A self-described grassroots organization wants to force a sample manual count of ballots cast by electronic voting machines to ensure that these devices are accurately recording votes.
Arizona Citizens for Election Reform has filed the necessary papers to begin circulating petitions to put the issue on the November ballot. The group has until July 6 to get 122,612 signatures.
But Kevin Tyne, the deputy secretary of state, said the measure is not only unnecessary but actually would be a step backward. He said research shows that machines are more accurate than hand-counted ballots.
Stefan Silverston, vice chairman of the organization, said there is a concern that electronic machines — particularly the touch-screen machines that are becoming more popular — are subject to tampering and other problems that could change the outcome of an election.
One provision of the initiative would require that all machines produce a paper receipt that people can review to ensure the way they voted is being recorded properly. Tyne said that already is a requirement for the touch-screen machines the state is purchasing for use by the visually impaired; most other machines use paper ballots marked by individuals that are fed into optical scanners.
But the proposal also would require that a sample number of electronic votes be compared with actual paper ballots. If the difference in count at any polling place is more than four, then a full hand count is required. Silverston said that would ensure that the machines produce accurate results.
Tyne said that ignores evidence that shows hand counts are more likely to produce errors than the machines. And that, he said, would probably lead to having to count all ballots at all polling places by hand.
He acknowledged there have been problems with machines.
Most recently, that included a recount of ballots in the Republican primary in a Phoenix legislative race made necessary by the closeness of the vote. But the recount — in this case, feeding the same paper ballots back through scanners again — produced nearly 500 additional votes and changed the outcome of the race.
Maricopa County election officials blamed the disparity on the fact that different scanners were used the second time and that they had different sensitivity to some of the pens, pencils and markers that people voting by mail often use. They have since scrapped those scanners.
Douglas Jones, a computer expert who examined the machines — but was not given access to the actual ballots — said he could not determine whether the problem was machine malfunction or possible fraud.