I've been arguing there that FairVote's Instant Runoff Voting(IRV) isn't so bad,
even more so in a 2-party dominated system,
which isn't necessarily bad.
I've also argued that IRV can be improved upon
by hybridizing it with Approval Voting(AV)
as an IRV3/AV3 election rule that lets voters rank up to three candidates.
IRV3/AV3 takes a 2 stage approach.
The first stage treats the up to 3 rankings initially as approval votes. The three candidates who get the most rankings, or approval votes, go to the second stage.
In the second stage, the rankings are used first to eliminate the candidate who is top ranked by the least number of voters. Then, the rankings are used to transfer as many of the eliminated candidate's votes as possible to the remaining two candidates. Finally, the preferred candidate between the remaining two candidates is elected, most often by a majority of the voters.
It has been an intense and, now, we may have made a break through.
One of my most talented antagonists, Jameson Quinn, shared the results of a pseudo-experiment he designed to show that IRV3/AV3 doesn't necessarily work. It's worth sharing because it turned out to be very hard to design a situation where the 2nd best candidate beats the top candidate when IRV3/AV3 is used.
From: Jameson Quinn
To: election-methods list serve.
Date: Sun, 8 Jan 2012 19:03:29 -0600
Subject: [EM] A problem with IRV3/AV3
JQ: Imagine a scenario of an ABCD one-dimensional continuum:
There are 100 voters total.
41 voters with preferences: A>B>C
19 voters with preferences: B>A>C
20 voters with preferences: B>C>D
20 voters with preferences: C>B>D
If the [entire first group of] voters [strategically] vote 1st rank A and second rank D [with no other candidates ranked] then A [may] win. By raising the turkey D over the true Condorcet Winner B, they have stolen the win. Even if their strategy fails to keep B out of the top 3, they lose nothing; B will still win.
dlw: thanks for doing this. In the first stage wouldn't B and C tie for 3rd place if only the first set of voters all voted strategically together in the same way? They'd both get rankings from 59 of the voters. So if it came down to a coin-toss, there'd be a 50-50 chance of the CW winning vs the 2nd place candidate given a massive coordinated strategic vote by only a subset of the sample (We assume none of the 3rd or 4th set of voters decide to strategically leave off D rankings)?
JQ:To be honest, it was harder to tune this scenario than I thought it would be. Thus, having taken the time to write this down, I am no longer opposed to IRV3/AV3. (For IRV2/AV2[other hybrid], it's easier to get this problem. It's also easier to get the problem if there are clones involved, but real-world clones beyond 3 candidates are unlikely.)
dlw:Thank you again. The MSM+Blogosphere shd help identify clones in real world.
JQ: Since I'm now not opposed to IRV3/AV3, I consider it one of the 3 reforms (along with [JQ's creative multi-stage single-winner election rule] SODA and IRV) that would be most acceptable to incumbents, because it avoids the weak Condorcet winner problem [IOW, it doesn't trim the incumbency advantage as much as some other election rules.]. Still, it is basically just as bad as IRV for nonmonotonicity and spoilers; all the spoiler scenarios I consider realistic are essentially 3-candidate anyway. As such, I see no reason to believe that it would not lead to lesser-evil voting and 2-party domination, as IRV does. Since I see 2-party domination (as opposed to just having 2 strongest parties, a logical necessity) as a source of the most-serious problems with Plurality, I still feel that SODA is a much better option than IRV3/AV3.
dlw: And our difference is that I see the near exclusive use of Plurality voting rules as a key source of my country's current evils, since it's not hard to imagine a 2 party dominated system that is a lot better. All it takes is for there to be better checks and balances between the two major parties, along with increased scope for outsiders/dissenters to express themselves via minor parties and Local Third Parties (LTPs). The natural result of such a system would be two different than the current major parties who together would be a melding pot for the US's democracy.
dlw
ps, I'm going to repost [cleaned up version] this on my blog.
dlw
0 comments:
Post a Comment