[EM] Unicameral single-member + PR in Germany (Re: Competitive Districting Rule)

James Gilmour jgilmour at globalnet.co.uk
Tue Jul 18 06:21:43 PDT 2006


> Claes Wallin Sent: Monday, July 17, 2006 11:25 PM
> > James Gilmour wrote:
> > If you want a bicameral legislature, why would you want one chamber 
> > elected so that it is unrepresentative of those who voted for its 
> > members?  You can have both districts and PR for the same chamber. Of 
> > course, you cannot have single-member districts and PR, but STV-PR 
> > offers a good compromise of effective local representation (in 
> > modestly sized multi-member districts) and overall PR.
> 
> I may misunderstand your statement, but the German system for 
> the national parliament is exactly "single-member districts and PR".

Yes, I think you have misunderstood my original statement in its original context.  My statement related to voting systems based on
"single-member districts" and it is correct that if you have only single-member districts you cannot have PR (except by chance).

The voting system used in German Federal elections and Lander elections was designed as a PR system, and as such uses a mixture of
single-member districts and nation-wide or state-wide multi-member districts.  I know it well because we use a regionalised version
of it here in Scotland to elect the Scottish Parliament.  And some of us are campaigning to dump it and use STV-PR instead.  For
details, see
http://www.fairsharevoting.org/Fairshare%20Submission%20Arbuthnott%20Commission%2022%20Mar%2005.pdf
(NB This section of our website needs some narrative added to the list of documents.)


> This system could be subverted if two similar parties decided 
> to instruct their voter to vote for party A in the district and party
> B nation-wide. If A+B have 50% and C has 50%, this would yield
> (A,B,C) = (20%,40%,40%), since A would receive extra seats 
> worth 25% of the base seats. (25%,50%,50%), so to speak.

Yes, this is one of the most serious defects of AMS (as we call it  -  better known as MMP in EM circles).  Its exploitation has
been proposed here in Scotland to give an alliance of two parties (really the same party wearing two slightly different hats) an
absolute majority of seats (maybe 65%) for a minority of the votes (around 35%).


> Makes you wonder why the CDU voters didn't all vote for FDP 
> in the local elections.

Yes, I am surprised this has not been exploited more often in Germany, both in Federal and Land elections.  However, I recall that
there was an example of it being used effectively in one election in Baden-Wurttemberg (I think), many years ago, but I don't have
the details to hand.

James Gilmour




More information about the Election-Methods mailing list