"Jason Kitcat". Give me a break!
Digital rights activist, the Open Rights Group (ORG), says it will be sending 30 observers to monitor the UK's trial of electronic voting technologies in the May 2007 local elections. The team is meeting with the Electoral Commission this morning, where it says it expects its "observer status" will be accredited. The group is …
"Jason Kitcat". Give me a break!
There is no compelling reason to adopt e-voting, and plenty of reasons not to.
It's tempting to add complexity to fix a problem; but what often happens is that you end up adding layer upon layer of fixes for problems created by previous "fixes". The roof leaks, so you install a floor drain. The floor drain gets blocked and smells, so you install a plug-in air freshener. There's a power cut, so you light a joss stick. The joss stick sets off the smoke detector, so you change it for a kitchen-type smoke detector with a silence button. And all the while, the roof is still leaking.
If there's a system to be gamed, and the stakes are high enough, there'll be someone trying to game it. That's just human nature. If you can't make it impossible to cheat, then you have to make it easy to detect anyone attempting to cheat.
Hand-counting of votes has one big thing in its favour: It's universally comprehensible in a way that electronic voting will never be. That is important, because ultimately you cannot trust something that you do not understand. Hand counting basically has just two failure modes: ballot paper not put in correct pile, and number of papers in pile miscounted. And the way it is performed in practice is especially designed to minimise both these failure modes.
Even if the full schematic diagrams, mechanical blueprints and software Source Code of electronic vote recording machines are published (and I would expect no less; the principles of democracy must surely override any corporation's desire for secrecy), and machines were made available for public scrutiny whenever they were not required for an actual election, there will always only be a minority of people who can make proper use of this information. Everyone else just has to take it on faith that the machines do what they ate supposed to do -- they can't spot attempts to game the system.
If you know who knows who, you can even pretend to publish a list of everyone's name and address and for whom they voted; whilst in actual fact the election results are rigged, right down to the last vote. All you need do is make sure that everyone sees a subtly-altered version of the Big List, in which their own vote and those of their friends, family and neighbours are rendered correctly as they voted, with the rest altered to match the desired result. Not hard if you do it online with unique usernames and passwords (to make sure only those eligible to vote can see the results / identify whose votes to display correctly), add DRM (to prevent printing / annoy Mac and Linux users), and threaten severe punishment to anyone caught misusing the information or suspected of being about to misuse the information. It's a beautiful system, because it *looks* so well proof against interference. And nobody is going to ask a stranger how they voted!
It's much better all around to stick with what we all know. Pencil and paper, disabled voters get to choose a person whom they trust to help them vote, and hand counting. It's not perfect, but it's worked up to now. Why should we trust something even less perfect?