Governor Update Proposal: Removing Abstain Count from Quorum

Technical Proposals 📃 Protocol Upgrade
1

Executive Summary

This proposal outlines a targeted improvement to the Optimism Governor contract by removing Abstain votes from quorum calculations. The change resolves an edge case where Abstain votes could unintentionally reduce the vote threshold needed for proposals to pass. It has been fully audited, deployed, and is ready for governance approval.

Motivation

We are introducing this change to correct an edge case in the Governor logic where Abstain votes, despite being neutral by design, can unintentionally impact quorum dynamics. Specifically, when quorum has not yet been reached, each additional Abstain vote increases the total number of votes cast, which in turn lowers the number of Yes or No votes required for approval.
This has the unintended effect of making it easier to pass or reject proposals when there are a large number of Abstain votes. By excluding Abstain votes from the quorum calculation, we ensure that only active Yes and No votes contribute to reaching quorum, preserving the integrity of proposal outcomes and aligning voting behavior with intent. This change does not impact the outcome in the scenario where a majority of Abstain votes are cast, as the proposal would not be implemented under either logic.

Specifications

Contract Changed

The Optimism Governor would be changed as part of this upgrade. Please review the pull request here:

The deployed code can be found here: OptimismGovernor | Address 0x42436bB7BEA1E1E2De03f1223e2A3e0557F606E3 | OP Mainnet Etherscan There are no state changes to any governance contracts in this upgrade.

Security Considerations

This change has been audited by Trust and there were no issues identified given that it’s a minimal change which doesn’t affect the rest of the contract logic. In the unlikely scenario that we’ll need to make any changes to address issues, we’ll update this thread to inform the community on next steps.

Link to audit report: https://drive.google.com/file/d/1Q0cbH5CtnEAXKatUXdXfjxXJiwMmdal4/view?usp=drive_link

Impact Summary

Minimal to no impact.

Action Plan

If approved by governance, the Optimism Foundation admin will set the new implementation of governor proxy at 0xcDF27F107725988f2261Ce2256bDfCdE8B382B10 to new implementation deployed at OptimismGovernor | Address 0x42436bB7BEA1E1E2De03f1223e2A3e0557F606E3 | OP Mainnet Etherscan

If a critical security issue is discovered before upgrading, Agora will collaborate with the community to extensively communicate that the upgrade will no longer occur.

Conclusion

This Governor update improves quorum accuracy by excluding Abstain votes from quorum calculations. The change has been carefully audited, poses minimal risk, and has already been deployed to a new implementation. If approved, the Optimism Foundation will proceed with the upgrade with no expected downtime or impact to users.

8 Likes
2

Adding a procedural note that this upgrade will follow the same protocol upgrade process as the Collective has followed before. The new governance process for Protocol Upgrades will begin August 1st, once a new Developer Advisory Board has been elected. That means this proposal will require the usual 76% approval in the Token House, followed by a one-week veto period in the Citizens’ House (subject to veto by the Season 7 Citizens.)

Pending delegate approvals, this proposal will go to vote in special voting cycle 39a

3 Likes
3

I’m an Optimism delegate with sufficient voting power and I believe this proposal is ready to move to a vote.

4

I’m an Optimism delegate with sufficient voting power and I believe this proposal is ready to move to a vote.

5

If we had this upgrade implemented before this vote:

What would be the outcome of it? “No Quorum achieved?” “Approved” “Defeated”

Sorry, I need to fully understand this upgrade with an outcome example. Are there any other cases in the past where the vote would have reached a different outcome if this upgrade had been implemented before?

For Foundation: What is the difference between having a proposal defeated and not reaching quorum? Given that we are moving to optimistic approval, I don’t want to make proposals “easier” to pass

6 Likes
6

Cyber, as a member of the chain delegate program with sufficient voting power, believes this proposal is ready to move to a vote.

Removing abstain votes from quorum calculations makes sense.

7

In the vote you’ve linked to, there would have been no change to the outcome as the approval threshold is, and has always been, calculated as a percentage of yes/no votes (abstain votes are not considered.) Abstain votes have historically been counted in quorum, but this vote far surpassed quorum, even without counting any of the Abstain votes. This upgrade covers an edge case that would only occur if there is a majority of abstain votes, which occurs very infrequently. The edge case this upgrade protects against has never occurred in Optimism governance and so this upgrade, if applied retroactively, would not change the outcome of any historical proposals.

There is no practical difference between a proposal being defeated and not reaching quorum. The outcome is the same - the proposal is not implemented. This upgrade does not make it easier to pass proposals, it protects against a scenario in which a majority of abstain votes could result in a scenario where a proposal passes or fails by a lower threshold than intended.

7 Likes
8

The SEEDGov delegation, as we have communicated here, being an Optimism delegation with sufficient voting power, we believe this proposal is ready to move towards a vote.

