EDIT
Thanks everyone for initial thoughts. To avoid this topic spiraling into a 100 post monster that takes 90 min to read, I’m going to break out suggestions into individual topics. I’ll add to this list as we go:
Also, just as a point of order, my opinion is that all current treasury allocations should be subject to this next round, meaning that current delegates with votes from the treasury would be subject to have their votes reallocated subject to the parameters of the new program we design.
Original post below:
@Doo_StableLab 's proposal to delegate UNI from the treasury to active delegates has increased governance participation generally and empowered a new group of delegates. With two months to go, I though it’d be a good idea to get a thread together so we can aggregate ideas about how to further optimize this program going forward.
Some general things i’m thinking about:
qualification criteria: activity level in governance is a good first step but generally misses developers who are building on and around uniswap. This is a class of ecosystem participants who might be good delegates and provide a technical viewpoint that could be valuable. I think it’s worth experimenting w/ reserving some votes for developers.
expiring delegation: h/t @Getty as we’ve been chatting on and off about this for a couple months. Right now the delegates with treasury votes are in a position where they have to actively vote to revoke them. This is obviously a misalignment of incentives and relying on social consensus is not a scalable route forward. Expiring delegation solves this problem, but would need to be thoughtfully implemented so that active voting power never falls off a cliff.
lowering ceiling for allocated votes: when the original proposal passed, the threshold to post a governance proposal was 2.5m so the distribution algorithm was greedy based on that number. The threshold has dropped, so I’d suggest a corresponding drop in the max allocation per delegate (which also implies more delegates w/ non-zero voting power, which is probably +ev).
This is self interested, so should be taken with a grain of salt, but I want to voice strong agreement with your assessment that developers are currently an underrepresented stakeholder group in Uniswap Governance.
Therefore, it would be crucial to also ensure that even if such developers or contributors receive voting power, there should be accountability measures to ensure activeness (the same for current allocated ones)
I think it would be also important to think about how much voting power should be allocated for as well. Especially if we have to account that some previous delegates that used to participate might end up less active (which unfortunately has been the case)
Agree on accountability but would note that neither of those delegates are developers. I think whatever process we used to allocate votes should sort for developers who want to be involved… and it should be viewed as an experiment!
Agree that developers should have more votes. Give how it seems we’re still struggling to get to quorum, we think another round of Dao led delegation should be in the works. This time, maybe have the race be up to 1m (proposal threshold) and still have 10m delegation today, with x% being reserved for developers.
My recollection is that the UNI tokens issued to Uniswap Labs as part of startup compensation was that 21.5% were intended to be non-voting (obvious conflict of interest when % community was relatively low).
My general Socratic response is what is the behaviour you wish to incentivise?
the list is already filtered for activity so the 2.5M is not payment for voting …
should there be a minimum as delegate? It takes 1M to make a vote able proposal so if we want every delegate to be able to issue say a vote of no confidence censuring another for misfeasance, then you need sufficient group of minority to get to that threshold;
if you want to measure outputs, you could look at metrics such as length of post relative to holdings so if you got millions of UNI and not being visible compared with delegate with just a few thousand … so rewarding posting activity / holdings would encourage more feedback
however you also want the performance metric to be resilient to gaming … If some idiot starts throwing ChatGPT into the gov channels just to get more UNI, then quality of discourse suffers
how do you encourage thoughtful feedback (or even measure it)
normal voters should signal their “trust” in delegates so perhaps consider the number of supporting votes relative to holdings? (But then VCs can concentrate on a corporate lobbyist)
if you want diversity of views (eg LPs and Devs are relatively underrep) then you need to give UNI to delegate platforms which are not clustered near to existing ones
long-term sustainability … Without clawing back UNI from passive delegates, some future date the issuance may need to reduce;
Another way to look at it is like functional electoral college, and you are reallocating growth into seats which are more representative but this needs careful thought given perception that investors can swing the votes if they really wanted.
