[tlug] Re: Somewhat OT- open source software for US voting machines

["Jim Breen" <jimbreen@example.com>]

Edward Middleton <emiddleton@example.com> wrote:

> Dr. Breen, was talking about the Australia system,

Well, there are many systems: federal, state, territory, municipal,...

> which as it happens
> is were a trial was held for an e-voting system because the count of the
> paper ballot were found to be inaccurate enough to effect the election
> outcome.

That was in the ACT (a territory; not a state, although the article called it
"state" at one stage.) The ACT (sort-of the Australian equivalent of the
District of Columbia, but with a better climate and no violence) uses
a multi-member proportional system, which means quite long ballot papers
on which ranked preferences must be written, and later counted. As I recall
it was in the election of the last two candidates that the near dead-heat
occurred.

> "Two candidates were separated by only three or four votes, said Phillip
> Green, electoral commissioner for the territory. After recounting,
> officials discovered that out of 80,000 ballots, they had made about 100
> mistakes. They decided to investigate other voting methods."

And having done a proof-of-concept with nice open-source software, I
think they've
bone back to paper ballots. The trial only involved about 8% of the
ACT voters, and ISTR there were some problems not mentioned in the quoted
article.

In our usual single-member electorates if the final result is very close,
the paper ballots are recounted several times and every disputed paper
argued over. Usually it ends up in court, and the judge can order a new
election for the district if no clear result is possible.

Cheers

Jim
-- 
Jim Breen
Honorary Senior Research Fellow
Clayton School of Information Technology,
Monash University, VIC 3800, Australia
http://www.csse.monash.edu.au/~jwb/