9

I am an Optimism delegate with sufficient voting power and I believe this proposal is ready to move to a vote.

10

We are an Optimism Delegate with sufficient voting power and we believe this proposal is ready to move to a vote.

11

I am an Optimism delegate with sufficient voting power and we believe this proposal is ready to move to a vote.

12

This proposal makes the edge case scenario it is trying to address worse, please vote AGAINST this proposal

(Revised with AI to make it clearer)

TL;DR

Vote AGAINST removing Abstain votes from quorum.
This change increases the risk that bad proposals pass by discouraging “No” voters from showing up early.

:brain: Why the Proposal Backfires

Current Logic Proposed Logic What Really Happens
Yes + No + Abstain count toward quorum Yes + No count only (Abstain is ignored) If quorum is almost reached, extra No votes can push the proposal over quorum — helping it pass. Rational No voters will therefore wait until the last minute (or never vote), making proposals look better than they actually are.

:warning: Perverse Incentive for “No” Votes

Quorum not reached yet?

  • A “No” vote helps the proposal hit quorum.
  • The rational move is to stay silent until the end.

:cyclone: Late-Voting Spiral

When early “No” votes disappear:

  • Later voters see mostly “Yes.”
  • Humans herd.
  • More “Yes” votes follow.
  • A proposal that should fail squeaks through.

:chart_decreasing: Signal Distortion

Governance loses a neutral signal.
Right now, Abstain means:

“I showed up, but I’m not for or against.”

Removing Abstain from quorum makes it equivalent to not voting at all — which is not the same thing.

:white_check_mark: Better Design Patterns (Ranked)

Design Quorum Calculation Rationale
Best Yes + Abstain Encourages participation without punishing “No” votes
Okay Yes only Simple; treats quorum as pure turnout for support
Acceptable Yes + No + Abstain (with low quorum threshold) Works if quorum is rarely a deciding factor
Worst (this proposal!) Yes + No Creates the “hide your No vote” dynamic

:compass: Bottom Line

Keeping the status quo is safer than this change.
If we want to fix the edge case, a better solution would be to remove “No” votes from quorum — not Abstain.

Until then, vote AGAINST this Governor update to preserve balanced incentives and clearer signaling.

NOTE:

In other contexts, removing Abstain from counting towards quorum would make perfect sense, but in a DAO context with public voting & real time results, it is a bad idea.

6 Likes
13

Thank you for raising this concern, @Griff!

Agora Team,
On behalf of the ACC, we would appreciate hearing your thoughts on the issue raised by Griff. In order for us to support this proposal, we believe the concern should be properly addressed.

cc: @kent @yitong

2 Likes
14

Will do, sorry delayed given EthCC. Bringing in @lavande and Josiah too as they were championing on this from the foundation side.

1 Like
15

We will let @lavande and the Foundation continue to chime in but as we were thinking about this change we supported it for the following reasons:

Abstain still matters, it just shouldn’t move the outcome and given the edge case, specifically in OP where you could have a large number of Abstains hit quorum and then a very small number of Yes or No, move the vote. The goal is to count and incentive active participants. Abstain doesnt mean invisilbe, the vote is still there, and can be used a conflict of interest signal, no opinion signal, but it lets the “fight” happen between the Yes and Nos.

To be super clear, we are not remove Abstain votes, they are very important to a functioning governance system, we are just removing them from the Quorum calc.

You’re right that, under a Yes + No quorum, an early No ballot can tip us over the turnout threshold. In theory that gives a strategic ‘wait-and-see’ incentive to some No voters.

Imagine this scenario:
Scenario (token supply = 100, quorum = 40 votes)

Snapshot 1: Yes = 35 No = 4 (Quorum = 39 → proposal undecided)

So now, there could be a No, that will hold out and not vote, because they don’t want to tip the quorum, but, if they do go in an vote

Yes = 35 No = 5 (Quorum = 40 → proposal now decided)

However, the approval threshold will move and change accordingly too, ie: the denominator for approval threshold will decrease with each new no. So now the Yes votes will have to carry more of the vote to overpower the Nos.

We don’t agree with your ranked design patterns either, the main goal for Quorum is to make sure that the Yes and Nos have a representative place where they can make decisions. Both would need to be counted as is typical in corporate and traditional governance systems too.

You do raise a valid concern around hidden No votes for sure, but we still believe that the tail risk that @lavande originally designed this change for is worth the change.

3 Likes
16

The big difference is that in typical in corporate and traditional governance systems, real time results are not shown. It is a different context.

The herd mentality is too strong in this context… Disincentivizing “No” votes in a DAO context is a bad design decision.

Removing “Abstain” from Quorum further exasperates the edge case where quorum is an issue (stated as the reason change).

In the Current Design:

If there are a lot of Abstains, and abstain counts towards quorum than both Yes’s and No’s are at an equal footing… People will Vote no without any fear because Quorum will be reached.

