Election-machine problems spur call for study
Independent analysis requested after lack of explanation from Sequoia
Wednesday, March 12, 2008
BY DIANE C. WALSH
New Jersey's county clerks remain troubled by the errors uncovered in the February presidential primary election, and yesterday a statewide association representing the clerks called for an independent study of the state's voting machines.
"We want to make sure there's absolutely no question about the integrity of the election process," said Michael Dressler, the Bergen County surrogate who presides over the Constitutional Officers Association of New Jersey.
Union County Clerk Joanne Rajoppi, who was the first to discover the errors, urged her colleagues in the association to press for the analysis. Rajoppi has already contacted Edward Felten, a professor of computer science and public affairs at Princeton University, to test machines in her county.
Felten is a critic of electronic voting machines. Two years ago he demonstrated how a computer virus could alter the result on Diebold voting machines.
New Jersey uses machines produced by Sequoia Voting Systems, based in Colorado. The firm examined the affected machines from the presidential primary and concluded that poll workers pushed the wrong buttons on the control panels, resulting in the errors. The discrepancies involve voter turnout, not the tallies showing the winners of the races.
When Rajoppi double-checked the results from the presidential primary, she found on a handful of machines the number of Democrats and Republicans casting ballots did not match when the cartridge printouts from the machines were compared against the paper-tape backup inside the devices. At least four other counties, Bergen, Gloucester, Middlesex and Mercer, identified the same errors.
Dressler said the county officers "got little in the way of explanation from Sequoia" and an independent analysis is warranted.
"People have to be secure in knowing that the election process works," he said.
The constitutional officers association represents county clerks, sheriffs, surrogates, and registers of deeds and mortgages. The clerks have a critical role in the election process since they must certify the results. They do not control the election machines, however. The machines, polling books and poll workers are managed by the county boards of elections and superintendents of election.
Rajoppi said she was not satisfied with Sequoia's explanation for the errors.
"It did not address why the discrepancy showed up in 2008 and never before," she said.
Officials said the historic Feb. 5 primary was an unusual election since it was a simple ballot only asking voters to decide the Democratic and Republican nominees for president.
But Rajoppi searched her records and found a similar race in September 2006 when Elizabeth voters decided the Democratic nominees for two ward elections. Only 25 voting machines were used in the election and Rajoppi said no discrepancies were found in the results.
The constitutional officers unanimously passed the resolution seeking an independent analysis during an executive board meeting yesterday morning in New Brunswick. The association sent the resolution to the governor's office and the state attorney general's office, which oversees all elections.
David Wald, a state spokesman, said his office has been working with election officials to independently verify Sequoia's conclusions on the cause of the problem and how it could be resolved. Wald declined to comment on the clerks' calls for a new study.
Felten said there would be no charge for his analysis of the voting machines. The professor said he is in contact with Union County officials and expects to have several machines shipped to Princeton within a few days.