* Professors dissect e-voting
Longtime readers may recall that I have inveighed against paperless electronic-voting machines for some time. In recent weeks there’s been more bad news for proponents of such systems – and for the future of democracy.
The basic problem is that e-voting machines that provide no paper audit trail are a disaster waiting to happen. Any error in such systems, including deliberate tampering, that alters the vote counts cannot, even in principle, be prevented from altering election results. For the time being, we need systems that produce a paper record that a voter can compare with her vote; if the paper accords with her choice, she can then deposit the audit record in a secure repository for use in a recount if necessary.
On Sept. 13, professors Ariel J. Feldman, Alex Halderman and Edward W. Felten of the Center for Information Technology Policy at Princeton University released a paper entitled, “Security Analysis of the Diebold AccuVote-TS Voting Machine.”
Despite the efforts of Diebold to prevent independent study of its e-voting products, the researchers were able to buy an AccuVote-TS machine and subject it to a thorough security analysis.
Felten’s team found that malware inserted through a removable memory card with as little as one minute of access to the voting machine could cause all kinds of mischief, such as switching vote counts or otherwise damaging the election results.
Diebold’s response, attributed to Dave Byrd, president of Diebold Election Systems, attacked the study on specious grounds and was soundly refuted by Felten in a riposte on Sept. 20.
In an interview with Felten on the “Weekend Edition” program on Sept. 23, NPR’s Scott Simon raised many of Diebold’s criticisms, and Felten responded convincingly to each of them.
I think that we in the United States should be following the example of the Australian government. There, e-voting machine code is open for inspection by all and has continued to be improved over the years since its introduction and implementation in voting machines. See for example a 2003 report by Kim Zetter in Wired entitled, “Aussies Do It Right: E-Voting.”
Programmers and security experts interesting in supporting an open-source model for e-voting in the United States (and elsewhere) can volunteer at the Open Voting Consortium.
Make your vote count. Don’t let proprietary, secret software undermine what little democracy we still have left.
