[EM] Redistricting, now with racial demographics

[James Gilmour]

Brian Olson  > Sent: Sunday, July 19, 2009 6:25 AM
> Oh yes, I'm very much in favor of Proportional Representation 
> methods.  

We can agree on that!

> But for the foreseeable future we will have problems with drawing  
> districts and drawing better districts

Yes, I can see the attraction (necessity!) of that approach in the very short term.  But I sometimes wonder what might be achieved
if all the resources were really put behind some moves for PR for those city and state assemblies that are supposed to be
"representative".  The federal assemblies are, of course, the ultimate targets, but you have start where you will have a reasonable
chance of success.  And I do appreciate the significance of the very large number of single-winner elections for single-offices
there are in the USA.  So that dilutes the campaign, as it has to address both single-office elections and elections to
representative assemblies.  One problem I see is that some of the ardent campaigners for better voting systems for single-office
elections don't even begin to see the issues with unrepresentative assemblies, and think everything can be fixed within
single-member districts - which is, of course, impossible.


> and it's a kinda interesting  
> problem to tinker with too.

You may be interested to know that the Boundary Commission for Scotland experimented with some computer-based algorithms for drawing
district boundaries when fed the geographical locations of all our electors.  They abandoned them all because they found it was
taking longer to tidy up the machine-produced boundaries than it would to do the whole job "manually" by sensible amalgamations of
postal units (very small) on a good GIS.  The minimising algorithms have difficulty recognising and accommodating the physical
features like mountains and uncrossable rivers, which as Kathy pointed out, must be taken properly into account in looking for
sensible boundaries.  Incidentally, our Boundary Commissions are all independent of the politicians, so we don't have the "incumbent
problem" that is endemic in the USA and completely distorts the whole system.


> And I have this notion that it would be good to have a bicameral  
> legislature with one house elected at-large PR and one elected from  
> single member locality districts.

I thought you already had a bicameral legislature, together with a "separation of powers" for the Executive?  The main problem with
electing one house by PR and the other house from single-member districts is that, no matter what voting system you use within the
single-member districts, that house will not be properly representative of the voters, in contrast to the "PR" house.  I don't know
how long the electors would tolerate that, especially if there were real differences of political representation in the two houses
and both houses were of similar power in the overall political system.

I would just elect both houses by STV-PR but form the districts on a different basis, which should be easy if the "upper" house had
fewer members than the "lower" house.  Maybe the difference between the US House of Representatives (435) and the US Senate (100) is
too large, but it gives an indication of what I had in mind.

James
No virus found in this outgoing message.
Checked by AVG - www.avg.com 
Version: 8.5.392 / Virus Database: 270.13.20/2248 - Release Date: 07/19/09 05:57:00