STV for party candidate lists?

Herman Beun chbeun at
Mon Jul 27 03:21:03 PDT 1998

DEMOREP1 at schreef:

> Mr. Baum ...

Baum?! ;-)

> ... is suggesting the use of a mixture of instant runoff 
> and STV surpluses to elect a variable N persons.

Yes, that's a good summary. 

> The use of IRO is technically defective in a multi-member p.r. 
> election in the same manner that it is defective in a single 
> winner election.   For N seats to be filled each set of N 
> candidates should technically be matched against each
> other candidate (with the other remaining candidates being 
> deemed test losers)

That's an interesting point; is this why you proposed approval voting
at first?

On the other hand, IRO is _already_ a part of STV. After all you use
it to transfer votes to other candidates when not enough candidates
reach the quota by first preference votes alone (as is usually the

> Example--
> N = 20, 100 candidates
> Each set of 20 would be compared against 1 (with the other 
> 79 candidates being deemed test losers) to see if the 1 gets 
> a quota (3000/20) before each of the 20 gets a quota (or the 
> highest number of votes--- noting again that many voters might 
> not make enough choices- i.e. many of the voters in the 79 
> group might only vote for each other's candidates).

Yes, that is very well possible. I don't know to what extend this
happens, but the possibility has been suggested by other party

[ Finnish election system: ]
This sounds very good indeed! Yes, this is more or less how I
imagined implementation of an STV-like system in the Netherlands.

> 1. Yes but the Condorcet math should be done if possible.

Maybe. I'd be in favour of using Condorcet to elect the party leader
(the number 1 of the list). After all, it is important that this
person is acceptable to a large number of people. But I am a bit wary
of using Condorcet for lower places on the list. I do not want only
the candidates from the political middle, which Condorcet elects. I
want the "fractie" to be a broad representation of _all_ the opinions
among the party members. Only, say, 50% of the voters is somewhere in
the political middle. That means that the other 50% are on the
"extremes"; they deserve to be represented as well! 

> 3.  Most (if not all) have party elites choosing the candidates 
> who get on the party's list (which causes many of the problems 
> in parliamentary type minority coalition governments).

Yeah, most Dutch parties, if not all except D66, do it this way.
That's why I asked how other _democratic_ parties do it! ;-)

Herman Beun                                                    Arnhem                        Gelderland
CHBeun at                                       Nederland
**** Representative democracy is a contradiction in 4 year terms ****

More information about the Election-Methods mailing list