[EM] Request comments on MMP?

[Neal Finne]

Alex Small wrote:

>I've heard that the Libertarians tend to pull votes from the GOP rather
>than the Dems on a 2:1 basis.  This is supposedly according to scientific
>polls, not informal polls on websites or whatever.  Now, that is supposed
>to be nation-wide, so I have no idea if that is true in Washington state. 
>I'm certainly one of the exceptions (a Libertarian who prefers the Dems to
>teh GOP), but I also live in California so I'm completely irrelevant to
>the situation up there.
>
My guess was 60%/40%, so I made that 50%/50% for my own ease. 
Libertarians probably pulled 2%-3% of the vote statewide, and got 5% 
where they did best. Also, Nader pulled from Gore rather than Bush 
50%/20%, with the remaining 30% saying "would not vote." Still, my 
generosity to the Dems may be warranted, because I suspect many of those 
who voted for Dunn (R), whom I counted as Repub voters, actually prefer 
the Democratic Party to the Republican Party. The data I used are 
available at http://www.vote.wa.gov/2002/results/federal_sum.tpl . It 
appeared at a few times that Libertarians might have won a seat if I had 
been using CPO-STV, but I used SSTV, again for my own ease.

By the way, I'm one of those rare Democratic base voters who thinks 
highly of libertarianism.

>Since you have such a narrow majority for the Dems, I'm curious how it
>would change if you split the Libertarian votes 2:1 in favor of the GOP. 
>Since there are 9 seats, and even with a 1/(n+1) quota the LP didn't have
>enough votes to get a seat, they must have had less than 10%.
>
>Say they had 9.9% (optimistic, I know, but "worst case scenarios" can help
>establish limits on an answer).  You split them 50-50, so before adding LP
>votes the Dems would have had 48.85% and the GOP would have had 41.05%. 
>With 2:1 splitting in favor of the GOP, the Dems would get 3.3% added on,
>and they're above the 50% quota when using 1/(n+1) and 9 seats.  So they
>get 5 seats.  The GOP would be above the 40% quota (having 47.65%) and get
>4 seats.  Same difference.
>
>I don't know how things are done with a 1/n quota.  Both parties would be
>above the 44.44% quota for 4 seats.  Who gets the remaining seat?  The
>quota for 5 seats is 55.55%.  The Dems would be a lot closer to the quota
>than the GOP, so I'm guessing they'd still get 5.  A simple majority is a
>simple majority.  Hmm, the absence of any problem involving the final seat
>seems a powerful argument in favor of a 1/(n+1) quota instead of 1/n.
>
I allocated the first (n-1) seats normally, then kept eliminating the 
candidate with the least votes until one candidate had at least 50% of 
the remaining votes. I gave that candidate the last seat. On reflection, 
I don't know if that's actually the correct method.