It seems a forgone conclusion that, with everything going digital, voting is sure to follow. Especially after lost paper ballots triggered the recent rerun of Western Australia’s 2013 Senate election, sparking renewed calls to trash pencil-and-paper voting for an online alternative. And why not? Networks can transport data faster than vehicles. Machines can tally numbers faster and, arguably, more accurately than humans. And machines alone can’t be accused of manipulating votes. Estonia and Norway have, with the aid of cryptographic ID checks, launched internet voting without too much controversy. But are machines really any less fallible than error-prone humans?
Electronic voting, used to allow blind people and remote voters to participate in previous Victorian state elections and the 2011 NSW state election, could be opened to all voters in the state’s 2015 ballot through iVote. Recently endorsing a wider implementation of the system, the NSW Electoral Commission claimed it would “reduce systemic errors in current voting processes”, including informal votes, lost ballot papers and transposition and counting errors.
While it was well-received by eligible voters, 43 of around 43,000 votes cast via iVote were recorded by the system as ‘N’ “due to a javascript error”. That represents an error rate of 0.1 per cent, the same rate as the recent WA Senate lost ballot fiasco, where 1370 votes were lost in a state of 1.4 million enrolled voters.
“NSW iVote had about the same failure rate as Western Australia’s but it didn’t get anywhere as much attention as the WA vote because the [NSW] election was a landslide,” Vanessa Teague, a cryptographer at the University of Melbourne, told IT Pro. “This assumption that persists that computers will solve either the accidental error problem or the deliberate manipulation problem is completely untrue.”
Full Article: 10 ways e-voting could save or destroy democracy.