Interesting People mailing list archives
Re: BOUND TO CAUSE COMMENTS djf 50 Percent of Sequoia VotingMachines Flawed
From: David Farber <dave () farber net>
Date: Wed, 16 Jul 2008 13:03:23 -0700
________________________________________ From: jsq () internetperils com [jsq () internetperils com] On Behalf Of John S. Quarterman [jsq () quarterman org] Sent: Wednesday, July 16, 2008 3:57 PM To: Epstein, Jeremy Cc: John S. Quarterman; David Farber; ip; tfairlie () frontiernet net Subject: Re: [IP] Re: BOUND TO CAUSE COMMENTS djf 50 Percent of Sequoia VotingMachines Flawed For IP: Jeremy Epstein brings up some complications. Each of them has solutions. He mostly addresses my Indian example. See below for an Australia example.
(1) Most US elections have many races, not just one as in India (and some other countries). It's not uncommon to have elections with 50 or more races; I saw a ballot in Los Angeles that had more than 90 separate races. We can debate whether this is good public policy, but it's currently the law. (I don't think it would be nearly as simple to fix as the Slate article implies.)
We are indeed debating public policy. States that want to solve this problem can simplify their laws to simplify the problem. Although they wouldn't actually have to; see below.
(2) The Indian solution doesn't address any of the issues for voters with handicaps.
Even if you have to resort to human assistance in such cases, that's better than the current situation.
(3) The Indian solution is designed to allow a voter to pick exactly one candidate/party. It doesn't address issues like "pick up to 3 of the following 7 candidates", or "rank order the candidates", or "mark all the candidates that are acceptable". Yes, additional hardware could be created to allow these sorts of functions, but pretty soon you're reinventing in hardware the complexity of the software.
See (1).
I'm definitely not defending the vendors, just saying that the "simple" solution isn't the right one. Einstein said that everything should be made as simple as possible but no simpler; pointing to the Indian technology approach (or to all hand-counted paper) is an example of emphasizing the first part and ignoring the second.
We managed somehow with hand-counted paper for quite some time. Then we managed with paper plus optical scanners. Keeping a literal paper trail would solve many of the current problems. Simply requiring voting machines to emit a marked paper ballot to be stored would solve many of them.
Similarly, Tom Fairlie gets it partially wrong. Voting systems are amazingly complex, in part because every state has different, and sometimes contradictory requirements. I recall reading (but frankly can't recall the source) that some states require that the software allow IRV [Instant Runoff Voting], while other states explicitly prohibit it - not only do they prohibit the practice, but they prohibit the presence of such software in their voting systems. Pennsylvania still (I think) requires full-face voting machines (i.e., so you can see all races at one time, no paging), which is totally impractical for states with lots of races like California and Kentucky. Anyone who has tried to build a "universal" voting machine (even if the term "universe" really means "American") rapidly finds that it's a lot harder than it looks.
Again, not defending the vendors - but let's be realistic about the complexity.
Once again, if the states and the federal government wanted to solve this problem, they could simplify requirements to do so. If I'm not mistaken, Texas partly solves this problem by using home-grown voting machines adapted to the local voting laws. Besides, those are not the kinds of problems that make U.S. electronic voting machines insecure. The real problems are things like little or no password security, insecure network access, ease of changing or erasing votes after they are cast, and of course proprietary source code not visible to the public or anyone else for review. Not to mention, if the problem is so hard, how did the Australians manage to solve it more than five years ago? http://www.wired.com/techbiz/media/news/2003/11/61045 <blockquote> "specifications set by independent election officials, who posted the code on the Internet for all to see and evaluate. What's more, it was accomplished from concept to product in six months. It went through a trial run in a state election in 2001. ... Called eVACS, or Electronic Voting and Counting System, the system was created by a company called Software Improvements to run on Linux, an open-source operating system available on the Internet." </blockquote>
While Diebold has gotten the worst of the publicity, they're not alone. All of the vendors have similar problems. My guess is that if we looked at some of the less known vendors, we'd find that Diebold is pretty good by comparison. Of course, that's damning with faint praise!
No, without actual numbers, that's praising with hand-waving. Kind of like elections without real voting. What will it take before politicians and people decide the stakes are high enough that we actually have to solve this problem?
--Jeremy
-jsq ------------------------------------------- Archives: https://www.listbox.com/member/archive/247/=now RSS Feed: https://www.listbox.com/member/archive/rss/247/ Powered by Listbox: http://www.listbox.com
Current thread:
- Re: BOUND TO CAUSE COMMENTS djf 50 Percent of Sequoia VotingMachines Flawed David Farber (Jul 16)
- <Possible follow-ups>
- Re: BOUND TO CAUSE COMMENTS djf 50 Percent of Sequoia VotingMachines Flawed David Farber (Jul 17)
- Re: BOUND TO CAUSE COMMENTS djf 50 Percent of Sequoia VotingMachines Flawed David Farber (Jul 17)
- Re: BOUND TO CAUSE COMMENTS djf 50 Percent of Sequoia VotingMachines Flawed David Farber (Jul 17)