[EM] Request comments on MMP?

James Gilmour jgilmour at globalnet.co.uk
Mon Jul 28 10:06:03 PDT 2003

> >James, responding to Olli, had written:
> >There is confusion here between guaranteed seats on first preference 
> >votes, guaranteed seats after
> >all transfers, and de facto representation.  The results of the 
> >Northern Ireland Assembly election
> >in 1998 show well what happens in practice.  For a summary see this 
> >List's files archive at:
> >http://groups.yahoo.com/group/election-methods-list/files/STV-PR/
> >(NB You must have one continuous blue line of text from "http" to 
> >"PR/" for this live link to work.)

Olli wrote (> ):
> It required a Yahoo ID number which I don't have, as far as I know.

Yes, I thought you might need a Yahoo ID to get access to the FILES archive, ie you must become a
member of the Yahoo Group where this archive is stored.  (I could not test this as I already have a
Yahoo ID.)  The MESSAGES archive is open and can be read by anyone.  I have been having problem with
my own webpages, otherwise I would have posted a copy there.  I'll let you know when that site is
working again.

> Yes, there may be confusion. I was talking about the threshold of 
> exclusion, the lowest percentage of support that ensures a candidate 
> will win no matter what other voters do.
> >You'll see that the Northern Ireland Women's' Coalition with 2% of 
> >the first preferences won 2% of
> >the seats in the Assembly.  Of course, they used vote concentration 
> >in two constituencies
> >(districts) to win those two seats, but that's what all small 
> >parties to if they want to win.
> I presume you are not suggesting that all the groups that have 2% of 
> the votes get a seat.

No, I am not saying that all groups that can get 2% of the votes should get a seat.  But I was
drawing attention to the actual outcome of a real STV-PR election in evidence against those who
claim the threshold with modestly sized districts denies fair representation overall.  In this
particular case, the NI Women's Coalition contested eight of the 18 districts and won two. Where
they won, they had 9.6% and 4.8% of the first preference votes.  In both cases, after transfers
their share rose to 13% and they secured the last seat.  In the other six districts their share of
the first preferences ranged from 1.8% to 3.4%.


Re MMP (AMS) for the Scottish Parliament elctions:
> Are your constituency candidates allowed to stand on the lists?

Yes.  Some parties do, some parties don't.

> The Scottish system seems to allow more disproportionality than the 
> German system. 

The understatement of the year!  And of course, it was planned that way.

> In Germany they have as many list seats as there 
> constituency seats, so there's less chance of an overhang and to a 
> certain extent it can be corrected by adding seats.

The problem here is not just that our overall ratio is 73:56:  this gives a misleading impression
because of the integer effects when you allocate whole seats to whole candidates.  Because each
electoral region is an independent unit for the MMP (AMS) calculation, the actual ratios are 10:7,
9:7 and 8:7.  This increases the distortion.  And because the total numbers are fixed, there is no
mechanism to correct the overhang.  Of course, preserving that distortion was part of the deal!

> No doubt they had many compelling arguments for it.

Yes - like it will ensure we are the largest party even if the distorted results fail to give us an
overall majority.

> >This just what some of us want to get away from!!  We have so much 
> >"party discipline" that MPs and
> >MSPs just do what they are told, and we have a Prime Minister who 
> >can disregard the UK Parliament and behave like a president.
> He got a vote of confidence, didn't he? It is easier for a PM to 
> control a Parliament if his or her party has an absolute majority in it.

Yes, he did win the vote of confidence.  But that shouldn't be too difficult when your party has an
OVERALL majority of 165 seats for 41% of the votes.

> >  We also see the bad effects of excessive party discipline 
> in many of our local government councils.
> On the other hand parties provide predictability.

When a party with 46% of the votes wins 94% of the seats (17 out of 18), I think we need a little
less predictability and a little more representation!


More information about the Election-Methods mailing list