Secretary of State Jim Condos is embarking on an exercise to reassure skeptical Vermont voters that the state’s voting equipment counts votes accurately, by auditing voting machines in four randomly selected towns.
On Thursday in Montpelier, Condos will supervise a hand count of votes cast in the state treasurer’s race and the U.S. representative races, in Essex, Barre, Brandon and Newfane, comparing the hand count total to the machine count total.
Condos told VTDigger he hasn’t received specific complaints about these races or towns, describing the procedure as a standard check on the integrity of voting machines used in 108 municipalities, authorized under legislation passed in 2006. The audits can’t change the actual election results, unlike a recount, which has to be requested by a candidate and conducted by a court.
Barre and Essex were randomly selected, while Brandon and Newfane were selected purposefully afterwards to ensure a geographically diverse audit, evenly spread among towns with large and small populations. Only statewide races can be audited, and the two races selected were not chosen for any particular reason, said Condos.
“There are people out there who think the machines can be manipulated,” said Condos, who added that he presently isn’t concerned about serious vote manipulations or errors. “We’re doing this random audit as a continuing effort to ensure the integrity of the election.”
“Because this is done post-canvas, we can’t change the election results,” he explained. “So if we were to find a real problem, then we’d have to look at it the next time to fix it.”
Condos couldn’t put a hard figure to the vote difference between a hand recount and the original machine count which would trigger further action, saying the threshold depended on individual races and towns.
If there’s evidence of fraud, the Secretary of State is required to pass the matter to the attorney general. Condos doesn’t know of any past cases where fraud or error have been attributed to vote counting machines, which were first introduced in some Vermont towns in 1992.
In fact, he said, recounts in 2006 and 2010 for the auditor’s race and the governor’s race respectively indicated that voting machines counts were more reliable than hand counts.
The Secretary of State also conducted random election audits in 2006, 2008, and 2010. In the 2008 audit of the governor’s and lieutenant governor’s race, the largest vote change between a hand count and an electronic count came to 11 votes, in Stowe.
Towns must first pass a vote allowing the use of vote counting machines, before they’re adopted by local elections officials. Condos estimated that about 175 voting machines are in use throughout the state.
Permalink |
“There are people out there who think the machines can be manipulated,” said Condos, who added that he presently isn’t concerned about serious vote manipulations or errors.
He is clearly not concerned with vote manipulation or issues with the voting machines, as he was quick to dismiss my and my colleagues’ suggestions for improvements to the current administration of elections with these devices. He told the Times Argus that “he thought most of the recommendations were already being implemented by the state.” It’s not clear at all that the recommendations have been implemented.
Setting aside for a moment the issues of a pattern of security vulnerabilities in this same model and that citizens don’t have any way to inspect that the software running on these devices actually does what it is supposed to, we made these recommendations:
1. Require that the Elections Division be informed of problems with the optical scanner memory cards and/or
tabulator as such problems occur and prior to LHS [Associates, a corporation in Massachusetts] being
contacted.
2. Require that city/town election officials fill out Form A for problems with both memory cards and tabulators
and correct Election Day Procedures at page 5-3 to apply to both memory cards as well as tabulators.
3. Require that city/town election officials begin to fill out Form A for documenting the nature of problems with
the memory cards and/or tabulators immediately upon becoming aware of such a problem (rather than later
when memory may not be precise).
4. Require that LHS staff complete Form B contemporaneously with their work.
5. Require that both city/town elections officials and LHS staff transmit Forms A and B to the Secretary of State
elections staff immediately either by FAX or electronically.
6. Require that elections division staff have a procedure for suspending the use of optical scanners in a
municipality and/or across the state, when problems with memory cards and/or tabulators reaches a threshold
which warrants suspension.
7. Require that you, as Secretary of State, be contemporaneously informed when problems with memory card
and/or tabulators reaches a certain level.
8. Require that the elections divisions submit a report to you by the Friday after an election regarding the
functioning of memory cards and optical scanners and set out in this report any problems experienced and the
correction undertaken.
Jeremy A. Hansen, PhD, CISSP
Assistant Professor of Computer Science, Norwich University