Thursday, March 20, 2003

E-voting banter between scientists

There was voluminous and heated discussion on the cryptography mailing list about the dangers of the paper audit trail for e-voting that is being pushed by the e-voting academic experts. The instigator and perpetuator of the discussion was Ed Gerck.

His main criticism was that the paper audit trail does not address the problems of massive external vote tampering by extortion (vote this way and prove you voted this way or I'll kill you) or vote selling (vote republican, prove it to me, and I'll pay you $$). He is afraid that the paper audit trail will be just the thing that can be photographed as proof of your vote to enable these system.

Rebecca Mercuri replied:

"The whole idea of photographing paper ballots is a straw man. It is akin to saying that people
will just run through red lights anyway so we shouldn't place them at intersections."

This seemed to sum up my thoughts on the complaint. He seemed to be arguing for throwing the baby out with the bathwater, saying "[printing paper receipts] creates problems that are even harder to solve than the silent subversion of e-records"

He included criticism later on that a paper audit trail does not really make e-voting systems any better than existing paper-based systems and seemed to argue that it is academically uninteresting. I think that this is exactly the point though: nobody has yet come up with an entirely electronic voting system that solves the fundamental problem that a paper audit trail solves. It may be unsatisfying, but what I think is far more unsatisfying are the voting districts that are ignoring this academic result and swapping out systems with unverifiable ones. People need to understand the limits and risks of electronic systems.

Rebecca's most interesting statement for me was:

"The salient requirement of Democratic elections is that the voters must be assured that their ballots are recorded and tabulated as cast. If the process is such that it can only be understood by a team of
scientists with Ph.D.'s, the average citizen can have no confidence that their voice is being heard."

She ended her posting with a response to the criticism:

"I have never said that the paper balloting solution is a perfect one, but it provides assurances in a human-accessible format that is a considerable improvement over both the black-box systems and the chad-based ones.If you can devise a system that is equally user-friendly and has the same ability for independent auditing, then please do so."

The discussion ended with that.

No comments:

Post a Comment