[EM] IRV vs. Plurality

[Forest Simmons]

On Sun, 14 Sep 2003, Bart Ingles wrote:

>
> I think it would be hard to get the proxy concept into public elections,
> but maybe for party primaries.  As a sort of "approval purist", I would
> lean toward making "No Proxy" the default, replacing the checkbox with
> one labeled, "Allow Proxy".

I prefer this default, too.  The only reason I suggested the other was
that it would make it easier to tabulate the ballots using existing
machinery:

You would just pass the ballots through the precinct counting machines
twice, once with the single winner FPTP setting, and once with the
Multiple Winners at Large setting.

In the first pass all of the ballots with more than one mark (i.e. the non
proxy approval ballots) would be ignored, since they would be "scratch"
ballots according to the rules of FPTP.

In the second pass all of the ballots (including proxy) would be counted
as approval ballots.

To get the true approval ballot results, subtract the first pass results
from the second pass results.

> One way to judge the applicability of the proxy idea is to consider a
> Proxy/FPTP hybrid.  If proxy voting adds anything to an approval
> election, then the benefit when added to a first-past-the-post election
> should be even greater, shouldn't it?

Definitely.  Personally, I think that this hybrid, though not as good as
the Proxy/Approval hybrid, would beat IRV in performance, as well as in
public appeal.

Furthermore, it would work well for Proportional Representation, either as
is or as Proxy/Cumulative, since Cumulative and Plurality are
strategically equivalent, and Cumulative doesn't do that bad for PR when
the polls are good and the voters can coordinate, both conditions being
satisfied in the Election Completion Convention under Candidate Proxy
methods.

Forest