Temp Check: Adopting Delegated Voting

[hirokennelly]

Title: Temp Check: Adopting Delegated Voting

Author: HiroKennelly

Squad: Governance Department

Date submitted: November 6, 2023

Working Doc: Delegated Voting Temp Check

Summary

This is a temp check on whether BanklessDAO should adopt a delegated voting model. As an extension of consent-based decision making, delegated voting will enable broader participation in the DAO’s governance.

Background

BanklessDAO has one of the most widely-held social tokens, yet, on any given Snapshot proposal, the votes cast represent only a small percentage of the BANK tokens in circulation. Delegated voting can help to increase the amount of BANK cast in each vote, providing an avenue broader BANK-holder participation in the DAO’s decision-making process.

Delegated voting is the act of giving trusted, governance-minded community members the ability to vote with another token holder’s voting power. The basic idea is that those who do not have an interest in participating in a DAO’s governance process can delegate their vote to those who do. Delegation can be changed or revoked at any time and would not be mandatory.

Delegated voting is used by most major projects, including Compound, Aave, Uniswap, and Maker, and more recently, Optimism and Arbitrum.

Specification

Onchain or Offchain Voting?

There are many options for delegated voting, but the core distinction rests on whether the platform is onchain or offchain. Onchain voting is the gold standard, as it leverages the blockchain’s attributes of transparency and certainty, but many organizations choose offchain voting because its traditionally been more cost effective – although this is changing with the rise of voting protocols on L2s.

For onchain voting, the Governor contracts developed by Compound are the most popular, and used by most major DeFi projects, including Uniswap, Compound, and PoolTogether. There are also independent blockchain-based platforms, such as those created by Agora and Aragon.

There are really no direct competitors to Snapshot, and it remains the most developed offchain voting solution.

To access onchain and offchain voting, Most projects rely on a mix of Snapshot, Boardroom, Tally, Commonwealth, and Agora. These frontends enable people to both be informed about and participate in ecosystem governance.

Considering our treasury and revenue streams, it would be premature to move to onchain voting, unless it could be executed on a Layer 2. Delegated voting is also available through Snapshot, including the use of the Governor Bravo contract.

TL;DR

We can do delegated voting onchain or offchain.

Sybil Resistance

An important consideration in collective governance is the effects of Sybil attacks and to take measures to minimize their impact. A Sybil attack is when a single entity creates multiple identities to gain disproportionate and unfair influence in the outcome of an election. To mitigate this possibility, we could consider adopting comprehensive Sybil resistance measures like those offered by BrightID, Proof of Humanity, and Gitcoin Passport as part of the implementation of delegated voting.

Delegate Compensation

After years of iteration, many protocols incentivize their delegates’ continued participation by offering various financial rewards for governance work. Likewise, BanklessDAO could consider adopting incentives to drive governance participation by delegates, incentives that could be in the form of BANK or tlBANK payments and linked to a particular participation threshold for Forum and the voting platform of choice.

Delegator Compensation

We’ve seen some protocols, such as Optimism, reward token holders with additional tokens if they delegate their voting rights. BanklessDAO could similarly reward those who delegate by airdropping them additional BANK or granting access to an exclusive Discord channel, among other options.

Obstacles to Implementation

The main obstacle to implementation involves the technical process of delegating votes to a different DAO member. There are a number of ways to overcome this:

1. Build a custom frontend to enable delegation.
2. Use the ‘write’ to contract function in Etherscan.
3. Use a third-party platform, such as Agora or Aragon.

These options have efficiency and cost tradeoffs, as well as practical educational hurdles.

KPIs

Successful delegated governance would include:

• High participation by delegates on Forum.
• 80% vote rate for delegates.
• Roadmap to make delegated governance more Sybil resistant.
• 30% of total circulating supply of BANK delegated by Season 11.

Polls

Should BanklessDAO adopt a delegated voting framework?

• Yes
• No
• Abstain, please comment

If BanklessDAO adopts delegated voting, the percentage of circulating BANK delegated to any one delegate should not exceed:

• 10%
• Greater than 5%
• Less than 5%
• No limit
• Abstain, please comment

Should BankessDAO adopt measures to minimize Sybil attack vectors when determining eligibility to serve as a delegate, such as BrightID, Proof of Humanity, or Gitcoin Passport?

• Yes
• No
• Abstain, please comment

Should we incentivize BANK token holders to delegate?

• Yes
• No
• Abstain, please comment

If yes, how should BANK holders be incentivized to delegate their BANK?

• Airdropped BANK rewards
• Access to a gated Discord channel
• Other, please leave in comments
• Abstain, please comment

Should delegates be rewarded for their governance work?

• Yes
• No
• Abstain, please comment

Should delegates be rewarded in tlBANK or BANK?

• tlBANK
• BANK
• Abstain, please comment

[Ornella]

gm @hirokennelly thank you for the proposal.

Here are my concerns for voting ‘no’. I don’t believe it is fair that someone that is not involved in the operations or function of the DAO today, has the proper information and context to select a delegate that can vote for them.

