[Election-Methods] A Better Version of IRV? (Forest)

[James Gilmour]

Chris Benham  > Sent: Thursday, July 17, 2008 4:55 PM
> In the multi-winner STV races, normally the candidates of the 
> same party are grouped together on the ballot paper under the 
> name of the party. It isn't that difficult to find them and 
> number them in order of preference.

How the candidates are "normally" listed on the ballot paper depends of the relevant election rules.  In all UK public elections
that use STV-PR, the candidates are listed in alphabetical order of family name, followed by alphabetical order of given name for
those of the same family name, as they are in all UK non-party list elections.  There has been NO grouping by party on STV ballot
papers.

There were calls for party grouping before STV-PR was introduced for local government elections in Scotland in 2007, but these were
rejected because the opposition politicians and smaller parties thought it would favour the largest party (then in government - a
coalition).

There has been renewed discussion about the ordering of candidates on STV ballot papers because the VERY large alphabetical effects
that were so obvious in the Scottish local government elections in 2007.  The only real solution to such "list ordering" effects is
Robson rotation, but that does raise other issues, especially about disability equality.


> The STV algorithm can readily handle truncation.

The STV algorithm should be written to match exactly whatever election rules are specified for the particular election.  It would be
my hope that ALL election rules for STV-PR would allow truncation because it is idiotic (and possibly unconstitutional) to demand
that any voter should express preferences that he or she does not have.

James

No virus found in this outgoing message.
Checked by AVG. 
Version: 7.5.526 / Virus Database: 270.5.0/1557 - Release Date: 17/07/2008 05:36