February 23, 2008
What to believe …
The numbers from the cartridges that print out vote tallies and the paper-tape backup didn’t match. http://tinyurl.com/2odsco
This is the dilemma faced by election officials in several New Jersey counties including Union, Bergen, Gloucester, Middlesex and Ocean counties during the recent Presidential primary. The voting machines in question are paperless DRE’s from Sequoia. Not machines that have paper for voters to check to see that their vote was recorded correctly but machines where votes are recorded invisibly somewhere in the depths of the machine’s memory.
Now they (and we) find out that even the computer speaks with forked tongue when asked what the election results are.
The discrepancies involved the political-party turnout reporting. Sequoia Advantage machines in several counties showed different figures between the result tape from the machine and the records of a secondary memory cartridge, for the number of Democratic and Republican voters. http://tinyurl.com/269edr
The Sequoia machines print out a summary results tape at the end of each election. In addition the memory cartridges can be removed and loaded into another machine for printing. What happened in these five New Jersey counties is that election officials compared the two printouts they got a different tally of the number of ballots cast in the Democratic and Republican primaries.
The discrepancies have rekindled concerns over the reliability of 10,000 Sequoia Voting Systems machines used in New Jersey. http://tinyurl.com/2odsco
This raises a really basic question about the integrity of the election results and makes it challenging for election officials in these five counties to officially certify the election in a timely fashion.
Just because Sequoia technicians hypothesize that the cause of the discrepancy might be a corrupt computer chip does not mean that that their explanation is correct. It could well be a software programming error — the wrong version of software installed on some machines. Since the software is secret and proprietary there is no way to check out whether this might be the cause of the discrepancy. Ditto the computer chips inside the voting machine — also a proprietary secret.
“It’s a reminder that it is not possible to depend on software alone in elections,” said Verified Voting Foundation president Pamela Smith. http://tinyurl.com/269edr
Furthermore there is no quick fix and the county clerks are caught in the middle with no good solution in sight.
Counties were under deadline this week to certify the election results, despite being unable to reconcile or explain the non-matching results. Penny Venetis, a law professor at Rutgers University who represents citizens suing to have the touch screens scrapped, was quoted in yesterday’s Star-Ledger, “I realize the clerks are caught in the middle here,” she said. “If you can’t certify an election, I feel you shouldn’t certify it. Period. Why is it that the citizens of this state can’t be protected?” http://tinyurl.com/269edr
Why indeed certify under these circumstances? There should be no rush to certify under these circumstances. Let’s wait for a final explanation of the discrepancy, not just a theory from a voting machine company anxious to protect itself, not the voters.
Meanwhile, this situation in New Jersey demonstrates the need for voter verified paper ballots, not over-reliance on paperless computers. Then once paper ballots are in place, we can have routine mandatory post-election audits. Right now there is still discussion in New Jersey about whether to retrofit their Sequoia voting machines with printers that would produce a paper record of each vote cast. This is a half-step in the right direction that has been delayed by the fact that the printers failed during testing. But the better and more cost-effective solution would be to opt for voter-marked paper ballots that could be counted on optical scan machines.
The good news about this situation is that New Jersey has careful election officials who are thorough in the performance of the duties. We should all be grateful for their diligence.
Meanwhile, the Sequoia Advantage voting machines produce these problematic results — and their printers fail to meet reasonable standards. The judgment of the New Jersey clerks is more to be trusted than that of Sequoia. The Union County Clerk Joanne Raioppi, who first noticed the discrepancy says
“Initially, when I called Sequoia they said it was an anomaly. And I said, ‘Excuse me. It’s not. It’s an error.'” http://tinyurl.com/25oc4o
She’s right, of course. It is not the fault of the county clerks that they found an error and called an attention to it — the responsibility lies squarely in the lap of Sequoia, the vendor which supplied the problematic voting machines.
Rajoppi said the discrepancies between the numbers were small and would not have changed the outcome of the Feb. 5 primary, but a few votes could mean the difference in future elections if the problem isn’t found and corrected.
“Every year there’s at least one election that’s decided by one or two votes,” Bergen County Clerk Kathleen Donovan said. “I want them to tell me why this happened, and I want them to fix it.” http://tinyurl.com/25oc4o
Let’s give the Bergen county clerk Donovan the final word:
“It’s not widespread, but it’s a problem. We want every single vote to be correct.” http://tinyurl.com/25oc4o