Cleveland study questions accuracy of Diebold voting machines

The Columbus Dispatch
Wednesday, August 16, 2006 1:11 AM

http://www.columbusdispatch.com/news-story.php?story=205502

A new study raises serious questions about whether Diebold touch-screen voting devices used in more than half of Ohio's counties produce accurate results, but Diebold insists the machines can be trusted and that the study is flawed.

A three-month review of Diebold electronic voting machines used in Cuyahoga County during the May primary concluded that the votes recorded electronically and on paper receipts did not always match.

“(T)he election system, in its entirety, exhibits shortcomings with extremely serious consequences, especially in the event of a close election,” concludes the study, released yesterday by Election Science Institute of San Francisco, Calif., or ESI.

“These shortcomings merit urgent attention. Relying on this system in its present state should be viewed as a calculated risk,” the study said.

Steven Hertzberg, ESI's founder, said last night researchers can't yet explain the discrepancies and that more study should be done among researchers, Diebold and election officials.

“We need to understand much better why this occurred, whether it was human error, machine error or a combination of both,” he said.

But the findings no doubt will fuel suspicion among critics of electronic voting and could become a political issue because of past criticism of Diebold and Ohio Secretary of State J. Kenneth Blackwell, the Republican candidate for governor who oversaw the rollout of new machines to replace punch card ballots and other antiquated voting systems.

The $300,000-plus study, paid for by Cuyahoga County, examined figures in the May 2 gubernatorial primary from 467 voting machines used at 50 polling locations in the county. It compared the four ways votes are recorded: a paper printout of the votes cast in each race, a summary of votes cast on that printout, a computer memory card and a separate electronic storage device in each unit.

Researchers found that the four sources did not always match. For example, a comparison of the electronic results with the paper totals showed discrepancies of more than 25 votes in 36 to 46 percent of the precincts -- some varied by more than 200 votes.

But Diebold spokesman Mark Radke argued that ESI did not account for improperly handled memory cards or include all votes cast in the election, such as from 17 year-olds who can vote for candidates and not issues. He said the conclusions “simply are wrong.”

“The discrepancies they found were not discrepancies,” Radke said, arguing the same system was has been used without problems in other Ohio counties and many other states.

Blackwell spokesman James Lee blamed inadequate poll worker training in Cuyahoga County, noting the machines were exhaustively tested before they were deployed.

“No matter what system is used, if there is not proper training and adherence to procedures, there are going to be problems,” he said.

Still, the study concluded that its review showed the problems uncovered in Cuyahoga County are “systemic and therefore faced by other counties nationally and across Ohio.”

Licking and Fairfield counties are among the 47 counties in Ohio that use the touch screens made by Diebold, based in North Canton, Ohio. ESI also has reviewed the touch-screen system made by Election Systems & Software used in Franklin, Delaware, Pickaway, Union and five other counties and found no major problems.

Daniel P. Tokaji, an expert on election systems and assistant professor of law at Ohio State University, said he hasn't fully reviewed the report but that the researchers are respected and that the findings should be taken seriously.

Tokaji noted the findings of numerous blank, smudged or missing paper receipts and said the discrepancies don't necessarily mean the electronic votes are wrong.

The state passed a law in 2004 mandating use of the paper receipts to verify electronic votes, but Tokaji said that law needs to be revisited, the machines upgraded, or both before the November election.

With suspicions of mischief in Ohio's 2004 presidential election still percolating, Democrats yesterday argued the report raises concerns about how Blackwell implemented the new machines and the state's ability to accurately count or recount votes in a close election.

“This is historic because the report gives us empirical evidence of faulty procedures that affect the integrity of our most Democratic process, which is voting,” said Jennifer Brunner, Democratic nominee for secretary of state. “It should give every voter pause and great concern.”

Ohio GOP Chairman Robert T. Bennett, chairman of the Cuyahoga County Board of Elections, said last night that he considers ESI's findings serious, but he also raised the possibility that politics might be involved.

But Tom Hayes, who will resign Friday as director of the Ohio Lottery to join a consulting firm hired by the commissioners to assist the county elections board with the November vote, rejected Bennett's notion of political mischief.

“I always tend to operate under the assumption that people are operating in good faith,” Hayes said last night.