The only thing we’d like to add is that we would be in favor of increasing the delegation pool from 10M to 20M to support bringing in existing Uniswap ecosystem partners and providing votes to new meaningful voices.
If we consider the voting blocs to be akin to a functional electoral college, there is one class that is conspicuously missing … regulators.
Reasoning
As treasury starts to take shape, a legal structure will emerge which cannot escape existing FATF rules and many hope Uniswap aims to be law-abiding, || even if certain tokens are alegal at the moment (see the section on pg2 I wrote for Wyoming) ||
If Uniswap DAO wants to go down the path of self/co-regulation, then you need a way to negotiate with regulators to assure them that regulatory equivalence is mutually desirable goal but without the onerous burdens (and SecOps risks)
It is easier to familiarise competent authorities with the principles and ethos of Uniswap DAO via the proposal/temp check than in adversarial setting … || Ideally moving towards a no-action letter on the mechanism of UNI as a non-economic governance token ||
Regulators as Delegate
nothing in the rules against this … If pick 5 jurisdictions each with 500k UNI they’d still need a seconder to get the 1M threshold to make formal voting proposals
legitimacy is in the eyes of the beholder … good faith bargaining allows for alignment with public policy such as addressing asymmetric information which leads to fraud
building in protocols such as computational forensics for evidence recordat allow for dispute resolution such as binding arbitration which can be enforced via existing legal system
Rational - want Robocop not Rambos
The adversarial legal mode is not appropriate for DeFi as it pits a centralised top-down control&coerce against scattered individuals. An alternative inquisitorial mode, akin to international aircraft crash investigation which aims to determine root-cause of failure and not fault-finding allows for course correction whilst preserving jurisdictional neutrality (if diverse regulators agree to be delegates). There are some darknet state-actors which are outside commercial sphere so rather than relying on self-help which can lead to … extreme outcomes … better to build bridges rather than barriers. If investors want safeguards, then they would vote in favor with the regulator delegates for any reasonable proposition.
Allocating too much UNI to this program is not so much the bottleneck—it’s more so the number of qualified candidates. Even if we have a 10M UNI pool spread out across 10 new delegates, I struggle to list 10 new parties that should qualify for more delegation just in the governance camp. Yes, more delegates have begun participating in governance since you can monetize gov participation. But I feel uneasy prioritizing new delegates who recently got involved in governance over alternative contributors. Since the pool of long-term, active delegates with lackluster voting power is relatively small at the moment, I think a larger portion of the total pool should actually go to devs, researchers, grantees, etc. Maybe a 60-40 split of the 10M is warranted.
Expiration makes things tricky. Too much VP from the treasury could just lead to collusion and permanence of existing delegates, so perpetual delegation is a valid concern. But this may make more sense to explore when the amount of voting power is a certain degree above the quorum, that way, even if we lose some delegates due to simultaneous expiration, proposals are still able to pass (ideally expiration is staggered…but this adds subjectivity as to how long delegation terms last for different delegates). That’s why it’s important for the community to elect parties that have a strong reputation in the community from the get-go, and as mentioned above, there are not enough of those simply on the governance side.
All of the candidates from the previous delegate race also have high voting participation rate and do a good job of communicating their rationales. So determining who loses delegation, or if it’s a ubiquitous removal of voting power, will be key. It may be easier to frame treasury delegation as a “seat” that you are voted into for x duration of time, rather than something you hold in perpetuity unless you become an adversarial actor. And after the delegate’s tenure is up, they are up for re-election against a pool of other candidates.
^I’d much rather allocate resources to delegates and other contributors that have clearly partook in the growth and development of the protocol as opposed to regulators. This is a gray area that I don’t think we are in a position to address today. Plus, there’s no explicit indication of interest on this front, partly due to liabilities that regulators may incur from partaking in governance. Note—I do not have expertise in this area…so this is just my hunch.
True, but that’s why the DAO spent $10M on the DEF this year. So, it’s not like the DAO has not considered its impact in the political space. If a an individual who happens to be privy to the political sphere wants to partake in the DAO, I think we’d welcome them. And once they demonstrate good participation, they can perhaps get voting power.
