The impression I got, was that collusion between likeminded people created an “indirect democracy” where causes supported by the most people could most efficiently advocate their position.
If that is the case, then this system does punish parties that are less willing (or able) to cooperate, which could feasibly be a bad thing, if it means that unpopular results occur because one side is less nuanced on it’s position (e.g. a 40% group beats three 20% groups who cannot cooperate).
One way around this, maybe, is a Negative Vote (allowing a united method opposition), but that has foreseeable issues, especially if Negative Voting is as efficient or more efficient than Positive.
A 40% group will (and IMO should) beat 3 non-cooperating 20% groups in pretty much any voting system.
tl;dr: I disagree. Other than first-past-the-post, which is terrible, 3 non-cooperating 20% groups with similar preferences will and should beat a co-operating 40% group. This is also true for quadratic voting.
Here is a detailed scenario matching your 40/20/20/20 example. Suppose we have the following voters:
Alice prefers apples to other fruit, and strongly prefers fruit to vegetables.
Bob prefers bananas to other fruit, and strongly prefers fruit to vegetables.
Charlie prefers cherries to other fruit, and strongly prefers fruit to vegetables.
Yasmine prefers yams to other vegetables, and strongly prefers vegetables to fruit.
Zebedee prefers zucchini to other vegetables, and strongly prefers vegetables to fruit.
Y and Z are able to coordinate. A, B, and C are not. This is not because Y and Z are more virtuous, nor because vegetables are better than fruit. It’s for the prosaic reason that A, B, and C do not share a common language. All voters have similar utility at stake, for example Charlie is not allergic to yams.
In a first-past-the-post voting system, with apple, bananas, cherries, yams, and zucchini on the ballot, Y and Z can coordinate to gets yams and zucchini on alternating days. This is good for Y and Z, but does not maximize utility.
However, in a (good) ranked voting system, we instead get a tie between apples, bananas, and cherries, which we break randomly. This is good for A, B, and C. Proportional representation would get a similar result, assuming that the representatives, unlike the voters, can coordinate. Approval voting would get a similar result in this example.
Quadratic voting calculations are a bit harder for me, and I had to experiment to get a near-optimal voting strategy.
Let’s suppose that A votes as follows:
$30 for Apples (+5.48)
$15 for Bananas (+3.87)
$15 for Cherries (+3.87)
$20 against Yams (-4.47)
$20 against Zucchini (-4.47)
B and C vote similarly but according to their own preferences. Naively this maps to A preferring apples to other fruits (by $15) but strongly preferring fruits to vegetables (by $35). I don’t have good intuitions for whether A would vote this way in practice.
Meanwhile Y and Z coordinate and vote as follows:
$10 against Apples (-3.16)
$10 against Bananas (-3.16)
$10 against cherries (-3.16)
$70 for yams (+8.36)
On alternate days Y and Z coordinate to vote for zucchini, as in the first-past-the-post coordination example. Again, I don’t have good intuitions for how they should vote, but I experimented with a dozen strategies and this one was best I found.
The results are:
Apples, Bananas, and Cherries: 6.90
Yams: 3.32
Zucchini: −13.4
So although coordination/collusion allowed Y and Z to boost their effective voting power, they are not able to enforce rule-by-minority in this example.
OK, by “cannot cooperate”, you meant “unable to coordinate communication about their already-shared values” rather than “unable to agree to support each others’ unrelated interests”. Got it.
Okay, I accept your point that a cooperating 40% group will beat three non-cooperating 20% groups with unrelated interests in pretty much any voting system. That doesn’t change whether A+B+C are physically incapable of communicating, or they lack sufficient trust to make an agreement stick, or there is a law against voting agreements that they are following (and Y+Z are not), or something else.
So it’s not the case that QV is vulnerable to collusion/cooperation when other voting systems are not. I think the remaining debate is whether QV is more vulnerable, or vulnerable in a worse way. I’m not sure what the answer is to that.
(I’m not brgind or EdgyCam, I can’t speak to what they meant)
The impression I got, was that collusion between likeminded people created an “indirect democracy” where causes supported by the most people could most efficiently advocate their position.
