Martin Cullen must find some way of reassuring the public about e-voting, writes Simon Nugent.
It is time the Government faced reality. The credibility of the new electronic voting system has been successfully challenged.
First we witnessed a spiral of increasing complex technical argument which is beyond the grasp of most voters and, indeed, most legislators.
Now the debate has moved on to a second phase. It is no longer about technology. It is about the credibility of the electoral process. The Minister for the Environment, Mr Cullen, must come up with a way of reassuring us that the system works. Reassurance can only come through an independent review commission which may recommend additional guarantees.
Before Christmas it was already clear the reassurance game was going horribly wrong. A meeting of the Oireachtas committee assessing the system got so bogged down that Opposition TDs ended up suggesting (in vain) that the system's technical critics should be allowed directly to interrogate the Department of the Environment's experts. If TDs whose careers depend on their grasp of the electoral system could not frame their own questions and make sense of the arguments, should average voters feel the system is secure?
One has some sympathy with the Department officials responsible for procuring and rolling out the chosen system. They selected what is, probably, the most secure and proven technology. (Dutch voters have been using Nedap machines for years). They ran two pilots, without anybody raising any substantive technical critique. They also sought out six technical evaluators and added sundry belts and braces, thanks to their recommendations.
On the other hand, some of the arguments now being spun to defend the system have done more harm than good, and Irish voters do not take kindly to being misled about their voting system.
For example, despite Mr Cullen's claim last week, it is not possible to spoil one's vote anonymously with the new system. It is also not possible to conduct a "full paper recount" as has been claimed by the Government. All a returning officer wishing to conduct a recount can do is print out the ballots the computer claims it has been asked to remember, and count them.
What he cannot do is count what the voters wished the computer to register.
More importantly, the extent to which researchers have had to resort to Freedom of Information (FOI) legislation to drag technical evaluation reports out of the Customs House has raised suspicions. One researcher is said to have had to spend €1,200 on FOI requests. Also, the fact that the "source code" is not publicly available is regrettable, and is like a red rag to many of the most IT-literate critics of the system.
However, the biggest problem with confidence in the system is entirely outside the control of the Minister. Electronic voting systems in the US are taking sustained and justified criticism. For example, the New York Times has now run four editorials in the space of a few weeks arguing that electronic voting without a "voter verifiable \ audit trail" is unsafe. To make matters worse, the largest US company offering this technology, Diebold Inc, has succeeded in inflaming hostility by (a) attempting to gag critics of its code and (b) being closely associated with the Bush re-election campaign. This story seems likely to run and run.
It matters little that the machines we in Ireland have bought for €40 million are vastly superior to those being used in a range of US states. Controversy about e-voting elsewhere will constantly refresh scepticism in Ireland which a government minister and his officials will be unable to allay on their own.
Anyway, the new campaign, Irish Citizens for Trustworthy E-voting (ICTE), is correct in its main arguments:
A paper trail allowing voters to review a print-out of their expressed preferences would add confidence to the system.
Without such a trail, random checks to ensure the new system is working effectively are not possible.
End-to-end (rather than piecemeal) testing of a complex system like this is necessary.
Without access to the source code we can not be 100 per cent confident that the expert reviewers have not missed significant errors.
Yet, how great is the risk of undetected corruption of the elections this summer? To my mind, it is very slight. A software error (in, for example, the count process) is most likely to throw up a result that is sufficiently incredible to trigger a recount (which will be possible using the original electronically recorded votes). A deliberate, subtle manipulation of the code by someone who also has physical access to the voting machines is implausible, though not impossible. Would it be worth the risk of detection in order to subvert either a local authority or European election?
Besides, there are genuine benefits of e-voting quite apart from administrative efficiency. The Department's secretary general, Niall Callan, told the Oireachtas committee last December that officials had undertaken a study of rejected ballots and concluded that only a small fraction appeared to have been deliberately spoiled. If an additional 1 per cent of all voters have their votes counted rather than discarded thanks to the introduction of e-voting, would that not be a democratic gain to set against the slight risk of computer error or fraud?
We should continue the e-voting experiment by rolling out the system nationwide this summer. Simultaneously the Government should establish an independent commission, involving both technologists and political scientists, to review the system, consider what additional safeguards (such as, perhaps, some form of paper trail) might be warranted, and make recommendations which should be implemented before the e-voting system is deployed in another general election.
Simon Nugent has been conducting research on e-democracy at the Policy Institute in Trinity College, Dublin. simonnugent@dublin.ie