I don’t have time to map out properly but the principle component of your argument can be partitioned into 3
People
Commitment / Contribution / Competence
You desire thresholds for commitment beyond just monetary value of holding UNI … What are the signals?
uneasy prioritizing new delegates who recently got involved in governance
contribution - this is a trust issue which can be addressed via understudies, promotion of individual delegates to their own distinct platform, or allow prior experience in other AMMs to count. This is observable but takes time …
Competency is measurable … And upskilling courses exist such as MBAs modules on asset allocation or duties of independent director/trustee. It may be that part of the understudy component is subsidised training for the role of full participation. It may be that unsuccessful seats go into a random draw for observer status so you build up a pool of institutional knowledge despite the lack of social trust or commitment.
Platforms
Philosophy / Performance / Participation
Delegates would have certain philosophies or belong to different schools of economic thought from degens to crypto-anarchists. Whilst you don’t want to head towards political infighting, you do want ideas to bubble up … Perhaps some language to point out that non-traditional approaches are not a barrier so long as rational and willingness to engage in evidence-based reasoning
Tracking the performance of delegates and scorecard for delegations may be an automated process if suitable metrics are sought. Frankly, some of the proposals can be rather technical or complex so having some clue as to policy positions act as short-cut to digesting the details.
Development - focused platforms … So since the temp check is that certain groups (eg Devs) are underrepresented perhaps revisit the original presumption that UNI awarded to employees of Uniswap Labs be excluded be reconsidered? For example 5 year cliff, means UNI vested 5 years ago become delegate able or ex-employee, waivers granted. Technical directions or desired projects akin to existing “positions”. This starts bringing in the 20% which so far have been ineligible to vote but are savvy and may want to switch from a technical to business track.
Process
selection, election, rejection
Points raised re how selected/elected, criteria for retention, mechanism for removal. These are all details to be debated to get the desired balance between stable continuity and fresh blood. Obviously the treasury working group have examined other DAOs so there should be some clue as to good practices elsewhere.
100% agree on getting more developer representation. especially with uni v4 coming out, and with more hook developers entering the ecosystem, ensuring their interests are sufficiently represented will be an important input to their long term commitment to building on uniswap
@Doo_StableLab above has a great point. will note though that I believe the issue here is more on the side of passive delegators who have delegated and are not tracking the activities of their delegates. a trickier problem to solve particularly in the case of anonymous delegators
re other comments in the thread - I generally believe that, if the long term personal interests of the delegate are in line with the interests of the protocol, for instance, if a hook developer is building a project on top of uni v4, and they benefit as uni v4 grows in market share / activity, I we can care less about measuring “thoughtfulness” of feedback or length of post – their feedback matters simply because their interests/opinion matter, as contributors to long term protocol success
(this comment represents my personal opinion and not that of the UF!)
We believe that the delegation of UNI to developers, the delegation expiration, and the cap on delegation allocation proposed by Erin are important considerations that should be discussed and implemented. However, we also believe that an expansion of the program itself is crucial.
Due to the low voter participation rate for the large delegates in 2023, a proposal was passed to Delegate 10M UNI from the Foundation to active but under-represented delegates. This period is set for one year and will expire this December. After reviewing this program, we found that it is working well, but think that the current governance is not stable yet, thus we think it’s worth considering and evaluating an expansion of the program.
Detailed analysis
Regarding the activities of the selected delegates
Check the voting percentages of the selected delegates before and after this program.
Regarding the 10 Onchain votes before being selected
There were a few selected delegates that had been active in the onchain proposals before the delegations but some had not participate in all the proposals. Since the delegation, all delegates participated in all votes, which made the DAO healthier.
Program Effectiveness
Calculation of VP (Voting Power) contributed based on post-election activities
Since all delegates cast all onchain votes, we can say that the program contributes 10M UNI for every vote.
