Experts divided over accuracy of voting machines

Thousands of people ? between 30,000 and 300,000 ? might have trouble casting their votes on the touch-screen voting machines Tuesday.

“The whole technology is troubled,” said Linda Schade, spokeswoman for TrueVoteMD, a nonpartisan voter advocacy group in Maryland.

“They say it?s the most accurate,” but there are still allegations of candidates being left off of the ballot or machines not counting votes, she said.

The November 2004 general election in Maryland had a more than 10 percent error rate, said Lillie Coney, associate director of the Electronic Privacy Information Center.

But that?s a “very misleading number,” said Jessica Goon, spokeswoman for Diebold Inc., which provides the touch-screen machines for Maryland.

The error rate iscloser to 1 percent, she said.

Between 2000 and 2004 when Maryland switched from paper ballots to electronic machines, there was a dip in the so-called residual error rate, or the human error from voting for more than one candidate or leaving part of the ballot blank, Goon said.

The error rate dropped to 0.3 percent, from 0.5 percent, which shows people are more comfortable with the machines, she said.

During the primary in Maryland, Diebold tested machines at random and found 100 percent accuracy, Goon said.

The error rate with electronic voting machines “is pretty high,” said California-based security expert Bruce Schneier.

“It?s a myth that it?s 100 percent accurate,” he said, adding that because it?s such a political issue, it?s hard to get definitive numbers on the accuracy.

No technology is completely accurate, he said, but “those machines are so bad that I recommend that anybody refuse to use them.”

One way to increase accuracy is to create a voter verifiable paper trail, or have the machines print a paper ballot that verifies the voters? choices.

[email protected]

Related Content