Epistemic Status:I'm confused! Let's go shopping! (...for new political systems <3)
I want to write an essay about the actually best voting system, but before I do that I want to get clear on what the desiderata should even naturally or properly or wisely be...
Participation?
Sometimes it is illegal to not vote. You could create a two day holiday, and have 24 hour emergency workers do shifts but have some time off to go in and be fingerprinted and register their preferences and so on. There could be free money at the polling station for voting, and voting assistants hunting down the people who haven't voted yet.
If you have this system, then "refusing to vote" can never happen.
But also, certain voting systems fail the Participation criteria such that some people might wish, in retrospect, to have turned in a ballot that says NULL (and makes it possible for the election to fail quorum?) rather than turning in a ballot.
On the other hand, if a polity uses a system that FAILS the Participation criteria AND ALSO it forces everyone to vote, then maybe it would be unethical to have forced people though the puppet show of pretending to be able to express their civic preferences without them actually being able to express their civic preferences?
On the gripping hand, maybe if you're trying to boot up a new polity from scratch (as was attempted in Iraq, after George W Bush invaded that country in 2003) maybe you really really really want to incentivize people to vote for a bit just to "get the thing started"? Maybe Participation is super important for building and merging rather than shrinking and splitting? Maybe Metcalfe's Law is relevant to polities? Is bigger always better?
Forking?
Sometimes a country's citizenship is very valuable (like the US has a citizenship like this, but it isn't the most valued-in-practice citizenship from the "cost to become citizen" estimates I can find) and other country's citizenship is net negative, with people trying to escape. Sometimes a lot of people want to escape all at the same time. Also, maybe certain election results will cause some large faction of citizens to want to exert their right to revolution, and break away? (Or maybe there is no moral right to revolution? Or maybe whether there is a right to revolution is culture dependent?) And so maybe it is a positive feature of an election if "None Of The Above For A Single Polity / Break The Polity In Two With These TWO Leaders" is a possible outcome? Or not? How would we know?
According to The CAP Theorem, if you refuse to allow Forking then you MUST choose between Availability and Consistency in your system design... but when is Forking really bad and when is Forking actually maybe kinda good?
Something I notice: there is very very little attention paid to the "polity merge operation" where two polities might be separate, and both hold elections, and then end up merged at the end, and it somehow goes very smoothly and nicely, because they were, in some sense, already "running the same civic operating system" and that civic operating system is able to fork and merge by design. Maybe if all the US states were running civic operating systems that support this behavior somehow, then maybe the state boundaries wouldn't be fucked beyond belief and very very very far from the naturally good places for them to be?
Maybe there are systematic insanities latent in human nature, and the median leader preferred by almost everyone in head-to-head pairwise comparisons would turn out to be someone who is "objectively" very evil, and wants to do something like commit genocide on 15% of the population (or whatever... if you are personally in favor of genocide then imagine I said some other "clear moral evil" that you would see as a violation of Natural Law (or whatever standard you use for deciding if something is ethical or unethical based on a coherent conscience that is distinct from "whatever the fuck you merely feel like you want right now"), but which might also be predictably something that a majority of people in some country would simply want).
If human's just really love to do evil a lot in practice (or certain humans in certain situations?) then their collectively most preferred outcome "in the middle" where it "seems common-sensically preferable to most of them" with the Condorcet Criterion might misfire, and reliably generate one evil leader after another.
In practice, in the US, with our POTUS elections, it seems like we reliably get a POTUS that some large fraction of the country really really dislikes but also, if you look at the polling data, and the third party options, if POTUS elections reliably selected the Condorcet Winner from among the top 1000 people who got enough signature to be in the election, then... NONE of the recent past Presidents would have won, most likely? It would have been a bunch of namby-pamby libertarian environmentalists who believe in civic virtue, self defense, small government, and prosperity, over and over and over.
Maybe "namby-pamby libertarian environmentalists who believe in civic virtue, self defense, small government, and prosperity" is an objectively evil platform for a leader to adopt, and something America should not want to reliably elect over and over? So maybe we shouldn't have POTUS elections that fulfill the Condorcet Criterion? Or maybe I'm wrong about what Condorcet Criterion satisfying leaders would look like here?
Also, maybe different cultures are more or less "objectively good or evil", and only the "evil cultures" should avoid the Condorcet Criterion, whereas the "good cultures" should adopt it? (This would assume some pragmatically relevant variant of moral realism is true, of course, and maybe no variant of moral realism at all, in any form, is true?)