In 2023, before this measure was implemented, there were three votes that were QUORUM NOT REACHED, but since its implementation, all proposals have been EXECUTED (except for the latest one)
However, regarding this year’s proposals, there were many proposals in the 40M-50M range and some were near the quorum. The average number of onchain votes since the program went into effect was about 47.46M UNI.
Current risk of missing proposals
Out of the 11 proposals since the program went into effect, 6 proposals that would not have reached the quorum IF one delegate had not missed voting. In addition, we saw the latest proposal just missed the quorum as well. While we saw an improvement of the voting activities, the DAO is still at risk of missing important proposals to be passed (or rejected).
Detailed analysis of which proposal would be missed if a delegate had not voted
The following 7 delegates are considered under-represented with a VP of 250k or less, similar to the previous application requirements:
Curia, Tané, SEEDGov, Bobbay, Arana, DAOplomats, and Argonaut.
Conclusion
This program is working well, but the governance is not stable yet since some of the past proposals are close to missing its quorum, therefore we believe this program should be expanded further. As @GFXlabs suggested, we would also consider an additional 10M UNI suffices to achieve a stable and healthy governance state.
Options for expansion
The options that were voted on last time are as follows
Split equally among Qualified Delegates
Top 4 Delegates get delegation till 2.5M is achieved. The rest of UNI are split among the rest proportional to their votes won.
Top 4 Delegates get delegation till 2.5M is achieved (So likely first 4 delegates achieve 2.5 Mil. 5th gets some)
If the process is similar to the previous one, we believe that we should do the following steps: delegate’s candidacy, determination of the distribution method, Snapshot for the delegates (or additional parties like developers) to get delegations, and Snapshot for its approval.
Great analytics @Tane … There’s ~60+ topics under the Delegation category so after deducting pinned messages, there may be 40+ platforms. Elsewhere@jengajojo identified the most active quintile “deserving” of reward.
Stepping back, the decision about treasury delegation is a variant of a triage. There is a pool of UNI which is “lent” (OK it appears to be a absolute grant but shouldn’t be treated as irrevocable) so from a triage perspective,
delegates who have sufficient “weight” to swing the dial
semi-active delegates who may consider their paltry holdings not worth the “effort”
newly established platforms who lack street cred to delegators
non-performers (who may be weeded out … )
So triage is to allocate the limited resources where it may mobilise more delegators to choose those delegation platforms. What data would help make a more informed choice?
distribution of delegations VP sizes … Is it bimodal, lumpy or long-tailed?
have there been active points of discussion which subsequently are not reflected in the actual vote … This the “abstain” vote which is not easily captured unless a subsequent explanation given post-vote;
minority opinions … have there proposals where the larger VPs “oppress” the concerns raised by the minority;
have certain delegators gone dark due to Uniswap Foundation being the source of grant money - before and after analysis?
unknown is that due to original commitments, Uniswap Labs employees are not eligible to vote … If this was changed and conflicts of interest can be addressed, how many would delegate?
What other known unknowns are there to help make a decision?
Broad stakeholder participation: makes sense to include or even give a majority of this round’s share to developers/developer teams, possibly also researchers and grantees as suggested by @AbdullahUmar, who makes some good points on this.
Expiring delegation is an interesting concept. But it wouldn’t make sense to let active voting power expire off a cliff. Could you trigger a countdown after x days of inactivity by which the delegation then after another month expires? I also thought of having a built-in slow bleed where vp is drained over time with inactivity and gets restored after certain activities. Not sure how difficult it would be to implement.
1m delegations to reach prop threshold also makes sense
Seems very reasonable to me to allocate some to the 7 delegates mentioned by @Tane (great analysis, thank you!) - how much? Not sure. Another option instead of pushing 4 of the 7 delegates to 2.5m, you could also raise all to 1m and delegate the rest to those stakeholder groups we feel are underrepresented, be it developers or whoever?
