Voting without a receipt
Are electronic votes less secure?
By Courtney Hunter The Guilfordian
Published: Friday, October 1, 2004

The presidential election is a little over a month away, and the controversy over electronic voting is in full swing. Computer scientists, civil rights groups, and public officials are banding together to protest the use of electronic voting machines, which, according to The Economist, will be used by 50 million people (one third of all voters), this November.
Their chief complaint: electronic voting machines do not produce a receipt for each vote. Bev Harris, the founder of the watchdog group Black Box Voting, posted a paper on her group's Web site, Blackboxvoting.org, that says, "Counting votes is a form of bookkeeping. You must show your work, and be able to prove how you came up with the numbers."
Robert Duncan, a Guilford professor of political science, is also wary of electronic methods with no paper trail. He said he is a fan of paper-and-pencil voting methods, but, if that were not an option, "would be a lot more confident if there was a paper backup so in case there was a question, we've got a hard copy."
Others have confidence in electronic voting. In a recent telephone interview, Guilford County Board of Elections director George Gilbert, said electronic voting is safer.
"Paper ballots are much more prone to tampering," he said. "We can serve a much more diverse community (with electronic voting). Not everybody has 20/20 vision, perfect dexterity, and reads and writes English."
Guilford County has used electronic voting machines since 1988.
Machines produced by Diebold Election Systems in particular have come under fire in recent years.
Aviel Rubin, a professor of computer science at Johns Hopkins University, led a team of researchers in a study of paperless electronic voting methods. They presented their findings July 2003 in a paper titled "Analysis of an Electronic Voting System".
According to an abstract of that paper, available at Avirubin.com/vote/, a poll worker could violate voter privacy by matching votes with the person who cast them, or even modify the votes cast.
"Our analysis shows that this voting system is far below even the most minimal security standards applicable in other contexts," the abstract says. "We identify several problems including unauthorized privilege escalation, incorrect use of cryptography, vulnerabilities to network threats, and poor software development process ... we conclude this voting system is unsuitable for use in a general election."
Diebold Election Systems responded to Rubin's investigation by issuing a rebuttal. According to that document, "Every critical process throughout an election cycle is verified by third parties to ensure the integrity of the electoral system."
Diebold says its equipment is certified by the National Association of State Election Directors, and an independent committee evaluates and chooses election equipment.
According to The Economist, however, standards for testing voting machines are outdated. The Help America Vote Act of 2002 included provisions for a technical committee responsible for testing and certification standards. The committee was appointed this June, but will not announce new standards until next June. The earliest these standards could be implemented is June 2006.
According to Legalnewswatch.com, Harris and computer programmer Jim March filed lawsuit against Diebold Election Systems and Diebold Inc. in November 2003, alleging that software bugs in Diebold's equipment leaves California elections vulnerable to would-be hackers.
Outcomes of that lawsuit and of the governmental testing standards will not be finalized in time for this year's presidential election.
However, according to Reuters, Harris said it's not too late to add a paper trail to the machines in time for
Nov. 2.