Preference Strengths?
Right now the federal minimum wage in the United States is $7.25 and so working fulltime for two days would earn $116 which we can round to $100 for ease of mental math.
Hypothetically, people could go to polling stations and be given $100 to show up and vote "I'm a sheep and I don't even care but I know I like money and so I'm not voting but I'm just gonna take the money".
Then you'd have to refuse the $100 to actually vote at normal strength.
Then you could pay $100 to vote with 2X weight.
And then for $333 you could vote with 3X weight, and pay $1000 to vote with 4X weight, and pay $3333 for 5X and pay $10,000 for 6X, and so on all the way up to paying billions of dollars in optional taxes?
Quadratic Voting was the new hotness for a while in mechanism design but it fundamentally presumes an "allocation of goodies to whoever wants the goodies the most" mindset. Some people want high taxes and large handouts because they are poor, and other people want low taxes and few handouts because they are rich, for one example, and presumably these "selfish motivations" in BOTH directions are "not really about ethics and fairness"? It probably connects to deep questions like the moral issue of compulsory charitable giving.
One nice thing about soliciting preferences, is that revolutions are very very costly, and if you have 60% of the population who wants to murder and eat the other 40% just a little bit as one of many possible things they could eat (and the 40% would instantly choose to revolt against the government if the government tried to implement this policy) then letting the 40% pay the 60% a little bit of money to control the government despite being in the minority, and use their government control to make the government NOT try to kill them thereby, it would be cheaper and better for everyone overall?
Truth Solicitation?
A different frame would be that everyone is assumed to be enlightened, and wanting to know the truth, and express the truth, but uncertain.
Maybe people lean towards truth on average, and then we can use the Condorcet Jury Theorem to aggregate uncertainty into higher quality beliefs about what the best way for the polity to proceed would be?
Then again... if you seriously wanted to get the truth, then presumably there are better ways to do this than force everyone to vote (ooh! the Participation criteria showed up again!) but instead hire experts, and use Bayesian Truth Serum, and have betting markets for a lot of it.
Maybe it depends on the complexity of the questions being faced? Maybe if the issues are very simple then everyone already knows the right answers and truth solicitation is pointless to optimize for, but if the issues are very complex, and being wrong would hurt a lot, then maybe an electoral system being performant on this dimension could be The Ultimate Thing To Get Right?
Moral Maze Resistance?
Something that often happens in organizations that exist for more than about 8 years (which is roughly how long someone is a CEO in most for profit companies, and also the term limit for President) and have more than about 150 people (such that anonymity can creep in above that number) is that it turns into a Moral Maze ruled de facto according to the Iron Law Of Oligarchy at the top, and patrimonial bureaucratic norms in the middle.
When this happens, it is very common for the humans at the top to be there because they want to abuse their power for personal gain, deriving joy and wealth and reproductive success from the unbalanced exercise of social power, rather than engaging in servant leadership.
When political scientists look at polities, they find that if there is a single-party unicameral parliament with no proportional representation (especially not the kind that is resistant to gerrymandering), then you almost certainly will end up with rampant corruption. Forcing there to be >1 parties somehow helps reduce corruption. Making two different Houses have to agree on legislation that is finalized before either of them votes helps reduce corruption. Proportional representation might be a sufficient solution all by itself? Except when I searched again for new papers on this topic it apparently matters A LOT whether the proportional representation is "open list" vs "closed list". The closed list option is the bad one.
If you look at Wikipedia's awesome and good "Comparison Of Electoral Systems" you will not find "resistant to Moral Mazes and conducive to Low Corruption Multiparty Outcomes" as one of the criteria, even though this might be literally the most important thing?
But also the need for this might be of very low importance for a city state full of philosophically wise saints?
But also, if you're trying to reduce Forking, and trying to get people to Participate, then maybe no one will want to participate if they can't have a little bit of corruption... as a treat?
Anyway, there's a huge literature on this stuff, figuring out empirically what systems have the most coups, and most corruption, and so on and so forth. I'm not an expert on this literature, and that's why I'm asking a question rather than writing an essay <3
I honestly don't know.
Other Factors?
Surely I'm missing a lot of factors.
This is, after all, a post that is marked as a question.
What are the important factors to look at in a polity to help that polity even decide what the right desiderata are for picking an electoral system?