I feel that this process could work in the future, but at the moment we have a low number of contributors involved in the DAO that a ‘whale’ with X amount of BANK could eventually move forward with a decision that its not the best for the DAO.

I guess I worry about the fact that wallets with millions of BANK can vote right now and the community has no idea of who that person is due to the nature of governance right based on token holdings.

Maybe you can bring some clarity to my thoughts and concerns. That would be very much appreciated.

Thank you!

[Trewkat]

@hirokennelly

If BanklessDAO adopts delegated voting, the percentage of circulating BANK delegated to any one delegate should not exceed:

The wording of this question creates ambiguity, as do the options, so the end result is that only one viable option is included.
The options ‘greater than 5%’ and ‘less than 5%’ do not make sense so I’m not sure why they have been included along with abstain.
How can a percentage not exceed greater than 5%? That essentially means it can’t be higher than anything above 5%, which means it could be 100%.
How can it ‘not exceed no limit’?
Given this is a temp check there could have been multiple options or ranges given as options which would have resulted in a much clearer indication of community sentiment.
e.g. should not exceed 5%, 10%, 15%, 20%, and so on.

[Trewkat]

Some further questions occurred to me which influence whether I would vote yes to the first poll:

• Is it envisioned that delegates would self-nominate, or be elected and need to accept nomination?
• Will it be necessary for all token holders to delegate, if only to themselves, in order to vote? This could have the opposite effect on voting participation.
• Would delegates need to hold a particular percentage of BANK themselves?
• Will there be minimum duration as L1 before someone could act as a delegate?
• Would delegates be expected to communicate their voting intention and rationale prior to each vote?
• Why is the participation KPI related to Forum when token voting happens on Snapshot?

[links]

Personally I am for delegated voting, and I think we should take a small, iterative approach to this.

One assumption that affects my votes: in the absence of activity incentives, token-weighted voting is pretty sybil-resistant on its own. Those who have tokens either bought or earned them, and so there’s no outcome-changing reason for people to create multiple wallets. Of course this falls apart when you start to reward delegates/delegators.

For that reason I think we should start with a very small experiment where we simply allow our token-holders to delegate if they wish and make it easy for them to do so, and take some baseline metrics (i.e. BANK/vote, wallets/vote, etc). Then we can add in other incentives and see how they change the baseline.

I also think it’s too early to try to set a cap on delegation. What does it matter if a single delegate has >10% of the BANK or even >50%…will it increase or decrease participation? It’s impossible for us to know without tracking baseline metrics (above). Let’s start with something small and if vote centralization becomes and issue we can add additional rules.

[hirokennelly]

gm gm! Thanks for sharing your concerns!

@Ornella

I don’t believe it is fair that someone that is not involved in the operations or function of the DAO today, has the proper information and context to select a delegate that can vote for them.

Typically delegate statements help with that, and imo, BANK holders should have the autonomy to delegate to whomever they choose, but def understand where you’re coming from.

@Ornella

I feel that this process could work in the future, but at the moment we have a low number of contributors involved in the DAO that a ‘whale’ with X amount of BANK could eventually move forward with a decision that its not the best for the DAO.

I guess I worry about the fact that wallets with millions of BANK can vote right now and the community has no idea of who that person is due to the nature of governance right based on token holdings.

I actually think this is where delegated voting helps, because rn we don’t know who is voting unless they have reverse records set on an ENS, but with delegated voting, we’d know.

As for whales having outsize influence, that’s why I personally would like to see a cap on max delegation. We ofc have whales rn who don’t vote (Genesis members) and it’s by their good graces that we have functional governance. With delegated voting, we can help to minimize the future impact of now-inactive whales.

[hirokennelly]

gm gm! Agreed the wording is terrible idk why I thought it made sense. You even tried to warn me!

For future temp checks, I’ll be more clear, something like greater than 10%, 5-10%, under 5%, or no limit. Is that better?

[hirokennelly]

gm gm again! Many of these questions should make it into the next temp check, thanks for them Trewkat.

@Trewkat

Is it envisioned that delegates would self-nominate, or be elected and need to accept nomination?

IMO, self nomination is appropriate

@Trewkat

Will it be necessary for all token holders to delegate, if only to themselves, in order to vote? This could have the opposite effect on voting participation.

I would def prefer to not require self-delegation to vote. Agreed it could have a negative affect on participation.

@Trewkat

Would delegates need to hold a particular percentage of BANK themselves?

I would personally say no since we should have the freedom to choose delegates, and related to your next question:

@Trewkat

Will there be minimum duration as L1 before someone could act as a delegate?

I do think any L1 should be eligible to be a delegate, but perhaps limiting to L2s would be better.

@Trewkat

Would delegates be expected to communicate their voting intention and rationale prior to each vote?

I would want to include an option for rationale but not make it mandatory (I’m not sure the tech supports that tbh - in Agora for ex, it’s optional). My sense is that delegates who provide rationales make for better delegates and would be hopeful that potential delegators would choose such delegates.

