DRC - Private Voting for dYdX Governance [2-month Trial]

BristolBlockchain · June 4, 2023, 10:17pm

This proposal was inspired by [ARFC] Private Voting for Aave Governance [2-month Trial], a collaboration between (@LBS Blockchain Society), Michigan Blockchain (@Michigan_Blockchain), and Flipside Crypto ( @fig ,@fig ).

Simple Summary:

Facing challenges of early vote influence, potential manipulation by large stakeholders, and voter apathy in the current open voting system, we propose the integration of Shutter Governance’s shielded voting mechanism to enhance the privacy, and engagement.

Shielded voting allows voters to maintain vote confidentiality, hide their votes and vote counts during the duration of the voting process and reveal their choices and results after the vote concludes. This encourages unbiased voting and increases community participation.

This proposal seeks to integrate Shutter Governance’s shielded voting mechanism into dYdX’s governance system for a 2-month trial period.

Motivation:

The dYdX governance process currently utilises a combination of off-chain (DRC) and on-chain (DIP) voting. The initial stage takes place off-chain, using Snapshot for sentiment signalling for future on-chain DIPs and binding votes for variables controlled off-chain.

However, a recurrent observation is that many votes are primarily decided by a small number of participants, with limited divergent viewpoints. This concentration of influence could potentially compromise the goal of achieving a truly decentralised and inclusive decision-making process.

The inspiration for this proposal comes from the successful implementation of Private Voting within the Aave Governance system, as outlined in the [ARFC] Private Voting for Aave Governance [2-month Trial] proposal. The adoption of private voting brought notable improvements to Aave’s governance, encouraging broader participation and reducing the potential for undue influence by major token holders.

To highlight the potential benefits of adopting Shutter Governance’s shielded voting, we turn to two examples put forth by Shutter:

“Consider a contentious vote in which a minority is pushing the poll early in one direction. Voters that don’t have a strong opinion already formed might see this and think the outcome is already decided. Thus they’ll be discouraged from voting and also from researching/forming an opinion. This could then lead to the vote going in favor of the minority due to voter apathy, which would not be a good outcome, given that the goal of the poll is to represent the majority opinion.”

With Shutter’s shielded voting mechanism, the trend of the vote remains undisclosed until the voting period ends. This feature can motivate users to research, form an opinion, enhancing overall voter turnout and representation.

“Consider a whale with malicious intent observing and waiting for a vote to play out. Only to come in at the last minute, borrowing/buying just the right amount of tokens needed to sway the vote. And doing this at a time when there’s no more time for the rest of the community to react.”

By maintaining vote confidentiality until the end of the voting period, shielded voting can reduce the potential for such manipulative tactics. The whale would not have visibility into the vote’s current state, and hence would not know how many votes might be required to influence the outcome.

Shielded voting has four keys benefits:

Prevents voter intimidation or coercion
Reduces Herd Mentality: Promotes independent decision-making and discourages voters from being swayed by the trends or decisions of early voters.
Upholds Voter Confidentiality: Safeguards voter privacy by keeping individual voting preferences undisclosed, until the voting period ends.
Ensures Voting Process Security: Enhances the overall integrity of the voting process, making it more difficult for malicious actors to interfere or manipulate the outcome.

Please look at some ongoing proposal to see the implementation: Aave - Snapshot

Also, a demo shows that a minimal change in the Snapshot UI: Shielded voting is live! — Snapshot Labs

Risks/Concerns:

A notable comment form Figue, Paladin:

“Hey guys, Figue from Paladin. We’ve been studying private voting quite a lot internally.*

Our current stance on it is it is more dangerous than useful in the sense that it totally cancels out discussions during the vote.

In some very specific cases like elections, this could be relevant, but it also means it won’t be possible to detect collusion until vote is over.

Shielded voting will also prevent to know the stance of actors until the end of the vote unless they have publicly expressed it. This means that all the discussions with VCs who haven’t followed the governance closely are impossible. If this happened, Uniswap wouldn’t have been deployed on Uni v3, since A16Z nuked the vote and no one would have known about it until too late.

Removing the live outcome of the vote kills part of its interest.”