Epistemic Status: I'm confused! Let's go shopping! (...for new political systems <3)
I want to write an essay about the actually best voting system, but before I do that I want to get clear on what the desiderata should even naturally or properly or wisely be...
Participation?
Sometimes it is illegal to not vote. You could create a two day holiday, and have 24 hour emergency workers do shifts but have some time off to go in and be fingerprinted and register their preferences and so on. There could be free money at the polling station for voting, and voting assistants hunting down the people who haven't voted yet.
If you have this system, then "refusing to vote" can never happen.
But also, certain voting systems fail the Participation criteria such that some people might wish, in retrospect, to have turned in a ballot that says NULL (and makes it possible for the election to fail quorum?) rather than turning in a ballot.
On the other hand, if a polity uses a system that FAILS the Participation criteria AND ALSO it forces everyone to vote, then maybe it would be unethical to have forced people though the puppet show of pretending to be able to express their civic preferences without them actually being able to express their civic preferences?
On the gripping hand, maybe if you're trying to boot up a new polity from scratch (as was attempted in Iraq, after George W Bush invaded that country in 2003) maybe you really really really want to incentivize people to vote for a bit just to "get the thing started"? Maybe Participation is super important for building and merging rather than shrinking and splitting? Maybe Metcalfe's Law is relevant to polities? Is bigger always better?
Forking?
Sometimes a country's citizenship is very valuable (like the US has a citizenship like this, but it isn't the most valued-in-practice citizenship from the "cost to become citizen" estimates I can find) and other country's citizenship is net negative, with people trying to escape. Sometimes a lot of people want to escape all at the same time. Also, maybe certain election results will cause some large faction of citizens to want to exert their right to revolution, and break away? (Or maybe there is no moral right to revolution? Or maybe whether there is a right to revolution is culture dependent?) And so maybe it is a positive feature of an election if "None Of The Above For A Single Polity / Break The Polity In Two With These TWO Leaders" is a possible outcome? Or not? How would we know?
According to The CAP Theorem, if you refuse to allow Forking then you MUST choose between Availability and Consistency in your system design... but when is Forking really bad and when is Forking actually maybe kinda good?
Something I notice: there is very very little attention paid to the "polity merge operation" where two polities might be separate, and both hold elections, and then end up merged at the end, and it somehow goes very smoothly and nicely, because they were, in some sense, already "running the same civic operating system" and that civic operating system is able to fork and merge by design. Maybe if all the US states were running civic operating systems that support this behavior somehow, then maybe the state boundaries wouldn't be fucked beyond belief and very very very far from the naturally good places for them to be?
Sauce.
Objective Evil?
Maybe there are systematic insanities latent in human nature, and the median leader preferred by almost everyone in head-to-head pairwise comparisons would turn out to be someone who is "objectively" very evil, and wants to do something like commit genocide on 15% of the population (or whatever... if you are personally in favor of genocide then imagine I said some other "clear moral evil" that you would see as a violation of Natural Law (or whatever standard you use for deciding if something is ethical or unethical based on a coherent conscience that is distinct from "whatever the fuck you merely feel like you want right now"), but which might also be predictably something that a majority of people in some country would simply want).
If human's just really love to do evil a lot in practice (or certain humans in certain situations?) then their collectively most preferred outcome "in the middle" where it "seems common-sensically preferable to most of them" with the Condorcet Criterion might misfire, and reliably generate one evil leader after another.
In practice, in the US, with our POTUS elections, it seems like we reliably get a POTUS that some large fraction of the country really really dislikes but also, if you look at the polling data, and the third party options, if POTUS elections reliably selected the Condorcet Winner from among the top 1000 people who got enough signature to be in the election, then... NONE of the recent past Presidents would have won, most likely? It would have been a bunch of namby-pamby libertarian environmentalists who believe in civic virtue, self defense, small government, and prosperity, over and over and over.
Maybe "namby-pamby libertarian environmentalists who believe in civic virtue, self defense, small government, and prosperity" is an objectively evil platform for a leader to adopt, and something America should not want to reliably elect over and over? So maybe we shouldn't have POTUS elections that fulfill the Condorcet Criterion? Or maybe I'm wrong about what Condorcet Criterion satisfying leaders would look like here?
Also, maybe different cultures are more or less "objectively good or evil", and only the "evil cultures" should avoid the Condorcet Criterion, whereas the "good cultures" should adopt it? (This would assume some pragmatically relevant variant of moral realism is true, of course, and maybe no variant of moral realism at all, in any form, is true?)
