We appreciate the proactive approach to fortify our governance against potential threats. However, excluding ‘Abstain’ votes from quorum calculations may have unintended consequences as explained by @juanbug.
We agree with @Chris_Areta that changing threshold requirements on critical proposals seems a better alternative to increase resilience to outside attacks.
If we want to move forward with changing requirements, we recommend that the Arbitrum Research and Development Council (ARDC) conduct a thorough analysis to assess the potential impacts of this change on past and future proposals. This would ensure that any modifications to our governance parameters are effective and inclusive.
In conclusion, while we recognize the intent to safeguard the DAO, we advocate for a measured approach that considers all potential outcomes and maintains the integrity of our governance processes.
As mentioned by @Atomist the ARDC V2 is currently working with Nethermind & Castle Cap on this. We aim to have a first version of this report available within the next 2 weeks - if then there is the desire to dive deeper on certain aspects e.g., exploring scenarios and its impacts on past and future proposals we are happy to further investigate it.
Personally I feel like having “No”-votes not count towards quorum in the first place is counter-intuitive, but I can understand the intent behind it. Having “Abstain”-votes also not count towards quorum seems like it would align with that intent, so for that reason I don’t oppose the suggestion.
However as has been pointed out, I’m not sure this will have a great degree of impact towards increased resilience. Some impact, granted, but not a great degree.
Perhaps a more impactful change would be:
“Yes”-votes count towards quorum proportional to weight
“Abstain”-votes have no effect on quorum
“No”-votes actively decrement quorum proportional to weight
It is a bit of a shift in how things work, granted, but it would certainly go a long way towards enforcing “controversial proposals require a higher degree of participation to pass”, which is the proposal intent.
To play devil’s advocate on my own proposal, this would create a new kind of vulnerability in that a saboteur with significant voting power could just vote “No” on every proposal to raise the bar on quorum, so that may need to be addressed. On the other hand such a vulnerability already kind of exists, in that said saboteur could already vote no to every proposal if they had an overwhelming majority, so perhaps it just lowers the bar on that kind of attack rather that introducing a new issue.
This is a misunderstanding of how governance works.
If we want to make it harder to pass proposals, then we can require a supper majority for support (e.g. 75% in favour vs against).
But eliminating Abstain from quorum is compounding over the mistake already of not including Against for quorum.
Quorum is there to ensure proposals can’t be sneaked through e.g. during a holiday break. And that’s not what this is about.
What we need are solutions so more people vote, and Arbitrum Tokens are more valuable. Not breaking down governance.
We think that there needs to be considered several vectors here, these are some of the most popular comments within the community
Removing the “abstain” option - hurts a lot of stakeholders because they can no longer voice opinion due to conflict of interests and some other reasons like not being technical enough to properly understand the proposal but still willing to participate etc.
There are plenty of examples within the traditional voting system on all levels where “abstain” not only exists, but counts towards quorum. In a way how the system is designed right now is aligned with global best practices.
More importantly it is vital to understand what kind of problem we are trying to solve, and it appears that we are going in the wrong direction. The definition of the problem is to prevent malicious attacks on DAO. Governance does not exist in DAO space only, it is all around us in tradefi space, the key differences here are mechanisms. Typically traditional tradefi corporate governance is a very “red taped” system, which relies on courts to enforce the mechanisms. DAO’s operate in a completely different reality where “code is the law” and no courts, if someone maliciously takes control over your entity / assets you are pretty much on your own. For example “airdrop” is a very new instrument which only works on blockchain.
Keeping that in mind we should not be trying to break some of the basic best practices of governance, but rather think about new innovative ways how we can protect our DAO.
Admittedly in my eyes it’s not the most elegant solution, but it works quite well - ENS implemented a security council which is a multisig which has power of veto over any proposal passed.
We should be working more closely with a community of delegates to create a large diverse base of voices with voting power spread more or less evenly - that reduces the probability of successful governance attack significantly.