In the Proposed Design:

If there are a lot of Abstains, and abstain doesn’t count towards quorum than Yes’s have an advantage over No’s… People will hold back their No votes hoping the vote fails via quorum… but then this makes the herd have a unfair bias towards YES…

IMO its more dangerous to give extra advantage to the YES voters, as that will bring change and risk… We shouldn’t make low turn out votes more likely to pass… if anything they should be designed to be more likely to fail.

1 Like
17

Hey Griff! Appreciate the thought you’ve put into this. I work at the Foundation with @lavande and helped study this issue so figured I’d share my two cents.

I understand that you’re flagging the scenario where NO voters withhold their vote in hopes that the proposal fails by not reaching Quorum. This causes early votes to favor YES, and you’re also highlighting that this may then impact vote herding because preliminary results are displayed and some voters will simply follow-on with the current majority consensus.

The dynamic you’re describing is theoretically possible, but I’m not sure that it’s likely to play out. A rational NO voter would need to weigh the impact of any perceived herding effect due to early voting against the likelihood of a vote not reaching quorum– a dynamic which already exists with today’s calculation methodology. For this to have a real impact, NO voters would have to believe that a proposal is unlikely to reach Quorum without their votes. The overwhelming majority of governance proposal exceed the quorum requirement with a healthy margin. A quick analysis of our past proposals on vote.optimism.io indicates removing Abstain from the calculation would’ve historically impacted only a single Code of Conduct vote from Oct 2023 and even then it would not have changed the outcome of this vote since the vote was defeated by NO voters.

I appreciate you highlighting that dynamic and agree that we want to ensure our voting system is resistant to such logical traps. We have to weigh the scenario you’re describing against the problem that this proposal intends to address wherein Abstain votes have the unintentional effect of allowing a potentially very small majority to determine the entire outcome of a proposal– which I think is a more first-order effect. I do think that your comment highlights the issue of vote herding, and there is likely some further experimentation and analysis that could be done to understand and address that phenomenon independent of this change.

18

If this change is going to have any impact, it would be that proposals with a lot of “Abstain” votes fail to meet quorum… and it is exactly in those votes that “No” voters would want to hold back their votes.

We are describing the same scenario, they are not different.

If this proposal passes, and the scenario it is supposed to prevent happens, and there are a lot of abstains so it might fail on quorum, then the voters that think the proposal should fail, should hold back their votes… and because vote herding is very real thing we need to expect, this makes proposals with a lot of abstains more likely to pass. This is bad, proposals that have a lot of Abstains should not be easier to pass… if anything they should be harder.

What am I missing?

1 Like
19

The following reflects the views of L2BEAT’s governance team, composed of @krst, @Sinkas, and @Manugotsuka, and it’s based on their combined research, fact-checking, and ideation.

We voted AGAINST the proposal.

First off, we believe there’s no ‘right’ answer to whether or not ‘Abstain’ should count towards quorum. We can understand both takes, and we believe both are subjective and heavily rely on the given context. That said, we are more aligned with the opinion that ‘Abstain’ should count towards quorum.

We understand @Griff’s points and we largely agree with them. Apart from the dynamic where voters who are against a proposal are disincentivized from voting ‘No’, thus skewing the appearance of a vote’s actual opposition, removing ‘Abstain’ also takes a valuable option away from delegates.

Right now, ‘Abstain’ is a way to help the governance process without necessarily having a strong opinion about the direction of a particular proposal. By removing ‘Abstain’ from counting towards quorum, we push those people to either not vote at all, and thus not help the governance process, or to take a position, even if they’re not comfortable doing so.

2 Likes
20

This is already the case. Currently, if the Yes + Abstain votes are close to meeting quorum, then a No vote could push it over the line.

If the No voters are strategic enough to withhold their vote in the hopes that quorum won’t be met, we can assume they also understand the issue you described; game theory would dictate they should vote early in order to prevent the Yes votes from setting the narrative.

I believe it is the opposite. If Yes voters can’t rely on Abstain votes to meet quorum, they must find more Yes votes in order to pass the proposal. The proposed change would indeed make it harder for proposals with a lot of Abstains to pass.

I don’t think this is the case. We are actually preserving the neutral signal by making it possible for the proposal to fail if there is too much neutrality (i.e. not enough voters have expressed a opinion).

If a proposal needs to rely upon Abstain votes to meet quorum, then it is not a strong proposal.

Final thoughts

By not counting neutral votes, the quorum requirement forces more involvement from additional participants, ideally to reach consensus.

Consider the following:

Scenario (token supply=100, quorum=40)
Snapshot 1: Yes=15 No=20 Abstain=10

  • Existing system: (Quorum is met and resolves as Defeated)
  • Proposed system: (Quorum not met and requires further input)

The proposal should not resolve because quorum has not been met. We actually need more representation to find a definitive resolution.

1 Like