[EM] Gerrymandering

[Joe Weinstein]

Short of forcing everyone into a single district, with resulting guaranteed 
huge campaign costs for small parties or obscure candidacies, it's NOT 
necessarily easier to maximize overall geographic fitness or 'utility' of an 
apportionment scheme by using PR.

By the way, usual PR presumes that voters want to be proportionally 
represented ONLY according to political party, not other criteria, including 
geographic proximity.  That's as mistaken as the present one-rep 
one-locality fiasco.  Also by the way, we would get much better 'PR' using 
PAV applied to individual candidates, not parties.

Given a set of social and political and geographic criteria, including 
district compactness - measured e.g. as Alex suggests or by sum of some 
function of distance of each voter from a minimiziing 'central' point in 
her/his district - it would probably be impossible even with current 
computing power to choose the VERY best apportionment scheme.

However, each voter, or each of a large class of undeniable experts, could 
each be allowed to nominate an apportionment scheme, and then computer 
checking could verify which scheme among those nominated would be best or 
near best.  Or, a legislature or court would be mandated to choose from 
among the top 100, or top 5, or maybe instead the top 1%, of the nominated 
schemes.

----Original Message Follows----
From: "Alex Small" <asmall at physics.ucsb.edu>
Reply-To: election-methods-list at eskimo.com
To: <election-methods-list at eskimo.com>
Subject: Re:[EM] Gerrymandering
Date: Mon, 18 Mar 2002 18:42:56 -0800 (PST)

Probably the simplest measure I can think of, and the most easy to justify
(at least in a simple case) is the average perimeter of each district.

Simple case:  Suppose that we have a square region with uniform population
density.  Draw a straight line that divides it in half.

Here's an obvious gerrymander:

  __________
|        | |
|   _  _|  |
|  /|_|    |
|  \  ____ |
|   | |  | |
|   |_| _| |
|______|___|

(I don't know if I got the area exactly 50-50, but you get the point.)

Those two districts give a much higher average perimeter.  Now, I realize
that people want districts to respect city and county lines, give fair
representation to ethnic minorities, represent distinct economic interests,
and solve every other problem under the sun, so maybe simple rectangles
aren't the best solution.  But trying to solve all those problems with
single-member districts requires King Solomon.  PR is much easier.

(Yes, I realize that I just preached to the choir.)

Alex Small



_________________________________________________________________
Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp.