If this page does not print out automatically, select Print from the File menu.

Kewney: Will IT ever get our vote?

The recent elections in London provided yet more evidence of problems with electronic voting kit

Guy Kewney, IT Week 21 May 2008

Elections held five years ago saw electronic voting systems being piloted “as part of the government’s efforts to get everyone and everything online by 2005”, wrote Madeline Bennett at the time. Of course, we knew about “hanging chads” then. But did we wonder about dazzled scanners?

It would seem that complacency about the likely future effectiveness of voting equipment is a reflex. “We’ve thought of five problems, so we’ve covered everything” is the mindset. But in reality, as we’re now seeing with scanners, the opportunity to perpetrate fraud doesn’t require technology to be used in a sophisticated way.

Encryption was a problem five years back. There was an option to log onto the voting site with your web browser. Obviously, the link had to be safe – so 128bit crypto was fed in. And most people, it seemed, discovered that they didn’t have a browser capable of supporting that, or hadn’t set it up right to use it, so they couldn’t vote.

This year, in London’s elections, monitors were astonished to discover that the system didn’t need to be rigged to be dubious. At one counting centre, boxes of completed ballot papers showed up overnight. Nobody knows when they arrived, but when the counting people appeared to start work, a majority of those boxes were already open.

Had someone tampered with them? It seems unlikely, but it’s definitely possible. Equally possible, and far easier, would be to shine a diffusing light on ballot scanners in areas where your rival is expected to poll well. The light increases errors, causing a closer result than expected. “Do you know what colour light works for this purpose?” I asked one monitor. “No idea” was the response.

The surprise this year was that someone deemed it to be illegal to perform a manual check of voting scanners. The argument was too daft to repeat, but nonetheless, it was accepted. So if the result was crooked, we’ll never know. Not even in a vaguely statistically significant way, never mind specifically.

I suppose it’s a good sign that we trust our technology – and our officials – so completely. I am right, aren’t I?

www.computeractive.co.uk/2217159
This article was printed from the Computeractive web site
© Incisive Media Ltd. 2008
Incisive Media Limited, Haymarket House, 28-29 Haymarket, London SW1Y 4RX, is a company registered in the United Kingdom with company registration number 04038503
Close this window to return to the website