New Election Auditing Protocol Proposed

By Warren Stewart, VoteTrustUSA
August 12, 2006

http://tinyurl.com/nqqow

In a newly released paper, “Random Auditing of E-Voting Systems: How Much is Enough?”, Howard Stanislevic argues that “Auditing protocols proposed and implemented at the federal and state levels that rely on small-percentage random sampling without replacement are unlikely to detect miscounts sufficient to change the outcome of Congressional or smaller local races, even if such races initially appear to be decided by relatively wide margins.”

While accepting that a random sample comprising a small percentage of thousands of voting systems may be adequate to confirm the outcome of all but the closest statewide races, Stansislevic recommends that a much larger percentage of systems must be audited than commonly suggested for Congressional elections. "Each race must be considered a separate auditing process taking into account the vote margin, the number of precincts in which the race appeared on the ballot and the possibility of miscounts concentrated in a relatively small number of large precincts."

Stanislevic is quick to add that he does not discourage legislation requiring routine election audits, but rather, that the limitations of such audits must be acknowledged by those who promote them so as not to engender a false sense of security. “Routine methods of error and fraud detection must be developed and employed to supplement small-percentage audits, particularly when there is no obvious trigger for additional auditing or full recounts.”

The entire paper is attached below in PDF format.