tamper with the vote count, would they? They wouldn't falsify ballots and doctor vote totals on a grand scale. Would they?
For some of us, this possibility no longer seems beyond the pale, and we must ask, If they did, how would we know? Election recounts are rare, because they are expensive and because the press has been hard on the "sore losers" who ask to see the evidence first-hand. The one large-scale recount in recent years, in Ohio after the 2004 Presidential election, was conducted in secret, and widely reported to be fabricated by the same suspects who reported the tally the first time round.
Exit polls, designed to serve the news media with early forecasts of election results, have been cited as a check of accuracy in the past. And indeed exit polls have diverged alarmingly from the official results in recent election cycles. The pollsters and the news services that hire them have taken steps to assure that this will not happen in the future, because the data will be kept from the public.
Exit polls were our last independent check, holding the vote tallies accountable to the collective voices of the people who voted. We cannot afford to lose them.
Not a moment too soon comes a project by Steve Freeman, Kenneth Warren, and Stephanie Frank Singer of Philadelphia, to conduct an independent exit poll this November. Dr. Freeman wrote the original Internet article in November, 2004, documenting the wide disparity between exit polls and official counts for the Presidential election. He is a professor at the Center for Organizational Dynamics of the University of Pennsylvania, and author of the recent book, 'Was the 2004 Presidential Election Stolen'. Dr Warren teaches political science at St Louis University, has authored texts on polling practices
('In Defense of Public Opinion Polling'), and has run diverse polling operations for over twenty years, with extraordinary accuracy and success. Dr Singer is a math professor at Haverford and author of two books on mathematical physics, who has been moved by conscience to direct her formidable talent with numbers toward the collection and analysis of voting data.
The exit poll they plan will be a major improvement over those traditionally sponsored by the news media. With no deadlines, the poll will be designed for greater accuracy, with redundant checks at every step. Even more important, Warren and Singer have devised statistical techniques to separately measure the traditional sources of polling bias that have made exit polls suspect in the past, and will be able to detect errors in the vote count without ambiguity.
Complementing this effort will be Singer's data analysis, using state-of-the-art data mining techniques to sift through vote counts at every level and flag anomalies of all sorts. All the data and the analysis for both projects will be posted on a public web site, and alternative, independent analysis will be invited. This policy stands in stark contract to the media pollsters of the past, who have held their data close to their chests, even when great issues of public policy hung in the balance.
The exit poll and data collection projects are budgeted at $200,000. With commitments in hand for $70,000, the trio is already working full-time on the project, while seeking donations for the balance.
"We can't guarantee an accurate count," said Singer, "but for once the public will know exactly what happened, if there are problems, and where the problems are."
Read more about the project at ElectionIntegrity.org.