Kilroy Was Here
February 08, 2003
Computer Scientists Fear Voting Via Computer
When the experts raise an alarm, we should listen. And as reported by by the San Jose Mercury News, the experts are weighing in.

David Dill, a professor of computer science at Stanford University, originated the petition. His statement is simple:

If you are interested in this issue, you should take a look at Cal Tech-MIT Electronic Voting Project.

However, David's critique reminds me of another computer scientist's critique of the security system in airports. Bruce Schneier, author of the classic book Applied Cryptopgraphy, points out a major flaw in security thinking in an Atlantic Article entitled Homeland Insecurity by Charles C. Mann :

Bruce Scheier gives an illustrative example of this when he encounters security troubles at the local airport:

Similarly any electronic voting scheme must "fail smartly." If after election day, fraud is suspected at a polling place, there must be a way for human beings to first, verify that there was fraud, and, more importantly, verify the true count.

A system that only stores votes in a single, proprietary, computer-readable format would have no way to verify the count. A system that stored votes in a variety of formats, including a human readable one, would not only have checks and balances, it would have an easy to verify and official way of hand-counting votes, should a manual recount be necessary.

My idea of an ideal voting system would be one that looks like the following:

At least this would fail better than say, a system that has no accountability and may allow a candidate with strong ties to the manufactor of voting machines to stuff an electronic ballot box and fradulently win an election.

Comments: Post a Comment

Powered by Blogger