Preference Strengths?
Right now the federal minimum wage in the United States is $7.25 and so working fulltime for two days would earn $116 which we can round to $100 for ease of mental math.
Hypothetically, people could go to polling stations and be given $100 to show up and vote "I'm a sheep and I don't even care but I know I like money and so I'm not voting but I'm just gonna take the money".
Then you'd have to refuse the $100 to actually vote at normal strength.
Then you could pay $100 to vote with 2X weight.
And then for $333 you could vote with 3X weight, and pay $1000 to vote with 4X weight, and pay $3333 for 5X and pay $10,000 for 6X, and so on all the way up to paying billions of dollars in optional taxes?
Quadratic Voting was the new hotness for a while in mechanism design but it fundamentally presumes an "allocation of goodies to whoever wants the goodies the most" mindset. Some people want high taxes and large handouts because they are poor, and other people want low taxes and few handouts because they are rich, for one example, and presumably these "selfish motivations" in BOTH directions are "not really about ethics and fairness"? It probably connects to deep questions like the moral issue of compulsory charitable giving.
One nice thing about soliciting preferences, is that revolutions are very very costly, and if you have 60% of the population who wants to murder and eat the other 40% just a little bit as one of many possible things they could eat (and the 40% would instantly choose to revolt against the government if the government tried to implement this policy) then letting the 40% pay the 60% a little bit of money to control the government despite being in the minority, and use their government control to make the government NOT try to kill them thereby, it would be cheaper and better for everyone overall?
Truth Solicitation?
A different frame would be that everyone is assumed to be enlightened, and wanting to know the truth, and express the truth, but uncertain.
Maybe people lean towards truth on average, and then we can use the Condorcet Jury Theorem to aggregate uncertainty into higher quality beliefs about what the best way for the polity to proceed would be?
Then again... if you seriously wanted to get the truth, then presumably there are better ways to do this than force everyone to vote (ooh! the Participation criteria showed up again!) but instead hire experts, and use Bayesian Truth Serum, and have betting markets for a lot of it.
Maybe it depends on the complexity of the questions being faced? Maybe if the issues are very simple then everyone already knows the right answers and truth solicitation is pointless to optimize for, but if the issues are very complex, and being wrong would hurt a lot, then maybe an electoral system being performant on this dimension could be The Ultimate Thing To Get Right?
Moral Maze Resistance?
Something that often happens in organizations that exist for more than about 8 years (which is roughly how long someone is a CEO in most for profit companies, and also the term limit for President) and have more than about 150 people (such that anonymity can creep in above that number) is that it turns into a Moral Maze ruled de facto according to the Iron Law Of Oligarchy at the top, and patrimonial bureaucratic norms in the middle.
When this happens, it is very common for the humans at the top to be there because they want to abuse their power for personal gain, deriving joy and wealth and reproductive success from the unbalanced exercise of social power, rather than engaging in servant leadership.
When political scientists look at polities, they find that if there is a single-party unicameral parliament with no proportional representation (especially not the kind that is resistant to gerrymandering), then you almost certainly will end up with rampant corruption. Forcing there to be >1 parties somehow helps reduce corruption. Making two different Houses have to agree on legislation that is finalized before either of them votes helps reduce corruption. Proportional representation might be a sufficient solution all by itself? Except when I searched again for new papers on this topic it apparently matters A LOT whether the proportional representation is "open list" vs "closed list". The closed list option is the bad one.
If you look at Wikipedia's awesome and good "Comparison Of Electoral Systems" you will not find "resistant to Moral Mazes and conducive to Low Corruption Multiparty Outcomes" as one of the criteria, even though this might be literally the most important thing?
But also the need for this might be of very low importance for a city state full of philosophically wise saints?
But also, if you're trying to reduce Forking, and trying to get people to Participate, then maybe no one will want to participate if they can't have a little bit of corruption... as a treat?
Anyway, there's a huge literature on this stuff, figuring out empirically what systems have the most coups, and most corruption, and so on and so forth. I'm not an expert on this literature, and that's why I'm asking a question rather than writing an essay <3
I honestly don't know.
Other Factors?
Surely I'm missing a lot of factors.
This is, after all, a post that is marked as a question.
What are the important factors to look at in a polity to help that polity even decide what the right desiderata are for picking an electoral system?