[EM] Re to Mike O wrt 3 seat LR Hare.

[David L Wetzell]

MO:1. Proportional Representation is obsolete, now that we have
technology to easily implement Proxy Direct Democracy. (I discussed
Proxy DD in a fairly recent post).

dlw: I will look into it if you ask me kindly to do so and provide me
a link to a good summary of it.


MO: 2. Largest Remainder, with the Hare quota, doesn't favor small parties. It's


unbiased with respect to party-size. But it's also not very proportional.

dlw: 3 seat LR Hare is the form of PR I prescribe for "more local"
elections that are typically never competitive and consequently rarely
interesting.  It is almost like 1 seat LR Hare, or First-Past
the-Post.  It has one candidate per party and one voter per voter.
Most of the time, the top three vote-getters will get one seat each.
However, if the top vote-getter were to beat their third place
vote-getter by more than 1/3rd of the total vote then (s)he would get
to appoint a vice-candidate to the 2nd seat for their party and the
2nd place vote-getter would get the third seat.  If the top
vote-getter were to beat the 2nd place vote-getter by more than 2/3rds
of the vote, then (s)he would get to appoint two vice-candidates to
hold the other two seats.  It's that simple.  Unlike most forms of PR,
it doesn't require quotas.

MO:It has lots of random deviation from proportionality.

dlw:*Random?*   If the vote percents were 40-30-20-10 then the top
three would win one seat each.  If they were 50-35-10-5 then the top
vote-getter would get two seats.  But this means that if the top vote
getter were to get 43.3 or less of the vote then the third place
candidate could win a seat with as little as 10% of the vote.  That is
what I mean when it favors small parties.  It is not a random
deviation from proportionality.

And, I don't think proportionality matters.  So long as we continue to
use single-winner elections that favor bigger parties then we should
be willing to use multi-seat elections that favor smaller parties.
It's karma.  The opposite biases will tend to even out over time...

So if this was just using 3-seat LR Hare, like for a state
representative election with 40,000 potential voters and 20,000
typical voters.  Let's say there were two major parties, D and R who
get 50 and 40% of the typical vote or 10,000 and 8,000 die-hard
supporters.   By virtue of these supporters, the major parties are
guaranteed that their candidates will win a seat.  However, since the
third seat is in play.  So let's say there are 2 third parties vying
for the third seat.  And let's say they succeed in grass/net-working
among family/friends to persuading 20% each of the non-typical voters,
or 4,000 voters to vote for them and they get split the remaining 10%
of the typical vote 60-40%.  Then, the Percent totals among the five
parties would become

50/140=35.7% - 40/140=28.6% - 26/140 =18.6% - 24/140%=% - 17.1%.

As a result, the third party that does the best wins a seat and gets
to help decide which of the two major parties is effectively in power.
 But if there was a single-seat election that was going on at the same
time as this election and the two third parties voted strategically in
that "less local" election, they would be the swing voters.  In both
cases, they would get attention to their issues by the two major
parties and influence would be decentralized.

MO: Sainte-Lague is the proportional PR.

dlw:1. LR Hare is mathematically designed to minimize the absolute
value of the difference between the percent of the vote received and
the percent of the seats won by all of the parties in an election.


2. As stated above, if PR is used in conjunction with single-winner
elections that are biased in favor of bigger parties then it does not
need to hug proportionality.  As I understand from "Choosing an
electoral system", the best predictor of proportionality in practice
is the number of contested seats, but greater numbers of contested
seats in PR elections make for fewer competitive seats and less
interesting elections...


MO:PR is unwinnable in the U.S, where electoral reform, in addition to
efforts for Proxy DD, should be about a better
single-winner method.


dlw: why is PR unwinnable in the US?  In 1870, we adopted 3-seat
quasi-PR for IL state representative elections.  It's an easy way to
kill lots of birds at once.

MO:Of course, with Proxy DD, all decisions will be single-winner
decisions, among all kinds of sets of alternatives.


dlw: Why then is Proxy DD what makes PR obsolete?
dlw
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20111118/f2491c66/attachment-0003.htm>