If that is the case, then this system does punish parties that are less willing (or able) to cooperate, which could feasibly be a bad thing, if it means that unpopular results occur because one side is less nuanced on it’s position (e.g. a 40% group beats three 20% groups who cannot cooperate).
One way around this, maybe, is a Negative Vote (allowing a united method opposition), but that has foreseeable issues, especially if Negative Voting is as efficient or more efficient than Positive.
I don’t understand. A 40% group will (and IMO should) beat 3 non-cooperating 20% groups in pretty much any voting system.
A system that encourages groups to work together for their collective benefit seems like a solution for that situation, not a problem.
tl;dr: I disagree. Other than first-past-the-post, which is terrible, 3 non-cooperating 20% groups with similar preferences will and should beat a co-operating 40% group. This is also true for quadratic voting.
Here is a detailed scenario matching your 40/20/20/20 example. Suppose we have the following voters:
Alice prefers apples to other fruit, and strongly prefers fruit to vegetables.
Bob prefers bananas to other fruit, and strongly prefers fruit to vegetables.
Charlie prefers cherries to other fruit, and strongly prefers fruit to vegetables.
Yasmine prefers yams to other vegetables, and strongly prefers vegetables to fruit.
Zebedee prefers zucchini to other vegetables, and strongly prefers vegetables to fruit.
Y and Z are able to coordinate. A, B, and C are not. This is not because Y and Z are more virtuous, nor because vegetables are better than fruit. It’s for the prosaic reason that A, B, and C do not share a common language. All voters have similar utility at stake, for example Charlie is not allergic to yams.
In a first-past-the-post voting system, with apple, bananas, cherries, yams, and zucchini on the ballot, Y and Z can coordinate to gets yams and zucchini on alternating days. This is good for Y and Z, but does not maximize utility.
However, in a (good) ranked voting system, we instead get a tie between apples, bananas, and cherries, which we break randomly. This is good for A, B, and C. Proportional representation would get a similar result, assuming that the representatives, unlike the voters, can coordinate. Approval voting would get a similar result in this example.
Quadratic voting calculations are a bit harder for me, and I had to experiment to get a near-optimal voting strategy.
Let’s suppose that A votes as follows:
$30 for Apples (+5.48)
$15 for Bananas (+3.87)
$15 for Cherries (+3.87)
$20 against Yams (-4.47)
$20 against Zucchini (-4.47)
B and C vote similarly but according to their own preferences. Naively this maps to A preferring apples to other fruits (by $15) but strongly preferring fruits to vegetables (by $35). I don’t have good intuitions for whether A would vote this way in practice.
Meanwhile Y and Z coordinate and vote as follows:
$10 against Apples (-3.16)
$10 against Bananas (-3.16)
$10 against cherries (-3.16)
$70 for yams (+8.36)
On alternate days Y and Z coordinate to vote for zucchini, as in the first-past-the-post coordination example. Again, I don’t have good intuitions for how they should vote, but I experimented with a dozen strategies and this one was best I found.
The results are:
Apples, Bananas, and Cherries: 6.90
Yams: 3.32
Zucchini: −13.4
So although coordination/collusion allowed Y and Z to boost their effective voting power, they are not able to enforce rule-by-minority in this example.
OK, by “cannot cooperate”, you meant “unable to coordinate communication about their already-shared values” rather than “unable to agree to support each others’ unrelated interests”. Got it.
Okay, I accept your point that a cooperating 40% group will beat three non-cooperating 20% groups with unrelated interests in pretty much any voting system. That doesn’t change whether A+B+C are physically incapable of communicating, or they lack sufficient trust to make an agreement stick, or there is a law against voting agreements that they are following (and Y+Z are not), or something else.
So it’s not the case that QV is vulnerable to collusion/cooperation when other voting systems are not. I think the remaining debate is whether QV is more vulnerable, or vulnerable in a worse way. I’m not sure what the answer is to that.
(I’m not brgind or EdgyCam, I can’t speak to what they meant)