David Griscom, PhD - Presentation outline below.

AAAS Annual Meeting
American Association for the Advancement of Science

Science and Technology for S
Sustainable Well-Being

15–19 February 2007 • San Francisco
Hilton San Francisco, Hotel Nikko, and Renaissance Parc 55

Sociology and Politics of Sustainability

Are We a Democracy? Vote Counting in the United States

Organized by: Stephanie Singer, Campaign Scientific LLC, Philadelphia, PA

The past few years have seen major changes in the way U.S. votes are cast and counted. In particular, the federal Help America Vote Act of 2002 spurred states and counties to implement electronic voting and tallying. With hundreds of thousands machines, hundreds of thousands of poll workers, hundreds of millions of voters and high, high stakes, the U.S. electoral system deserves careful quality assurance. Are U.S. votes recorded properly? Are they counted properly? Answering these questions involves several branches of science. Computer science can evaluate the accuracy and security of the existing machines. Statistics can detect irregularities in actual results. Psychology, political science, and other social sciences can evaluate the crucial 'soft' factors, such as the training of poll workers and voters. The goal of the symposium is to present progress and describe open problems in this scientific field.    

FORENSIC STATISTICAL MECHANICS APPLIED TO PUBLIC DOCUMENTS PROVE POLL-WORKER FRAUD  

David L. Griscom, Naval Research Laboratory (retired), Washington, DC; John R. Brakey, Americans United for Democracy, Integrity and Transparency in Elections (AUDIT)

As Cluster Captain on 2 November 2004, John Brakey returned to one of his assigned Tucson polling stations an hour after the polls had closed, surprising poll workers apparently altering the poll books. Brakey began an audit of this precinct (#324) based on copies of all public records: (1) a list of all voters registered in precinct, (2) all Signature Rosters (SRs), (3) the Consecutive Number Register (CNR) with 884 poll-worker-printed voter names, (4) the Official Ballot Report and Certificate of Performance signed by all 7 poll workers, and (5) a list voters who signed envelopes conveying Provisional Ballots (PBs) to the county Recorder. Brakey recovered from the morning-after trash (6) the poll- worker-annotated 'Advice to Voter' slips. Records (2) and (6) indicated which voters were required to vote on PBs (which are only accepted by the Recorder if she ascertains that the voter is registered and had not mailed in an Early Ballot). Record (4) could not be reproduced by from the public data without assuming 39 PBs were illegally fed into the optical-scan ballot box on Election Day. The CNR contained 11 fewer unique names than the number of ballots in the ballot box according to (4), implying 11 felony double votes. The poll workers issued 11 extra ballots as alleged spoil replacements, possibly to cover up (but failing to disprove) these double votes. There were also exactly 11 voters who signed a 'regular' SR but whose names are not listed on the CNR, 11 voters who signed the PB SR but are not on the CNR, 11 voters who signed both the 'regular' and PB SRs, 11 registered voters listed on CNR who failed to sign any roster at all, and 11 phantom voter names appearing on 11 of the signed envelopes of PBs received by the Recorder that do not match any signature on any SR - nor any entry on the CNR. The probability of any one of these irregularities occurring 11 times is <1/11. The odds of all 7 occurring exactly 11 times as independent random accidents (e.g., due to incompetence) are <<one chance in 11 to the 7^th power = 19.5 million. Three voters had their names inscribed a second time on the  CNR exactly 100 places after the first, with one-chance-in-131-million probability. Despite their control of the CNR, the poll workers wouldn't have been able to contrive such statistical rarities without a 'system.' Indeed, (6) revealed a non-standard hand numbering scheme which would have fit the purpose. We conclude that 22 valid Kerry votes could have been discarded (as 11 allegedly spoiled ballots and the 11 PBs rejected by the Recorder) and 61 Bush votes could have been forged (as 39 PBs illegally fed into the ballot box on Election Day, 11 double votes, and 11 alleged spoil replacements) -- a shift of 9.4%. Still, the inferred 'system' would have deposited paper ballots in the ballot box exactly matching the number claimed in (4), and voter choices on these ballots would match the official tally, thus appearing honest in the event of a hand recount -- and thereby covering up  demonstrably possible hacking the 1.94w memory cards in optical-scanner precincts where  the poll workers were honest.   

**********************************

RISKS AND BENEFITS OF DIFFERENT VOTING TECHNOLOGIES  

Barbara Simons, IBM Research (retired), San Jose, CA  

As a result of Florida 2000, some people concluded that paper ballots simply couldn't be counted, even though businesses, banks, racetracks, lottery systems, and other entities in our society count and deal with paper all the time. Instead, paperless computerized voting systems (Direct Recording Electronic or DREs) were touted as the solution to 'the Florida problem'. By 2006 DREs were widely deployed and causing problems in elections such as the very close Sarasota, Florida House of Representatives race, where an abnormally large number of votes were not recorded, and the close Virginia Senate race, where a recount would have been impossible -- because in many areas of the state there was nothing to recount. Election officials were told that DREs in the long run would be cheaper than alternative voting systems. They were told that DREs had been extensively tested and that the certification process guaranteed that the machines were reliable and secure. They were also told that DREs would allow people with disabilities to vote independently. No mention was made of the significant costs of local testing and of secure storage of DREs; no mention was made of the inadequacy of the software testing and certification processes or of the obstacles to creating bug-free software. And no mention was made of the existence of or potential for much less expensive systems for people with disabilities, to say nothing of the large number of shortcomings of the current crop of DREs for people with disabilities. We will examine some of the technical issues relating to DREs and to Internet voting, describe some alternative systems currently available, review some horror stories from the 2006 elections, and consider legislative efforts to repair the damage caused by the purchase of voting machines that have major security, accessibility, and reliability problems. If there is time, we'll also discuss some of major meltdowns that occurred in the 2006 elections with newly deployed statewide databases of registered voters.

**********************************

MASS SCALE ELECTION FRAUD IN RECENT U.S. FEDERAL ELECTIONS  Steven F. Freeman, University of Pennsylvania, Philadelphia, PA  (No abstract)  

________________________________________________________________

N.B. Speaker Steven F. Freeman (author of the book 'Was the 2004 Presidential Election Stolen?)  and discussant Joshua Mitteldorf did not submit abstracts. The Griscom abstract is slightly corrected from the version on the official AAAS abstract CD  from which the other materials were copied.