Reference: Figue’s comment

Specification:

Community Discussion: This proposal should be open for discussion among the dYdX community. If the majority of the community members express their favour towards the implementation of shielded voting, the proposal will proceed to the next step.
Proposal (DRC) Creation: Anyone holding more than 10,000 dYdX tokens please contact Bristol Blockchain Society to create this proposal on Snapshot. Due to the lack of voting power, Bristol Blockchain Society alone cannot initiate this proposal. We encourage stakeholders to come forward in order to represent the interests of the community.
Implementation: If the vote passes with a majority, the next step would be to initiate the process of implementing shielded voting.

i. An admin or controller of a space and go to the settings page
ii. In the “Voting” section, choose Shutter as a privacy option and click “Save”
iii. Once you’ve done this change, all proposals from then on will use shielded voting.

The community may disable this feature anytime.

The current list of Snapshot admins is as follows:

0x410210ef65F69E1ac515974604323794FAb6ae1c

References:

[ARFC] Private Voting for Aave Governance [2-month Trial]:

Shutter and Snapshot:

dYdX:

DIP Proposal Lifecycle

Copyright

Copyright and related rights waived via CC0

0xcchan · June 5, 2023, 12:08am

Thanks for suggesting this! It’s definitely a novel and important concept in governance.

My primary concern would just be that if this trial were successful, can Shutter be integrated with dYdX v4 (COSMOs) as a continuation? Hopefully this won’t be a one off idea.

And in this case, how does the team intend to define the success criteria (will it be the same as lbsblockchain’s suggestion?)

Finally, will the team be able to shed some insights and learnings/improvements on this implementation on AAVE since it’s been nearly a month?

Thank you!

Kagan · June 5, 2023, 6:41am

Hi @BristolBlockchain,

This is an intriguing proposal, and I fully support your endeavour to tackle issues of early vote influence, potential manipulation by large stakeholders, and voter apathy.

Given their apparent benefits, I’m surprised to see that such mechanisms aren’t already prevalent in more decentralised governance systems.

However, I also believe that we should consider this innovation as part of a larger, more comprehensive restructuring of the governance system. There are many facets to this process, and shielded voting could potentially play a crucial role in many of them.

In this light, I’d like to encourage everyone to think about the broader structure this innovation could fit into. For instance, a council system could potentially be a beneficial restructuring of the governance model, and this shielded voting mechanism might be seamlessly incorporated into such a system. You can find some ideas for such a structure on this forum thread.

Thank you for the thought and effort you’ve put into this proposal. It’s initiatives like this that drive forward our collective efforts to improve our decentralised governance processes.

RealVovochka · June 5, 2023, 10:51am

In my personal opinion, such a voting system has its drawbacks without offering sufficient benefits. I agree with “Figue’s” position.
If anything should be changed in the voting system, it is the attitude of large holders towards transparency and clear communication of the reasons why a vote was cast in a certain way. Given the current distribution of tokens, this system essentially won’t change anything

fig · June 5, 2023, 7:38pm

Ha, this is pretty cool to see in the wild!

We are honored that @BristolBlockchain is referencing our work and experimentation in Aave - and exploring another implementation and trial in dYdX.

I’ll attempt to be unbiased while considering this proposal - some important background:

Private voting is a way to promote more healthy, honest decision-making.

It removes the concern of intimidation or peer pressure – or a general lack of effort. It encourages voters to form their individual opinion versus mirroring ones of the top voters or most active ones.

This was a problem in Aave where the top voter controlled the voting power by a factor of 2x; there was a political incentive to align with this person’s vote if you were unsure or split.

Another reason we advocated for this:

To promote more contentious votes.

We believe more often than not DAO’s reach a consensus quickly.

That’s great - but there are moments in the ecosystem where delegates must vote honestly. It is healthy to have votes which result in NO, or even ABSTAIN as a leading option.

This thesis is still being tested - and we can further evaluate when Aave’s trial has finished.

Some thoughts so far from its implementation in Aave:

it forces you to form and define a clear opinion, not just defer to the largest / most passionate voter

it is more difficult to anticipate the outcome and initiate the next steps; i.e you must wait til the Snapshot vote has concluded to understand the results and begin the next steps

it’s pretty easy to turn on (and off) and creates a new, less familiar UX for Snapshot users

We have yet to tally the participation rates as the trial has not finished.

For dYdX, this could be an interesting implementation if the community desired it. The trial may have to last longer as there are infrequent Snapshot votes / to compare a larger pool of data.

Before implementing private voting, it may be best to define the role of Snapshot in V4.

Is dYdX still going to use it? Will the space remain the same?

This feels like the most important first step - then debating the merits (or lack thereof) of private voting.