CIP-48: CityDAO Conversion to Quadratic Voting

In recent Snapshots, there have been many people who hold multiple citizen NFTs having a large impact on voting outcomes.

In order to limit this, I would like to start a discussion around implementing quadratic voting into our governance system.

This will reduce the impact of large holders on votes but not restrict it completely.

I would also like to discuss using this same concept for the Parcel 0 drop as I believe those holding multiple NFTs will likely spread across wallets anyway. Using a quadratic drop formula, those who hold multiple citizen NFTs can claim a certain amount more Parcel 0 NFTs than if they held one, while discouraging them to go to the effort of spreading across multiple wallets.

22 Likes

finally, QV is coming!100% support

1 Like

Can you please provide or propose more details on the mechanics of how this would be implemented?

1 Like

Quadratic voting - Wikipedia.

This page from Snapshot describes the voting types (not to be confused with the voting strategies, which are on top of this): Voting types - Snapshot

1 Like

I am 100% supportive of moving to quadratic voting - even if its before we move into a quadratic voting with proof of humanity (I hope will be our ultimate goal). We might need to have a more robust proposal for it though. I’d be happy to help put something together.

@MemeBrains … any chance we can get Glen Weyl on the podcast and ask him for advice? Or some kind of Discord AMA?

As for some things to consider… Quadratic voting will require us to decide how many voting credits per wallet, per term. We’d have to decide the number of credits, AND the length of the term. I think ideally, we’d be able to give a sense for all citizens about how many votes would come up in the term.

Another option would be to say everyone will get 100 voice credits per 10 CIPs on snapshot. After 10 CIPs are complete, 100 more credits will be given.

How would we want to handle inactive members? let them build up voice credits even though they aren’t paying attention? Zero out the voice credits after 10 CIPs on snapshot for everyone, unless they cast at least one vote?

Also, I assume that if someone sells a citizen NFT, that the votes DO NOT go with the NFT…

2 Likes

Yea lets do this but what about KYC ? There is no point using quadratic voting unless every voter is precisely identified as person. I am fine with KYC by the way but there are people prioritizing their privacy

2 Likes

I do not believe we are there yet, but something for the future for sure

1 Like

Fungible token allows for multiple votes today; votes can be split out to multiple wallets

We’re getting close to needing KYC/PoH type solutions if we’re going to expect Sybil resistance, or at least some type of registry that identifies which tokens are owned by which individuals

1 Like

This is just the first step. A PoH solution would be next.

I don’t understand why quadratic voting requires any credit tracking at all. Can’t we just use the same tracking thing we do but instead of using 1:1 with NFTs in your account, when you vote, it does a quick square-root calculation of the number of NFTs in your account, floor it, and count that number. It doesn’t really need to be more complicated than a simple calculation in my opinion.

If the goal is a step in the right direction, then absolutely this would make sense.

However, if the goal is to ensure that this happens, it gets a lot trickier.
There’s nothing that ties N tokens to an individual, so multiple wallets = multiple votes.

Not saying square rooting the holdings isn’t a good step in the right direction (it is). Maybe the fees of transferring to different wallets or having to vote N number of times will be a deterrent, but its still relatively easy to bypass if that is the only mechanism.

1 Like

Ah, yeah, got it. That makes sense due to the way the 1155 is being used (ie citizenship is a fungible token right now).

hmm… looks like @DenverCitizen9 answered your question. But I did want to point out that it sounds like there might be two different perspectives on the value of Quadratic Voting (QV).

The first, which you and others seem to be talking about is more about decreasing (but not eliminating) the voting power of people who hold multiple NFTs. I support this perspective and think its a useful outcome of QV. In this case though, the QV formula is just applied to citizen NFTs at the start of every vote

The second perspective is more closely aligned with the ideas in the book Radical Markets, where the main beauty of QV is that everyone has equal voting power, but they can choose to spread their voting power across multiple choices (1 vote for choice A, 2 votes for choice B) or over multiple votes (I don’t vote in CIP 60, 61, or 62, but use all my vote credits for CIP 63 which I feel strongly about).

The second perspective has a lot to offer CityDAO as well, in my opinion, so I don’t want to lose sight of that. In either case though, we’ll need to figure out a mechanism for linking NFTs to individuals.

1 Like

To avoid confusion, I would like to propose we split these concepts out and not merge them into one CIP:

  1. Quadratic Voting, meaning that we determine the number of votes a single wallet has by computing the square root of the number of citizen NFTs held by that wallet, rounded down.
  2. Vote splitting, meaning that if there are more than 2 choices on a proposal, and a single voter is allowed multiple votes, then a voter can split their votes across the various choices

I believe this CIP should only discuss #1 above as we are talking about how many votes a single wallet has access to given the number of citizen NFTs they own.

Once we agree on the number of votes an individual has based on their citizen NFTs, then we can discuss how we should handle proposals with more than 2 choices. These are orthogonal concepts in my opinion and should be decided separately.

1 Like

I think we are talking about #1 - a quadratic effect where the voting power decreases based on the number of citizenships a wallet holds (the square root of citizenships, but at least 1)

FYI #2 has been trialed by the Colorado legislature, but I agree I don’t think is something we want to try and add as part of this proposal

1 Like

I am 100% supportive of moving to quadratic voting!

1 Like

Limiting monopolies without enabling plurality is simply authoritarian. In this case it’s a restriction on Tokenholders who were playing by the rules; you just don’t like the rules.

The goal of QV is to enable the expression of graduated nuance, not coercion of the rich.

How should CityDAO be thinking about this in terms of strategy/execution @will?

From RadicalxChange:

Quadratic Voting (QV) is a redesigned voting method reflecting the intensity of people’s preferences in collective decisions. It greatly mitigates tyranny-of-the-majority and factional control problems.

Voters receive budgets of “voice credits,” which they allocate to different questions on the ballot to signal the intensity of their conviction. Their voice credits convert to “counted votes” according to their square root. So if you put one voice credit on an issue, that is one vote; four credits are two votes; nine credits are three votes, and so on.

QV is not treating each membership as a voice credit but rather allocating credits proportional to held membership tokens, which diminishes whale influence by enabling individuals to outvote whales en masse.

1 Like