I think gamifying forum activity is complex and, in reality, only a part of delegate governance. Length of post is, as you recognise, a poor metric as time and attention is precious and we want to encourage clear and concise discussion. Not sure about the posting activity / holdings thing either, to me it doesn’t seem reasonable
I don’t see why we would delegate to an official regulator just for the sake of it. What would it aim to achieve? a goodwill bribe for looking the other way or providing legal advice? Imo better to invite established crypto legal teams then that are already contributing to related ecosystems and working towards the betterment of this industry, like Axis Advisory (no affiliation) for example.
Partially agree. I’m not sure if removing votes from the better performing delegates is the right thing to do. We have to be careful to ensure that the quorum is easily reachable for non-contentious votes.
On the other hand, not having enough votes to post a proposal can be a source of friction. Since the threshold is now 1M, optimizing for that makes sense.
Last time we had a forum poll about the distribution options. The results were a close call:
Specifically I voted for the second option (who ended up winning) because of the 2.5M threshold, so that at least a few of the competing delegates would get proposal posting rights.
Hey @eek637- Thanks for bringing this up ahead of the end of the program!
TL;DR
SeedGov supports the continuity and expansion of the program. Raising voices and giving VP to small delegates is crucial for the resilience of the DAO.
We are open to finding a fair way to include unheard developers.
Increasing the delegated amount to 20M will reduce the risk of proposals not reaching quorum and allow delegates to abstain when appropriate.
Maintaining the 12-month timeframe with quarterly reporting will continue to yield positive outcomes for the program.
Lowering the criteria for receiving foundation delegation from 2.5M to 250K will increase focus on underrepresented delegates.
Thoughts on the ongoing discussion:
1) Allocations:
a)
We acknowledge this risk and therefore suggest increasing the allocation to 20M lent UNI. This increase would not only widen the margin of error for proposals to be passed but also allow delegates the flexibility to abstain and not vote merely because they feel obligated to do so.
However,
If this is the case, we have a concern about this overshadowing new delegates. We believe it would be fair to prioritize new unrepresented delegates, both tech and non-tech.
b) Regarding first Round criteria: “Must have voted on 80% (or more) of proposals (minus the proposals that were canceled) over the last 3 months” - We recommend sticking with the 80% because it worked well in the first cycle and aligns with the percentage used for the delegate reward incentive.
c) Regarding first round application criteria: “Voting Power is less than 2.5 million UNI” - According to Jengajojo’s list, there are currently five delegates participating in the rewards cycle with less than 250,000 Voting Power (VP), which represents than 10% of the 2.5 million VP threshold if the same criteria were to be used. We suggest lowering the bar to a more realistic level of VP, which would better align with the current distribution of voting power.
d) How do we feel about being 2.5M currently the 6,27% of the quorum?
Alternative ideas are valuable, and we agree this approach could be viable. Additionally, we’re wondering if increasing the pool is feasible. It might be worth considering a shift in perspective - moving away from ‘2.5M to 4 delegates’ and instead focusing on the current situation: We have X delegates, and based on Y criteria, they will receive Z amount of lent UNI
Technical Representation:
Implementing a developer quota would help support and encourage developer-focused delegates to participate. We think a more balanced approach could be more beneficial overall.
While the number of delegates with this characteristics is small, it’s still greater than the 4 or 5 delegates who would receive 2.5M UNI if we adhere to the original allocation. Also, the percentage split between technical and non-technical delegates could be determined after we know how many potential UNI recipients exist. If we set a quota beforehand, we might end up in a situation where developer delegates apply and receive UNI due to the quota, potentially leaving out other highly participative delegates without allocated UNI.
2 Expiring Delegation :
We believe setting expiration dates benefits the program. We think renewing the program for 12 months could still align with the original idea, as it has shown positive results in the first round. The concept of allowing renewal if delegates meet certain criteria also seems feasible and aligns with Kfk’s comment: ‘Longer periods would reduce campaigning overhead for re-election and other operational burdens for the DAO.’ Setting aside ideas that require developer resources, implementing a quarterly reporting requirement to maintain the participation rate could contribute to the program’s success.
