5 ms·

One thing that I think gets lost in political debate around voting is that RCV also changes the behavior/selection of the candidates. Its not just that the cand

One thing that I think gets lost in political debate around voting is that RCV also changes the behavior/selection of the candidates. Its not just that the candidates are identical and you vote with a different mechanism - but candidates have an incentive to moderate. This may mean the outcome is the same (in terms of which candidate wins) but their positions represent the median voter more. I would imagine approval voting and other voting systems have this too (I'm not going to pretend to know how this would turn out for all of them) and I think that gets lost a bit when analyzing the systems.

ajmurmann 4y ago
This is also where it would be great to skip primaries, since primaries favor more extreme candidates.
gamegoblin 4y ago
This is actually not the case - RCV suffers from the "center squeeze" effect [1], in which moderates are squeezed out in the early rounds. That said, things like Approval have the opposite problem, in which moderates are given an advantage (some approval people will say this is a feature, not a bug). For an edge case, consider an election with 3 factions: socialists, libertarians, and fascists. There is also a 4th candidate, a boring status quo moderate, running. Imagine the populace is split 1/3 between the first 3 factions as their 1st choice, but everyone selects boring status quo moderates as their 2nd choice, because everyone in the first 3 factions hates the other ideologies. The boring status quo moderates get eliminated in round 1, and eventually 1 of the 3 extremes will win, resulting in 2/3 of the country being very angry. Not ideal. Obviously this is just an unrealistic edge case, but it demonstrates the effect. [1] https://electowiki.org/wiki/Center_squeeze https://electowiki.org/wiki/Center_squeeze
- godelski 4y ago
  > That said, things like Approval have the opposite problem, in which moderates are given an advantage (some approval people will say this is a feature, not a bug). This is a weird way to say that representing the population is a bug. Your example election would be a compelling argument if population preference was more of a multimodal distribution rather than closer to a unimodal distribution. Despite the US's current division, it is nowhere near as divided as your scenario.
  - gamegoblin 4y ago
    To be clear, I personally prefer approval to IRV. But raw approval gives undue advantage to moderates. Approval-to-runoff being clearly superior than raw approval, for this reason. And STAR being a little bit better than that. The RCV people's main criticism of Approval is that it doesn't allow you to express different levels of support, and they aren't totally wrong (even though their preferred mechanism of IRV doesn't get them there, either). I think (without data, happy to be proven wrong) that the US population is fairly strongly bimodal on most issues. Many issues that people care a lot about (for better or worse...) have two clear centers of gravity, e.g. for gun control and abortion most people are in the orbit of a "mostly ban" or "mostly allow" group, with very few people falling into some nuanced middle ground.
    - godelski 4y ago
      > But raw approval gives undue advantage to moderates. If by "moderates" you mean "most liked" then I'm not sure what the problem is here. The point of voting is to find the most universally agreed upon ruler. You may call this unfair, but I would say that it is MORE unfair to give advantage to fringe groups rather than giving advantage to the masses. I thought the point of democracies was about having rule by the masses. If you think fringe groups should have the advantage I'd like to hear your argument.
      - gamegoblin 4y ago
        > If you think fringe groups should have the advantage I'd like to hear your argument. I'm not sure what gave you the impression I think that. My comments clearly suggest that my view is that STAR > Approval-to-runoff > Approval > RCV-IRV > Plurality. My reasoning is that I place a very high value on the condorcet and favorite betrayal criteria.
        godelski 4y ago