@Trewkat

Why is the participation KPI related to Forum when token voting happens on Snapshot?

The “80% vote rate for delegates” language should be clarified to mean on Snapshot. I do think Forum participation should be a KPI too tho, as it’s an essential part of governance.

[hirokennelly]

gm gm! I generally agree with measured, iterative approaches to things like this, but disagree about setting limits on delegation. Those should be set from the outset to limit outsize influence and impact. IMO it makes a very big difference if one delegate holds 50% of BANK voting power. This is related to the sybil resistance part, which is that if you cap the amt of delegated BANK, folks could spin up another wallet to get around that. Granted, that’s a long game, but folks are happy to play a long game if the rewards justify the efforts.

[homie]

Regarding this…

What would be the reason to incentivize people to delegate?

a pessimist view is that you’d be paying people to give away their decision making power BUT I can’t see that opinion coming from you, @hirokennelly.

So. What would the optimist view of incentivizing delegation be?

[homie]

Hm. Okay I get that.

Where would the Dormant bank be? Is it because the bank that is sitting in wallets, with contributors that have not participated in the past, would be now put to use on a sense?

[hirokennelly]

yeee i just mean BANK owned by holders who don’t participate on Forum or Snapshot

[homie]

sure, i understand what you’re saying here.

[hirokennelly]

uh oh i deleted this by accident lmao. it said i get the pessimistic take, but the optimistic take is that it incentivizes greater gov participation by getting dormant BANK (not voting) off the sidelines and into the decision-making process

[homie]

Thanks!

I get the idea behind incentivizing people who may not have had much interest in governance before.

Very curious as to how you would incentivize. What would be the reward? Would that be “wasting” (for lack of a better word) bank? I suppose it would be better to reward in tlbank IF you wanted to reward.

[hirokennelly]

Could be one of the below, or something else. Curious what people think tbh.

[0xZFi.eth]

@hirokennelly

If BanklessDAO adopts delegated voting, the percentage of circulating BANK delegated to any one delegate should not exceed:

The options are confusing as worded, so I’m abstaining from a selection here.

A 5% limit passes a gut check for “sufficiently decentralized” but I haven’t given this threshold any serious consideration or projection testing. 10% seems risky. 10% of circulating VP could be more than the total current voter participation rate. Do we have the exact number for what percentage of the circulating VP is currently participating on Snapshot votes, on average?

@hirokennelly

If yes, how should BANK holders be incentivized to delegate their BANK?

Abstaining here because I voted against incentivized delegation. I suppose there wouldn’t be any way to prevent privately incentivized delegation, though, so maybe it would be better to get ahead of the system being privately gamed and globally gamify it from the beginning?

To whom a holder delegates might be incentivized by the delegates themselves (delegate X has promised to pay me Y amount for my VP in their pursuit of amassing personal power in the ecosystem) or through an individual values alignment (delegate X has best represented my opinions on past matters and I want their voice to represent mine in future votes).

Either way, the holder delegating their VP is being rewarded in a free governance service, in their own interests. Additional external incentive to delegate one’s VP should not be needed, as incentives are already aligned.

@hirokennelly

Should delegates be rewarded for their governance work?

Yes. This is where the time-consuming in-depth analysis work gets put in.

The investment to immerse oneself in the constant discussions and stay up to date with governance and operational changes is no small ask. BANK holders who don’t want to take the time to do that themselves may prefer to offload that responsibility to a delegate. This is a service being provided to the disengaged BANK holder. Given that it’s healthier for the DAO at large if more of the circulating VP is engaged in regular voting, an incentive alignment exists for the DAO to reward delegates.

[hirokennelly]

gm gm ser, thank you for the questions/comments and apologies for the late response, was traveling.
Re: Caps on Delegation:

@hirokennelly

If BanklessDAO adopts delegated voting, the percentage of circulating BANK delegated to any one delegate should not exceed:

@0xZFi.eth

The options are confusing as worded, so I’m abstaining from a selection here.

Very confusing as it turned out, see previous comment to Trewkat:

@hirokennelly

For future temp checks, I’ll be more clear, something like greater than 10%, 5-10%, under 5%, or no limit. Is that better?

To answer your question:

@0xZFi.eth

Do we have the exact number for what percentage of the circulating VP is currently participating on Snapshot votes, on average?

It looks like the last handful of Snapshot votes have averaged around 15m-20m BANK cast. Total circulating supply is 825m, so votes cast are about 2% of circulating supply.

Also appreciate your thoughts on incentivizing delegation and delegates. Thank you!!

[0xZFi.eth]

Yikes! Only 2% of circulating supply is voting right now?! I’m suddenly VERY curious where that other 98% is distributed…

Is there a rabbit hole somewhere around here I can jump down? I’d like to see a map of BANK token and tlBANK NFT distribution. Has anyone gotten a head start on this in Governance? or maybe Treasury?

[Icedcool]

I want this too!

We will be getting some analytics together for it soon.