[EM] Consistency, Truncation, etc. (was CR ballots, etc.)

[Blake Cretney]

On Mon, 1 Oct 2001 11:59:47 -0700 (PDT)
Forest Simmons <fsimmons at pcc.edu> wrote:  
> > 
> > All methods will have some results that are counter-intuitive, at
> > least to some people.
> 
> What are the counter-intuitive results of Approval?

The fact that it doesn't allow enough ballot expression to determine
if one candidate has a majority of first preferences is generally
considered a bad thing by people who are first introduced to the
method.  Some people can be convinced that this characteristic isn't
undesirable, but that's what I mean by counter-intuitive.  It takes
some convincing.

There are other issues.  A question like "Who is your favourite
candidate?" or "Do you prefer this candidate to this one?" are clearly
answerable.  But "Do you approve of this candidate?" has no obvious
meaning without relating it in some way to alternatives.  That seems
odd.

If you take the view that approval indicates that the candidate is of
above average utility, then introducing a really bad candidate can
change who a voter approves of.  In fact, the best way to reduce the
influence of a minority is to introduce a candidate that is really
appalling to it.  If they are sincere, this will force them to approve
of more of the other candidates.  That seems like an unfortunate
result.

---
Blake Cretney