        My apologies then. It was the italicized "undue" that made me think that this was your position. I read it as "undeserving". As for the condorcet winner, that is quite hard actually. But if we consider a weak condorcet winner (where it is true 90+% of the time), that isn't as hard. I do think simplicity, transparency, and reduced vulnerability to strategy is a worthwhile trade-off for a strong condorcet winner. But of course, this is just my personal opinion and what's the preferred optima depends a lot on personal preferences.
godelski 4y ago
That's kinda the entire point. One of the (many) metrics for voting systems is called Voter Satisfaction Efficiency (VSE). This is basically a global L2 distance from each voter to each candidate. To maximize VSE we want to minimize that distance. In other words, we want winning candidates to be most representative of a population's opinions and values. You can see a comparison of different systems here[0]. Note that there are other criteria that matter, but there are plenty of methods that have high VSE and the other criteria we want. Also note that the spread in different types of strategies represents the method's vulnerability to manipulation. If a strategy still yields a high VSE then it is not an effective strategy. [0] https://electionscience.github.io/vse-sim/ https://electionscience.github.io/vse-sim/
- fasthands9 4y ago
  If I'm not mistaken this link is exactly what I am describing. It simulates how close the winning candidate is to voters based on how people would vote. But it explicitly doesnt simulate how candidates would shift due to changing incentives. In other words - it treats elections as a one side game where voters pick strategy (voting methods/preferences) and the candidates are stable. But it is a two-sided game where politicians change opinions/policies based on system as well.
  - godelski 4y ago
    That is a hard thing to capture, but strategy is a decent way to approximate it. I do think this is fair criticism for what it is worth, but I don't think anyone has been able to create an accurate model for this (for any voting system) and I'm not sure you can accurately model this tbh.
RyanCavanaugh 4y ago
RCV-IRV doesn't encourage moderate candidates because, in a polarized environment, moderate candidates are much more likely to be no one's first choice and get eliminated as a result. This isn't hypothetical -- you can look at the results in Seattle's City Attorney race, which strangely ended up electing a Republican to a citywide office for the first time in 20 years, and where the outcome under RCV would have been the same because the moderate candidate (the reasonably well-liked incumbent) didn't secure enough first-place votes, either in FPTP or in RCV-IRV.
- fasthands9 4y ago
  That is interesting but I don't know if that's always true. In Alaska, its pretty clear that Senator Murkowski, who was always somewhat moderate, has really leaned into being even more moderate since RCV was introduced there. Presumably, she knows that she will face little penalty for rebuffing Trump or McConnell on the issues they have for which the right position is unpopular - since there isn't much of a chance of her being dethroned by a right person. I do take your point that the dynamics can be weird and aren't always "go to the median". And perhaps having a weird dynamic where the introduction of a different voting strategy or other party changes things dramatically - is not a good thing for trust.
  - jcadam 4y ago
    Murkowski would have been eliminated in the GOP primary this year under the traditional system. Most republicans here in AK despise her. Whether she could pull off another highly questionable win in the general via write-in ballots is doubtful. McConnell has thrown his support behind her, rather than the candidate with far more republican support within the state, because of course he has. Murkowski probably wins with RCV. The one D on the ballot, Chesbro, is a placeholder nobody expects to win, but her voters will surely rank Murkowski #2.
- zajio1am 4y ago
  > moderate candidates are much more likely to be no one's first choice and get eliminated as a result. If i look at my country's upper chamber, which is elected with two-round system, then it is quite the reverse. Extremist candidates rarely pass to second round and if they do, they are eliminated there.

