Up: Home page
Previous: Issue 19
Next: Issue 21
Voting matters - for the technical issues of STV
Issue 20, June 2005
(Complete issue in PDF, 164Kb.)
There are 4 papers in this issue:
- I. D. Hill and C. H. E. Warren: Meek versus Warren.
This article compares two computer-based STV counting algorithms. Although
the Meek version seems to be the only version which is widely used, readers
of Voting matters should surely appreciate the differences and
draw their own conclusions.
- I. D. Hill and Simon Gazeley: Sequential STV - a further modification.
This paper considers a variant of STV in which later preferences are used
to exclude candidates. The modification described here has proved necessary
due to two issues which are described in the paper.
- Earl Kitchener: A new way to break STV ties in a special case.
This short paper considers one special case in which the proposal is surely
non-controversial. This is followed by summary and moderated debate on
breaking ties produced by the editor with assistance from those listed.
- P Kestelman. Apportionment and Proportionality: A Measured View.
The author's abstract reads: Apportionment (allocating seats to multi-member constituencies
equitably) can illuminate proportionality (allocating seats to
parties fairly) and its quantification. Sainte-Laguë (Webster)
is the fairest method of apportionment - and electoral principle.
Several disproportionality measures have been proposed: among
which the Loosemore-Hanby Index straightforwardly measures Party
total over-representation. UK general elections (First-Past-the-Post)
have clearly proved non-PR; and even nominally PR elections
of British MEPs and Regional Assemblies have yielded only semi-PR
(`broad PR'). Allowing for vote transferability, multimember
STV in Ireland has mediated full PR (despite low District Magnitude);
while Alternative Voting in Australia has arguably proved semi-PR.
The New Zealand STV elections
A Parliamentary investigation (Justice and Electoral Committee) is under way into the delays in producing
the results. It has not yet reported.
Steve Todd reported in the last issue that the ballot data should be available.
In fact, the electoral officers were divided on the provision of this data
so that complete data is only available for 15 of the 79 elections.
(There were 81 STV elections, but two were not contested.) A table giving
the availability of the data is available on http://stv.sourceforge.net/.
The British Columbia Referendum for STV
The Referendum produced a majority for STV, but not the 60% to ensure that
the necessary legislation will be passed. It is unclear at this stage
what will happen.
Readers are reminded that views expressed in Voting matters by
contributors do not necessarily reflect those of the McDougall Trust
or its trustees.
Papers with citations
- I. D. Hill and C. H. E. Warren: Meek versus Warren. (p1-5, PDF 44Kb)
- I. D. Hill and Simon Gazeley: Sequential STV - a further modification. (p6-8, PDF 32Kb)
[20, 9-11, PDF].
[23, 3-9, PDF].
- Earl Kitchener: A new way to break STV ties in a special case. (p9-11, PDF 28Kb)
[22, 1-6, PDF].
- P Kestelman. Apportionment and Proportionality: A Measured View. (p12-22, PDF 80Kb)
[27, 27-36, PDF].