Connecticut provides a new example of why election audit and recount procedures need to be legislated.
Connecticut races that are to be recounted are exempt by law from manual audits. Yet the CT Secretary of State apparently plans to do recounts by machine, not manually.
Connecticut: Bysiewicz To Consider Elimination Of Manual Recounts
http://www.votetrustusa.org/index.php?option=com_content&task=view&id=2671&Itemid=113
Recounts in any state would be unnecessary if elections were manually audited using scientific sample sizes that give probability of at most
5% for certifying an incorrect election outcome caused by the most well-hidden vote fraud. The need to recount close races is eliminated entirely with scientific manual election auditing because 100% manual counts automatically occur whenever necessary in close races.
2. The National Election Data Archive Releases a Model Legislative Election Auditing Proposal
Unfortunately, federal election audit legislation is not likely to pass in time to protect the November 2008 elections because Democratic House leaders, Representative Steny Hoyer and Nancy Pelosi, have been persuaded by groups funded in part by voting machine vendors such as The Election Center and the National Association of Counties, not to pass the federal election reform legislation (HR811 and S2295) that would require independent manual audits of machine counts and require that paperless electronic-ballot voting systems in states like MD, GA, DE, NJ, SC, and other states be replaced with less costly, more reliable, less susceptible, voter-verified paper ballot systems. That leaves the upcoming primaries and the 2008 November elections susceptible to undetected vote miscount - again.
Manual audits could ensure that at most 5% of incorrect election outcomes were certified (assuming the most difficult to detect vote fraud), by requiring manually auditing on average, over time, a mere 2% to 3% of precincts in federal (US House and Senate) and in statewide (Governor) races - more auditing for close races and less (1%) for wide-margin races.
For state-level legislative races, prior election data could be analyzed to know on average over time how many vote counts would need to be manually audited to ensure that no more than 5% of incorrect outcomes produced in ways that are difficult to detect would be certified.
The costs for auditing elections is minimal in comparison to the budgets that winners of elections control. New Mexico audits 2% of its election results for only $200,000 per election; and CT uses volunteers to manually audit 10% of its precincts.
This is model election auditing legislation:
Mandatory Vote Count Audit proposal
http://utahcountvotes.org/legislature/ElectionAudits4Utah.pdf
Please work with your own state legislators (call and ask to meet with your own State Senator, Representative or Assembly Person) to sponsor legislation to require routine scientific independent manual election audits.
Although US HR811, S2295, and an election audit bill for NJ, S507, are better than any states' current election auditing procedures and should be supported, they contain some loopholes which would allow vote fraud to be undetected in some specific situations.
This model legislation, written for Utah, closes the loopholes in other election auditing bills -
Mandatory Vote Count Audit proposal
http://utahcountvotes.org/legislature/ElectionAudits4Utah.pdf
The Technical Guidelines Committee of the US Election Assistance Commission, and the National Institute of Standards and Technology recommend "software independent" voting systems where a mistake in the software cannot translate to a mistake in the election results. However, no voting system is software independent without manual auditing, or full hand counts, of voter-verified ballot records. Paperless voting systems cannot be software independent and it is difficult, although not impossible, for paper roll DRE voting systems to be software independent (although no state which currently uses paper roll DRE systems employs procedures that make their DRE voting systems "software independent").
This 16 page model election auditing proposal is written for general use with any type of voting system for any state. Drop references to paper roll canisters for states not using paper roll DREs; as appropriate replace "Lt. Governor" by "Secretary of State" or "State Election Director"; and review its definitions and text for consistency with particular state election statutes.
As citizens, we are supposed to control our government. We are entitled to statistically sufficient audits of the invisibly-counted election results produced by machines, not just some ad hock percentage audits that may certify incorrect election outcomes.
I want to thank the experts who helped write this election auditing legislative proposal such as activist Joycelynn Straight, LWV Member, Idaho, and recent reviewers such as Philip Stark, Ph.D., U.CAL at Berkeley, and Ronald Rivest, Ph.D. M.I.T.; and thank the legislators in Utah who have encouraged me to pursue this.
Mandatory Vote Count Audit - legislative proposal
http://utahcountvotes.org/legislature/ElectionAudits4Utah.pdf
Kathy Dopp
Executive Director, National Election Data Archive
http://electionarchive.org
phone 435-658-4657
History of Confidence Election Auditing Development & Overview of Election Auditing Fundamentals
http://electionarchive.org/ucvAnalysis/US/paper-audits/History-of-Election-Auditing-Development.pdf
Vote Yes on HR811 and S2295
http://electionmathematics.org/VoteYesHR811.pdf
Voters Have Reason to Worry
http://utahcountvotes.org/UT/UtahCountVotes-ThadHall-Response.pdf
"Truth comes as a conqueror only to those who have lost the art of receiving it as a friend." --Rabindranath Tagore
http://freepage.twoday.net/search?q=Kathy+Dopp