ajmurmann 4y ago

This is also where it would be great to skip primaries, since primaries favor more extreme candidates.

gamegoblin 4y ago

This is actually not the case - RCV suffers from the "center squeeze" effect [1], in which moderates are squeezed out in the early rounds. That said, things like Approval have the opposite problem, in which moderates are given an advantage (some approval people will say this is a feature, not a bug). For an edge case, consider an election with 3 factions: socialists, libertarians, and fascists. There is also a 4th candidate, a boring status quo moderate, running. Imagine the populace is split 1/3 between the first 3 factions as their 1st choice, but everyone selects boring status quo moderates as their 2nd choice, because everyone in the first 3 factions hates the other ideologies. The boring status quo moderates get eliminated in round 1, and eventually 1 of the 3 extremes will win, resulting in 2/3 of the country being very angry. Not ideal. Obviously this is just an unrealistic edge case, but it demonstrates the effect. [1] https://electowiki.org/wiki/Center_squeeze https://electowiki.org/wiki/Center_squeeze

godelski 4y ago
> That said, things like Approval have the opposite problem, in which moderates are given an advantage (some approval people will say this is a feature, not a bug). This is a weird way to say that representing the population is a bug. Your example election would be a compelling argument if population preference was more of a multimodal distribution rather than closer to a unimodal distribution. Despite the US's current division, it is nowhere near as divided as your scenario.
- gamegoblin 4y ago
  To be clear, I personally prefer approval to IRV. But raw approval gives undue advantage to moderates. Approval-to-runoff being clearly superior than raw approval, for this reason. And STAR being a little bit better than that. The RCV people's main criticism of Approval is that it doesn't allow you to express different levels of support, and they aren't totally wrong (even though their preferred mechanism of IRV doesn't get them there, either). I think (without data, happy to be proven wrong) that the US population is fairly strongly bimodal on most issues. Many issues that people care a lot about (for better or worse...) have two clear centers of gravity, e.g. for gun control and abortion most people are in the orbit of a "mostly ban" or "mostly allow" group, with very few people falling into some nuanced middle ground.
  - godelski 4y ago
    > But raw approval gives undue advantage to moderates. If by "moderates" you mean "most liked" then I'm not sure what the problem is here. The point of voting is to find the most universally agreed upon ruler. You may call this unfair, but I would say that it is MORE unfair to give advantage to fringe groups rather than giving advantage to the masses. I thought the point of democracies was about having rule by the masses. If you think fringe groups should have the advantage I'd like to hear your argument.
    - gamegoblin 4y ago
      > If you think fringe groups should have the advantage I'd like to hear your argument. I'm not sure what gave you the impression I think that. My comments clearly suggest that my view is that STAR > Approval-to-runoff > Approval > RCV-IRV > Plurality. My reasoning is that I place a very high value on the condorcet and favorite betrayal criteria.
      - godelski 4y ago
        My apologies then. It was the italicized "undue" that made me think that this was your position. I read it as "undeserving". As for the condorcet winner, that is quite hard actually. But if we consider a weak condorcet winner (where it is true 90+% of the time), that isn't as hard. I do think simplicity, transparency, and reduced vulnerability to strategy is a worthwhile trade-off for a strong condorcet winner. But of course, this is just my personal opinion and what's the preferred optima depends a lot on personal preferences.

godelski 4y ago

That's kinda the entire point. One of the (many) metrics for voting systems is called Voter Satisfaction Efficiency (VSE). This is basically a global L2 distance from each voter to each candidate. To maximize VSE we want to minimize that distance. In other words, we want winning candidates to be most representative of a population's opinions and values. You can see a comparison of different systems here[0]. Note that there are other criteria that matter, but there are plenty of methods that have high VSE and the other criteria we want. Also note that the spread in different types of strategies represents the method's vulnerability to manipulation. If a strategy still yields a high VSE then it is not an effective strategy. [0] https://electionscience.github.io/vse-sim/ https://electionscience.github.io/vse-sim/

fasthands9 4y ago
If I'm not mistaken this link is exactly what I am describing. It simulates how close the winning candidate is to voters based on how people would vote. But it explicitly doesnt simulate how candidates would shift due to changing incentives. In other words - it treats elections as a one side game where voters pick strategy (voting methods/preferences) and the candidates are stable. But it is a two-sided game where politicians change opinions/policies based on system as well.
- godelski 4y ago
  That is a hard thing to capture, but strategy is a decent way to approximate it. I do think this is fair criticism for what it is worth, but I don't think anyone has been able to create an accurate model for this (for any voting system) and I'm not sure you can accurately model this tbh.

RyanCavanaugh 4y ago

RCV-IRV doesn't encourage moderate candidates because, in a polarized environment, moderate candidates are much more likely to be no one's first choice and get eliminated as a result. This isn't hypothetical -- you can look at the results in Seattle's City Attorney race, which strangely ended up electing a Republican to a citywide office for the first time in 20 years, and where the outcome under RCV would have been the same because the moderate candidate (the reasonably well-liked incumbent) didn't secure enough first-place votes, either in FPTP or in RCV-IRV.

fasthands9 4y ago
That is interesting but I don't know if that's always true. In Alaska, its pretty clear that Senator Murkowski, who was always somewhat moderate, has really leaned into being even more moderate since RCV was introduced there. Presumably, she knows that she will face little penalty for rebuffing Trump or McConnell on the issues they have for which the right position is unpopular - since there isn't much of a chance of her being dethroned by a right person. I do take your point that the dynamics can be weird and aren't always "go to the median". And perhaps having a weird dynamic where the introduction of a different voting strategy or other party changes things dramatically - is not a good thing for trust.
- jcadam 4y ago
  Murkowski would have been eliminated in the GOP primary this year under the traditional system. Most republicans here in AK despise her. Whether she could pull off another highly questionable win in the general via write-in ballots is doubtful. McConnell has thrown his support behind her, rather than the candidate with far more republican support within the state, because of course he has. Murkowski probably wins with RCV. The one D on the ballot, Chesbro, is a placeholder nobody expects to win, but her voters will surely rank Murkowski #2.
zajio1am 4y ago
> moderate candidates are much more likely to be no one's first choice and get eliminated as a result. If i look at my country's upper chamber, which is elected with two-round system, then it is quite the reverse. Extremist candidates rarely pass to second round and if they do, they are eliminated there.