Vote of no confidence – how to improve electronic voting systems
August 8, 2007 There is no doubt that sophisticated computers have the potential to handle election data better than the card ballot method, yet contrary to what one might expect, the introduction of electronic voting systems across the world has been plagued with damaging allegations of incompetence, corruption and severe exploitation. The University Voting Systems Competition aims to raise the credibility levels of electronic voting systems and encourage people to get involved in the democratic process to such an extent that, not only will they vote for a leader, but they’ll design the very voting system itself. With some gaping design flaws and a lack of public confidence holding electronic voting machines back from making a significant impact in global democracy, it’s an initiative that’s sorely needed.
A certain amount of error is expected in elections, no matter what method is used to collate the votes. This error is acceptable only because, with one delivery system, it will impact the candidates more or less equally. When errors in vote tallying show bias, however, they become unacceptable. The problem with electronic voting systems is that, for a number of reasons, they are prone to the latter type of miscalculation.
The first reason for this is implementation. Due to financial or bureaucratic concerns, many countries are making the transition to electronic voting systems on a state by state basis. This is a cause for concern because if the error levels between the competing systems don’t match up, it could alter the result of the election. However, making an instant transition to a single computer voting system also carries a fatal flaw – the company in charge will have a monopoly, and therefore is in a position to remain unchallenged and unimproved.
The second reason electronic voting is treated with suspicion is because of the inherent weakness of a computer system. A paper ballot system leaves a hard copy record which can be scrutinized in a time consuming, but accurate, audit. But electronic voting systems sell themselves on the fact that they make paper trails unnecessary and redundant. Not only is there no hard evidence of the voting tally, but the computer records themselves can be erased or tampered with. Unlike with paper ballots, the electronic voting record is neither independent from the voting system, nor verifiably true.
Which leads to the third reason electronic voting is mistrusted. It increases the ease with which individuals in positions of power can manipulate the election results. Systems in use in the US and Europe have been questioned in terms of being vulnerable to software alteration that can effectively hack the system. In paper ballots, it’s not possible to simply press some buttons and change what people have written. In electronic voting, this is a frighteningly realistic scenario.
The University of Surrey won the best system in the VoComp tournament with their Prêt-à-Voter model. David Lundin, a PhD student and member of the team comments: “We want to make it possible for all voters to check that their votes were actually counted. This is called end-to-end verifiability and my bet is that this is the way we will safeguard democracy in the future.”
The electronic voting research conducted at Surrey and other institutions around the world aims to create systems that not only are able to count millions of votes quickly but also makes elections truly verifiable. Michael Shamos, a computer science professor responsible for testing the accuracy of voting machines, says of their many problems “If smart people put their minds to solving this, it will be cracked very quickly.” Lucky, then, that VoComp has the attention of the best and brightest.