[E-voting] Novel approach to reducing electoral fraud

Michael McMahon Michael.McMahon at Sun.COM
Tue Jan 15 12:36:55 GMT 2008

It certainly is an interesting approach, but
being able to select a ballot randomly (to be copied and taken away)
sounds simple enough in principle, yet it could be quite tricky
in practice. The requirements on it, are that the ballot papers need
to be kept secure, but at the same time, each voter has to be able to
see one of them, chosen at random, and somehow a copy gets made of this
ballot. It strikes me that the authors of this system may have avoided
the cryptographic (software) complexity of other e-voting systems, but 
have replaced
that with a degree of "mechanical" complexity, with all the cost and 
implications which that entails. I could be wrong about that, but that's
my immediate impression. You would also have to be careful, that voter's
couldn't choose the vote just posted by their spouse/family 
who has just voted before them. ie. some of the randomness has to be
imposed by the system, and some by the voter.

Also, this system is clearly designed for first past the post type systems.
It doesn't solve the problem, that we all know about, for multi-preference
(eg STV) elections, because this system requires all plain-text votes to be
published on a web site. And that leads to problems with ballot secrecy.

Incidentally, this work was published at EVT-07 last Summer. I submitted
a paper to that conference, which wasn't accepted unfortunately. The main
reason was that my proposal (which was for a DRE based system that would
work with STV) had some similarities to work that one of the co-author's
of this scheme had done (independently) but which wasn't peer-reviewed
or published widely.

- Michael

Casey, Dermot (GE Money) wrote:
> * Picked up from the RISKS-FORUM Digest 25.02 ( http://www.risks.org ) *
> Subject: Novel approach to reducing electoral fraud
> *The New York Times*, 7 Jan 2008
> http://www.nytimes.com/2008/01/07/opinion/07poundstone.html?th&emc=th
> The idea, proposed by Ronald L. Rivest of MIT and Warren D. Smith, is
> that votes are cast on paper and tallied by scanner or by hand.  After
> casting their vote, each voter is given a photocopy of a randomly
> selected ballot **cast by another voter**.  (A serial number, but no
> name, is on each
> ballot.)
> At the end of the day, all votes cast are entered on a web site.  The
> holder of each copy connects to the site and confirms that the ballot
> whose copy they hold is present and correct, or not.  The theory is
> that, even with a low proportion of web confirmations, any electoral
> malpractice will be revealed with a high degree of confidence, and that
> the knowledge that the scheme is in force will, in any case, deter any
> attempt to rig the ballot.
> Comments on the article put forward most of the obvious objections,
> which are answered by the author or by Smith.  There are links to the
> papers in which Rivest and Smith describe their method in detail.
> _______________________________________________
> E-voting mailing list
> E-voting at lists.stdlib.net
> http://lists.stdlib.net/mailman/listinfo/e-voting
> http://evoting.cs.may.ie/

More information about the E-voting mailing list