3 ms·

Thanks for your post, I'll dig into your links. I hadn't heard of STAR in particular. Two counterpoints: - I'm aware of Arrow, but I file it under "perfect be

Thanks for your post, I'll dig into your links. I hadn't heard of STAR in particular. Two counterpoints: - I'm aware of Arrow, but I file it under "perfect being the enemy of the good". :) Whatever the flaws of IRV, it still represents an order of magnitude improvement over FPTP at capturing political preferences. (We suffer so much status quo bias, it takes a lot of activation energy just to communicate to the average voter that better options exist at all!) But I happily cede the point: IRV is not the best choice of all the available options. - My favorite is actually Approval, and I don't always bring it up because it has lower mindshare and is less self-descriptive / "sticky-branding" compared to Ranked Choice. My reasons for the preference: (a) There's no nitpicking over counting methods: whoever gets the most votes wins. (b) Easy to explain to voters (including the counting system), and degrades gracefully for those who still want to vote for only one candidate. (I've heard objections that some voters hit a cheater-detection trigger, that "some people get more votes than others"; this can be solved by reframing as "Y/N" per candidate, so X candidates = X Y/N votes per voter). (c) I believe it largely sidesteps Arrow by redefining the success condition: rather than attempt to perfectly measure ideal preferences, it definitionally represents "Consent of the Governed": not who's your favorite, but who are you willing to live with? This would make it more difficult for bolder and more unconventional candidates, but at the benefit of rewarding coalitions and consensus-building. Polarization and sowing distrust would cease to be a viable strategy. (As an aside, in a perfect world, a blank ballot would be counted as a meta-vote for "None of the Above"; and if it wins, a whole new election of all-new candidates must be held.)

godelski 6y ago
> I file it under "perfect being the enemy of the good" This is 100% how you should think about it. But what I'm trying to say is that we already have substantially better methods, so why not use them? Perfect is the enemy of good, but if I have the choice of eating acid, a bowl of oatmeal, and chocolate pudding, I'm going with the pudding. Perfect is the enemy of good when we don't already have the choices sitting right in front of us. > [IRV] still represents an order of magnitude improvement over FPTP at capturing political preferences And here's where I'm disagreeing, but only slightly. It is definitely better, but not an order of magnitude. > Approval voting I love approval voting. It is great because of its simplicity. That's why they use it in systems like Netflix. Easy to collect and gather results from. Honestly I use it all the time. It is how I solve the "where do you want to eat" problem with friends. You just list things until you get a unanimous vote or if you exhaust your list you just take the most approvals. Simple and easy. The reason I don't like it for elections is that it makes tying easy and I do not believe it has enough precision. The interview I linked Arrow talks about this so I'll deffer to him (he's clearly smarter than me). In elections you need a clear winner. Why I like score voting (and Condorcet methods, like RP, which is again RCV) is because I actually hate party systems, but I recognize that they can't be destroyed. But these methods drastically reduce their power (I don't think IRV does). And I highly encourage you to dig more into the subject and am happy to provide links. I also highly suggest looking deep into everything that the Fair Vote chart discusses, mainly because I believe you will be as frustrated as me (and others, like Condorcet proponents) when you learn what each of those topics are. Everyone I know that is well read in the subject is frustrated by Fair Vote's mischaracterizations.
pdonis 6y ago
> in a perfect world, a blank ballot would be counted as a meta-vote for "None of the Above"; and if it wins, a whole new election of all-new candidates must be held I think this, all by itself, would be a great thing to have in elections. If this had existed in every election in which I have voted, I think I would have used it at least half the time, quite possibly more.
- lukifer 6y ago
  I think the argument against would be pragmatic: the cost and time it takes to run a new election, and the mad scramble for new candidates, presumably leaving the incumbent in office in the interim. (For more local elections, where running unopposed is not uncommon, there's also an interesting conundrum: should that candidate be rejected if they only receive a plurality, ie under 50%? What if literally no one else wants the job?) At any rate, I'm very much in agreement, I think it's a sound policy across the board, even in the context of FPTP. One thing our election system fails to do is disambiguate between voter apathy, and the implicit protest vote of staying home (Australia's mandatory voting being another solution, albeit lacking the teeth of a "None of the Above" reboot).
  - pdonis 6y ago
    > presumably leaving the incumbent in office in the interim An alternative for many posts would be to have no one in the office once the incumbent's term expires. For example, suppose no one wins a race for a seat in the House of Representatives: when the incumbent's term expires, that seat could simply be vacant, and could be deemed to have voted "abstain" for all votes during the term in which it is empty (or until a new special election is held to hopefully fill the seat). (For an office like the President, there might have to be some other provision made.) This would also handle the other case you mention: > For more local elections, where running unopposed is not uncommon, there's also an interesting conundrum: should that candidate be rejected if they only receive a plurality, ie under 50%? I would say yes: that counts as no one winning the election, so the seat remains vacant until a new special election is held (assuming someone does win that). > What if literally no one else wants the job? Then I would say the will of the people is that they prefer to have no one doing it, if they do not approve of anyone who wants it.

