When politicians rush in

First published by Linux User & Developer.

Unless you slept through November, you’ll recall that the Democrats trounced Bush’s Republicans in America’s mid-term elections. You may also remember the problems Americans voters suffered with their evoting equipment: votes were counted twice or not at all, machines broke down, and memory cards got corrupted.

You might not have heard, however, about the UK’s own evoting plans. You probably weren’t expecting any evoting trials at all here, given that in September 2005 ministers “shelved plans to test electronic voting in local elections next year”. But Harriet Harman did warn us, “We are not ruling out piloting e-voting in the future and any future plans will be taken forward at the appropriate time.”

It seems that time is now.

The deadline for evoting pilot applications was 17 November, and systems must be developed, tested and ready for polling day on 3 May. The Government is giving just six months to local authorities and vendors to prepare for the pilot, despite past protests about such timescales by the Electoral Commission.

Given the rush, you’d assume that there must be some pressing need for an evoting solution. And you’d be wrong. Apparently the reason we need evoting is because pencil and paper isn’t ‘modern’ enough: we all use ATMs now, so we should use voting machines too. It’s also claimed that evoting will improve voter turnout, which on the surface seems like a better argument but has been debunked by the Electoral Commission itself, based on the results of a number of other pilots held between 2000 and 2005.

Over the last century and a half, improved oversight of the paper ballot system has minimised election fraud to the point where it’s no longer seen as a significant problem, despite the potential financial rewards of stealing an election. The introduction of electronic voting could reverse this.

Elections have, traditionally, been observed by candidates and their supporters to ensure that procedures are followed and that the election is fair. Polling place evoting machines are opaque to human observers – you can watch the machine to prevent physical tampering, but you cannot observe the software at work, which is where the fraud would occur. Internet voting, of which the Government is quite enamoured, is worse, because you cannot observe the machines, there is no paper trail for recounts, and the ballot’s secrecy cannot be guaranteed.

The scale of fraud could be dramatically increased compared to traditional voting systems. It’s very difficult to tamper with hundreds of thousands of paper ballots, but electronic systems remove these logistical barriers and open the way for widespread undetectable vote rigging.

Unlike financial transactions such as withdrawing money from an ATM, voting must be secret as well as secure. Most financial transactions are secured through audit – your name is recorded along with the transaction and so it can be reversed if fraud does occur. In an election, your name should not be linked to your vote, so it would be very difficult to apply the anti-fraud lessons learnt from finance to electoral systems.

Indeed, none of the identity verification techniques we are used to, such as taking your name, address, date of birth, or credit card number, would be allowable if the system is to meet the basic requirements of a secret ballot.

And then there are the non-malicious, but equally disruptive, flaws demonstrated by existing systems. Common Cause’s preliminary report on the US mid-term elections showed that 21% of problems were with registration, 17% were mechanical, and only 4% were billed as ‘coercion/indimidation’. The wild claims of conspiracy made in the media aren’t supported by the facts, but the US did see a 17% rise in problems with voting machines, compared to 2004.

This may seem to undermine objections over potential fraud, but just because fraud hasn’t yet been brought to light does not mean it isn’t going to happen, or isn’t going on undiscovered.

If candidates and voters cannot be sure that the declared result accurately reflects the voting public’s intentions, then confidence in the system, and those to whom it awards power, is severely damaged. And voter confidence is at the heart of the success of any democracy.

Despite the technical challenges posed by evoting, and the lack of any compelling evidence to show we really need it, the British government is pressing ahead with another round of trials. Rushed and ill-conceived, these pilots will probably show the same results as past trials: turnout won’t significantly improve, and neither will the technology. And, just like the past trials, these results will be ignored by a government determined to ‘modernise’ no matter the cost.