[EM] Re: PR not representing median voter, and a system using best of PR and single seat.

Clinton Mead cryptor at zipworld.com.au
Thu Jul 24 09:39:25 PDT 2003


James Gilmour wrote:

>Clinton sang the praises of electing in single-member districts
>so loudly
>
I refute this statement. I'm fairly sure I made it that I don't like the 
idea of single house proportionally elected or singled seat elected, 
calling them both parts of a "lesser of two evils". For the record, here 
is my personal preference.

1. One house single seat, other house proportionally represented.
2. One house proportionally represented.
3. One house single seat represented.

I only praised single-member districts in combination with a 
proportionally represented house. Otherwise their terrible, worse than 
proportional representation by themself, as they give to much power to 
the leading minority party.

> that I was prompted to dig out the figures for recent
>elections to the Australian Federal House of Representatives
>(Adam Carr's archive).  As he said, they use IRV in these
>single-member districts. In this table the parties are arranged
>in descending order of their first preference votes in the 2001
>election.
>
>	1996	1996	1998	1998	2001	2001
>Party	%FPV	Seats	%FPV	Seats	%FPV	Seats
>ALP	39%	49	40%	67	38%	65
>LP	39%	76	34%	64	37%	69
>NP	8%	18	5%	16	6%	13
>AD	7%	0	5%	0	5%	0
>AG					5%	0
>ON			8%	0	4%	0
>Others	8%	5	7%	1	4%	3
>						
>Total		148		148		150
>  
>
Thanks for digging out the figures, as it supports my argument. I said,
"Note that minor parties gained no seats in the lower house, but enough 
seats to hold the balance of power in the upper house." refering to my 
state parliment. This is echoed in federal parliment, where, as your 
data shows, minor parties hold no seats in the lower house. Also, minor 
parties do hold the balance of power in the upper house.

Your figures also support something I said about independents.

"7 independents won lower house seats, no independents won upper house 
seats." - Refering to the state parliment.

In both 1996 and 2001, independents won seats roughly equal to or better 
than their proportion of voters, however, not in 1998. The general trend 
is better than the number of independents in the upper house, which is 
one out of about 75 from memory.

I believe the number of minor party candidates and particularly 
independents would increase if condorcet voting was implimented, as the 
major parties tend to suggest to voters preference each other behind 
non-extreme minor parties and independents. Hence it would be possible 
for more minor party and particularly independent candidates to beat 
major parties in pairwise elections.

Evidence of this is in the recent wollongong bi-election for federal 
parliment. The seat was controled by labor (opposition) for years. The 
government decided not to run a candidate in the bi-election. The 
greens, a minor party, won that seat, making them the first minor party 
to win a seat in the lower house. Had the government of ran a candidate, 
a likely situation would be that the greens candidate would of been 
eliminated early, and the opposition would of beat the government. The 
government used its knowledge that the greens were strong enough to beat 
the labor party pairwise, to deliver a blow to the opposition (even 
though the greens are further away politically from the government than 
the labor party). With condorcet voting, this sort of thing won't only 
happen when parties play strategy games, but in many seats with strong 
local candidates.

>I would not consider these results acceptable in terms of
>representation or stability.  There is serious distortion of the
>voters' wishes (as measured by first preference votes) and the
>distortion switches in a way that cannot be controlled by the
>voters.
>  
>
I think this result is stable. Since the government has such a strong 
and even unfair majority in the lower house, theres no way for changing 
alliances in the upper house to rapidly change policy. However, due to 
the proportionally represented upper house, the government can not do 
anything that does not satisfy the majority of voters wishes.

In terms of representation, I quite freely admited that the lower house 
may not consist of parties which closely match the votes of the first 
preferences of people. Which I think is good, quite frankly. I don't 
like the idea of parties. I believe representitives first priority 
should be the people, not satisfying the party room. And single seat 
elections seem to be the only way to gain independents. Hopefully, with 
condorcet, and these are just numbers out of thin are, at least 20% of 
the lower house would be independents, at least another 10% minor 
parties. In the upper house, you'll lose most the independents 
unfortuantly, but they'll ideally be 30% minor party seats, in line with 
their votes.

>The problem is not IRV, but single-member districts.
>James
>
I disagree. The greens winning the bi-election verses a major party, 
despite the strong green support for their anti-war stance and the weak 
support for the opposition for having not much of a stance, shows that 
minor parties can beat major parties pairwise. And independents do win 
some IRV single-seat elections, and I believe are robbed of even more 
seats when they beat all other candidates but are not elected due to 
IRV. Single-member districts may not produce proportional 
representation, I never said they would. What they do I outlined in my 
previous post. Copied below.

- Geographical representation (lower house) AND ideological 
representation (upper house).
- Potential for independents to gain seats (lower house).
- Potential for minor parties to gain seats (upper house).
- Protection from extremist forces (due to centralist lower house).
- Protection from power hungry government (due to diverse non-government 
controled upper house)
- A strong stable centralist government (lower house).
- An effective house of review, due to diverse opinion (upper house).

The upper house provides proportional representation. If I had to choose 
one, I'd go for upper house. But I feel a lot of the above points are 
important, from both houses, thats why I'd much rather have both. If you 
think geographicl representation, independents and some protection from 
extremist forces is completely uninportant, well, theres no need for a 
lower house. I personally think they're a little important, not as 
important as things like protection from power hungry government, but 
still an important part, thats why I'd prefer to see both houses in action.

Clinton Mead





More information about the Election-Methods mailing list