[EM] Need IRV examples; voting show

[Bart Ingles]

James Gilmour wrote:
> 
> Bart wrote:
> > For example, instead of precincts, suppose the division is between
> > walk-in and absentee votes, or between election-night and recount
> > results.  Imagine candidate A being declared the winner, with a recount
> > turning up additional votes supporting A, thereby causing A to lose.  I
> > think this would undermine public acceptance of the outcome, especially
> > given the recent hysteria surrounding the U.S. electoral college.  The
> > answer might be to keep all preliminary results secret, but I'd be more
> > comfortable if that weren't necessary.
> 
> I think you describe two quite different scenarios here.
> It may be current practice in the USA to declare walk-in and absentee votes
> separately, but I would suggest there is no possible justification for it.  All
> the votes, however cast, count towards one single result to produced one single
> winner.  How the respective candidate totals were made up is completely irrelevant
> to that result.  Of course, the political parties, professional psephologists and
> interested academics would all like precinct by precinct information, and much
> more, but none of that is of any relevance to the result.  So why should any of it
> be made public?
> 
> I do agree that your recount scenario could cause problems.  But going back to the
> original dataset which produced this "problem", isn't the answer to adopt
> Condorcet's rules?  That option is not available to us in the UK because we must
> be able to count public elections manually (except in approved pilots of "new"
> technology), and so Condorcet is impractical.

Or the answer could be to adopt approval voting, especially if the only
practical alternative is IRV.  If you acknowledge that voter rankings
will be utilized in such a haphazard way that you would prefer to keep
information about subsets of the vote secret, wouldn't it be better to
avoid collecting information you can't use reliably?  In return,
approval ballots contain information not present in ranked ballots,
namely an indication of the voters' strength of preference.

In computer models conducted by Merrill and others, approval voting
produced results more in line with Condorcet's method than did IRV,
especally when there are many candidates.  This is even more true when
the IRV variant is a restricted one, such as the "supplemental vote"
method used in London (where the voter is only allowed a first and
second choice).

Bart

----
For more information about this list (subscribe, unsubscribe, FAQ, etc), 
please see http://www.eskimo.com/~robla/em