[EM] re: Brazil

[Olli Salmi]

I was just going to unsubscribe when I noticed this interesting thread.

At 18:36 +0200 16.7.2003, Kevin Venzke wrote:
>The worst I think I've heard of is Brazil's open-list PR.  (You can 
>vote either
>for a party or for a single candidate within a party, which also 
>counts as a vote
>for the party.  The seats are divided up based on each party's share 
>of the vote, and
>the seats are filled in the order of who got the most individual votes.)

We use a similar system in Finland, but you can't vote for a party.

At 02:28 +0200 17.7.2003, Kevin Venzke wrote:
>It's possible I'm confusing the method for the effects of the method 
>in a certain
>(Brazilian) context.  In that context, the method seems to me quite 
>bad because
>the proportionality is among parties which may not stand for any particular
>position.

I think you are. The method works quite well in Finland, although 
sometimes there are calls for closed lists because the parties feel 
forced to nominate well-known candidates that are not the normal 
party apparatchiks, like our present minister of culture, a former 
Miss Finland and lingerie model, who was the top vote getter in her 
district in her second election.  To the surprise of reporters, she 
seems to be good at her job, or at least skilful with publicity. 
There's also a former colleague of Mr. Ventura's, a top vote getter 
in areas of usually low turnout in Helsinki, now in a coma in 
hospital (recovering, according to today's tabloids) and suspected of 
firing an illegal gun in his flat. Such MPs, whether good or bad, are 
a check on the party leadership and a useful safety valve.

If you feel sceptical, you can see here that he system does no 
overall harm for democracy in Finland.
http://www.worldaudit.org/democracy.htm
http://www.transparency.org/pressreleases_archive/2002/2002.08.28.cpi.en.html

There is more than enough cohesion within Finnish parties. This tends 
to be so even in local government. Proportionality within lists has 
been a problem when parties have fractions or two parties present a 
common list. This could be remedied with sublists but that approach 
was ruled out by a former president who was afraid that it would 
result in two permanent blocks.

In electoral systems there's always a tradeoff between refinement and 
manageablility. The Finnish system is at the manageability end. It's 
very quick to count paper ballots with only the number of the 
candidate. The Swiss system where you have as many votes as there are 
seats may give the voter more choice but the count is much more 
complicated. It's even more complicated to count preferential ballots 
and they have to be counted centrally, but James Gilmour has assured 
us that STV is not so slow as I believe.

The system where you can vote for a number of candidates and cast the 
rest of the votes in the order determined by the parties usually 
means that voters do not actually affect the order of the election. I 
feel a strong aversion against closed lists but they are very easy to 
implement and the actual performance is not  bad. I'm mainly thinking 
of Sweden, which effectively had closed lists. With closed lists 
making nominations easy is an important check to the power of 
parties. In Illinois you need 25000 signatures to get on the ballot, 
which is too many. In Finland it used to be 30 per candidate for 
Parliamentary elections, 10 for local elections, in Britain it was 
7+deposit when I last read about it. I can't see why standing for 
election should be restricted.

It is possible to have proportionality on two levels if one wants to 
have small districts and accurate proportionality between parties. In 
Switzerland the Federal Court has decided that districts smaller than 
about 10 seats are unconstitutional, so Zürich is planning that 
proportionality between parties is determined within regions 
consisting of several districts. The districts are guaranteed the 
number of seats that their population entitles them to which means 
that seats are transferred after the initial allocation. Such a 
system is in use in Basel-Land. Bern also had one and professor Henri 
Carnal has shown that it could produce paradoxes and problems. Here's 
the old Bernese law in French.
http://www.sta.be.ch/belex/f/1/141_1.html
The law has just been changed.

When judging country-wide proportionality, I think one could also 
look at the Netherlands, with 1/150 as the legal threshold. I haven't 
heard of any particular bad effects. I don't like legal thresholds 
because they can be used arbitrarily to exclude certain groups. 
Moreover, the jump between no MPs to, say, 5% of MPs may cause nasty 
instability.

If one wants to give a chance to more parties than two, I think some 
form of PR is the only chance. I can't see how any other system could 
do that. If the districts have around 5 seats, as with STV, you would 
probably get a three-party system ( about 2+2+1 in every district). 
This is a step forward if you have two parties, but often important 
minorities are smaller so I think the districts have to be bigger 
than is usual with STV. James Gilmour disagrees with this.

One of our problems is that the district magnitude varies a lot, from 
about 6 to 31. The effective threshold varies accordingly. The 
parties can't agree on redistricting. If the 12 to 18 districts were 
all of the same size our effective threshold would be between 5 and 8 
percent in districts of 11 to 17 seats.

Olli Salmi