Suggesting to break down the “abstain” mechanism, seems like a wrong solution to the posed problem. Since blockchain is a completely new world, we should invent new ways to protect DAO’s from attacks.
After listening to @cupojoseph’s arguments during this week’s call, I am heavily leaning towards supporting this change. Abstain votes within a DAO serve a very different purpose as opposed to you country’s Parliamentary process where votes are often counted 1 to 1.
This is why some company bylaws require that the sum of YES votes outnumbers NO and ABSTAIN in certain circumstances.
Amassing large voting power lets anyone indirectly back unpopular proposals without compromising themselves or their stances.
The sole fact that we can reference examples where Abstain votes’ effect on quorum could have influenced voting outcome is enough to warrant at least revising this topic.
Also, NOBODY SAID WE ARE REMOVING THE ABSTAIN OPTION
Having looked at all the pros and cons, I would also like to speak out:
Indeed, there may be such a situation (in theory) that 99% will abstain, and 1% will accept some proposal. However, in this case, I think it is obvious that the vote is not important, since everyone does not care and votes by abstaining.
The problem, however, exists, and it can be solved in another way.
For example, for critical votes, where key decisions are made and/or where a large amount of funds are sent (for example, from 10-20 million ARB), you can apply not changes to the quorum, but set requirements for 60-70% of votes FOR. Such a solution exists in other DAOs (for example, Jupiter on Solana). Also, this is used in state parliaments for key decisions (from 2/3 to 3/4 of the votes)
If we are going to set such a vote, then I would like this option to appear in it as well.
A simple analysis shows that the OpCo and Hackathon votes would not have passed (and if there was a limit on funds, only OpCo). I believe that the DAO would have benefited from this, since both proposals brought out a lot of debate in the community, and the goal of the community is not to divide, but to come up with the best solution.
With a threshold of 70% FOR, we know for sure that the majority of the community accepts this solution.
Agree with this. As mentioned in my previous comment, if a malicious proposal were to be posted, delegates still have the option to either not vote or vote against. I don’t see a situation where a malicious proposal is posted and it is passed because of delegates voting abstain.
Again, as mentioned in my previous post, the real situation we’re tackling here is controversial proposals. For that, I think higher thresholds is a much better approach.
Appreciate @cupojoseph for bringing up such an important topic. With Arbitrum DAO sitting on a $5B+ treasury, we totally get the worry about governance attacks. The OpCo example really drives it home—18.5M “Abstain” votes got the quorum to 140M, leaving just a 620k ARB gap for “Yes” votes. That’s only 0.5% of quorum, and with ~35M ARB up for grabs on Aave, someone could borrow enough to swing it for a few grand. That’s a bit scary, right?
But I’m not sure excluding “Abstain” votes from quorum is the best fix like what @Curia pointed out .
Here’s what we’re thinking instead. First, let’s set a minimum “Yes” vote rule—like, 40% of quorum has to be “Yes” to pass. For OpCo, that’d mean 48.72M “Yes” votes (with a 121.8M quorum), so the 620k margin wouldn’t cut it. Second, how about higher quorum for big spends? Say, proposals over 10M ARB need a 5% quorum and 60% “Yes” votes, but smaller ones stick to 3%. It’s kinda like what @Chris_Areta suggested, just with clearer numbers. To get more people voting, why not reward delegates who vote in 75% of monthly votes with a small ARB bonus from treasury yield? That could help with turnout.
This way, we tackle the security issue without ditching “Abstain,” which a lot of folks like for staying neutral. Thanks again, @cupojoseph, for getting this convo going!
This is certainly an interesting topic to ponder on. We would like to update all delegates that Castle Labs is currently collaborating with Nethermind and we are already looking into governance challenges and risks within the research topic for ARDC.
The impact of not including ABSTAIN votes to quorum is being explored along with other metrics to have a better gauge of governance in its totality. The research is currently in progress.
