In the May 2007 local elections Open Rights Group observers, accredited by the Electoral Commission, took part in the monitoring of pilot electronic voting and electronic counting schemes. We observed serious failings in the process. Since then, further problems have come to light in other countries leading to many electronic voting solutions being banned or withdrawn. In light of this, yesterday's Government response to an Electoral Commission report is of great concern.
While the Government acknowledges some of the Electoral Commission recommendations for extending implementation periods for systems, it has ignored the fundamental failings observed in trials so far. It has ignored the analysis by computer security experts that shows the technology for secure computer-mediated voting does not currently exist, let alone a secure system for remote electronic voting. Remote voting systems also threaten the privacy of voting, allowing third parties to coerce and influence other voters, particularly within their household.
The only bright spot is that there is a reliable method that permits the secure operation of electronic vote-counting machines: this requires the performing of hand counts on statistical samples and comparing results with the machine tally to detect errors or fraud. Far less brightly, our experience last May was that e-counting didn't have these checks, was rather expensive, and even turned out to be considerably slower than a manual count would have been.
But back to yesterday's Government response. It states that:
"All the pilots supported successful elections... all pilots had comprehensive contingency plans to ensure that electors were not disenfranchised and retained their option of a paper ballot."
This is not the considered view of the observers present at these elections who saw signifcant problems which included disenfranchisement, with voters turned away from the polling station when they found themselves unable to vote by telephone or online (see Open Rights Group, May 2007 Election Report [pdf], page 25).
The Electoral Commission's report made a great deal of sense, in that it made clear their desire to see "a robust, publicly available strategy that has been subject to extensive consultation" before any further pilots took place.
The Government's refusal to halt its pilots is therefore of great concern and reflects a disconnect between Government policy, the evidence and current expert thinking in the field. If the Government's goal of "evidence based policymaking" is to be upheld, then a public debate about the role technology has to play in our electoral process is long overdue. A scheduled public consultation on the introduction of e-voting would be a welcome development.
However, the Government's response to the Electoral Commission's report makes it clear that, from their point of view, this question has already been answered: e-voting is the way forward and the imperative now is to "support [the] implementation of a modernisation strategy", ie to make it work. What public consultation they will engage in will be focused around not if but how e-voting should be introduced. They refuse to accept, despite evidence from the UK and from abroad, that e-voting may not be a viable or desirable area of pursuit.
Elections are one of the most complicated areas it is possible to conceive of to which to apply digital technology. Not only must the system be robust and easy to use, it must ensure voters' anonymity and privacy, yet be transparent and auditable, and be completely secure against both external tampering and fraud by employees, consultants and the outsourced workers often used to develop components of the system.
A single software or hardware engineer can bias marginal seats a percentage point or two and there is a low probability of a professionally executed fraud being detected. In comparison, while fraud is possible with traditional voting systems any large scale fraud would require huge manpower and be difficult to conceal. We are told that e-voting will increase participation, yet the pilots tell a story of voter turnout increased marginally, if at all. The risks posed to our democracy by the introduction of e-voting outweigh these unproven benefits considerably.
Every voter expects their vote to count, and to count once. Until there is consensus that that expectation can be met, remote electronic voting should be reserved for the purposes for which it is fit - naming cats on Blue Peter and voting on the X factor.