godelski 6y ago

> I file it under "perfect being the enemy of the good" This is 100% how you should think about it. But what I'm trying to say is that we already have substantially better methods, so why not use them? Perfect is the enemy of good, but if I have the choice of eating acid, a bowl of oatmeal, and chocolate pudding, I'm going with the pudding. Perfect is the enemy of good when we don't already have the choices sitting right in front of us. > [IRV] still represents an order of magnitude improvement over FPTP at capturing political preferences And here's where I'm disagreeing, but only slightly. It is definitely better, but not an order of magnitude. > Approval voting I love approval voting. It is great because of its simplicity. That's why they use it in systems like Netflix. Easy to collect and gather results from. Honestly I use it all the time. It is how I solve the "where do you want to eat" problem with friends. You just list things until you get a unanimous vote or if you exhaust your list you just take the most approvals. Simple and easy. The reason I don't like it for elections is that it makes tying easy and I do not believe it has enough precision. The interview I linked Arrow talks about this so I'll deffer to him (he's clearly smarter than me). In elections you need a clear winner. Why I like score voting (and Condorcet methods, like RP, which is again RCV) is because I actually hate party systems, but I recognize that they can't be destroyed. But these methods drastically reduce their power (I don't think IRV does). And I highly encourage you to dig more into the subject and am happy to provide links. I also highly suggest looking deep into everything that the Fair Vote chart discusses, mainly because I believe you will be as frustrated as me (and others, like Condorcet proponents) when you learn what each of those topics are. Everyone I know that is well read in the subject is frustrated by Fair Vote's mischaracterizations.

pdonis 6y ago

> in a perfect world, a blank ballot would be counted as a meta-vote for "None of the Above"; and if it wins, a whole new election of all-new candidates must be held I think this, all by itself, would be a great thing to have in elections. If this had existed in every election in which I have voted, I think I would have used it at least half the time, quite possibly more.

lukifer 6y ago
I think the argument against would be pragmatic: the cost and time it takes to run a new election, and the mad scramble for new candidates, presumably leaving the incumbent in office in the interim. (For more local elections, where running unopposed is not uncommon, there's also an interesting conundrum: should that candidate be rejected if they only receive a plurality, ie under 50%? What if literally no one else wants the job?) At any rate, I'm very much in agreement, I think it's a sound policy across the board, even in the context of FPTP. One thing our election system fails to do is disambiguate between voter apathy, and the implicit protest vote of staying home (Australia's mandatory voting being another solution, albeit lacking the teeth of a "None of the Above" reboot).
- pdonis 6y ago
  > presumably leaving the incumbent in office in the interim An alternative for many posts would be to have no one in the office once the incumbent's term expires. For example, suppose no one wins a race for a seat in the House of Representatives: when the incumbent's term expires, that seat could simply be vacant, and could be deemed to have voted "abstain" for all votes during the term in which it is empty (or until a new special election is held to hopefully fill the seat). (For an office like the President, there might have to be some other provision made.) This would also handle the other case you mention: > For more local elections, where running unopposed is not uncommon, there's also an interesting conundrum: should that candidate be rejected if they only receive a plurality, ie under 50%? I would say yes: that counts as no one winning the election, so the seat remains vacant until a new special election is held (assuming someone does win that). > What if literally no one else wants the job? Then I would say the will of the people is that they prefer to have no one doing it, if they do not approve of anyone who wants it.