I can't resist calling your attention to some very thoughtful letters to the editor of the New York Times, on the topic of voting technology. For starters, the NYT had very straightforward editorial with trust as the keyword How to Trust Electronic Voting that started with the opinion that "Electronic voting machines that do not produce a paper record of every ballot cast cannot be trusted" and which advocated the adoption of the voting method that is variously called "hybrid" or "machine independent" or "optically scanned paper ballots." More recently, the NYT published a letters article Searching for a Reliable Voting System (note the shift from trust to reliability). All five comments are worthy, but span a fascinating range of views that I have to share briefly, especially since some of them are a great tee for a future posting ;-)
- Mitch Trachtenberg pointed out that paper ballots are helpful, but are much more helpful if the paper ballots are available for independent counting "whether by hand of by computer assisted projects like the Humboldt County Election Transparency Project."
- Daniel Cannon also pointed out that paper ballots by themselves are not sufficient for trust. Cannon wants measures for trust in the counting technology, including technology transparency (full disclosure from vendors) to address concerns of technical methods for stealing elections.
- James Stevens, by contrast, says that since the paper ballots are not actually hand counted, we're still inherently trusting the technology, which can be rigged. Stevens calls for hand count, with no technology. (Interestingly, the hybrid scheme provides a half-way point in, where partial hand counting is used to detect flaws in the electronic counts -- which is something that Cannon calls for as well.)
- Henry Finkelstein points out that paper ballots are easily tampered with. Hand counting created tampering concerns that led to voting machines in the first place -- the mechanical "lever machines" that work fine without computers.
- John Smith agrees with the need for paper audit trails, but isn't sold on optically scanned paper ballots because of the cost of paper ballots, especially the need to print enough to avoid running out in polling places.
What a range of opinions! And the oddest part is that each of these 5 remarks has merit, despite seeming contradicting one or more of the others -- yet another poignant testament to the inherent squirrelly-ness of U.S. election practices and technology.
And with such poignancy, I will have to save for another day the pleasure of doing a Stevens-vs.-Finkelstein face-off posting, or a posting of counterpoint to Smith explaining why dollars per voter per election for paper ballots isn't so terrible compared to existing alternatives, to say nothing of cheaper near term alternatives.