Why we need random mandatory audits By Owen Mulligan
Many Vermont municipalities currently use machines designed by Diebold Election Systems, Inc. to count their paper ballots. But can these machines, known as optical-scanners, be completely trusted? Here’s how they work: each machine has a memory card which is like an electronic voting box where all the votes from the paper ballots are recorded and tabulated. Before each election, these memory cards must be programmed to reflect all the races with the candidates names, party affiliation, etc. This is done by a private company, LHS Associates, based in Massachusetts. On election day, these memory cards are inserted into the machines and then the machines ‘scan’ the paper ballots. This is certainly convenient for election officials and voters alike but it is important for Vermonters to know that these optical-scanners are by no means perfect.
On June 28, 2006, the Brennan Center for Justice at New York University (NYU) School of Law released a report by its Voting System Security Task Force on the security of electronic voting machines. The Task Force spent over a year analyzing three of the most popular voting machines, which included the Diebold AccuVote optical-scanners used in Vermont elections. The Task Force was composed of internationally renowned government, academic, and private-sector scientists, voting machine experts and security professionals.
The Task Force concluded that the optical-scan machines, even though they use paper ballots, are just as prone to errors and software attacks as the controversial Touch Screen Voting machines. The Task Force also stated in their report that, "Almost everything that a malicious attacker could attempt could also happen by accident."
It was a local election in 2006 when our neighbors over in Grafton, New Hampshire, who use the Diebold optical-scanners as well as LHS Associates for the memory card programming, experienced the Task Force's findings first hand. The state's Attorney General seized two vote tabulators after they malfunctioned during elections held on March 14. The incident was a faulty equation on a warrant article vote with 193 "yeas" and 198 "nays". But the vote total recorded by the machine was 369. Obviously, that did not add up.
It doesn't end there. Incidents with the Diebold optical-scanners have been reported in towns located in Massachusetts, Washington, Michigan, California, North Carolina and Arizona.
This may be why Diebold recently changed its name to Premier Election Solutions, Inc. after they failed to find a buyer for their elections division.
These incidences with Diebold’s optical-scanners are enough to make one wonder why Vermont has not experienced any malfunctions when almost every other state that uses optical-scanners has.
The reality is Vermont may have had incidences that either were not reported or were not discovered because rarely do our election officials do a manual hand count of the paper ballots. While Vermont did conduct a statewide audit and recount in recent elections with no abnormalities reported, it still does not change the fact about the vulnerabilities of these machines and what numerous independent studies have shown.
Because of these proven vulnerabilities, independent studies, including the Brennan Report, highly recommend random mandatory audits of all elections. Currently, there is no law requiring random mandatory audits in Vermont. Audits are at the sole discretion of the Secretary of State. A law requiring mandatory audits would certainly be in the best interest of democracy and would be a small price to pay to make absolutely sure every vote is accurately counted.