[EM] Amnon Rubinstein's Proposal for Electoral Reform in Israel

Antonio Oneala watermark0n at yahoo.com
Wed Apr 26 16:46:47 PDT 2006

  James Gilmour <jgilmour at globalnet.co.uk> wrote:
  > Antonio Oneala> Sent: Wednesday, April 26, 2006 1:13  AM
 You said:
 > I've never really been a fan of STV-PR. It's still a system 
 > that's not independent of irrelevant alternatives, so in most 
 > states it squeezes out third parties and moderates, unless 
 > you expand the districts to an enormous size
 I asked:
 On what evidence do you base this assertion?
 : It certainly does exist in IRV, and since STV is simply a version of IRV that elects multiple person I wouldn't see why it wouldn't exist there.  The problem is, moderates are more likely to make the majority of the populace the happiest, but unlikely to gain more first place votes.  The likely result of this is that they will be eliminated and their votes transferred to the more radical ideology that the voters placed as second.  This can be seen in any system that relies on eliminations to do their work; you can't be put back into the rounds because somebody put you as their second choice, after all.
 You replied:
 > > Besides that fact that practically every country that has 
 > > used STV has boiled down to a more or less two-party system?
 So I asked: 
 > Would you care to list the countries and states around the 
 > world that use STV-PR together with the numbers of parties
 > represented in their parliaments before they used STV-PR and 
 > now after many years of using STV-PR?
 Ireland has boiled down to a coalition of parties against one party; basically a two party system.  Australia is a two party system, whith a few minor third parties in the senate where STV is used (although their seats aren't in proportion to their votes).  Malta is a classic two party system.  While STV certainly will allow more third parties in, these third parties are likely to be based around a few factions, since the natural effect of the eliminations is to transfer the votes to the fringes of the political spectrum.  After a while, this will naturally result in two factions, one left, one right.  It also suffers from vote-splitting, just like SNTV, IRV, and plurality.  Besides this, there's the massive problem of non-monotonicity.
 You now reply:
 > It is simply a fact that STV is not a system that is 
 > independent of irrelevant alternatives. If X and Y are 
 > competing, adding A to the list could cause one of them to 
 > lose. 
 In other words, you have NO evidence to justify your damning assertion. We are well aware of the defects in STV-PR,
 including the lack of independence from irrelevant alternatives (in all versions except Meek STV). But that is not
 evidence that where real voters have used STV-PR in real public elections, STV-PR "has boiled down to a more or less
 two-party system". Indeed, there is good evidence to the contrary.
  -  What evidence?  And no, Meek is not independent of irrelevant alternatives.  It's simply the only STV system that doesn't involve an element of randomness (besides Warren, but no one uses that method so i won't go into it). It also requires a computer.  So I say, if your going to need a computer anyway to have non-random proportional representation without relying on group voting, why not just use PAV, which doesn't suffer from these problems, or any problems more than normal approval?  It's not like I'm pulling anything out of thin air here, IIA is a very well known criterion, and you aren't going to find people who say that STV is invulnerable to it.  If A were going to win over B, and you added C (who's not going to win), it could make B win.  Approval is just about the only largely recongnized voting system I'm aware of that doesn't suffer from it and doesn't immediately devolve into a two party system (besides, possibly Condercet), and PAV is the only proportional
 method I've heard of that doesn't suffer from non-monotonicity or IIA (it's not ABOLUTELY IIA, as there is an indirect effect there, but it comes the closest).  The simple fact is that if everbody cast their votes honestly in an STV election the results wouldn't be fully proportional, they'd have many innacuracies (as has been demonstrated many times, such as the fact that while the Australian Democrats have the largest voting base in the senate outside of the two party system, and yet they have only 1 seat, compared to the 4 the Greens have, and the fact that the National party of Malta was able to win only a minority of seats with 51% of first place votes).  If everyone cast their votes honestly in a PAV election, the results would be fully proportional to the votes cast.
 "It is right and proper to explore and expose the theoretical defects in all voting systems, but for practical reform
 those concerns need to be tempered by the evidence from experience. We have more than 80 years of practical experience
 for STV-PR, so its well-known theoretical defects can be set in context. STV-PR will create a two-party political
 system ONLY if that is what the voters want."
 This sort of reminds me of what they say in America: it's a two-party system because that's the way the voters want it!  Well, I try to reason with them, because it's ludicrous to take a system obviously designed only to aggregate the best out of two choices, lump 10 in, be suprised whenever the voters do the obvious thing and vote tactically for the top two, and then say, "well, that's just how they want it!".  STV-PR has many, many technical defects that keep it from aggregating the true will of the electorate, and while PAV isn't perfect in that regard, it certainly does a better job.
 "The adoption of STV-PR could make an effective contribution to the resolution of some of the political problems in
 Israel that arise directly from its current voting system."
 I'll have to agree with that.  STV-PR is the best hope for a politically feasible option at the moment, but I'm trying to say that PAV would most likely do a better job.  PAV is a rather esotric system invented just 5 years ago, though, which is why it has no hope right now.
 James Gilmour
 election-methods mailing list - see http://electorama.com/em for list info

Celebrate Earth Day everyday!  Discover 10 things you can do to help slow climate change. Yahoo! Earth Day
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20060426/79548f19/attachment-0003.htm>

More information about the Election-Methods mailing list