A comparison of two different methodologies of voting. Which do you think is more just?

[Debrikishaw]

Here is a comparison of two different methodologies of voting.

In a hypothetical election:

43% vote for Candidate 1
37% vote for Candidate 2
20% vote for Candidate 3


In the typical FPTP (First Past The Post) system, Candidate 1 would have won.
But is he rightfully the winner?

Here's how it would work in the IRV (Instant Runoff Voting) system.
Voters are allowed to rank their preference in candidates. No candidate got the majority of the vote, so votes must be redistributed to see who actually holds the support of the majority of the population. Candidate 3 had the least amount of votes, so he is eliminated. Candidate 3's votes are redistributed. 80% of Candidate 3 voters had Candidate 2 as their second place choice, while 20% had Candidate 1 as their second place choice. That means Candidate 2 gets the 16% (80% of 20%) of general voters that voted for Candidate 3 and had Candidate 2 as their second choice, and Candidate 1 gets the 4% (20% of 20%) of general voters that voted for Candidate 3 and had Candidate 1 as their second choice.

The results are now:
Candidate 1: 47%
Candidate 2: 53%

Candidate 2 has the majority and is declared the winner in the IVR system.

FPTP Perspective:
Candidate 1 had the most votes out of anyone, therefore it is only fair to elect him.

IRV Perspective:
Candidate 1 only had a minority of the vote. The majority voted against him, but the majority happened to split up their vote between two other candidates. The actuality is 53% of the voters would have rather had Candidate 2 over Candidate 1. Candidate 2 had the largest and broadest support, therefore it is only fair to elect him.


Which system do you think is more fair?

[Xenophage]

I remember this as a topic in '08 on the forums. I think there were some issues with IRV that some other, more exotic voting styles sought to address, but I don't remember it at all.

Any way, IRV seems a lot better to me than our current system. Doesn't Australia use it?

[Acala]

There is a whole field of advanced study called Public Choice that addresses this and much more. See, for example, the Calculus of Consent. In summary, every method of voting can be rigged to give different results.

The only fair method is decision by consensus with complete power of secession at every level.

[RonRules]

Brian Dunning who runs the Skeptoid podcast made an episode on the subject:

The Science of Voting
Are democratic elections actually as free and fair as we think they are?
http://skeptoid.com/episodes/4281

You can listen to it here:
http://hw.libsyn.com/p/4/1/d/41d56e82b4091951/skeptoid-4281.mp3?sid=c3b549b7b5622d3b88687c4fc707fca3&l_sid=17974&l_eid=&l_mid=2756543